2019-08-24 01-07-36 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	Firefox (https://chat.inkscape.org/file-upload/ZkeEDZqnh2RfictE6/Firefox)
2020-08-22 06-08-11 PDT	Valentin Ionita <vanntile@protonmail.com>	and Paint Servers
2019-07-10 02-26-23 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	I think the first step needed was to connect
2019-06-09 00-59-53 PDT	Valentin Ionita <vanntile@protonmail.com>	I'm looking at it
2019-09-01 09-23-52 PDT	Valentin Ionita <vanntile@protonmail.com>	Thank you, I'll remind you tomorrow
2019-07-06 05-40-30 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	ok, thank yuo, I'll update
2019-06-24 09-51-20 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	Screenshot from 2020-06-05 18-51-23.png (https://chat.inkscape.org/file-upload/h6Dv7BB5qcnsthLz8/Screenshot%20from%202020-06-05%2018-51-23.png)
2019-07-30 07-36-56 PDT	Valentin Ionita <vanntile@protonmail.com>	However I fixed the issue I was working on (its pipeline is running now)
2019-07-10 02-27-16 PDT	Valentin Ionita <vanntile@protonmail.com>	[ ](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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	Mornin'. Ok, I will do that
2019-07-04 08-59-57 PDT	Valentin Ionita <vanntile@protonmail.com>	Screenshot from 2019-07-04 18-59-43.png (https://chat.inkscape.org/file-upload/9jSLdJmr978PwBhRQ/Screenshot%20from%202019-07-04%2018-59-43.png)
2019-07-18 09-26-29 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	I'm not home at the moment
2019-08-18 07-50-16 PDT	Valentin Ionita <vanntile@protonmail.com>	Yes, tonight I will
2019-07-26 20-37-24 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	What do you think would be specific enough and would bring value to Inkscape?
2020-08-21 23-56-06 PDT	Valentin Ionita <vanntile@protonmail.com>	Our conversation starts on this thread on the UX issue
2020-08-17 07-26-43 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	Either that or I'm not doing the right interactions
2019-06-27 02-22-32 PDT	Valentin Ionita <vanntile@protonmail.com>	Screenshot from 2019-06-27 12-21-29.png (https://chat.inkscape.org/file-upload/dfYoy5PFSJkqt5WRQ/Screenshot%20from%202019-06-27%2012-21-29.png)
2019-06-20 06-52-20 PDT	Valentin Ionita <vanntile@protonmail.com>	Thank you, I'll use this for testing.
2020-08-22 00-16-24 PDT	Valentin Ionita <vanntile@protonmail.com>	And then linked in the DialogBase::update() method
2020-08-14 02-21-34 PDT	Valentin Ionita <vanntile@protonmail.com>	Honestly, it could be much worse. It's okay, I get the code.
2019-06-30 09-14-35 PDT	Valentin Ionita <vanntile@protonmail.com>	No problem, thank you.
2019-06-29 10-08-14 PDT	Valentin Ionita <vanntile@protonmail.com>	ok. I will
2019-07-04 05-31-23 PDT	Valentin Ionita <vanntile@protonmail.com>	One problem, if I may
2020-03-18 03-16-31 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	Thank you, no problem, you won't need to worry today.
2019-06-26 10-58-14 PDT	Valentin Ionita <vanntile@protonmail.com>	Should I do it from gitlab's own interface?
2019-08-17 08-28-47 PDT	Valentin Ionita <vanntile@protonmail.com>	yes, I will pull from master then make a MR
2019-07-05 04-40-35 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	thank you
2019-07-10 03-07-34 PDT	Valentin Ionita <vanntile@protonmail.com>	It would be nice. I'll make a make a method for the meantime.
2020-08-22 06-10-37 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	How so?
2020-03-18 03-08-44 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	[ ](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 <vanntile@protonmail.com>	Hello, just pushed the fix for the above bugs
2020-08-14 01-47-49 PDT	Valentin Ionita <vanntile@protonmail.com>	I have a question, are you on Wayland?
