2019-08-24 01-07-36 PDT Valentin Ionita Hello, Tav, may I ask you for something? Can you leave me a recommendation on LinkedIn? It would be really helpful on finding an internship as a result of GSoC. 2019-06-27 10-24-38 PDT Valentin Ionita Firefox (https://chat.inkscape.org/file-upload/ZkeEDZqnh2RfictE6/Firefox) 2020-08-22 06-08-11 PDT Valentin Ionita and Paint Servers 2019-07-10 02-26-23 PDT Valentin Ionita Hello, Tav. Here it's pretty awful, I can't get much free time or internet. However, Inkscape was on my mind. 2019-09-03 07-33-07 PDT Valentin Ionita I think the first step needed was to connect 2019-06-09 00-59-53 PDT Valentin Ionita I'm looking at it 2019-09-01 09-23-52 PDT Valentin Ionita Thank you, I'll remind you tomorrow 2019-07-06 05-40-30 PDT Valentin Ionita I have just pushed the current version but can't make it out to work. Line 224 gave me an `GtkDialog mapped without a transient parent.` error 2019-06-09 01-20-00 PDT Valentin Ionita let's say there is only one patch. it's the limit case for the use, isn't it?, then, the colored points would be the circled ones in the next picture, wouldn't they? 2019-07-05 02-51-16 PDT Valentin Ionita No matter what we choose, the direction is important. Choosing the actual hatches and other assets will come as the last step 2019-08-19 10-48-48 PDT Valentin Ionita ok, thank yuo, I'll update 2019-06-24 09-51-20 PDT Valentin Ionita I have commited my latest code that has fixed the above green-red stripes issue and support for rotations and transformations. However, the latter are not working as intended, because I can't seem to find the right transform origin. As I understand it, it must be (x,y). Anyway, this is just a status update 2020-06-05 08-52-06 PDT Valentin Ionita Screenshot from 2020-06-05 18-51-23.png (https://chat.inkscape.org/file-upload/h6Dv7BB5qcnsthLz8/Screenshot%20from%20...) 2019-07-30 07-36-56 PDT Valentin Ionita However I fixed the issue I was working on (its pipeline is running now) 2019-07-10 02-27-16 PDT Valentin Ionita [ ](https://chat.inkscape.org/direct/Tav?msg=y98wQhbqdnyzoEYaE) Yes, this was on my target list, I would have liked to use the same model as for swatches. I mean, clicking would set the fill and Shift+click would set the stroke. () 2019-06-09 00-51-13 PDT Valentin Ionita So, I'm still struggling with the bicubic. All in all, I can't get how you interpolted 8x8 patches from just 4 points using bicubic. Can you explain in more detail? 2019-07-03 11-02-08 PDT Valentin Ionita For the rest, I assumed that the file is preformatted to respect those rules. However, I will add those checks. 2019-08-07 20-17-42 PDT Valentin Ionita Mornin'. Ok, I will do that 2019-07-04 08-59-57 PDT Valentin Ionita Screenshot from 2019-07-04 18-59-43.png (https://chat.inkscape.org/file-upload/9jSLdJmr978PwBhRQ/Screenshot%20from%20...) 2019-07-18 09-26-29 PDT Valentin Ionita Right now I am trying to connect the new load_document function with the modified connection in the current's document's defs 2019-08-18 07-50-28 PDT Valentin Ionita I'm not home at the moment 2019-08-18 07-50-16 PDT Valentin Ionita Yes, tonight I will 2019-07-26 20-37-24 PDT Valentin Ionita Well, I can't say I have an update. I've hit a wall, on having a gradient file to add just as we add the hatches and patterns. I couldn't find a solution that doesn't crash in some testcases. 2019-12-01 22-42-02 PST Valentin Ionita Hello, Tav, I know it's not your responsability anymore, but can you make me a short update on the current status of Inkscape? I know we are on the brink of a beta, but I'd like to know the direction we're going right now. I'd like to restart contributing :smile: 2019-07-04 08-53-18 PDT Valentin Ionita Yes, after GSoc, I would like to work on some of the documentation/compilation warnings. Just to improve some code quality. :smile: 2020-03-17 08-44-54 PDT Valentin Ionita What do you think would be specific enough and would bring value to Inkscape? 2020-08-21 23-56-06 PDT Valentin Ionita Our conversation starts on this thread on the UX issue 2020-08-17 07-26-43 PDT Valentin Ionita Honestly, that sounds like a brilliant idea. And the work you have done with the CanvasItem is important. I have been following the before&after 2020-03-18 02-58-02 PDT Valentin Ionita Either that or I'm not doing the right interactions 2019-06-27 02-22-32 PDT Valentin Ionita Screenshot from 2019-06-27 12-21-29.png (https://chat.inkscape.org/file-upload/dfYoy5PFSJkqt5WRQ/Screenshot%20from%20...) 2019-06-20 06-52-20 PDT Valentin Ionita Thank you, I'll use this for testing. 2020-08-22 00-16-24 PDT Valentin Ionita And then linked in the DialogBase::update() method 2020-08-14 02-21-34 PDT Valentin Ionita Honestly, it could be much worse. It's okay, I get the code. 2019-06-30 09-14-35 PDT Valentin Ionita No problem, thank you. 2019-06-29 10-08-14 PDT Valentin Ionita ok. I will 2019-07-04 05-31-23 PDT Valentin Ionita One problem, if I may 2020-03-18 03-16-31 PDT Valentin Ionita The first two are clear enough and the last thing is obviously an UX must-to. 2020-08-13 01-45-25 PDT Valentin Ionita Thank you, no problem, you won't need to worry today. 2019-06-26 10-58-14 PDT Valentin Ionita Should I do it from gitlab's own interface? 2019-08-17 08-28-47 PDT Valentin Ionita yes, I will pull from master then make a MR 2019-07-05 04-40-35 PDT Valentin Ionita But thank you for all the advice. I'll surely have questions, I hope you'll be around to answer 2019-06-26 11-15-05 PDT Valentin Ionita thank you 2019-07-10 03-07-34 PDT Valentin Ionita It would be nice. I'll make a make a method for the meantime. 2020-08-22 06-10-37 PDT Valentin Ionita it must be something related to the changes I made that fixes it. So it's about signals... 2020-08-22 00-01-32 PDT Valentin Ionita How so? 2020-03-18 03-08-44 PDT Valentin Ionita I have to be honest, this might be a bit too much for someone with my level of experience in GTK. However you know I'm eager to learn and help. What would be in your view an end-of-gsoc target for this (obviously a complete version is as important, but that should get to be into a proposal)? 2019-07-04 05-16-27 PDT Valentin Ionita oh, well, that's bad news. If it were to be my mistake it would have been easier to fix it 2019-07-06 06-12-53 PDT Valentin Ionita [ ](https://chat.inkscape.org/direct/Tav?msg=yH9uZjdC44hrm5oPb) I was going to ask about GC at some point in the future... () 2019-06-27 08-20-04 PDT Valentin Ionita Hello, just pushed the fix for the above bugs 2020-08-14 01-47-49 PDT Valentin Ionita I have a question, are you on Wayland? 2020-06-05 08-52-51 PDT Valentin Ionita these two callbacks are only used to connect to signals for the dropzones 2019-07-04 05-16-46 PDT Valentin Ionita I'm trying to test event taking to make it faster when we get past this error 2019-06-13 11-11-10 PDT Valentin Ionita Starting from Monday, I will dedicate a lot more time to Inkscape, it has been a quite time-consuming with the exam period, but I promise to make up for it and to push forward 2020-08-22 00-07-50 PDT Valentin Ionita Oh, the prototype dialog right now is a good example of how things should work. I updated/tweaked it myself for a better grasp of concepts. But I think the general scheme is the following: we have an update() method that gets triggered on relevant events through the callstack. Good. Now, in the update, trigger the dialogs internal callbacks, eventually checking if the reference to the thing you're tracking has changed (document, selection, whatever). This removes the need of signals most of the time. More than this, most dialogs with signals kept references anyway, so the complexity can only go down. 2019-06-27 00-59-11 PDT Valentin Ionita Oh, that's great. Then I'll push the license, to keep it in check 2020-08-22 00-10-17 PDT Valentin Ionita here 2020-08-14 01-43-08 PDT Valentin Ionita ok, then I'll work on reproducing it, and then look for a fix 2019-07-26 20-40-37 PDT Valentin Ionita [ ](https://chat.inkscape.org/direct/Tav?msg=am97t9rfdwA5KQdSW) @Tav I was afraid this would affect me. However, I received the payment just fine, so there's no problem. No need to worry. :smiley: () 2019-07-21 02-23-56 PDT Valentin Ionita Fixed it by adding inkscape:collect="always" to hatches.svg 2020-08-21 23-57-22 PDT Valentin Ionita And his blocker is no. 4 2019-09-03 07-34-48 PDT Valentin Ionita At least this is how it is on the mobile app 2019-08-06 20-22-12 PDT Valentin Ionita Yes, I would need to ask you what should the scaling knot handle should do to a hatch. Rotation is obvious, position is obvious, scaling isn't, and I have written them both. If you actually want to scale the hatch, then I would need some help with the specific maths 2019-08-15 20-47-36 PDT Valentin Ionita Hello, Tav, and good morning. I have received the final heads-up email from Google and I'll have time to submit my article and code between 19th and 26th. 2019-08-15 20-49-46 PDT Valentin Ionita About the hatchDialog branch (handling the paint server dialog), it's obviously on the community decision, but, when I have tried, I couldn't get their attention. ๐ 2019-06-26 04-33-36 PDT Valentin Ionita I have fixed the bug (it was a null check). should I push or wait another recommendations from you? Also, nice test :))) 2019-06-27 09-26-55 PDT Valentin Ionita Hello, again, Tav. I have something to remind you. If you remember, when I made my GSoC application, I said that I have applied to a summer camp too. In the meantime, I have been accepted to both, and that means I will be abroad from 7th of July to 17th of July. There are ten more days until then, and I thought I should tell you to be prepared. With these two features almost ready, I think I am on track. After I get back, I will double my efforts into Inkscape. I hope all it's well with this break and I know you will understand. If you think I should push harder these days, just give me some pointers about the development direction I should pick off and I'll do it. 2020-03-18 03-17-05 PDT Valentin Ionita I kinda like how this sounds. So I've decided to take upon this. I'll write the proposal by tonight. 2020-06-05 07-44-14 PDT Valentin Ionita I have gone through the first paned example (at last), and fixed a bug which stopped adding the same dialog after closing a notebook in which it was before 2019-07-05 11-58-26 PDT Valentin Ionita Sure, will think about that. Surely, that will evolve somehow into an editor (mostly color, probably) 2019-06-27 01-08-05 PDT Valentin Ionita in a moment 2019-08-11 20-29-50 PDT Valentin Ionita Hello, Tav, no, unfortunately. In my issue with updating the hatch, I couldn't find a solution. I have added a few missing interface things for hatches (now it appears in the status bar if you have an object with a hatch fill or stroke).
Another thing is that I have started my article. I have done about half of it. Soon, I will be ready for a few rounds of feedback. 2019-06-17 06-29-28 PDT Valentin Ionita Hey, Tav, I am back to work from the workshop. I have pushed some changes I have done today, now the performance of pepper.svg is at 9.9s . I would have a few questions that I need your pro opinion
1. In `sp-mesh-array.cpp` at lines 1714 (and a few others) we have an if based on the condition that a distance is non-positive. Would that ever happen? While running pepper, I have added some debugging messages to know when this happens, but they are never triggered. It would be nice to cut the branching out. 2. In the `Patch` class, there are three functions you have used for debugging, that I have not used and assume they are safe to cut out. 3. Should I add my name in the head of the file as a contributor? 2019-07-03 11-05-42 PDT Valentin Ionita Ah, I haven't actually done ninja install. I only run from the executable 2019-07-08 03-23-01 PDT Valentin Ionita Hello, Tav, finally got a moment to talk. I don't have any internet at the dorms, but I will try to work on the last bugs I talked about and push whenever I can. 2019-06-09 01-27-45 PDT Valentin Ionita yes, because when you have at least 2x2 patches you have the necessary number of points for an interpolation 2020-03-17 08-42-03 PDT Valentin Ionita I've seen the headless Inkscape and that is interesting. 2019-06-30 07-15-38 PDT Valentin Ionita Nevermind, was just too distracted by the code to realise it was your own sandbox 2019-06-26 04-38-50 PDT Valentin Ionita Actually, there is a visual artifact (a color inversion). I will let you know when I fix it 2019-07-06 08-48-48 PDT Valentin Ionita Tomorrow I will have a few preparations to do and after that my flight, so I will be unavailable for code. But I will read any messages you give me at night. I will tell you tomorrow if I have my laptop with me and can work in the camp too. 2019-07-18 06-11-55 PDT Valentin Ionita Is there any signal triggered when the definitions of a document are changed? I am thinking of that being the moment where I should rescan the current doc 2019-06-26 10-59-19 PDT Valentin Ionita Ok, I'll ping you if I encounter issues 2019-07-29 02-17-19 PDT Valentin Ionita As I haven't given you an update lately: here it is. I have continued testing with gradients but, while I haven't found a real solution there (but maybe a lead), I have found a bug, if I could call it that way (it's actually a behavior I haven't given much tought till now, actually) - let's say you turn an object into a pattern, it gets added to current document's definitions, however, it's not synced to the TreeModel for the current document in the paint server dialog.