2020-06-05 08-52-51 PDT	Valentin Ionita <vanntile@protonmail.com>	these two callbacks are only used to connect to signals for the dropzones
2019-07-04 05-16-46 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	Oh, that's great. Then I'll push the license, to keep it in check
2020-08-22 00-10-17 PDT	Valentin Ionita <vanntile@protonmail.com>	here
2020-08-14 01-43-08 PDT	Valentin Ionita <vanntile@protonmail.com>	ok, then I'll work on reproducing it, and then look for a fix
2019-07-26 20-40-37 PDT	Valentin Ionita <vanntile@protonmail.com>	[ ](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 <vanntile@protonmail.com>	Fixed it by adding inkscape:collect="always" to hatches.svg
2020-08-21 23-57-22 PDT	Valentin Ionita <vanntile@protonmail.com>	And his blocker is no. 4
2019-09-03 07-34-48 PDT	Valentin Ionita <vanntile@protonmail.com>	At least this is how it is on the mobile app
2019-08-06 20-22-12 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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. 😅
2019-06-26 04-33-36 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	in a moment
2019-08-11 20-29-50 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	Ah, I haven't actually done ninja install. I only run from the executable
2019-07-08 03-23-01 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	I've seen the headless Inkscape and that is interesting.
2019-06-30 07-15-38 PDT	Valentin Ionita <vanntile@protonmail.com>	Nevermind, was just too distracted by the code to realise it was your own sandbox
2019-06-26 04-38-50 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	Ok, I'll ping you if I encounter issues
2019-07-29 02-17-19 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	For the moment, this is what I have
2019-08-20 02-19-19 PDT	Valentin Ionita <vanntile@protonmail.com>	Thank you! I have completed the evalutation this morning.
2020-08-17 07-19-53 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	https://gitlab.com/inkscape/ux/-/issues/38#note_399684778
2019-07-03 02-58-11 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	In the actual app, what is the ink dialog?
2020-08-14 01-48-06 PDT	Valentin Ionita <vanntile@protonmail.com>	I think that's why I can't reproduce your bug.
2019-07-02 05-58-43 PDT	Valentin Ionita <vanntile@protonmail.com>	I am okay with everything above. I think
2020-06-05 07-46-21 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	Got it, will take care of all that later today 😀
2019-08-11 20-59-46 PDT	Valentin Ionita <vanntile@protonmail.com>	Great, thank you. It's gonna be enough, I think
2019-07-05 04-43-37 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	https://drive.google.com/file/d/1-hkNbkBhqGxev-RWArv-lLRwryKlNZgl/view?usp=drivesdk
2019-07-03 10-44-00 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	Screenshot from 2019-07-04 19-52-02.png (https://chat.inkscape.org/file-upload/2btqvKzyiXpSe5jMA/Screenshot%20from%202019-07-04%2019-52-02.png)
2019-07-03 01-35-29 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	``` ```
2019-07-22 08-37-02 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	nevermind. fixed
2019-07-26 20-35-55 PDT	Valentin Ionita <vanntile@protonmail.com>	Hello, I've just woken up.
2019-07-26 20-38-55 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	Wooow, that sounds great - the Alps.
2019-06-29 10-02-39 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	I pushed
2019-07-05 04-39-34 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	Hey, it compiled and ran correctly, but I have the following console warning/error
2019-06-25 10-38-30 PDT	Valentin Ionita <vanntile@protonmail.com>	One last thing from me tonight (and I promise I'll stop :joy: :joy: :joy: ) – 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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	[ ](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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	I have fixed the behavior with groups and pushed my changes
2019-07-03 02-58-17 PDT	Valentin Ionita <vanntile@protonmail.com>	I can click on them...
2020-08-22 00-10-14 PDT	Valentin Ionita <vanntile@protonmail.com>	https://gitlab.com/vanntile/inkscape-gsoc-2020/-/blob/master/RFC.md#dialogbaseupdate
2020-08-14 01-48-50 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	https://gitlab.com/inkscape/inkscape/issues/309
2020-08-22 00-02-02 PDT	Valentin Ionita <vanntile@protonmail.com>	Oh, well, I feel that's only normal
2019-06-11 06-20-34 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	(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 <vanntile@protonmail.com>	Oh, got it then. No problem, I will change it, I wanted just to understand
2019-07-18 09-25-45 PDT	Valentin Ionita <vanntile@protonmail.com>	It might be solved by the formatting of the Patterns file
2019-06-27 01-28-42 PDT	Valentin Ionita <vanntile@protonmail.com>	?
2019-07-03 02-57-25 PDT	Valentin Ionita <vanntile@protonmail.com>	I tried IconView too.
2019-07-18 09-25-58 PDT	Valentin Ionita <vanntile@protonmail.com>	I'll try to figure it out next days.
2020-09-15 10-51-46 PDT	Valentin Ionita <vanntile@protonmail.com>	[ ](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 <vanntile@protonmail.com>	Screenshot from 2020-06-05 18-51-45.png (https://chat.inkscape.org/file-upload/x7W7uC8ZNffNYuRGz/Screenshot%20from%202020-06-05%2018-51-45.png)
2019-06-28 03-18-59 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	Screenshot from 2020-08-22 15-49-46.png (https://chat.inkscape.org/file-upload/TCixLvdPRRvPdZtM2/Screenshot%20from%202020-08-22%2015-49-46.png)
2020-03-18 02-58-54 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	it happens just before finishing loading.
2019-07-18 06-08-31 PDT	Valentin Ionita <vanntile@protonmail.com>	Glad you had a good flight
2020-08-18 01-02-56 PDT	Valentin Ionita <vanntile@protonmail.com>	And thank you for keeping me up to date
2019-06-20 06-55-41 PDT	Valentin Ionita <vanntile@protonmail.com>	great, will use
2019-06-20 05-57-47 PDT	Valentin Ionita <vanntile@protonmail.com>	Well, I should report that then. Maybe later today
2019-06-27 11-38-54 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	Actually, a question
2019-06-26 10-56-30 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	The bicubic interoplation is working
2019-06-26 10-20-36 PDT	Valentin Ionita <vanntile@protonmail.com>	[ ](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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	Wondering if you have a moment
2019-06-09 01-30-40 PDT	Valentin Ionita <vanntile@protonmail.com>	exactly right
2020-06-07 07-46-28 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	No need to rush today, hope things will ameliorate.
2019-07-03 01-33-36 PDT	Valentin Ionita <vanntile@protonmail.com>	https://developer.gnome.org/gtkmm/stable/classGtk_1_1FlowBox.html#ac54c1f423d594ad7b7a985f1c1eeac36
2019-07-04 09-55-24 PDT	Valentin Ionita <vanntile@protonmail.com>	Anyway, I'll stop for today.
2020-08-18 01-01-34 PDT	Valentin Ionita <vanntile@protonmail.com>	Yep, I am aware about his high number of requests/suggestions.
2019-07-04 05-34-45 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	I haven't checked master yet
2020-08-22 05-35-26 PDT	Valentin Ionita <vanntile@protonmail.com>	If it builds, I guess it's okay merging your branch.
2019-07-18 06-08-03 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	Got news for you
2019-06-26 10-26-01 PDT	Valentin Ionita <vanntile@protonmail.com>	Screenshot from 2019-06-26 20-25-40.png (https://chat.inkscape.org/file-upload/4duPAyiZLdYHgbDoc/Screenshot%20from%202019-06-26%2020-25-40.png)
2019-06-27 02-23-41 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	[ ](https://chat.inkscape.org/direct/Tav?msg=8jdGQDqb9jLsBg2o3) Tav, can you help me with this? ()
2019-06-20 05-58-11 PDT	Valentin Ionita <vanntile@protonmail.com>	Do you have a moment for me to ask a few questions?
2020-08-22 05-28-09 PDT	Valentin Ionita <vanntile@protonmail.com>	I'll try rn and come back with an answer
2020-08-22 06-08-54 PDT	Valentin Ionita <vanntile@protonmail.com>	oh, ok
2020-08-14 01-57-37 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	Cool
2019-07-04 06-14-18 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	https://www.w3.org/TR/2016/CR-SVG2-20160915/pservers.html#Hatchpaths
2019-06-27 10-22-47 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	I'm gonna stop for today
2019-07-03 21-56-18 PDT	Valentin Ionita <vanntile@protonmail.com>	Anyway, I have pushed my changes
2019-06-09 00-55-01 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	[ ](https://chat.inkscape.org/direct/Tav?msg=F4fm9pH5xtdKRj35i) Similar to HTML dom selectors ()
2019-07-04 05-43-05 PDT	Valentin Ionita <vanntile@protonmail.com>	I tried like this *->getRepr()->content()* but failed...