That's what I am working at now. 2019-06-20 06-52-29 PDT Valentin Ionita For the moment, this is what I have 2019-08-20 02-19-19 PDT Valentin Ionita Thank you! I have completed the evalutation this morning. 2020-08-17 07-19-53 PDT Valentin Ionita Hey, Tav, lately I have been fixing the most annoying bugs and crashes (including a couple gtk warnings) 2019-07-05 05-29-53 PDT Valentin Ionita A few questions on the above tasks. 1. I am aware that a TreeModel is dynamically synced with an IconView, so there is the problem of how do we filter the icons for the selected document. You proposed having an IconView for each document. However, I believe it would be better to have a different store for each document, and changed the attached store of the IconView when the ComboBox state is altered. This way we could easily have a view for all icons. 2. Would a label be needed if I have the name of the document in the ComboBox? I could have the string in the svg's title inserted in the document root instead of the filename... 3. How is the fill connected with the hatch color, for example? I mean, it wouldn't change a thinng, that string. What it matters is each hatchpath's fill attribute, which I haven't worked with for this panel, yet... 4. Yeah, I find it rather difficult working with regex in Glib, so I would appreciate some help or a link to a tutorial (the documentation is vague...) 2020-08-21 23-56-07 PDT Valentin Ionita https://gitlab.com/inkscape/ux/-/issues/38#note_399684778 2019-07-03 02-58-11 PDT Valentin Ionita If you can, I could push my latest version. The pixbufs get added into the iconview but they don't show up... 2019-06-30 07-08-21 PDT Valentin Ionita In the actual app, what is the ink dialog? 2020-08-14 01-48-06 PDT Valentin Ionita I think that's why I can't reproduce your bug. 2019-07-02 05-58-43 PDT Valentin Ionita I am okay with everything above. I think 2020-06-05 07-46-21 PDT Valentin Ionita That doesn't mean I'm not doing anything, just that I don't know if what I'm doing is what I'm supposed to be doing or even if I won't delete later what I'm writing now 2019-07-06 02-16-50 PDT Valentin Ionita Got it, will take care of all that later today ๐ 2019-08-11 20-59-46 PDT Valentin Ionita Great, thank you. It's gonna be enough, I think 2019-07-05 04-43-37 PDT Valentin Ionita I'll keep you posted, anyway (I want to get the most advice while we're still not constrained by both our schedules) 2019-06-09 01-22-29 PDT Valentin Ionita https://drive.google.com/file/d/1-hkNbkBhqGxev-RWArv-lLRwryKlNZgl/view?usp=d... 2019-07-03 10-44-00 PDT Valentin Ionita Also, I should save the node representation of the actual hatch element in the store, shouldn't I? 2019-06-20 06-02-01 PDT Valentin Ionita From what I understand, objectBoundingBox means that for the x,y,pitch and rotation attributes, the values are expressed as [0, 1], as a mapped interval to the calling object's [0, boundingBox.width] 2019-07-04 08-59-02 PDT Valentin Ionita Which function doesn't need any arguments? Right now I have
``` for (auto &filename : get_filenames(Inkscape::IO::Resource::PAINT, {".svg"})) { std::cout << filename << std::endl; SPDocument *document = SPDocument::createNewDoc(filename.c_str(), FALSE); std::vector elements = document->getObjectsByElement("rect"); document->getRoot()->requestDisplayUpdate(SP_OBJECT_MODIFIED_FLAG); document->ensureUpToDate(); renderDrawing.setRoot(document->getRoot()->invoke_show(renderDrawing, key, SP_ITEM_SHOW_DISPLAY)); ``` 2019-07-04 09-52-12 PDT Valentin Ionita Screenshot from 2019-07-04 19-52-02.png (https://chat.inkscape.org/file-upload/2btqvKzyiXpSe5jMA/Screenshot%20from%20...) 2019-07-03 01-35-29 PDT Valentin Ionita This is how my list column looks like ``` class HatchesColumns : public Gtk::TreeModel::ColumnRecord { public:
Gtk::TreeModelColumn hatch_id; Gtk::TreeModelColumn< Glib::RefPtr hatch_image;
HatchesColumns() { add(hatch_id); add(hatch_image); } }; ``` 2019-06-20 05-41-47 PDT Valentin Ionita ```
``` 2019-07-22 08-37-02 PDT Valentin Ionita For the moment I see it's not open. I looked both this morning and now and it looks it hasn't started yet 2019-06-29 00-52-32 PDT Valentin Ionita nevermind. fixed 2019-07-26 20-35-55 PDT Valentin Ionita Hello, I've just woken up. 2019-07-26 20-38-55 PDT Valentin Ionita Maybe you have a special idea on how to add a single gradient for each. Another possibility would be actually modifying the code, but that might not be garbage collected 2019-09-01 09-23-43 PDT Valentin Ionita Wooow, that sounds great - the Alps. 2019-06-29 10-02-39 PDT Valentin Ionita Hello, Tav, I have a question. I have already been wasting too much time on this and I need help. How do you call a function on an event. I have been trying to call a function when a button is clicked (a simple test case to prepare me for later), but seems like everything I do is not even compiling. :cry: :cry:
I have tried both `g_signal_connect()` and `button->signal_clicked().connect()`. Any pointers for me? 2019-07-03 21-41-43 PDT Valentin Ionita More than this, SPObject's getAttribute method has an assert inside and I don't know how to set at compile time to stop the assert so I could see if my try-catch works. 2019-06-09 01-04-28 PDT Valentin Ionita What are the inputs and the outputs of this function? As an input, we get 4 points (coordinates) and 4 colours and as outputs are 8x8x4colours ? 2020-06-05 07-45-34 PDT Valentin Ionita Now I am trying to integrate the new multipane in the demo, but that means I need to detect signals and you have added a number of 'Deprecated' comments (and I think you meant you want the interface to the class to be different) and I have no idea where I should start from. 2019-08-09 20-14-30 PDT Valentin Ionita Yes, that's pretty much it. Now that you mentioned the missing signal, I know what to try. Thank you! 2019-06-27 01-17-30 PDT Valentin Ionita Okay, I'll go for two flags for now. Gonna ping you when I'm stuck 2019-08-19 10-51-14 PDT Valentin Ionita I pushed 2019-07-05 04-39-34 PDT Valentin Ionita That's quite a lot what you're throwing at me (still a bit unfamiliar with gtk), but I am sure I can make it through, one step at a time. I will begin in 15min. 2020-08-22 05-50-17 PDT Valentin Ionita Hey, it compiled and ran correctly, but I have the following console warning/error 2019-06-25 10-38-30 PDT Valentin Ionita One last thing from me tonight (and I promise I'll stop :joy: :joy: :joy: ) โ I pushed a somehow finalised versionof the hatch polyfill on my branch, the hatch_test file is also there (but a "fixed" version), so you can check it out whenever you have time. If You approve of it, I will create a MR and start thinking about the next task... 2019-06-27 02-13-32 PDT Valentin Ionita I have fixed the first one, the second is about objectBoundingBox units, which I am unsure how to fix yet. Will do till the end of today, now I have to go get some paperwork done. 2020-08-20 07-44-21 PDT Valentin Ionita I did, a few times, but I'm not consistently testing there, because, as a user, I enjoy using my graphics card which doesn't work under wayland 2019-07-04 05-41-43 PDT Valentin Ionita [ ](https://chat.inkscape.org/direct/Tav?msg=6AgMrj9KPYyZJm4jY) A XML::Node, some actual printable code of the representation of the node... () 2020-08-17 07-22-22 PDT Valentin Ionita And I'd like to ask if you could point me where the code for the canvas rulers and scrollbars is, if you know 2019-07-10 04-19-25 PDT Valentin Ionita I have fixed the behavior with groups and pushed my changes 2019-07-03 02-58-17 PDT Valentin Ionita I can click on them... 2020-08-22 00-10-14 PDT Valentin Ionita https://gitlab.com/vanntile/inkscape-gsoc-2020/-/blob/master/RFC.md#dialogba... 2020-08-14 01-48-50 PDT Valentin Ionita And I am aware that dialog popping behavior is inconsistent in Wayland (and on Windows sometimes) 2019-07-18 07-24-21 PDT Valentin Ionita I'll work on turning the extraction to a separate function and implementing the things necessary 2020-06-07 11-20-55 PDT Valentin Ionita I have rewritten completely the on_size_allocate function (with the help of a well algorithm-versed friend), fixing all the bugs, improving performance (theoretically) and reducing the code almost 3x. You can see it in my latest commit. 2019-06-22 23-48-33 PDT Valentin Ionita https://gitlab.com/inkscape/inkscape/issues/309 2020-08-22 00-02-02 PDT Valentin Ionita Oh, well, I feel that's only normal 2019-06-11 06-20-34 PDT Valentin Ionita Hey, Tav, just a ping with my progress. Bicubic interpolation is almost done, I have a symmetry problem that I'm gonna solve soon (shouldn't be that hard). Now, I gotta start studying for my last exam. Have a great Tuesday! 2019-06-27 09-41-38 PDT Valentin Ionita (part of the problems are because I haven't updated the minified version of the script. Will fix the others) 2019-07-06 03-50-44 PDT Valentin Ionita Oh, got it then. No problem, I will change it, I wanted just to understand 2019-07-18 09-25-45 PDT Valentin Ionita It might be solved by the formatting of the Patterns file 2019-06-27 01-28-42 PDT Valentin Ionita ? 2019-07-03 02-57-25 PDT Valentin Ionita I tried IconView too. 2019-07-18 09-25-58 PDT Valentin Ionita I'll try to figure it out next days. 2020-09-15 10-51-46 PDT Valentin Ionita [ ](https://chat.inkscape.org/direct/Tav?msg=DRMfqWu7XuuhqQ5zh) I think it's only on my branch, because of the number of the conflicts I fixed for a proper rebase () 2020-06-05 08-52-17 PDT Valentin Ionita Screenshot from 2020-06-05 18-51-45.png (https://chat.inkscape.org/file-upload/x7W7uC8ZNffNYuRGz/Screenshot%20from%20...) 2019-06-28 03-18-59 PDT Valentin Ionita I think that I can answer 3 myself. They both inherit Widget::Panel so probably no 2019-06-09 01-23-38 PDT Valentin Ionita and we calculate the inner derivatives for the four inner points and the outer derivatives for the eight boundary points, don't we? 2020-08-22 05-50-27 PDT Valentin Ionita Screenshot from 2020-08-22 15-49-46.png (https://chat.inkscape.org/file-upload/TCixLvdPRRvPdZtM2/Screenshot%20from%20...) 2020-03-18 02-58-54 PDT Valentin Ionita From a UI standpoint of view, it generally looks okay, pretty similar with the GIMP behavior (as you started from that idea). From the code, I am yet to pronounce myself. 2019-06-26 11-55-11 PDT Valentin Ionita This is for you to check, whenever you have time. I couldn't wait for it to merge, so I pulled from the mesh branch the latest changes (I needed the readme). 2020-08-22 05-50-37 PDT Valentin Ionita it happens just before finishing loading. 2019-07-18 06-08-31 PDT Valentin Ionita Glad you had a good flight 2020-08-18 01-02-56 PDT Valentin Ionita And thank you for keeping me up to date 2019-06-20 06-55-41 PDT Valentin Ionita great, will use 2019-06-20 05-57-47 PDT Valentin Ionita Well, I should report that then. Maybe later today 2019-06-27 11-38-54 PDT Valentin Ionita Yes, I know, but then Chrome renders strangely two consecutive M commands by putting a new line between them. In FF it doesn't happen 2019-07-04 05-31-32 PDT Valentin Ionita Actually, a question 2019-06-26 10-56-30 PDT Valentin Ionita For the moment I have a problem wit rebasing. I can't seem to mae it work. How should I update the merge specifically? 2020-08-14 02-20-10 PDT Valentin Ionita Great, this helps. However, it's more an UX decision what should be in the tooltip. 2019-06-28 03-53-05 PDT Valentin Ionita ok, well, it will probably take me a few days to get knowledgeable with the topic. Until then, have a great day! 2019-06-13 11-02-08 PDT Valentin Ionita The bicubic interoplation is working 2019-06-26 10-20-36 PDT Valentin Ionita [ ](https://chat.inkscape.org/direct/Tav?msg=w5dtAQXRfREvDLAkJ) Surprised to hear that the green areas are fine. Moreso, blinear intrpolation is fine, but I must've missed something, because I default to bicubic interpolation whenever there are more than 2x2 patches, so bilinear is actually not used there. () 2019-06-26 10-26-40 PDT Valentin Ionita Glad to say that this was my mistake. Now everything renders correctly, with the small precision artifacts 2020-08-21 23-56-54 PDT Valentin Ionita It's reduced to three issues, one which is completely opinion-based which I will ignore (no 5) 2019-07-06 05-39-40 PDT Valentin Ionita Hey, Tav, any advice on how do I add the actual paint server to the defs section? 2020-08-14 01-50-08 PDT Valentin Ionita ok, I'll choose one day when I'll work only on Wayland until I have a better answer than "inconsistent" 2019-06-22 23-47-53 PDT Valentin Ionita Apart from this update, I wanted to ask if I should get any attention for my merge request and to whom 2019-06-09 00-47-50 PDT Valentin Ionita Wondering if you have a moment 2019-06-09 01-30-40 PDT Valentin Ionita exactly right 2020-06-07 07-46-28 PDT Valentin Ionita Totally understand, don't have to stress about it. I'll try to fix it myself and will give you an update tomorrow 2019-06-26 11-55-44 PDT Valentin Ionita Anyway, they're probably gonna be some changes tomorrow and tonight's getting late. Have a good night! 2019-08-17 07-09-05 PDT Valentin Ionita No need to rush today, hope things will ameliorate. 2019-07-03 01-33-36 PDT Valentin Ionita https://developer.gnome.org/gtkmm/stable/classGtk_1_1FlowBox.html#ac54c1f423... 2019-07-04 09-55-24 PDT Valentin Ionita Anyway, I'll stop for today. 2020-08-18 01-01-34 PDT Valentin Ionita Yep, I am aware about his high number of requests/suggestions. 2019-07-04 05-34-45 PDT Valentin Ionita For now, I load a single file (or all the files in the hatch folder) in the constructor. I was thinking on doing something similar as to what's in the Symbols Dialog, with a grid widget with multiple section, the first that would be a selector (dropdown) and the second the preview.
The question is: should I load all the files possible in the constructor, destroy it in the destructor and only show the relevant data based on which is the selected option in the dropdown? 2020-09-15 10-51-52 PDT Valentin Ionita I haven't checked master yet 2020-08-22 05-35-26 PDT Valentin Ionita If it builds, I guess it's okay merging your branch. 2019-07-18 06-08-03 PDT Valentin Ionita At the moment, I have added a comboboxtext to select fill or stroke, and fixed a bug when using ustring::replace 2019-06-13 11-01-43 PDT Valentin Ionita Got news for you 2019-06-26 10-26-01 PDT Valentin Ionita Screenshot from 2019-06-26 20-25-40.png (https://chat.inkscape.org/file-upload/4duPAyiZLdYHgbDoc/Screenshot%20from%20...) 2019-06-27 02-23-41 PDT Valentin Ionita As the definition is this, how should I scale pitch and rotate? because I don't see why rotate would be affected by the hatchUnits. x would be scaled by the shape width and y by the shape height, am I right? 2019-07-10 04-14-41 PDT Valentin Ionita [ ](https://chat.inkscape.org/direct/Tav?msg=8jdGQDqb9jLsBg2o3) Tav, can you help me with this? () 2019-06-20 05-58-11 PDT Valentin Ionita Do you have a moment for me to ask a few questions? 2020-08-22 05-28-09 PDT Valentin Ionita I'll try rn and come back with an answer 2020-08-22 06-08-54 PDT Valentin Ionita oh, ok 2020-08-14 01-57-37 PDT Valentin Ionita 4) Yes, I was going to ask about this on the dev channel. I don't know how to do it. If anyone has any idea, I could guide them where that icon needs to be inserted. 2019-06-27 01-35-26 PDT Valentin Ionita Cool 2019-07-04 06-14-18 PDT Valentin Ionita Wait, you mean just to copy, let's say, the patterns file and format it accordingly in the specific directory for paint servers? 2020-08-17 07-21-47 PDT Valentin Ionita Now I'd like to focus on a non-critical UI bug - when opening a dialog from a clear, no-dialog, window, the rulers and scrollbars get turned on. 2019-06-27 12-08-30 PDT Valentin Ionita Found the error. Chrome doesn't render the last point, so my script doesn't calculate the width right. 2019-06-24 04-31-07 PDT Valentin Ionita https://www.w3.org/TR/2016/CR-SVG2-20160915/pservers.html#Hatchpaths 2019-06-27 10-22-47 PDT Valentin Ionita Because of rounding errors, I assume, there appears one rendering artifact in hatches with contentUnits=objectBoundingBox 2019-08-02 09-04-04 PDT Valentin Ionita I'm gonna stop for today 2019-07-03 21-56-18 PDT Valentin Ionita Anyway, I have pushed my changes 2019-06-09 00-55-01 PDT Valentin Ionita and the input size is what? I guess that the bicubic() function is called for every patch, that has only for color values (the corners) 2019-07-04 05-30-06 PDT Valentin Ionita [ ](https://chat.inkscape.org/direct/Tav?msg=F4fm9pH5xtdKRj35i) Similar to HTML dom selectors () 2019-07-04 05-43-05 PDT Valentin Ionita I tried like this *->getRepr()->content()* but failed... 2019-07-10 02-56-39 PDT Valentin Ionita What code you'll push? 2019-06-09 00-47-38 PDT Valentin Ionita Hello, Tav 2019-06-18 08-49-31 PDT Valentin Ionita https://gitlab.com/vanntile/inkscape/tree/mesh-polyfill 2019-07-19 04-50-38 PDT Valentin Ionita A question, let me try to explain a behavior I have seen and I want to ask if it's a normal one. Let's say you have a rectangle, you set its fill on a pattern, then a second one. When you delete the rectangle, both the patterns are deleted from the definition section, not just the one that is currently set. Is this normal? 2020-03-17 08-45-25 PDT Valentin Ionita Currently, I just can't see what would help the most. 2019-06-24 04-28-30 PDT Valentin Ionita How should I go on? 2019-06-24 04-43-56 PDT Valentin Ionita Yes, I know, that's just how I think. Glad to be of common coding style 2019-06-24 23-26-06 PDT Valentin Ionita Just wanted you to know the first GSoC evaluation is live and I have already completed my part (I'm sure you already got an email, but I'm a bit OCD) 2019-06-18 22-43-16 PDT Valentin Ionita Oh, I really forgot that. Is a copy of the file sufficient? Or should I make a backup branch? 2019-06-24 04-35-34 PDT Valentin Ionita And that's one of the checks in the code 2019-07-03 21-40-41 PDT Valentin Ionita [NOT DONE] * Need to copy TestHatches.svg to the proper directory when installing. -> don't know how to do this, not even where to look for. [DONE] Test that fill string is not empty, if empty, print error message to std::cerr and continue. [DONE] Should test length of fill string to make sure it is long enough for finding the substring (otherwise Inkscape will crash). [DONE] * Errors should go to std::cerr and then use continue rather than if/then. [NOT DONE] * Test that document is not null. -> If the document is missing, there wouldn't even be returned by *get_filenames()*. How could even there be a null document? [NOT DONE] std::vector elements ... should be right before for loop. -> Do you mean the outer for or the inner one? if it's about the inner one, than okay, if it's the outer one, I don't understand why, the elements on that iteration are specific to that document. The store of documents is external. 2019-06-27 10-24-48 PDT Valentin Ionita Chromium (https://chat.inkscape.org/file-upload/bvXZWdiYSRfGSfcMb/Chromium) 2019-06-27 10-26-04 PDT Valentin Ionita These are pretty hard glitches, I have an idea to fix the first one, but nothing for the second 2019-06-20 05-41-40 PDT Valentin Ionita The code is 2019-07-22 08-33-18 PDT Valentin Ionita Don't rush with the code if you're busy. There is actually no problem. Better take care of things with attention then sooner. For the moment I have fixed the pipeline fails in my branch. 2019-06-13 11-13-36 PDT Valentin Ionita Have a great evening. Goodbye 2019-06-24 04-31-18 PDT Valentin Ionita Isn't this where I should be looking? 2020-08-20 07-44-43 PDT Valentin Ionita I'll check your reviews tomorrow 2020-06-05 08-55-42 PDT Valentin Ionita anyway, right now I merged the code from the new paned (without compile errors), and try to make functionality work properly 2020-08-17 07-25-13 PDT Valentin Ionita ok, great. I'll take it from here. Thank you 2020-09-15 10-51-14 PDT Valentin Ionita Also, after merging your code, I got a number of warnings and errors in the console (no crashes tho). I will try to fin out the pitfalls and edit the commits that need changing 2020-08-22 06-15-03 PDT Valentin Ionita line 45 2019-06-22 23-47-19 PDT Valentin Ionita Hello, Tav, I'm still working on hatches, I had some problems with deprecated APIs that worked in some browsers but not in others... Anyway, I have written from the bottom up the parsing of the path datapoints and will work further on rendering hatches. 2019-06-24 04-15-08 PDT Valentin Ionita As I understand the tests, on the left of each pair is how Inkscape renders the hatch, on the right how it should render it, the text on the side matches the actual hatch code. 2019-07-21 03-22-09 PDT Valentin Ionita I have pushed a working version of the thing I was working on. From my POV, it has a logical flow. If you could check the last couple of commits it would be great. 2020-06-07 06-19-52 PDT Valentin Ionita Hey, the algorithm for size allocation is broken from my POV. If you could have some time to make a live coding session for me to explain what is wrong with it, that would be great. If you don't, I'll try to write a detailed issue 2019-07-20 23-12-30 PDT Valentin Ionita Because you'll probably not see this very soon, I'll ask on the dev channel if I can add a definition, after I understand how the previous ones are defined 2019-06-26 11-06-06 PDT Valentin Ionita I think everything's fine, ready to merge 2019-06-28 02-59-33 PDT Valentin Ionita After looking at Swatches (I use them all the time in my design projects) and Symbols (I have never used them) and at the source code in `dialog-manager.cpp`, I have a few things to say/ask. 1. First of all, I think that the functionality we should be looking for is something between the two above. I imagine, let's say, for hatches, a rectangular preview of every hatch, arranged in a grid, that changes your fill of the current selected object when you click on a hatch (the swatches functionality). However, for the UI, I would go for something more like the Symbols dialog, with a hatches set dropdown and icons for removing hatches from the current document and the zoom level. I think that a color choice would be useful too. 2. Should I build a file resembling `share/patterns/pattern.svg` and the appropriate files? Or could I get a hatch collection (visually) or inspiration from a specific place? 3. In `dialog-manager.cpp`, is there any difference between swatches being a panel and symbols being a dialog? 4. Are there any more implementation files other than the `.cpp` and `.h` from `src/ui/dialog/`? 5. Any quick GTK guide or development process you would recommend me? I know that everyone from Inkscape can help me with this, but I just wanted to hear your take on it. I have Glade and gtk3-demo installed. 2019-06-13 11-10-08 PDT Valentin Ionita Thank you 2019-07-30 07-37-25 PDT Valentin Ionita Starting tomorrow I'll look at hatch handles. I guess that on another branch. :) 2019-06-28 03-43-51 PDT Valentin Ionita [ ](https://chat.inkscape.org/direct/Tav?msg=5bsNp4TKuHtjY8Zic) it's okay, found it () 2019-07-02 08-47-53 PDT Valentin Ionita I did try doing something similar without copy-pasting, because I find the symbols code convoluted over common sense. For showing them, I have tried using an IconView and it had no result, and I have pushed with FlowBox and it still has nothing showing up. I probably don't get anything in the pixelbuffers, but don't know how can I check if there is something useful in them. 2019-06-20 05-59-20 PDT Valentin Ionita I'll be asking in reference to this section https://www.w3.org/TR/2016/CR-SVG2-20160915/pservers.html#HatchElementAttrib... 2019-06-24 04-27-05 PDT Valentin Ionita Screenshot from 2019-06-24 14-26-54.png (https://chat.inkscape.org/file-upload/gF9NsJfBdqAB3tMpA/Screenshot%20from%20...) 2020-03-18 03-17-23 PDT Valentin Ionita Thank you for your time, once again. :smile: 2019-07-05 04-42-03 PDT Valentin Ionita no problem. 2019-06-30 09-10-12 PDT Valentin Ionita A question. What Gtk Widget should be drawn on? a DrawingArea? 2019-07-02 08-44-59 PDT Valentin Ionita Hi. Status update. I have pushed the first commit to the `hatchDialog` branch, with the changes necessary for the dialog appearing in the menu and everything else. Right now, nothing it's showing up, I have not succeeded rendering hatches as images. Don't know what to do for now. I'll finish for today. 2019-06-13 11-01-53 PDT Valentin Ionita Screenshot from 2019-06-13 21-00-56.png (https://chat.inkscape.org/file-upload/TLQbucTLZ6aLvdtGj/Screenshot%20from%20...) 2019-07-18 06-10-19 PDT Valentin Ionita As I understand it, you think of scanning the document whenever we load them and add a new hatch, don't you? 2019-06-27 01-28-40 PDT Valentin Ionita I think I edited all that is needed in this file. However, don't I need to edit something for the setting here `/options/svgexport/hatch_insertpolyfill` 2019-07-03 03-21-24 PDT Valentin Ionita I did. You can find it here 2019-06-27 01-39-44 PDT Valentin Ionita Don, I have pushed 2019-07-04 09-55-47 PDT Valentin Ionita Thank you very much for the help with the render and the new target, you've been invaluable. 2019-06-20 05-41-29 PDT Valentin Ionita Hey, Tav, do you have any idea why Inkscape renders the lines in this hatch with the same weight? 2019-06-26 11-54-33 PDT Valentin Ionita https://gitlab.com/inkscape/inkscape/merge_requests/754 2019-07-19 04-48-44 PDT Valentin Ionita Hello, Tav, I have added synchronisation between the current document tab and the current document's definition. Also, pushed the code from yesterday 2019-07-20 23-11-49 PDT Valentin Ionita Hello, Tav. I have been working on the previous buggy behavior, and I consider that I've fixed it for patterns. As far as hatches are concerned, there is a problem external to my code. In `src/widgets/paint-selector.cpp`, there are different `SPPaintSelector::Mode`s defined, but anything like hatch is missing, so there is a warning when I try to switch from a hatch paint server to, let's say, a linear gradient, and the hatch doesn't get removed from the code. 2019-07-04 06-09-33 PDT Valentin Ionita A question. If we want this paintservers dialog to get files from multiple folders (hatches and patterns directories, for starters, than it will need a different approach on setting resource variables). I'm gonna give my best try (I want to learn how to do it). If I get stuck, I'll message... 2019-06-27 01-15-15 PDT Valentin Ionita I think there's a bit more to be done than the hatch function.... 2020-08-22 00-16-07 PDT Valentin Ionita Well, maybe should be moved on the wiki. I think that would be best 2019-06-27 10-22-59 PDT Valentin Ionita Screenshot from 2019-06-27 20-21-06.png (https://chat.inkscape.org/file-upload/sXcHFoeiKtRLakR9W/Screenshot%20from%20...) 2019-06-24 04-28-20 PDT Valentin Ionita In the target, the first path is rendered on top, in mine the second is. 2020-08-14 01-41-51 PDT Valentin Ionita I am aware of the bias of already knowing something, but that doesn't make it easier to fix it. :cold_sweat: 2019-07-04 05-29-33 PDT Valentin Ionita [ ](https://chat.inkscape.org/direct/Tav?msg=g6KMj5NsJicKcQtpQ) That will be okay. () 2019-07-05 04-38-40 PDT Valentin Ionita Hey, Tav. 2020-06-05 08-51-09 PDT Valentin Ionita no, there are two signals that you specifically commented as "deprecated" let me show you a screenshot 2019-07-06 05-45-53 PDT Valentin Ionita Sure, no problem 2019-06-17 07-20-31 PDT Valentin Ionita great, thank you for the answers 2019-08-17 21-52-53 PDT Valentin Ionita Good morning, Tav, I have finished my article on the Inkscape work. You can read it here https://gitlab.com/vanntile/inkscape-gsoc-application 2019-06-13 11-01-39 PDT Valentin Ionita Hello, Tav 2019-07-22 08-34-00 PDT Valentin Ionita And thank you for remembering about the evaluation 2019-06-20 06-52-48 PDT Valentin Ionita left is my render, right is obviously inkscape 2019-07-21 03-23-13 PDT Valentin Ionita I think that something that could be done further is creating a file for a gradient store. However, now that I know that a target gradient is added with a second one that is actually used, I'm a bit unsure how should i pick this up 2020-08-22 06-02-43 PDT Valentin Ionita I might have been using a template, but I don't thinkk so. Let me check 2019-06-25 10-38-54 PDT Valentin Ionita Goodbye! 2019-07-05 06-38-17 PDT Valentin Ionita I have made a push that gets the ComboBox text from ` ` and filters the iconview just as I said in before #1 point. It works. Whenever you can take a look, it would be fine 2019-06-09 01-30-48 PDT Valentin Ionita it's somehow radial 2020-09-15 10-57-37 PDT Valentin Ionita I have just described the behavior and my first hypothesis. Will investigate when working on a fix 2019-07-05 02-31-49 PDT Valentin Ionita Screenshot from 2019-07-05 12-31-38.png (https://chat.inkscape.org/file-upload/RuMYa5JQaMP8WJbrE/Screenshot%20from%20...) 2019-07-03 01-33-28 PDT Valentin Ionita Hello, I would need some help with a function, I can't understand an argument. 2020-08-14 01-46-26 PDT Valentin Ionita Yes, that's true. However, I have a quite diverse teaching experience myself, with students in all learning cycles (primary, secondary, highschool and university) and I'm expecting to perform better. Or at least try to 2019-06-27 01-13-18 PDT Valentin Ionita Yes, I get that. :smile: 2019-06-13 11-08-13 PDT Valentin Ionita So, I got a mean score of 10.4s from 3 runs, that's an increase on the pepper of only 0.8s from the version without bicubic. About 9% increase in time 2019-07-04 06-12-33 PDT Valentin Ionita with paths to all the possibilities... 2019-06-22 23-48-15 PDT Valentin Ionita Also, I have submitted a bug report for the thing we talked about a couple of days ago 2020-08-22 05-28-51 PDT Valentin Ionita the branch is called canvasitems? 2019-06-27 10-23-24 PDT Valentin Ionita As you can see, there is one shifted path... 2019-06-27 01-16-04 PDT Valentin Ionita Yeah, it would make sense 2019-07-05 04-46-03 PDT Valentin Ionita I am sorry I couldn't push my latest changes before you took a look :sweat_smile:, but I will do that now 2019-07-19 08-41-02 PDT Valentin Ionita I'm back. After a few experiments, this is what I have: Gradients work just like you have said. There is a gradient created to be used and one that is actually set that references the target gradient. If there is a second gradient with the same target, only the second is created. The second gradient is deleted with the object that it is used. The target one is deleted if there is nothing pointing to it.
However, the patterns don't work that way. They are added directly and deleted if they are changed or there is no object pointing to it.
2020-06-05 08-59-23 PDT Valentin Ionita Anyways, I have something about X11, but that I'll ask on the dev channel later when I am more articulate about it 2019-06-22 23-47-55 PDT Valentin Ionita https://gitlab.com/inkscape/inkscape/merge_requests/746 2019-07-02 04-01-04 PDT Valentin Ionita Also, what is the Inkscape::Drawing class? 2019-06-29 10-10-41 PDT Valentin Ionita Well, if you would believe me, I really tried using exactly that. using bind to bind other parameters to mem_fun. Didn't work out. Anyway, I'll give it another go now I know it's the way I should do it. Thank you 2019-07-05 02-31-26 PDT Valentin Ionita Hello, Tav, have done a few changes to the widgets and stuff. Coming back to what I was saying yesterday, do you think we should better format the `svg`s like they have a background image and show up nicely no matter the gtk theme you have on your machine? This is how I have them right now 2019-06-18 09-38-26 PDT Valentin Ionita More than this, today i have studied hatches, using this ( https://www.w3.org/TR/2016/CR-SVG2-20160915/pservers.html#Hatches ) as a reference and also the code from `sp-hatch.cpp` and `sp-hatch-path.cpp`. Although I don't yet understand the implementation that well, I have a general idea of how hatches are meant to be and I have a few questions.
1. Do you have any advice on how should I render them? 2. Do you think I could use a pattern to render the hatches, adding paths into the pattern like the specification says? 3. Do you have an hour one of these days to have a call or discussin on how should I approach it? 2019-06-24 23-25-24 PDT Valentin Ionita Hello, Tav, and good morning! 2019-08-17 07-08-20 PDT Valentin Ionita I'm so sorry to hear all that, I hope you'll feel better and find a diet that works. 2019-06-09 00-52-08 PDT Valentin Ionita using finite differences and such. Didn't work out 2019-07-04 05-37-37 PDT Valentin Ionita (for debugging purposes) 2019-07-10 04-20-17 PDT Valentin Ionita I don't know how to detect shift-click. I'm gonna try looking it up 2020-08-22 06-14-32 PDT Valentin Ionita yes, I think there was something like a templlate inside 2019-07-29 09-14-10 PDT Valentin Ionita Oh, nice, will take a look after I figure my current issue out 2019-07-21 01-55-04 PDT Valentin Ionita I've avoided the warning, but I have found that hatches sre not garbage collected 2019-09-03 07-34-37 PDT Valentin Ionita Screenshot_2019-09-03-17-34-05.jpg (https://chat.inkscape.org/file-upload/vmWkYofaiff6e6AAv/Screenshot_2019-09-0...) 2019-07-10 02-42-05 PDT Valentin Ionita Also, can you explain me something? If an SPObject doesn't have an attribute and I use the getAttribute method https://inkscape.gitlab.io/inkscape/doxygen/classSPObject.html#abec49e6b36b9... it will throw a SPException. However, I couldn't catch it when I tested in the current code... 2019-06-24 04-17-22 PDT Valentin Ionita More specifically, I think that the last pair of points is *0,20*, not *20,0* 2019-07-05 04-40-08 PDT Valentin Ionita About the regex, I think I can do it myself. I have experience from Python and JavaScript. 2019-07-03 11-06-16 PDT Valentin Ionita [ ](https://chat.inkscape.org/direct/Tav?msg=T9ff6bpzD8WvvoMiN) @Tav Yes, I did it for myself while formatting the testfile and forgot about it () 2019-06-18 22-41-26 PDT Valentin Ionita I have found two bugs and fixed them 2019-06-29 11-05-55 PDT Valentin Ionita An update: thank you for the resources, I had a few issues in my code and retrying from scratch, helped me understand things. (Now it's working)
As promised, I have got the basics in these couple of days on hope to work with gtk, the terminology and approach to signal and tomorrow I will start implementing a FlowBox, like you recommended.