2019-07-10 02-56-39 PDT	Valentin Ionita <vanntile@protonmail.com>	What code you'll push?
2019-06-09 00-47-38 PDT	Valentin Ionita <vanntile@protonmail.com>	Hello, Tav
2019-06-18 08-49-31 PDT	Valentin Ionita <vanntile@protonmail.com>	https://gitlab.com/vanntile/inkscape/tree/mesh-polyfill
2019-07-19 04-50-38 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	Currently, I just can't see what would help the most.
2019-06-24 04-28-30 PDT	Valentin Ionita <vanntile@protonmail.com>	How should I go on?
2019-06-24 04-43-56 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	And that's one of the checks in the code
2019-07-03 21-40-41 PDT	Valentin Ionita <vanntile@protonmail.com>	[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 <vanntile@protonmail.com>	Chromium (https://chat.inkscape.org/file-upload/bvXZWdiYSRfGSfcMb/Chromium)
2019-06-27 10-26-04 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	The code is
2019-07-22 08-33-18 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	Have a great evening. Goodbye
2019-06-24 04-31-18 PDT	Valentin Ionita <vanntile@protonmail.com>	Isn't this where I should be looking?
2020-08-20 07-44-43 PDT	Valentin Ionita <vanntile@protonmail.com>	I'll check your reviews tomorrow
2020-06-05 08-55-42 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	ok, great. I'll take it from here. Thank you
2020-09-15 10-51-14 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	line 45
2019-06-22 23-47-19 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	I think everything's fine, ready to merge
2019-06-28 02-59-33 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	Thank you
2019-07-30 07-37-25 PDT	Valentin Ionita <vanntile@protonmail.com>	Starting tomorrow I'll look at hatch handles. I guess that on another branch. :)
2019-06-28 03-43-51 PDT	Valentin Ionita <vanntile@protonmail.com>	[ ](https://chat.inkscape.org/direct/Tav?msg=5bsNp4TKuHtjY8Zic) it's okay, found it ()
2019-07-02 08-47-53 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	I'll be asking in reference to this section https://www.w3.org/TR/2016/CR-SVG2-20160915/pservers.html#HatchElementAttributes
2019-06-24 04-27-05 PDT	Valentin Ionita <vanntile@protonmail.com>	Screenshot from 2019-06-24 14-26-54.png (https://chat.inkscape.org/file-upload/gF9NsJfBdqAB3tMpA/Screenshot%20from%202019-06-24%2014-26-54.png)
2020-03-18 03-17-23 PDT	Valentin Ionita <vanntile@protonmail.com>	Thank you for your time, once again. :smile:
2019-07-05 04-42-03 PDT	Valentin Ionita <vanntile@protonmail.com>	no problem.
2019-06-30 09-10-12 PDT	Valentin Ionita <vanntile@protonmail.com>	A question. What Gtk Widget should be drawn on? a DrawingArea?
2019-07-02 08-44-59 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	Screenshot from 2019-06-13 21-00-56.png (https://chat.inkscape.org/file-upload/TLQbucTLZ6aLvdtGj/Screenshot%20from%202019-06-13%2021-00-56.png)
2019-07-18 06-10-19 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	I did. You can find it here
2019-06-27 01-39-44 PDT	Valentin Ionita <vanntile@protonmail.com>	Don, I have pushed
2019-07-04 09-55-47 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	https://gitlab.com/inkscape/inkscape/merge_requests/754
2019-07-19 04-48-44 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	I think there's a bit more to be done than the hatch function....
2020-08-22 00-16-07 PDT	Valentin Ionita <vanntile@protonmail.com>	Well, maybe should be moved on the wiki. I think that would be best
2019-06-27 10-22-59 PDT	Valentin Ionita <vanntile@protonmail.com>	Screenshot from 2019-06-27 20-21-06.png (https://chat.inkscape.org/file-upload/sXcHFoeiKtRLakR9W/Screenshot%20from%202019-06-27%2020-21-06.png)
2019-06-24 04-28-20 PDT	Valentin Ionita <vanntile@protonmail.com>	In the target, the first path is rendered on top, in mine the second is.