Have a good night! ๐ 2019-06-24 04-18-21 PDT Valentin Ionita As I can read fromthe hatchpath specification, *The offset must be positive, a negative or zero offset value results in the hatch path not being rendered.*, so that's why, on my machine, Inkscape doesn't render the first three hatches 2020-08-14 01-41-21 PDT Valentin Ionita Thank you for taking a look. 2020-08-22 00-09-39 PDT Valentin Ionita [ ](https://chat.inkscape.org/direct/Tav?msg=HQkJAWK8HbrNpnRh6) no, but I think I mentioned it in the RFC document, linked in the MR itself () 2019-07-04 05-25-06 PDT Valentin Ionita Ok, than what should I do for now? I am testing some icon on click callbacks... Do you want me to try my hand at generalizing this dialog for the rest of the day? 2019-06-27 11-47-59 PDT Valentin Ionita Having fixed the first bug, I am thinking only of this last one. 2019-06-09 01-29-29 PDT Valentin Ionita but if I test it manually, it seems to somehow work. When in Inkscape, using a 1x1 mesh and setting it to cubic shows a different output than a linear, somehow implying that it works in interpolating 2019-09-03 07-32-04 PDT Valentin Ionita Hmmm, lemme see 2019-07-18 01-10-52 PDT Valentin Ionita As for my status, ass I can see it right now, my main goal is getting the paint server widget ready for integration. That would imply two proper secondary goals: - Adding a toggle button for changing the stroke or fill, as you mentioned. - Scanning the current document for paint servers to show in their proper tab, even if they have not been added through this widget's interface. 2019-07-18 07-23-54 PDT Valentin Ionita Ok, managed to go through your code and put into mine and things keep working properly. I understand most of it, with the exception of `get_url`, that my lack of C++ Regex makes a black box. 2019-07-18 01-08-14 PDT Valentin Ionita Hello, Tav, after an adventure with the airlines (they were delays and I missed my flights at both departures and arrivals), I am back, fresh and ready to work. I want to apologize for such a large communication gap, there was almost no reliable internet, and I wouldn't have liked to use roaming for development. Anyway, now I'm here and ready. I remember you must be in the US by now, so you probably won't be able to answer soon. However, I'll leave this here for whenever you can give a ping back. 2019-07-04 08-46-09 PDT Valentin Ionita For the moment, it crashes because patterns.svg is not formatted in the right form and I haven't modified the code per-se. I will, next 2019-07-06 05-45-03 PDT Valentin Ionita Ah, okay, my misunderstanding 2020-07-18 11-55-21 PDT Valentin Ionita Thank you, I believe so too 2020-08-14 01-56-42 PDT Valentin Ionita The third one is an inconsistency in Gtk, from my point of view. I have tested it on the lowest level and the part where I hide/show the label text is not getting executed. 2019-06-27 01-16-36 PDT Valentin Ionita So, do I make an incremental update or make them both? 2019-07-05 02-51-39 PDT Valentin Ionita Thank you for the feedback. I'll go on and let you know what I did from time to time 2020-07-11 11-08-11 PDT Valentin Ionita Ok, I was going to be there anyways, so I guess I can 2020-06-07 11-21-53 PDT Valentin Ionita In the meantime, I'll take up on the opportunity on learning glade from devs 2019-07-06 08-47-44 PDT Valentin Ionita Hey, Tav, just pushed my code changes. The dialog works, partially. The next two steps that I can think of are - fixing a bug that makes the code to crash if I try to use it on a group - changing string manipulation to regex, where I would need your help or guidance. 2019-06-26 05-06-13 PDT Valentin Ionita the two circles with greenish insides have the derivatives null because the nodes have the same coordinates. I don't know how to fix them or if this is the actual intended behavior for the readMesh() function 2019-06-25 09-28-34 PDT Valentin Ionita Hello, again, for today's update. I have added support for a hatch referencing another as a template, fixed the transform origin like you said and another corner case (a hatch with no hatchpaths). Also, I have a quick fix for relative paths that, although works, it's based on the assumption that the relative paths are having only one move command (at the beginning). I know that conceptually it doesn't fit the rendering, however, if you want to push for a more thorough approach, I would need some expertise on parsing hatches to shift them around.
More than this, where could I create a hatch in Inkscape? It would be helpful with testing, but I don't know how to use hatches as an user (although I see how they are rendered) 2019-06-09 01-33-51 PDT Valentin Ionita yeah, I think you're right 2019-06-13 11-04-10 PDT Valentin Ionita let me check it on chrome. in a minute 2019-07-04 08-59-35 PDT Valentin Ionita Even tho it has some transparency, like this 2019-06-20 06-52-37 PDT Valentin Ionita Screenshot from 2019-06-20 16-51-54.png (https://chat.inkscape.org/file-upload/dw2u4tyCa6zs2g5hs/Screenshot%20from%20...) 2019-06-09 01-36-45 PDT Valentin Ionita okay, thank you for now, I'll let you know how it goes further 2019-07-19 08-42-47 PDT Valentin Ionita I think that the Paint Server panel should work just as the patterns work (it does integrate patterns after all). For that, I will only need to delete a pattern when you change it with something else, because that is the part that doesn't work like that. 2019-06-26 11-20-54 PDT Valentin Ionita When it merges, I will create the merge for the hatch polyfill, if it's okay 2019-07-06 05-44-02 PDT Valentin Ionita Well, I forgot to mention. It makes Inkscape crash 2019-07-06 06-13-37 PDT Valentin Ionita [ ](https://chat.inkscape.org/direct/Tav?msg=3ywMFAvJ5m4EudTt6) I know, easy to check with the store[CURRENDOC] () 2019-07-02 05-59-06 PDT Valentin Ionita If I succeed adding them to a grid widget, I'll have an MVP 2019-07-03 21-33-55 PDT Valentin Ionita Hello, Tav. I'll take it step by step explaining a few things. 2019-08-18 09-22-51 PDT Valentin Ionita I have updated the branch 2019-07-18 08-37-34 PDT Valentin Ionita I think that right now there is some data replication between the id and paint columns in the TreeModel. 2019-06-26 10-21-36 PDT Valentin Ionita Now I see that bicubic ones have the bicubic attribute on them. Such a noob mistake. will fix asap 2019-06-27 01-15-01 PDT Valentin Ionita There is this flag... `insert_mesh_polyfill_flag` Does that link to an user setting? 2019-06-26 21-47-48 PDT Valentin Ionita That would be for both mesh.js and hatch.js 2019-06-30 09-14-38 PDT Valentin Ionita Goodbye! 2019-06-20 06-03-10 PDT Valentin Ionita Am I somewhere around it? 2019-06-09 00-53-11 PDT Valentin Ionita If you could explain in detail or advice me on how to debug inkscape while running the ::bezier() function so that I can analyse myself the inputs and outputs. 2019-06-26 10-55-57 PDT Valentin Ionita I'll think about it, but for no it eems to be working 2020-08-14 02-17-51 PDT Valentin Ionita I have pushed an update for the comments 2020-03-18 02-40-39 PDT Valentin Ionita Hey, Tav, so I'm looking through your sandbox ( https://gitlab.com/Tavmjong/gtk_sandbox ), have already run it in debug mode 2019-07-18 09-08-35 PDT Valentin Ionita While the function is evaluating the patterns file, I get the following message `SPIPaint::read: url is empty or invalid` 2019-07-04 05-37-28 PDT Valentin Ionita Also, I want to print to stdout a node representation. Can i do it? 2019-12-02 01-34-56 PST Valentin Ionita Thank you for the update. I'll build the code myself to test it and after the bug week I'll push on the wiki update 2020-08-20 07-48-33 PDT Valentin Ionita I've taken a quick look at your comments, they're all actionable, will fix them tomorrow. About the wayland bug, I wasn't aware, I will check it out. 2020-08-22 05-27-24 PDT Valentin Ionita hey, I don't know. I could try rebasing off it to check. 2020-08-18 01-02-44 PDT Valentin Ionita You don't have to worry too much about my code, relax. I would see it as a critical review around Saturday (before the last week of GSoC, I mean), from my POV 2019-06-27 02-13-42 PDT Valentin Ionita Thank you for mentioning the errors to me 2019-07-10 02-55-08 PDT Valentin Ionita So, from a couple of tests, I can see that when hings are grouped, it's getting the inner elements and sets their attributes, ungrouping things. 2019-06-24 04-35-25 PDT Valentin Ionita [ ](https://chat.inkscape.org/direct/Tav?msg=F6DLr7bCkve48prFG) yes, that was for a totally different thing. For the moment I cannot calculate the y Offset for a relative path () 2019-07-04 08-45-35 PDT Valentin Ionita Hey, I have done all the fixes and pushed. You can find it here. https://gitlab.com/vanntile/inkscape/commit/ed5748845b7c4c9323fb971850c027c1... 2020-08-14 05-41-29 PDT Valentin Ionita No, it is not a Wayland issue. I don't know how to manage state in Inkscape, it hasn't been built up that way and I have decided at the moment of implementation that I wouldn't pay attention of not opening a second window like the first. 2019-06-27 01-31-09 PDT Valentin Ionita something like `src/ui/dialog/inkscape_preferences.cpp`? 2019-06-27 10-25-32 PDT Valentin Ionita Because Chromium transforms M a b c d into M a b L c d, the rendering becomes the bottom one 2020-08-06 00-54-49 PDT Valentin Ionita Thank you for the advice. 2019-06-24 04-25-55 PDT Valentin Ionita thank you for confirming that. That means I can render the first three properly (as the second one has the same d value, I think it is wrong). 2019-07-02 03-59-28 PDT Valentin Ionita If you have any idea on how to get a pixbuf from an SPObject, I'm open to suggestions. I wouldn't like to have to create a new document for each hatch, as I find it using resources for nothing, and can't get that part of code completely 2019-07-04 08-51-59 PDT Valentin Ionita oh, is that so? so, it was my bad, all in all. I will test immediately. 2019-06-29 00-23-22 PDT Valentin Ionita Hello, Tav, I have a few questions. I can't get ninja build to work. Is this https://gitlab.com/inkscape/inkscape/blob/master/CONTRIBUTING.md page still up to date? 2019-07-03 01-34-13 PDT Valentin Ionita How can I write the second argument (it's some kind of generator function, but I can't seem to work it out) 2020-05-31 04-20-14 PDT Valentin Ionita I am going thoroughly through the sandbox code. I will create my own sandbox repo and try to practice/experiment with some of your code, starting with this newer `gtkmm_paned2` and then try to integrate the new multipane into the old experiment. Whenever I get to that point, I think it will be a good checkpoint for direction, do you agree? 2019-06-13 11-12-21 PDT Valentin Ionita So, that's about all for now 2020-03-17 09-01-16 PDT Valentin Ionita Thank you, I'll look around 2020-08-22 06-08-30 PDT Valentin Ionita Are you testing in a branch with both our code or just on canvasItems? 2019-07-04 05-32-03 PDT Valentin Ionita That's true. It's nice to have an excited and positive mentor. :) 2020-03-18 02-59-37 PDT Valentin Ionita If you think this has potential, I'll try making up a proposal today for everybody to review in the next couple of days. 2020-08-14 01-59-37 PDT Valentin Ionita 7) thank you for reminding me, I have been focusing on crashes for now 2019-06-30 09-08-23 PDT Valentin Ionita Hey, Tav, from what I understand, in `symbols.cpp` there is loaded a symbol file (don't know where is that part of the code) and for each symbol that has to be loaded we create a virtual file that is used to create a pixel buffer for the symbols shown. This is called every time the document to be loaded is changed (on search). Am I anywhere near the truth? 2019-07-05 02-33-10 PDT Valentin Ionita If you approve, I'll format patterns.svg in the same way and keep going with other functionality 2019-06-27 12-21-49 PDT Valentin Ionita Fixed every error above. hatch.js is clean for merge, from my POV. 2019-07-18 08-49-44 PDT Valentin Ionita [ ](https://chat.inkscape.org/direct/Tav?msg=TFotHoDbC5tmon7eQ) nevermind. it wasn't a good idea () 2019-06-27 09-38-05 PDT Valentin Ionita Ok, I'll think about it and let you know as soon as possible. Whatever the choice, I'll use the experience I get from working on one to implement the other. 2019-06-26 05-04-32 PDT Valentin Ionita Screenshot from 2019-06-26 15-04-17.png (https://chat.inkscape.org/file-upload/c3AeSqSD7cqzbacYg/Screenshot%20from%20...) 2019-06-28 03-47-22 PDT Valentin Ionita Good to know 2019-06-20 06-53-10 PDT Valentin Ionita it only works on a subset of all features, will further develop it 2019-06-20 06-00-33 PDT Valentin Ionita its values can be userSpaceOnUse or objectBoundingBox and I would be thankful if you could explain to me just how they work to be sure I am doing the right thing 2019-08-06 21-05-13 PDT Valentin Ionita That would be all, I'll wait for your answers. :smile: 2019-06-26 10-27-03 PDT Valentin Ionita I'm so sorry for this, but I'm glad it's fixed, thank you for the tip :))) 2019-07-30 07-36-28 PDT Valentin Ionita Yesterday there was a massive power outage at home (~6 hours) so not much development. 2019-06-27 10-24-27 PDT Valentin Ionita For another thing, I have seen there appears to be a rendering difference between Firefox and Webkit-based browsers (I have tested in Chromium and Vivaldi). Let me explain 2019-07-02 05-58-31 PDT Valentin Ionita Yeah, thank you. 2019-07-05 04-42-55 PDT Valentin Ionita Another thing I can say is that I have my plane on Sunday, but I hope I will have my laptop in the camp, so I'll try to work a bit in our free time on paint servers 2020-08-14 01-58-33 PDT Valentin Ionita 5) tooltips are not something I know how to handle right now. However, for compatibility, I think that a better information in the tooltip would be the dialog shortcut. 2019-07-10 02-28-08 PDT Valentin Ionita More than that, I want to ask what should happen in case of a group. If you try to set the paint to a group, should it set the paint to the inner element or silently fail? 2020-03-17 03-10-05 PDT Valentin Ionita Hey, Tav, I'm looking through the Inkscape potential GSoC projects yet again and have a few questions regarding on an assessment about P9 (Path Library Improvements) and a confirmation of P7 (SVG 2 Support with SVG 1.1 Fallback or Polyfills).
For P7, is there anything notable that comes to your mind about what else needs to be polyfilled at the current moment?
For P9, I see that it's a hard project, and I wonder if it actually could be done. 2019-09-03 07-34-29 PDT Valentin Ionita Screenshot_2019-09-03-17-33-48_1.jpg (https://chat.inkscape.org/file-upload/63k2RnJyZgvdNdtHb/Screenshot_2019-09-0...) 2019-07-19 07-12-58 PDT Valentin Ionita Thanks for the tip 2019-06-25 09-45-42 PDT Valentin Ionita I was thinking the same about the predefined hatches. I would highly agree 2019-06-09 01-32-16 PDT Valentin Ionita Upon further testing, I see that it happens with 1x2 and 2x1 patches too 2020-08-14 01-59-06 PDT Valentin Ionita 6) been requested. as far as I know, no chance for that 2019-07-10 02-57-51 PDT Valentin Ionita sorry, not ungrouping, just extracting the elements from all the nested groups 2020-09-15 10-49-03 PDT Valentin Ionita About the MR, I have seen your tag and put my two cents there 2020-06-05 08-52-31 PDT Valentin Ionita well, it was written as "depricated" 2019-06-09 01-04-07 PDT Valentin Ionita Let me try again with some specific questions 2019-07-19 07-12-44 PDT Valentin Ionita I think I will study the above matter and if I find a quick fix, I will do it and create a MR after that. 2019-06-09 01-21-44 PDT Valentin Ionita sorry, rocket doesn't seem to upload any picture, I'll try uploading to drive again 2020-08-22 06-12-40 PDT Valentin Ionita From what I can see, only pain servers does that. And the prototype dialog doesn't show up, but that might have been from one of my last changes and I have't tested the dialog 2019-06-27 11-48-45 PDT Valentin Ionita Let's take the path from the example. Do you think it's safe for me to remove the last M pair? 2019-06-18 09-29-03 PDT Valentin Ionita ok, then I think I am finished with the mesh gradient functionality. Should I minify the script for the compressed_include version and then rebase my branch for a push? 2019-08-06 21-04-49 PDT Valentin Ionita The end of GSoC is approaching and i have received the email that reminds me I should write about what I did, to send it to the final evaluation. Is it all right if i do it in the repository where I did my proposal also? I don't want to pollute any Inkscape repo more than necessary 2019-06-24 04-27-58 PDT Valentin Ionita This is how it's rendered in my polyfill. Why? because the second line, even if it's different from the first one, overlaps it because its offset is ewual to the first's offset + pitch of the hatch 2019-06-20 05-41-06 PDT Valentin Ionita Screenshot from 2019-06-20 15-40-06.png (https://chat.inkscape.org/file-upload/zRB5NZP4atnThmTea/Screenshot%20from%20...) 2019-06-24 04-31-21 PDT Valentin Ionita ? 2019-07-06 06-15-01 PDT Valentin Ionita okay, then I'll look at how to set a property. 2020-08-22 06-19-25 PDT Valentin Ionita So it probably needs a small update 2019-06-13 11-12-12 PDT Valentin Ionita Talking about performance. What I can say is that I don't think I'll be able to improve the time perofrmance of the script. However, I have a few things in mind that will improve the number of calculations and potentially the memory used by the script 2019-06-24 04-36-03 PDT Valentin Ionita [ ](https://chat.inkscape.org/direct/Tav?msg=4NntDhKLEiexZx5pW) ok! () 2019-07-06 08-54-14 PDT Valentin Ionita Thank you! 2019-07-06 06-15-48 PDT Valentin Ionita About the precedence... are you sure? I had a rectangle with a color fill, and when I have set the attribute, it used a black rendering (not actually working, but showing the precedence it's pretty strange) 2019-07-03 03-21-25 PDT Valentin Ionita https://gitlab.com/vanntile/inkscape/tree/hatchDialog 2020-03-18 02-57-42 PDT Valentin Ionita So, for the demo, I am not sure if all the functionality works on my Machine (Ubuntu 18.04, probably with X11 and as you said not everything works there). 2019-06-24 03-50-47 PDT Valentin Ionita hello, Tav, whenever you have a moment, it would be of great help if you could discuss some things about the hatch tests and specifications, I am pretty sure of some things, but I want a second set of eyes on it 2019-06-30 07-08-32 PDT Valentin Ionita :blush: 2019-06-27 09-38-32 PDT Valentin Ionita [ ](https://chat.inkscape.org/direct/Tav?msg=emjwZTkm8ZjKeShDN) I'll take a look right now () 2020-08-22 05-35-14 PDT Valentin Ionita I've rebase with minimal conflicts (~9) and now I'm building 2020-08-22 05-51-19 PDT Valentin Ionita But I think mergin is fine 2019-07-04 08-59-12 PDT Valentin Ionita And it works :))) 2019-06-24 04-39-07 PDT Valentin Ionita Well, thank you. I am writing and rewriting a lot of code as I get better insights in my tasks, and never commit if I don't think it has some value, so I thought it would seem I'm not progressing. Your last message brought my calm back 2019-07-10 03-02-33 PDT Valentin Ionita What would be an approach used in Inkscape for getting all the elements in a group (considering the elements from another nested group)? 2019-06-20 06-04-10 PDT Valentin Ionita That would be great. If you find them, it would be great, I could take it on my own from there. 2019-08-02 09-03-53 PDT Valentin Ionita Hello, Tav, another update: I am working on hatch handles, I have actually added them, however, I don't know what exactly I should do for scaling hatches. I will probably wait for you to come back with that.
Another thing is that the modifications of the handles don't get updated on the rendering (they โ position and rotation โ only get updated on the first modification), that shouldn't be any of my code's fault, so I will have a bug to spot. 2019-07-04 05-41-10 PDT Valentin Ionita [ ](https://chat.inkscape.org/direct/Tav?msg=nCStJHG6nQApjFZhW) No, it probably won't matter. () 2019-07-04 06-12-13 PDT Valentin Ionita Really? I was thinking of doing multiple calls to the ```get_filenames(Inkscape::IO::Resource::HATCHES, {".svg"})``` function 2019-06-11 06-19-16 PDT Valentin Ionita Screenshot from 2019-06-11 16-19-01.png (https://chat.inkscape.org/file-upload/jjwpjs7qAgNuMPEPP/Screenshot%20from%20...) 2019-07-04 09-55-16 PDT Valentin Ionita I have updated the patterns.svg formatting, and this is how the dialog looks like right now. Do you think I should add a kind of background to every paint server so there would be contrast on any kind of theme? I would add a parent with a white/black fill and create the pixbuf from the parent. I think. 2019-06-09 01-34-50 PDT Valentin Ionita that we should approximate in which way? linearily? 2020-03-18 03-03-00 PDT Valentin Ionita Actually, before I begin the propsal, I want to look in Inkscape to see where GDL is referenced 2019-06-27 10-21-52 PDT Valentin Ionita Small update: 2020-08-14 01-42-39 PDT Valentin Ionita [ ](https://chat.inkscape.org/direct/Tav?msg=exdsLrb4atrW9qZi9) Interesting. Were they initially in a floating window? () 2019-06-24 03-50-59 PDT Valentin Ionita It would be of great help, thanks! 2019-06-27 01-34-48 PDT Valentin Ionita I think I made the necessary edits to the interface too (the above file). If you don't think I missed something, I'll commit 2019-07-02 03-58-05 PDT Valentin Ionita yes, but there is a whole architecture behind Inkscape that I now tend to try using. Really hard to work without understanding the API fully (yes, I have the doxygen documentation open). 2019-07-19 06-58-36 PDT Valentin Ionita OK, thank you for clarifying, I will check things up. 2019-06-18 22-44-09 PDT Valentin Ionita ok, I will do it myself. 2019-06-27 01-12-29 PDT Valentin Ionita Ok, I will 2019-06-25 10-38-39 PDT Valentin Ionita That's all for today 2019-06-18 06-47-38 PDT Valentin Ionita Hello, Tav, have you got any moment to take a look on the previous #1 ? 2020-06-05 08-58-55 PDT Valentin Ionita It's okay, I can test around and try to figure it myself 2019-07-18 06-10-37 PDT Valentin Ionita But how should I remember the current's document state? 2019-06-20 05-59-42 PDT Valentin Ionita My questions are about the hatchUnits attribute 2019-07-18 08-37-45 PDT Valentin Ionita I will remove the first, as I can see it as obsolete 2020-09-15 10-48-24 PDT Valentin Ionita Hey, I think the verbs array in verbs.c and the enum in verbs.h are out of sync or something like that. Because now when I open a dialog, let's say Layers, I get the name of another dialog, let's say Objects. I will try to fix that from friday onwards (to have a correct state) 2019-06-27 09-29-27 PDT Valentin Ionita From my proposal, the following weeks I had this _Weeks 7-9: Paint server gtk widget for GUI use (rewrite or integration)_ I think this refers to some GUI for hatches, maybe? I'll wait for some guidance 2019-06-27 11-49-09 PDT Valentin Ionita If so, I will try it. I don't know if it's a good strategy tho 2019-06-09 00-51-56 PDT Valentin Ionita I understand that bicubic is possible using 16 points, even if you don't know slopes or derivatives, and that's what i was trying to do so far 2019-08-17 08-37-31 PDT Valentin Ionita It's done https://gitlab.com/inkscape/inkscape/merge_requests/838 2019-06-18 22-41-17 PDT Valentin Ionita Yes, I have tested them with a few on my own and all the meshes from your initial repo. 2020-08-22 06-10-18 PDT Valentin Ionita on my branch icon preview doesn't trigger it 2019-08-06 21-03-10 PDT Valentin Ionita The previous branch I was working on, hatchDialog, is complete, from my POV. If you agree with that, II'll grab the changes from the master and wait for a merge (ther already is a MR) 2019-06-27 11-59-53 PDT Valentin Ionita no, that doesn't fix it 2019-07-18 08-25-57 PDT Valentin Ionita Separated the code, the function is ready. 2019-06-20 05-59-06 PDT Valentin Ionita Thank you. 2020-08-22 06-03-59 PDT Valentin Ionita after you merge it I'll try to find out 2019-07-06 06-16-15 PDT Valentin Ionita [ ](https://chat.inkscape.org/direct/Tav?msg=SRHqFox3wgCN7hC5d) thaat I did. woops () 2019-07-01 22-21-51 PDT Valentin Ionita Hello, Tav, sorry for not responding yesterday, I felt really sick and couldn't do a thing all day. Now I'm better and I'll start by rereading `symbols.cpp`. I was thinking if it wouldn't be preferable to use the code from there for hatches too. I'm afraid there is a bit of a large gap between my understanding of gtk and the complexity of the task. That doesn't mean I'm giving up, but that I might need more help along the way. Anyway, I'll update you on this sometime today. 2019-06-24 04-16-51 PDT Valentin Ionita Looking at the overflow property hatches, I think the d property in the renderings is wrong, and it's not what was intended for the right images 2019-06-18 22-44-35 PDT Valentin Ionita Coming back one step, should I minify it myself with a regular javascript minifyier or does Inkscape do something special 2019-09-02 05-45-13 PDT Valentin Ionita Hello, Tav, this is just a reminder for yesterday. ๐ 2020-08-17 07-24-29 PDT Valentin Ionita Also, apparently my branch revealed an older one-line bug in PaintSelector and i fixed it and Martin merged it for me. That's a plus 2019-06-27 02-43-35 PDT Valentin Ionita Nevermind, I've found the commit that fixes this in inkscape. Will fix tonight 2019-07-06 01-28-23 PDT Valentin Ionita Hello, been busy with packing. But I've seen your comment, I understand now and will fix it. Thank you for reviewing for me 2019-08-18 07-50-36 PDT Valentin Ionita Thank you! 2019-08-06 20-22-33 PDT Valentin Ionita I haven't pushed that branch, yet, but I will, so you could look at it anytime 2019-08-06 20-21-07 PDT Valentin Ionita Hello, Tav, glad to hear from you, sorry for not answering earlier, I was sleeping. 2019-07-03 10-43-19 PDT Valentin Ionita Yes, that's exactly what happened to me. It didn't render the image, even though the data is going through. I can see that the icon widgets have the labels (when you hover). All other types of debugging works... 2019-06-24 04-26-23 PDT Valentin Ionita now, I have a question about the last overflow hatch 2019-06-24 04-34-45 PDT Valentin Ionita okay, so i think of something like incremental painting. I'll figure out a way to fix it 2019-08-15 20-48-44 PDT Valentin Ionita I want to ask you how will we be going forward with that. I believe that the branch I'm currently working on should be merged, as it does several small changes that improve hatches UIX. 2020-08-14 01-50-21 PDT Valentin Ionita Now going to take a look at your comments. 2019-06-29 10-07-21 PDT Valentin Ionita Okay, and if the function has parameters? 2019-06-18 21-22-55 PDT Valentin Ionita Morning, Tav, I have just received a heads-up email from GSoC for the first review. Just letting you know. I'm gonna have my morning reading session then I'll get to work for the day. Really excited to hear your opinions for yesterday's questions 2019-08-06 21-01-48 PDT Valentin Ionita Hello, I published the branch where you can look. https://gitlab.com/vanntile/inkscape/commits/hatchControl 2020-08-14 06-10-54 PDT Valentin Ionita Coming back to the thread above, loading the old state is done in `sp_namedview_window_from_document` from `sp-namedview.cpp`. I'll look over the fact that it's inside an always true flag, which is redundant. I don't understand the logic of the C style functions there to make design choices about how should I restrict the state loading to only one window. More than this, I know that moving this logic part into the load_state method itself is creating technical debt rather than advancing development. 2020-08-22 06-03-50 PDT Valentin Ionita Honestly, I have opened it twice more and it's gone. I think it might have been one of the dialogs, but don't ask me which one, I don't rememebr what was open 2019-07-06 06-16-58 PDT Valentin Ionita ok. will do in a minute. 2019-08-24 01-08-21 PDT Valentin Ionita I know you don't use LinkedIn much, but it's pretty popular in today's hiring environment 2019-06-13 11-10-15 PDT Valentin Ionita I'm gonna give another try for performance improvements on Monday. Tomorrow is my last exam, and after it I go to a whole weekend workshop as a trainer for teaching high-schoolers a web framework. 2020-08-14 01-54-55 PDT Valentin Ionita The first one has been addressed by more people and I have just fixed it. Only on local, tho 2019-06-26 21-47-35 PDT Valentin Ionita Hey, Tav, I've seen that I forgot to add a license in hatch.js, just fixed that. However, can you explain to me how to use clang-tidy? My last pipeline failed in the codequality (I can see that it wasn't blocking). I used for my JS code the standard.js formatter, but if Inkscape is set up with clang, I can change it. 2019-06-24 04-36-47 PDT Valentin Ionita Well, I think we've discussed my issues for now. I will get back to work. Thank you, you've been of great help! 2019-07-31 01-23-47 PDT Valentin Ionita Wow, I love the mountains. Hope you're having fun. Don't worry too much about me. I like to keep you updated 2019-07-03 11-01-41 PDT Valentin Ionita [ ](https://chat.inkscape.org/direct/Tav?msg=2NSWjk5ijjSTncDBw) @Tav I don't understand what you're saying here. () 2019-07-04 05-31-13 PDT Valentin Ionita okay. Well, then I scrape the dialog settings everywhere and modify them in the new concept. This could be a great feature (if we fix hatches somehow). 2019-07-18 06-07-17 PDT Valentin Ionita ok, I haven't got the notification for the code, but now that you mentioned it I found the issue and downloaded the file. I will check the code 2019-06-28 01-07-55 PDT Valentin Ionita Amazing, Thanks! 2019-07-19 07-12-52 PDT Valentin Ionita If not, I will create one nonetheless 2019-07-05 10-55-39 PDT Valentin Ionita For tomorrow, I plan to make clicking on a hatch to change the fill for the selected element. I have finally got how that works, interface and all. I plan it like this: -- getting the document name of the active paint server -- getting the document from the document map -- getting the element with the id of the active server -- adding it to the store for the current document, if it isn't there already and to the current document itself -- changing the fill of the selected objects to match that fill 2019-06-24 10-20-04 PDT Valentin Ionita Ok. I will try that 2019-08-15 20-50-04 PDT Valentin Ionita I'll be waiting for your advice. 2019-07-03 11-07-31 PDT Valentin Ionita Well, I guess I will do the sidework tomorrow while waiting for any solutions on why hatches are not being rendered 2019-09-03 07-33-44 PDT Valentin Ionita Now we're connected, I think you can recommend me using a button near the profile picture 2019-06-20 06-03-00 PDT Valentin Ionita And userSpaceOnUse doesn't change anything in the calling object's coordinate system 2019-06-24 04-31-55 PDT Valentin Ionita Screenshot from 2019-06-24 14-31-42.png (https://chat.inkscape.org/file-upload/ri7yabiEhjFeY4sKE/Screenshot%20from%20...) 2019-07-04 06-16-15 PDT Valentin Ionita Ok, will do. Sorry for bothering you so much. Glad we passed through this one 2020-08-22 00-09-15 PDT Valentin Ionita In my migration, I had to introduce the update method in all dialogs, but most of the time I did the minimumm to replace the setDesktop() method 2019-07-11 01-10-58 PDT Valentin Ionita Ok, great, got it 2020-08-22 06-14-40 PDT Valentin Ionita template*** 2019-07-06 03-27-39 PDT Valentin Ionita Hello, again. Now, looking at my code, I see that I have the following pattern. ``` Gtk::Grid *grid = new Gtk::Grid(); _getContents()->pack_start(*Gtk::manage(grid), Gtk::PACK_EXPAND_WIDGET); ```