2020-08-14 01-41-51 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	[ ](https://chat.inkscape.org/direct/Tav?msg=g6KMj5NsJicKcQtpQ) That will be okay. ()
2019-07-05 04-38-40 PDT	Valentin Ionita <vanntile@protonmail.com>	Hey, Tav.
2020-06-05 08-51-09 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	Sure, no problem
2019-06-17 07-20-31 PDT	Valentin Ionita <vanntile@protonmail.com>	great, thank you for the answers
2019-08-17 21-52-53 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	Hello, Tav
2019-07-22 08-34-00 PDT	Valentin Ionita <vanntile@protonmail.com>	And thank you for remembering about the evaluation
2019-06-20 06-52-48 PDT	Valentin Ionita <vanntile@protonmail.com>	left is my render, right is obviously inkscape
2019-07-21 03-23-13 PDT	Valentin Ionita <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	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 <vanntile@protonmail.com>	Goodbye!
2019-07-05 06-38-17 PDT	Valentin Ionita <vanntile@protonmail.com>	I have made a push that gets the ComboBox text from `) 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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	hatch01.svg (https://chat.inkscape.org/file-upload/KemGTBvwPRQyRHZJ5/hatch01.svg)
2019-08-11 20-57-26 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	Hi,
2020-08-14 02-19-07 PDT	Tavmjong Bah <tavmjong@free.fr>	```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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	Sounds good!
2019-06-29 10-08-04 PDT	Tavmjong Bah <tavmjong@free.fr>	You need to lookup the format on the gtkmm website.
2019-07-04 05-27-58 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	OK, that can be fixed later. Yeah, I really hate this whole SPNamedView stuff.
2019-06-24 04-43-15 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	1. I'll need to take a look.
2019-06-27 09-31-45 PDT	Tavmjong Bah <tavmjong@free.fr>	hatch01_with_js.svg (https://chat.inkscape.org/file-upload/x5LEawjf9h8dpFb5g/hatch01_with_js.svg)
2019-07-04 05-31-28 PDT	Tavmjong Bah <tavmjong@free.fr>	I'll work on fixing hatches.
2019-08-19 12-01-09 PDT	Tavmjong Bah <tavmjong@free.fr>	Thanks, merged.
2020-08-14 05-23-41 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	Merged!
2019-07-04 05-40-33 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	https://developer.gnome.org/gtkmm/stable/classGtk_1_1Button.html
2019-08-19 08-53-19 PDT	Tavmjong Bah <tavmjong@free.fr>	Hi, can you update one more time... I'll merge it immediately.
2019-06-20 06-55-31 PDT	Tavmjong Bah <tavmjong@free.fr>	Here's the test file the person who implemented hatches in Inkscape used.
2019-07-04 06-11-29 PDT	Tavmjong Bah <tavmjong@free.fr>	I would just move everything into one directory.
2019-06-23 00-22-42 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	Do you have any urgent things for me to look at?
2019-07-29 08-44-50 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	Yes, there is a preference in the Inkscape Preference's dialog under SVG output.
2019-06-27 09-35-29 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	* 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 <tavmjong@free.fr>	Probably...
2020-06-05 08-58-17 PDT	Tavmjong Bah <tavmjong@free.fr>	Ah, I don't remember why it is marked DEPRICATED.
2019-07-04 11-19-22 PDT	Tavmjong Bah <tavmjong@free.fr>	Great progress! I'm off for the night too.
2020-05-31 04-21-48 PDT	Tavmjong Bah <tavmjong@free.fr>	Sounds good!
2020-03-18 03-02-20 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	Questions are better than problems!
2019-07-05 02-50-18 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	Interesting....
2019-06-26 04-19-24 PDT	Tavmjong Bah <tavmjong@free.fr>	mesh_javascript_test.svg (https://chat.inkscape.org/file-upload/3K85YK2uBFPwMAFNL/mesh_javascript_test.svg)
2019-06-13 11-09-39 PDT	Tavmjong Bah <tavmjong@free.fr>	That's good!
2019-06-18 22-42-39 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	s/is/if/
2019-06-26 10-55-30 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	(insert_mesh_polyfill())
2020-03-17 08-35-45 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	Yes, the last point should be 0,20.