I have called `Gtk::manage` on every Gtk Widget, as far as I can see. My question is โ is there a difference between doing it at instatiation and the way I'm doing it now? 2020-06-07 11-21-35 PDT Valentin Ionita Next, I'll work on signals and a bit on interactions of functions then I'll start asking for feedback from the ux team and mentors 2019-06-11 06-35-51 PDT Valentin Ionita Thank you 2019-08-17 07-09-48 PDT Valentin Ionita I hope we can get to something actionable until tomorrow night. I'll have my article ready by then. 2019-06-24 04-13-52 PDT Valentin Ionita Thank you. For once, if you could open in inkscape the above hatch_test file and let me show you my train of thought I think it would be better 2019-06-09 01-08-08 PDT Tavmjong Bah Let me double check that. 2020-08-18 01-01-00 PDT Tavmjong Bah It will be a couple of days before I can take a good at your code as I'm away from home with my family. We'll be back home tomorrow evening. I wouldn't take adam.belis's requests too seriously. Sure we could have an undockable canvas... but at some point we must be practical. All requests need to be balanced against maintainability which as you have seen is a huuuuge problem for Inkscape. 2019-06-27 09-37-04 PDT Tavmjong Bah I'll let you choose. 2019-07-06 06-13-13 PDT Tavmjong Bah This will add the hatch to the defs section. One should check to see that it is not already there. 2020-08-14 01-44-56 PDT Tavmjong Bah It's actually quite hard for an "expert" to explain something to someone just learning. It's a skill many (most?) physicists lack from my experience. I've had really smart colleagues who really were poor teachers. 2019-06-26 04-20-40 PDT Tavmjong Bah This test file fails in Firefox and Chrome: mesh_javascript_test.svg:1201 Uncaught TypeError: Failed to execute 'createImageData' on 'CanvasRenderingContext2D': Value is not of type 'long'. 2019-07-03 10-52-12 PDT Tavmjong Bah Probably 2019-06-24 04-13-11 PDT Tavmjong Bah Sure, what questions do you have? 2019-06-24 04-35-31 PDT Tavmjong Bah Yes, it's incremental painting. The same thing happens for patterns and mesh patches, the order the tiles/patches are drawn is important. 2019-06-13 11-10-58 PDT Tavmjong Bah Looking forward to see what you can do on Monday! 2019-06-24 10-17-08 PDT Tavmjong Bah The transform origin for userSpaceOnUse is 0,0 in the current view (which is normally the upper left corner of the SVG). For objectBoundingBox it should be the 0,0 point of the bounding box. 2020-03-17 08-55-14 PDT Tavmjong Bah Otherwise, working on a headless Inkscape would be good. 2019-07-03 11-00-00 PDT Tavmjong Bah * std::vector elements ... should be right before for loop. 2019-07-06 02-03-12 PDT Tavmjong Bah I'm wondering if we can avoid building the document for the files that store hatches and patterns by just reading the XML. Since we'll paste the hatches and patterns into a dummy file, we don't really need to build the SPDocument. I'm not sure how to do this. I'll look into it this week. 2019-07-04 05-27-04 PDT Tavmjong Bah And a user could include their own set of fills by adding a file with them. 2019-06-09 01-15-41 PDT Tavmjong Bah This last step requires inverting the matrix 'v'. 2020-08-22 05-25-02 PDT Tavmjong Bah Would my merging the CanvasItem rewrite cause your merge trouble? I'm guessing that they are somewhat orthogonal. 2019-06-13 11-03-41 PDT Tavmjong Bah Excellent news!!! How long does it take to render the pepper? 2019-06-27 09-36-41 PDT Tavmjong Bah But before that, maybe one should look at implementing on-screen editing of the hatches transform, much like patterns. 2019-06-09 01-02-26 PDT Tavmjong Bah The derivatives at the edge are calculated so that the color function is parabolic, going to zero at the outer edge while matching the slope at the inner edge. 2019-07-18 09-24-50 PDT Tavmjong Bah That may mean that the paint is a color.... I'm not sure. 2020-06-07 07-28-54 PDT Tavmjong Bah Maybe tonight... it's a bit hard at the moment as I have kids and wife at home. 2019-06-26 11-37-48 PDT Tavmjong Bah Please do! 2019-06-09 01-31-59 PDT Tavmjong Bah It's calculating the derivative on one side first, and then using that result to calculate the derivative on the other side. 2020-08-14 01-48-51 PDT Tavmjong Bah That does help! I spent a couple years as a high school teacher, trying to teach a British syllabus to kids unfamiliar with many of the the things it assumed (try teaching forces inside an elevator to kids that have never seen one!). 2019-06-17 07-20-10 PDT Tavmjong Bah 3. Absolutely!! 2019-06-09 01-35-58 PDT Tavmjong Bah Although the other would result is less change is the user does insert a new row/column. 2019-12-02 00-56-10 PST Tavmjong Bah Hi, We're fairly far along in the release process, having had a couple(?) betas. I think we still have a lot to do. And there have been some fairly big code changes the last few weeks.... so I don't think we are too close to a final 1.0. This is bug week. We were hoping to have some people get together to work on bugs in Paris. Looks like it will just be Mc and me in person with hopefully others joining in remotely. It was a very busy week so I didn't get things as organized as I had hoped. 2019-06-09 00-53-59 PDT Tavmjong Bah The derivatives are being calculated. The key point is that at the edges, you have to guess what the derivatives should be. I forget which choice I made... I'll have a look. 2019-06-09 01-26-49 PDT Tavmjong Bah This does cause a problem with a mesh that is only one column wide or one row hide, as I see now. 2019-06-27 01-05-46 PDT Tavmjong Bah Can you rebase while you're at? 2019-07-26 19-17-09 PDT Tavmjong Bah I just realized I missed the second evaluation.... I thought I had till today to do it. It shouldn't effect you. You've been doing great! 2019-06-27 01-16-53 PDT Tavmjong Bah It's up to you. 2019-06-09 01-28-18 PDT Tavmjong Bah Perhaps a 1x1 mesh should fallback to bilinear interpolation. 2019-07-18 07-45-33 PDT Tavmjong Bah Sounds good! 2019-07-04 05-30-13 PDT Tavmjong Bah Yup 2019-07-06 05-45-44 PDT Tavmjong Bah I'll need to find an example of how to do it in the code... give me a minute. 2019-07-03 10-56-41 PDT Tavmjong Bah * Test that fill string is not empty, if empty, print error message to std::cerr and continue. 2019-06-28 03-44-54 PDT Tavmjong Bah 5. You can use Glade to prototype, I'm not sure about using it to build the dialog (we don't use it anywhere else yet, I think). We will be using .ui files eventually for menus and things... but it is a long process to get there. 2019-06-17 07-18-59 PDT Tavmjong Bah 2. Sure, you can cut them out. (We can always get them back from git.) 2020-08-22 00-16-31 PDT Tavmjong Bah Yup 2019-07-06 06-16-41 PDT Tavmjong Bah Maybe I have it backwards... but we should be setting the property in the style attribute. 2019-07-10 11-25-59 PDT Tavmjong Bah For shift-click you need to override on_button_release_event(). One probably also needs to enable this signal. 2019-06-17 07-18-16 PDT Tavmjong Bah Hi, 2019-08-18 04-26-57 PDT Tavmjong Bah Your write-up looks good! 2019-07-06 01-38-37 PDT Tavmjong Bah * Need to put the "standard" formatting section at the bottom of the files (see bottom of symbols.cpp). 2019-07-19 07-11-59 PDT Tavmjong Bah Could you make a merge request for your code, putting WIP at the start of the title? This would probably make it easier for other people to took at it. (This is not urgent so you can do it when you come to a natural break point.) 2019-06-29 10-06-48 PDT Tavmjong Bah It depends on if you are using C or C++ (and you should be using the later). Yes, you should use button->signal_clicked().connect(sigc::mem_fun(*this, &function). 2019-08-07 12-25-19 PDT Tavmjong Bah Please bring your hatchDialog up-to-date with master. Then we should get a review of the dialog by Inkscape developers and users. I think there is some issue with how it fits into the greater Inkscape GUI but that discussion can't happen until it's been more widely seen. 2020-08-13 01-40-56 PDT Tavmjong Bah Just a heads up, I'll be taking my family to my mother-in-law's today. If all goes well, I'll be up and reconnected to the Internet this evening. I'll give your code a workout tomorrow. Remember "Perfect is the enemy of the good"! Focus should be on crashes first, major bugs second. Keep up the good work! 2019-07-04 06-16-40 PDT Tavmjong Bah Don't worry about bothering me! 2019-06-27 09-32-38 PDT Tavmjong Bah A more complete test file. Mixing "hatchUnits" and "hatchContentUnits", the former applies to the attributes, the latter to the paths. There are still some problems. 2019-08-17 06-48-22 PDT Tavmjong Bah Actually, it was early Thursday morning we got home... 2020-08-14 01-41-41 PDT Tavmjong Bah I did find one major issue: With a notebook with two dialogs, dragging one dialog tab to a dropzone causes the other dialog to move to a new floating window. 2020-08-22 05-31-04 PDT Tavmjong Bah Yes 2019-06-18 08-52-58 PDT Tavmjong Bah If points are on top of each other you would get a divide by zero error if you don't check d > 0. 2019-07-04 06-15-33 PDT Tavmjong Bah Make a new directory: "paint" and move patterns.svg and hatches.svg to it. Remove the directories "gradients", "patterns", and "hatches". 2019-07-10 11-42-47 PDT Tavmjong Bah Ah, shift is used by IconView to allow selecting multiple icons. I guess we can fallback to a toggle button for fill vs. stroke. Otherwise, the IconView would need to be replaced by a ButtonBox which has quite different semantics. 2019-06-28 03-41-48 PDT Tavmjong Bah 2. Yes, we should have a set of canned hatches. You can search on the web for a selection (I'll have a look in a moment). 2019-07-03 10-58-52 PDT Tavmjong Bah * Test that document is not null. 2019-07-04 05-28-48 PDT Tavmjong Bah (You'll have to be a bit more clever to not select rectangles that are in patterns. You might be able to use a selector for that.) 2020-08-14 01-42-49 PDT Tavmjong Bah No. 2020-08-22 06-01-55 PDT Tavmjong Bah Were you loading a particular document? I'm wondering why you see it and I don't. I'm guessing the error message can probably be commented out. It is possible to call the drawing code without having a CanvasDrawingItem (the Eraser tool does it). 2020-03-17 08-37-27 PDT Tavmjong Bah There is also a student interested in working on a headless Inkscape (no GUI dependencies). This is actually a huge project but it can be divided up into smaller parts. 2019-06-09 00-59-12 PDT Tavmjong Bah The distance is the geometric distance between corner points. This is probably not optimal if the mesh is highly distorted but seems to work fine. 2019-07-05 04-36-06 PDT Tavmjong Bah Just tried out your code. Looking good. I think I would change a few things. The first is to create a new function load_source() which would create an icon_view and pack it with paints from one document. By passing in a document, one can find the paint servers in the current SVG document. For the moment, the icon_view's can be put into a Gtk::Box, maybe wrapped by a Gtk::Frame with a label from the title element of the SVG root ( Hatches) or "Current document". The second is to store the content of the "fill" attribute without extracting out the id. This can work for colors if we ever want to have the user be able to include their own colors into the list. The id can still be stored as it's needed for finding the paintserver in the section (we also need to store the file where it came from). It would be probably better to use a Glib::RegEx for extracting out the id. I can help write one. 2019-07-04 05-23-20 PDT Tavmjong Bah Your code is a good first step but I think we need to generalize it to include patterns and other fills. So, instead of a hatches dialog, it should be a "Paint Server" dialog. And instead of storing element_hatch, it should store the content of the fill attribute (or maybe the fill property) and the file which it came from. Then when the user chooses a fill we can copy it to the document def's section. (And we can include fills that are already in the user's document. 2019-06-20 06-50-46 PDT Tavmjong Bah hatch01.svg (https://chat.inkscape.org/file-upload/KemGTBvwPRQyRHZJ5/hatch01.svg) 2019-08-11 20-57-26 PDT Tavmjong Bah OK, I'll be back in France Wednesday and should have time Thursday and Friday to look at things closer. 2020-08-14 01-54-28 PDT Tavmjong Bah Here are a few more "nit-picky" observations:``` * The floating window title should contain the document name it's attached to. The current place the document name is shown would then be redundant. I think you said you are fixing this. * I like the way the tabs change from icon with name to just icon when there isn't enough room. * A new tab opened with just the icon even when there was enough room for the name, the name got added when the dialog lost focus(?). (Wayland bug?) * The in-between panel handle could be a little more distinctive (Gimp uses three dots). * I guess it was a design decision to lose the menu at the top of the dialog window. I wonder how right-clicking on the tab to pull up a menu might be made more discoverable. Maybe a tooltip? * I'm guessing that clicking on the area behind the tabs to pull up a menu is not possible? * There are a number of Gtk-CRITICAL messages (maybe Wayland related). These can be debugged by running: G_DEBUG=fatal-warnings gdb inkscape G_DEBUG=fatal-criticals gdb inkscape ``` 2019-08-07 12-16-40 PDT Tavmjong Bah Hi, 2020-08-14 02-19-07 PDT Tavmjong Bah ```auto page = notebook.get_nth_page(i); auto label = notebook.get_tab_label(*page); label->set_tooltip_text("A tooltip"); ``` 2019-07-04 05-42-35 PDT Tavmjong Bah Ah, an xml node. Yes, it's possible... somewhere in the code (I added it but can never remember where). Give me a minute. 2019-06-27 01-18-19 PDT Tavmjong Bah Sounds good! 2019-06-29 10-08-04 PDT Tavmjong Bah You need to lookup the format on the gtkmm website. 2019-07-04 05-27-58 PDT Tavmjong Bah The patterns.svg file could be adjusted to include the same 100x100 preview rectangles you have in the hatches.svg file. 2020-08-14 06-22-39 PDT Tavmjong Bah OK, that can be fixed later. Yeah, I really hate this whole SPNamedView stuff. 2019-06-24 04-43-15 PDT Tavmjong Bah I often rewrite code multiple times as I better understand the problem. I also am always looking for ways to simplify the code so the next person to work on it has an easier task. 2019-06-17 07-18-25 PDT Tavmjong Bah 1. I'll need to take a look. 2019-06-27 09-31-45 PDT Tavmjong Bah hatch01_with_js.svg (https://chat.inkscape.org/file-upload/x5LEawjf9h8dpFb5g/hatch01_with_js.svg) 2019-07-04 05-31-28 PDT Tavmjong Bah I'll work on fixing hatches. 2019-08-19 12-01-09 PDT Tavmjong Bah Thanks, merged. 2020-08-14 05-23-41 PDT Tavmjong Bah Just checked your updated code. Looks good. One more bug: for every new document window I'm getting a new floating dialog window with the same tabs as saved. Probably a Wayland issue. BTW, if you are using Wayland, you can test X11 by `GDK_BACKEND=x11 inkscape`. 2019-07-04 05-26-32 PDT Tavmjong Bah Yes, try to generalize it. I'm thinking that there should be one icon view per file. Then there would be a pattern section, a hatches section, a section with the fills from the current document, and maybe eventually a gradients section. 2019-06-28 01-07-34 PDT Tavmjong Bah Merged! 2019-07-04 05-40-33 PDT Tavmjong Bah I'm not sure that it matters. Loading the files shouldn't take a long time. (If it does, then you would only want to load what is selected in the option dropdown and then cache the results.) 2019-06-24 04-30-07 PDT Tavmjong Bah I don't see where in the spec it says that the offest must be positive. 2019-07-05 04-41-52 PDT Tavmjong Bah Yes, it is a lot.... but I've got great confidence in you! I'l be out for a little bit this afternoon. 2019-06-29 10-08-28 PDT Tavmjong Bah https://developer.gnome.org/gtkmm/stable/classGtk_1_1Button.html 2019-08-19 08-53-19 PDT Tavmjong Bah Hi, can you update one more time... I'll merge it immediately. 2019-06-20 06-55-31 PDT Tavmjong Bah Here's the test file the person who implemented hatches in Inkscape used. 2019-07-04 06-11-29 PDT Tavmjong Bah I would just move everything into one directory. 2019-06-23 00-22-42 PDT Tavmjong Bah I'll take care of the merge request this week. Good to know about your progress. 2019-08-06 15-22-49 PDT Tavmjong Bah Do you have any urgent things for me to look at? 2019-07-29 08-44-50 PDT Tavmjong Bah Sounds good... I hope to have time in the next few days to look at what you're doing. If you get stuck and looking for something to do in the meantime, you might have a look at adding the control nodes to hatches like patterns have. (Look at upper left corner when the node tool is enabled and a pattern is in use.) 2019-06-27 01-15-59 PDT Tavmjong Bah Yes, there is a preference in the Inkscape Preference's dialog under SVG output. 2019-06-27 09-35-29 PDT Tavmjong Bah Have a look at the symbols and swatches dialogs. What I would like to see is something similar where one can preview patterns and hatches. 2019-07-06 05-43-35 PDT Tavmjong Bah I think you can ignore the error.... it's usually harmless (but we should fix it eventually). 2019-07-06 01-47-33 PDT Tavmjong Bah * For error message, preface the error message with the name of the routine (std::cerr << "PaintServerDiakog::load_sources(): Fill attribute...). 2020-08-22 00-16-19 PDT Tavmjong Bah Probably... 2020-06-05 08-58-17 PDT Tavmjong Bah Ah, I don't remember why it is marked DEPRICATED. 2019-07-04 11-19-22 PDT Tavmjong Bah Great progress! I'm off for the night too. 2020-05-31 04-21-48 PDT Tavmjong Bah Sounds good! 2020-03-18 03-02-20 PDT Tavmjong Bah Yeah, I think this has lots of potential. The demo was well received at the last hackfest. 2019-07-04 05-31-42 PDT Tavmjong Bah Questions are better than problems! 2019-07-05 02-50-18 PDT Tavmjong Bah Yes, I would put a gray background (might be good to put a checkerboard... but we would need to experiment with that). 2020-08-22 06-10-50 PDT Tavmjong Bah Interesting.... 2019-06-26 04-19-24 PDT Tavmjong Bah mesh_javascript_test.svg (https://chat.inkscape.org/file-upload/3K85YK2uBFPwMAFNL/mesh_javascript_test...) 2019-06-13 11-09-39 PDT Tavmjong Bah That's good! 2019-06-18 22-42-39 PDT Tavmjong Bah Excellent! (Recall that GSoC will require you to submit your code so while it can be merged, you need to keep a copy somewhere.) 2019-06-27 01-12-53 PDT Tavmjong Bah We might want to always replace the polyfill to ensure that we always have an up-to-date version. 2019-06-09 01-33-24 PDT Tavmjong Bah So if a mesh has only one row or one column, then we should use linear interpolation in the short direction. 2019-06-09 01-36-13 PDT Tavmjong Bah s/is/if/ 2019-06-26 10-55-30 PDT Tavmjong Bah Great!! Can you update the merge request? If you have a better suggestion for handling cases where points overlap, I would be interested. I don't think the specification considers this possibility. 2020-08-17 07-24-50 PDT Tavmjong Bah SPDesktop contains the code for rulers and scrollbars which are used in ui/widgets/canvas-grid.h. 2019-06-27 01-11-13 PDT Tavmjong Bah (insert_mesh_polyfill()) 2020-03-17 08-35-45 PDT Tavmjong Bah For P7, I can't think of anything important that needs polyfills. There are few things we don't expose yet... markers on shapes for example that could use SVG2 -> SVG1.1 export functions. Probably not enough for a GSoC project. 2020-08-22 00-05-01 PDT Tavmjong Bah One thing I've been meaning to ask you, is if you've gained any insights on how dialogs in general should work. I found that each dialog seems to have its own way of doing things (which makes it difficult to track down bugs, etc.). I added the "Prototype" dialog at one point to try to demonstrate best practices but never got very far with it. I'm not sure if it would be useful to work on it further or to just delete it. 2019-06-24 04-25-16 PDT Tavmjong Bah Yes, the last point should be 0,20. 2019-06-18 22-47-10 PDT Tavmjong Bah Use can use whatever minifier you want... I don't remember what I did. 2019-06-18 22-46-24 PDT Tavmjong Bah Hatches: I would use patterns, making the pattern high enough to cover the object (including if the pattern is rotated). Width might be more complicated as patterns are not required to render outside their tiles while hatches are. 2020-08-20 07-43-26 PDT Tavmjong Bah Just had another look at your code and made a few trivial comments. Have you had a chance to try it under Wayland? 2020-08-22 05-27-53 PDT Tavmjong Bah It it's not too much work... 2020-09-15 10-54-25 PDT Tavmjong Bah Hmm, I don't see how things can be out of sync. 2019-07-06 08-53-52 PDT Tavmjong Bah OK, have a good trip. 2020-08-22 00-15-49 PDT Tavmjong Bah Very nice write up! Should be linked to from with in code. 2019-06-27 11-35-43 PDT Tavmjong Bah https://www.w3.org/TR/SVG/paths.html#PathDataMovetoCommands 2019-06-26 09-40-24 PDT Tavmjong Bah The greenish regions I think are fine, overlapping nodes will always be a problem. The meshes match the Inkscape rendering. There does seem to be a problem with the bilinear interpolation meshes (top left, top third from left). The images should have yellow "crosses". 2019-08-19 08-54-16 PDT Tavmjong Bah I've seen a bunch of bugs... but I want to get the code merged. We can always disable the dialog later. 2019-06-18 22-48-26 PDT Tavmjong Bah You might document which minifier you use (in folder README). 2019-06-09 00-59-35 PDT Tavmjong Bah Look for "// Calculate interior derivatives" 2019-06-20 06-03-45 PDT Tavmjong Bah I think so. It's been a very long time since I looked at this. Let me see if I can find some test files. 2020-03-17 08-54-02 PDT Tavmjong Bah There is another project I just thought of that could be quite interesting. I produced a demo program for redoing how Inkscape handles dialogs. (We currently rely on an unmaintained library.) It might make a good GSoC project to actually implement it in Inkscape. See gtk_paned in https://gitlab.com/Tavmjong/gtkmm_sandbox 2019-06-27 01-34-49 PDT Tavmjong Bah Yes, you need to edit src/ui/dialog/inkscape_preferences.cpp 2020-08-21 23-55-09 PDT Tavmjong Bah Hi, I don't understand exactly what ede123 thinks is a blocker. Where did he say this? 2019-06-09 01-00-17 PDT Tavmjong Bah The slopes are calculated per color channel. 2019-06-09 00-50-08 PDT Tavmjong Bah Yes 2019-09-03 06-52-15 PDT Tavmjong Bah OK, I admit I'm totally clueless, how do I leave you a recommendation on LinkedIn? 2019-08-11 16-01-53 PDT Tavmjong Bah Any progress to report? 2020-08-14 01-40-51 PDT Tavmjong Bah Hi, I've been through your code and did a little testing. I would say overall it looks good. The code could use a few more comments, I pointed out a few places on GitLab. One thing to remember is that things that seem obvious to you may not be to someone just learning the code. Anything you can do to help them will save them loads of time. 2019-06-27 01-55-12 PDT Tavmjong Bah Two errors with this file. (Inkscape also has an error. If zoomed up, the line thickness is correct, if zoomed down, it's too thin.) 2019-07-19 06-57-31 PDT Tavmjong Bah Hi, Not that is not normal. That is not how it's suppose to work. One might have a look at how patterns are added through the Fill ans Stroke dialog. Gradient and pattern elements can refer to other gradient and pattern elements so for example with gradients, Inkscape creates a gradient that contains the colors and their relative positions in the gradient, then Inkscape creates another gradient element that sets the start and end positions of the gradients and refers to the "color" gradient to get the colors. This allows one to define one color pattern and use it for multiple gradients attache to multiple shapes. Patterns probably do the same thing (I don't have time at the moment to check). If you use the "standard" way to set patterns this should be taken care of for you, again look at the pattern setting code in the Fill and Stroke dialog. 2019-07-08 06-42-41 PDT Tavmjong Bah One more point, the use will need to be able to set the paint of a stroke... 2019-08-06 15-22-08 PDT Tavmjong Bah Hi, I'll have some time to look at your work over the next few days. 2019-07-06 06-12-26 PDT Tavmjong Bah Inkscape::GC::release(repr); 2020-08-14 02-20-50 PDT Tavmjong Bah I hate editing in RocketChat! 2019-07-08 05-41-20 PDT Tavmjong Bah For the regex, src/extentions/internal/svg.cpp has exactly what you need at line 139. 2019-06-09 01-36-26 PDT Tavmjong Bah s/is/in/ 2019-06-20 05-58-28 PDT Tavmjong Bah Sure 2019-06-28 03-46-54 PDT Tavmjong Bah Speaking of prototyping, there is a "Prototype" dialog that you should probably use as a start. (It was my attempt a couple of years ago to understand how dialogs should be built, each dialog differs from the others in subtle but non-trivial ways. 2019-07-06 01-42-31 PDT Tavmjong Bah * Format the initialization with the punctuation first, so : Inkscape::UI:Widget::Panel(...) [new line] , current_store(ALLDOCS) [new line] 2019-06-27 01-16-44 PDT Tavmjong Bah Eventually, I want to redo all the SVG file saving stuff as we have a bunch of different save options that cannot be used together... 2019-07-02 03-56-22 PDT Tavmjong Bah Glad you are better. Might be a good idea to recycle the symbols code. I find it helps sometime to write a small, stand-alone Gtk program to try out new things. 2019-08-18 04-50-02 PDT Tavmjong Bah HatchControl branch merged. 2019-07-02 08-46-07 PDT Tavmjong Bah OK! When you start again, look at how the symbols dialog gets it's images. You can probably do something similar. I'll look more closely tomorrow. 2019-08-07 12-21-47 PDT Tavmjong Bah That is scaling the first two columns. You should be able to use the Geom::Scale for manipulating that. 2019-06-09 01-07-12 PDT Tavmjong Bah The input is an array of patches with (m*3 + 1) x (n*3 + 1) points, where m is the number of rows and n the number of colums. Every third point has a color value attached. 2019-06-28 03-49-17 PDT Tavmjong Bah People have created new dialogs by copying old ones but then introducing small changes and overtime all the small changes add up to big changes which makes debugging dialogs difficult (there is a really nasty bug in the XML dialog when it is floating and more than one document is open at a time). 2019-06-20 06-55-43 PDT Tavmjong Bah It's all userSpaceOnUse 2019-07-04 08-51-24 PDT Tavmjong Bah I've figured out the hatch rendering bug... document->ensureUpToDate() must be called before renderDrawing.setRoot(). It only needs to be called once with no arguments. 2020-09-15 10-51-29 PDT Tavmjong Bah Thanks! 2019-06-09 01-34-21 PDT Tavmjong Bah Or we could artificially insert "corner" points halfway. 2019-06-09 01-30-32 PDT Tavmjong Bah But the result is not "symetric". 2019-06-09 01-25-24 PDT Tavmjong Bah Yes to the first part. No to the second part. The derivatives are calculated from just the four corners. 2019-06-09 01-03-26 PDT Tavmjong Bah This was chosen as it would (probably) better match what would be expected if modeling a solid like a sphere or pepper. 2019-06-09 01-36-52 PDT Tavmjong Bah Great! 2019-06-20 06-51-53 PDT Tavmjong Bah I can't seem to find files using objectBoundingBox but I created one example. Left is userSpaceOnUse, right is objectBoundingBox. Rendering should be the same. 2019-06-25 10-51-33 PDT Tavmjong Bah Great. I'll dedicate some time tomorrow to reviewing your merge requests and getting them in. I'll also finish the evaluation (I'm having a hard time with finding something you can do better). 2020-08-14 02-00-57 PDT Tavmjong Bah Yes, crashes should be your top priority! The rest can be fixed post merge by you or someone else. 2020-08-22 05-53-30 PDT Tavmjong Bah Thanks, I'll try to figure out why you are getting the console warning. 2020-03-18 02-51-45 PDT Tavmjong Bah My Internet connection seems a bit flaky this morning... 2019-06-09 01-10-55 PDT Tavmjong Bah Yes, that's correct. A temporary array is then created with data from the patch corners to calculate the derivatives (d). 2019-06-24 04-37-15 PDT Tavmjong Bah Glad to be of help. I think you are making great progress! 2019-08-17 06-41-03 PDT Tavmjong Bah Hi vanntile, sorry for not replying sooner. We got back from home from the US very early on Wednesday morning after a long journey (partly our fault as we schedule a 10 hour layover in Dublin, never been to Ireland before). Needless to say we were quite jet-lagged. Then since Thursday I've been suffering from painful abdominal cramps which has made it impossible to concentrate. (I've been getting these about one a month for the past year and a half, no known reason but it seems to be diet related). Anyway, I will definitely look at your two branches. They probably both should be merged as soon as I do the review with perhaps the option of disabling the paint server dialog via an option. (Two reasons for this, having the code in the master makes it for people to see what you've done and give feedback and it also reduces the possibility of code rot.) 2019-07-04 05-30-10 PDT Tavmjong Bah Youp 2019-07-04 06-13-23 PDT Tavmjong Bah You could do that but I'm not sure that having multiple directories with just one file is that useful. 2019-06-30 09-14-11 PDT Tavmjong Bah I believe we render to a pixbuf and then use Gtk::Image (I've got to go out now, I'll look more carefully tomorrow). 2019-07-06 06-14-36 PDT Tavmjong Bah Your are setting the "fill" attribute. This won't work in general as Inkscape always converts fills and strokes to properties and properties have a higher precedents than attributes. 2019-06-09 01-26-11 PDT Tavmjong Bah We don't have color values for the other points. 2020-09-15 10-51-07 PDT Tavmjong Bah Oh, I'll have a quick look at the verbs. That's not good if things are out of sync. 2019-07-01 01-28-13 PDT Tavmjong Bah Ah, we use a Gtk::IconView to display a grid of buttons, each with an "icon" (Gtk::Pixbuf). This would be an alternative to using a Gtk::FlowBox. https://developer.gnome.org/gtkmm/stable/classGtk_1_1IconView.html#details 2019-06-09 01-14-20 PDT Tavmjong Bah Finally the "split" mesh is filled using the derived derivatives. 2019-07-06 01-37-50 PDT Tavmjong Bah Here are some more comments: 2019-06-27 01-35-19 PDT Tavmjong Bah I think that is probably all. 2019-06-26 09-40-51 PDT Tavmjong Bah BTW, I did your midterm evaluation. 2020-03-18 03-06-31 PDT Tavmjong Bah This proposal also changes the way dialog updates are triggered. It relies on the mouse moving into/out of windows (which triggers a GTK calllback). This should reduce signaling (or at least make it easier to understand). 2019-07-06 05-44-44 PDT Tavmjong Bah Yes, line 224 needs to be done differently but it's not related to the error message. 2019-07-26 19-19-34 PDT Tavmjong Bah I was also ill yesterday... which didn't help. 2019-08-09 12-47-25 PDT Tavmjong Bah Hi, The knot handling code doesn't seem to be the problem. It looks like a signal is missing. If you use the XML editor to change a pattern transform, the image updates. If you change the hatch transform, it doesn't. 2019-06-11 06-35-42 PDT Tavmjong Bah Excellent! Looking forward to see it done! Good luck with exam. 2019-06-09 01-13-02 PDT Tavmjong Bah Once the derivatives are found, a copy ("smooth") of the original mesh has each row/column "split". 2019-07-04 05-29-37 PDT Tavmjong Bah getObjectsBySelector() 2019-06-24 04-34-40 PDT Tavmjong Bah Ah, the non-negative "offset" refers to the offset in the 'y' direction and not to the attribute 'offset'. 2019-09-01 08-50-24 PDT Tavmjong Bah Yes, I can do that. Remind me if I haven't done it in the next day. (Got back from the Alps yesterday, had some great hikes.) 2019-07-06 05-45-30 PDT Tavmjong Bah You can't just add a repr from one document to another. 2019-06-09 01-04-19 PDT Tavmjong Bah ok 2019-07-10 03-04-04 PDT Tavmjong Bah You need to recurse down the tree. I don't think there is any method that does that (maybe we should add one). 2019-07-03 10-58-05 PDT Tavmjong Bah * Errors should go to std::cerr and then use continue rather than if/then. 2019-07-04 05-30-00 PDT Tavmjong Bah Or you could add a "preview" class and use getObjectsByClass(). 2019-07-08 05-36-22 PDT Tavmjong Bah Sounds good. 2019-07-10 02-56-51 PDT Tavmjong Bah It shouldn't be ungrouping, but it probably does set the property on child elements. Eventually it should probably set the property on the group, unsetting the property in the children. 2020-08-22 00-09-12 PDT Tavmjong Bah Ah, interesting. Is this explained with comments in the prototype dialog? 2019-07-30 19-58-37 PDT Tavmjong Bah Yes, another branch for hatch handles would be good... Spent the day hiking up some mountains (or what they call mountains in northern Minnesota), haven't had much time for Inkscape. Should be better next week. 2019-07-03 02-54-30 PDT Tavmjong Bah I don't know. I would need to take some time to figure this out. It seems that FlowBox uses a different setup than IconView. 2020-06-07 11-45-51 PDT Tavmjong Bah Brilliant! 2019-06-09 00-56-51 PDT Tavmjong Bah The SPMeshNodeArray::bicubic function is being called on the entire mesh. 2020-03-17 08-55-45 PDT Tavmjong Bah One could also look at https://wiki.inkscape.org/wiki/index.php?title=Refactoring_projects for additional ideas. 2019-07-05 11-28-27 PDT Tavmjong Bah Sounds good. One more thing to think about. A user document is not going to have nice rectangles with hatches or patterns waiting to be displayed. We'll probably need to generate the rectangles ourselves. One can use the same trick as the Symbols dialog of having a dummy document with a rectangle of the correct size, swap in the hatch or pattern, and then generate the image. 2019-07-02 05-58-09 PDT Tavmjong Bah Inkscape::Drawing class handles the on-screen display of the SVG drawing itself and all the editing aids that get drawn on top (handles, grid, guide, etc.). 2019-06-27 00-28-00 PDT Tavmjong Bah Good morning, don't worry about formatting. The clang-tidy process is advisory, not mandatory... and there are many things it doesn't do optimally. 2019-06-27 09-33-45 PDT Tavmjong Bah OK for the 10 day break, you'll just need to keep up the pace before and after. Note, I'll be in the US for four weeks from the 17th of July to 14th of August and won't be at a computer most of the time. 2019-07-02 06-01-00 PDT Tavmjong Bah Sounds good. 2020-08-22 00-01-01 PDT Tavmjong Bah OK, you seem to be handling things well! 2020-03-18 03-15-08 PDT Tavmjong Bah I'm not sure this really demands a lot of experience with GTK. There are several goals. The first is to remove our dependency on GDL. The second is to have all dialogs be laid out by the new "Multipane" widget. The third is to have updates managed through the new system. One thing missing from the demo is saving dialog state (which dialogs are open and where). 2019-07-10 02-57-29 PDT Tavmjong Bah This would be a major change in Inkscape's behavior so something not to be changed without thinking about it carefully. 2020-08-22 05-28-15 PDT Tavmjong Bah Thanks 2020-08-14 01-49-02 PDT Tavmjong Bah Ah, yes, I am on Wayland. 2020-08-22 06-07-46 PDT Tavmjong Bah Icon Preview triggers it... 2019-06-29 10-09-36 PDT Tavmjong Bah You can bind extra parameters or suppress them using sigc_bind. There are tons of examples in the code. 2019-06-18 22-40-47 PDT Tavmjong Bah Have you tested the mesh gradients with a variety of meshes (not just the pepper)? If so, then you can go ahead and prepare a push. 2019-07-10 02-49-39 PDT Tavmjong Bah HI, don't worry about Inkscape while your there. You can catch up after. I've got code that extracts hatches and patterns from files looking at the style attribute. I'll try to get it posted today. If you use the standard routines for setting properties, the group issue should just work (you'll need to follow the same model used elsewhere like in the Fill and Stroke dialog or in clicking on a palette entry). 2019-07-22 06-20-14 PDT Tavmjong Bah HI, Glad to see you making progress. I'll try to give our code a try today... might be hard as we're having a big family reunion this week at a house on a lake. I will make sure I get the GSoC evaluation done. 2019-08-18 09-26-31 PDT Tavmjong Bah Thanks 2019-06-27 01-57-24 PDT Tavmjong Bah (Checked with Firefox.) 2019-06-27 01-15-37 PDT Tavmjong Bah And as a final refinement, we probably want to change the "insert_mesh_polyfill_flag to insert_polyfills_flag 2020-06-05 08-50-27 PDT Tavmjong Bah Are you talking about test-application? (That's the only one `grep` shows has the word `deprecated`.) That's a toy project to investigate how to use actions. The more complete `paned` prototype is in `gtkmm_paned`. 2019-08-07 12-26-11 PDT Tavmjong Bah I think it is fine if you write about what you did in the repository. Once you have finished, you can post a link to it to the developer's list so everyone can see your great work. 2020-03-17 08-36-23 PDT Tavmjong Bah For P9, it is hard! There is another student who is applying to work on it. 2019-06-24 04-33-20 PDT Tavmjong Bah The last overflow example should be green. Hatches are rendered left to right, so the the first hatch generates a green line at "5" and a red line at "20", the next hatch is at +15, so it's path at "5" is on top of the previous hatchpath's red line. 2019-07-18 06-05-20 PDT Tavmjong Bah Hi, Glad you finally made it back. Yes, I'm in the US. Fortunately my flights went smoothly. The code I sent can be used to scan the current document. One just has to extract out the scanning code into a separate function so it can be used to scan both external files (e.g. patterns.svg) and the current document. 2019-06-20 06-55-07 PDT Tavmjong Bah hatch_test.svg (https://chat.inkscape.org/file-upload/zyjaBjNYWEC779Zgg/hatch_test.svg) 2019-08-07 12-19-57 PDT Tavmjong Bah The scaling knot handle should effect the hatchTransform attribute. 2019-06-25 09-44-41 PDT Tavmjong Bah Hi, We don't have hatches in the GUI. The idea would be to start with a dialog that shows patterns and hatches and allow one to choose one. We have a set of predefined patterns and should do the same for hatches. Right now, one can choose a pattern under the Fill and Stroke tabs of the Fill and Stroke dialog if the pattern option is selected but there is no visual preview. 2019-07-26 18-00-49 PDT Tavmjong Bah Hi, Any updates? 2019-06-09 01-32-50 PDT Tavmjong Bah Yes, that would be expected. 2020-08-17 07-23-45 PDT Tavmjong Bah Yeah! (To fixing the most annoying bugs and crashes!) 2019-06-26 10-59-02 PDT Tavmjong Bah You can try to do it from gitlab's interface. 2019-07-03 10-55-36 PDT Tavmjong Bah * Should test length of fill string to make sure it is long enough for finding the substring (otherwise Inkscape will crash). 2019-07-06 05-40-48 PDT Tavmjong Bah I'll have a look. 2019-07-02 05-58-57 PDT Tavmjong Bah Good! 2019-07-03 10-38-39 PDT Tavmjong Bah I don't know why, but the hatches are not being rendered. If you add a stroke to the rectangles, the stroke will be shown. 2019-06-29 10-11-34 PDT Tavmjong Bah Take a look at https://developer.gnome.org/gtkmm-tutorial/stable/sec-connecting-signal-hand... 2019-06-18 08-48-39 PDT Tavmjong Bah Hi, Where are you pushing your changes? 2019-06-09 01-35-13 PDT Tavmjong Bah Start with linear... that is the easiest. 2020-08-17 07-25-49 PDT Tavmjong Bah The ruler/scrollbar setting code should be moved to canvas-grid.h... but that's another refactoring project. It's my long term goal to get rid of SPDesktop. 2019-06-27 01-54-37 PDT Tavmjong Bah hatch01_with_js.svg (https://chat.inkscape.org/file-upload/B3gbsvi8EcwCNdkAy/hatch01_with_js.svg) 2019-08-09 12-48-46 PDT Tavmjong Bah The update works the first time probably because a new hatch is created to hold the transform. 2019-06-18 22-43-53 PDT Tavmjong Bah I would check the GSoC documents... I don't recall what is exactly required. 2019-07-03 03-13-54 PDT Tavmjong Bah Why don't you push what you have and then I'll try to have a look this afternoon. 2019-06-27 01-10-53 PDT Tavmjong Bah Oh, you'll need to modify src/extension/internal/svg.cpp to add the javascript to the file. Just follow what is done for meshes. 2020-08-06 00-53-57 PDT Tavmjong Bah I know how frustrating bug fixing can be. I just spent three days on a bug in my SPCanvasItem rewrite... finally got it fixed. Might be a good idea to take a break on that bug and work on something else for awhile then come back to it. I actually get some of my better coding ideas working in the garden or during a bike ride. 2019-07-06 03-48-06 PDT Tavmjong Bah In function no, but in clarity yes. All the rest of the code uses the pattern where Gtk::manage is used at the construction point. 2019-07-06 06-15-49 PDT Tavmjong Bah You also forgot to wrap in it a "url()". 2020-09-15 03-48-52 PDT Tavmjong Bah I was wondering what problems you were having in rebasing after my shortcuts stuff was merged. Also, there is a merge request that you're probably the best person to review: https://gitlab.com/inkscape/inkscape/-/merge_requests/2315 2020-08-14 01-42-11 PDT Tavmjong Bah That's for sure! 2019-07-03 10-52-51 PDT Tavmjong Bah There is a lot of weird stuff going on. I'll have some time tomorrow to look closer. 2019-07-03 11-05-34 PDT Tavmjong Bah One more thing, there is really no need for hatches.js to be in the share/inkscape/hatches directory. 2020-07-11 11-00-41 PDT Tavmjong Bah @vanntile Next Saturday's "Summer Hackfest" is going to focus on User Experience. Maybe you can demo what you are doing? 2019-06-24 04-18-21 PDT Tavmjong Bah Yes, the hatch on the left is the test, the one on the right the expected rendering. Give me a moment to check the image in question. 2020-07-18 11-46-19 PDT Tavmjong Bah Nice presentation. I think people are quite enthused about this! 2019-06-26 04-34-37 PDT Tavmjong Bah If the test passes (matches Inkscape's rendering) push the fix. I'll then check it and commit the MR (if you could update to current master, it would be helpful). 2019-06-28 03-41-13 PDT Tavmjong Bah 1. I agree. Have a look at Gtk::FlowBox as the container widget. 2019-07-03 10-54-20 PDT Tavmjong Bah * Need to copy TestHatches.svg to the proper directory when installing. 2019-06-27 11-33-56 PDT Tavmjong Bah M a b c d should produce the same rendering as M a b L c d. 2019-06-26 11-15-39 PDT Tavmjong Bah Looks good to me... I've set it to merge automatically once the pipeline succeeds. 2019-07-04 08-52-38 PDT Tavmjong Bah Well, I wouldn't say it was you bad, it's our horrible, undocumented code. 2019-06-29 11-11-57 PDT Tavmjong Bah Good to know you got it working! Good night! 2020-08-22 00-01-50 PDT Tavmjong Bah Responding to his concerns, asking for more info. 2019-07-03 10-53-56 PDT Tavmjong Bah A few problems I've found: 2019-06-26 10-58-40 PDT Tavmjong Bah It's really only altering three files... so you could just to a fresh checkout, add them, and make a new merge request. (I could just copy the files and commit them but I think it's better for you to create the merge request so you get the credit. 2019-06-28 03-43-16 PDT Tavmjong Bah 4. If I understand the question correctly, you'll need to add the new dialog in the dialog manager. I'm not sure where that code is. 2019-07-10 11-33-29 PDT Tavmjong Bah Hmm, that doesn't seem to work. 2020-08-22 06-08-46 PDT Tavmjong Bah Just canvasitems. 2020-08-22 06-18-56 PDT Tavmjong Bah The .h file has `Inkscape::Drawing renderDrawing;` which creates a drawing without a linked CanvasItemDisplay. So I think it's OK. 2019-07-04 05-40-57 PDT Tavmjong Bah What do you mean by a "node representation"? 2020-03-18 02-50-47 PDT Tavmjong Bah Any comments? 2019-07-03 11-04-54 PDT Tavmjong Bah "ninja install" needs to copy the file to share/inkscape/hatches 2020-08-22 06-14-11 PDT Tavmjong Bah Hmm, why would paint servers being trying to draw an SVG snippet? Maybe rendering a hatch or pattern. 2019-08-18 07-32-02 PDT Tavmjong Bah Can you bring the paint server dialog branch up-to-date with master? 2019-07-18 06-06-14 PDT Tavmjong Bah The new function should just take an SPDocument* as an argument. 2019-07-06 06-12-01 PDT Tavmjong Bah OK: SPDocument *document_target = desktop->getDocument(); Inkscape::XML::Document *xml_doc = document_target->getReprDoc(); Inkscape::XML::Node *repr = paint_server->getRepr()->duplicate(xml_doc); document_target->getDefs()->appendChild(repr); 2019-07-04 05-15-54 PDT Tavmjong Bah I'm still working on why hatches are not showing up. Patterns work (if the code is modified to include them). 2020-08-22 06-05-35 PDT Tavmjong Bah OK, thanks! (I could imagine a dialog with a preview might trigger it.) 2019-06-20 05-57-29 PDT Tavmjong Bah Ah, no. That appears to be a bug! 2019-07-04 05-43-50 PDT Tavmjong Bah Oh, for an individual node, so all the attributes, etc. For that, one can use the XML dialog. 2019-08-17 08-28-21 PDT Tavmjong Bah Can you make a merge request or the hatch UIX branch?