2019-06-18 22-47-10 PDT	Tavmjong Bah <tavmjong@free.fr>	Use can use whatever minifier you want... I don't remember what I did.
2019-06-18 22-46-24 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	It it's not too much work...
2020-09-15 10-54-25 PDT	Tavmjong Bah <tavmjong@free.fr>	Hmm, I don't see how things can be out of sync.
2019-07-06 08-53-52 PDT	Tavmjong Bah <tavmjong@free.fr>	OK, have a good trip.
2020-08-22 00-15-49 PDT	Tavmjong Bah <tavmjong@free.fr>	Very nice write up! Should be linked to from with in code.
2019-06-27 11-35-43 PDT	Tavmjong Bah <tavmjong@free.fr>	https://www.w3.org/TR/SVG/paths.html#PathDataMovetoCommands
2019-06-26 09-40-24 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	You might document which minifier you use (in folder README).
2019-06-09 00-59-35 PDT	Tavmjong Bah <tavmjong@free.fr>	Look for "// Calculate interior derivatives"
2019-06-20 06-03-45 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	Yes, you need to edit src/ui/dialog/inkscape_preferences.cpp
2020-08-21 23-55-09 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	The slopes are calculated per color channel.
2019-06-09 00-50-08 PDT	Tavmjong Bah <tavmjong@free.fr>	Yes
2019-09-03 06-52-15 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	Any progress to report?
2020-08-14 01-40-51 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	Inkscape::GC::release(repr);
2020-08-14 02-20-50 PDT	Tavmjong Bah <tavmjong@free.fr>	I hate editing in RocketChat!
2019-07-08 05-41-20 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	s/is/in/
2019-06-20 05-58-28 PDT	Tavmjong Bah <tavmjong@free.fr>	Sure
2019-06-28 03-46-54 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	* 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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	HatchControl branch merged.
2019-07-02 08-46-07 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	It's all userSpaceOnUse
2019-07-04 08-51-24 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	Thanks!
2019-06-09 01-34-21 PDT	Tavmjong Bah <tavmjong@free.fr>	Or we could artificially insert "corner" points halfway.
2019-06-09 01-30-32 PDT	Tavmjong Bah <tavmjong@free.fr>	But the result is not "symetric".
2019-06-09 01-25-24 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	Great!
2019-06-20 06-51-53 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	Thanks, I'll try to figure out why you are getting the console warning.
2020-03-18 02-51-45 PDT	Tavmjong Bah <tavmjong@free.fr>	My Internet connection seems a bit flaky this morning...
2019-06-09 01-10-55 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	Glad to be of help. I think you are making great progress!
2019-08-17 06-41-03 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	Youp
2019-07-04 06-13-23 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	We don't have color values for the other points.
2020-09-15 10-51-07 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	Finally the "split" mesh is filled using the derived derivatives.
2019-07-06 01-37-50 PDT	Tavmjong Bah <tavmjong@free.fr>	Here are some more comments:
2019-06-27 01-35-19 PDT	Tavmjong Bah <tavmjong@free.fr>	I think that is probably all.
2019-06-26 09-40-51 PDT	Tavmjong Bah <tavmjong@free.fr>	BTW, I did your midterm evaluation.
2020-03-18 03-06-31 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	I was also ill yesterday... which didn't help.
2019-08-09 12-47-25 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	Excellent! Looking forward to see it done! Good luck with exam.
2019-06-09 01-13-02 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	getObjectsBySelector()
2019-06-24 04-34-40 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	You can't just add a repr from one document to another.
2019-06-09 01-04-19 PDT	Tavmjong Bah <tavmjong@free.fr>	ok
2019-07-10 03-04-04 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	* Errors should go to std::cerr and then use continue rather than if/then.
2019-07-04 05-30-00 PDT	Tavmjong Bah <tavmjong@free.fr>	Or you could add a "preview" class and use getObjectsByClass().
2019-07-08 05-36-22 PDT	Tavmjong Bah <tavmjong@free.fr>	Sounds good.
2019-07-10 02-56-51 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	Ah, interesting. Is this explained with comments in the prototype dialog?
2019-07-30 19-58-37 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	Brilliant!
2019-06-09 00-56-51 PDT	Tavmjong Bah <tavmjong@free.fr>	The SPMeshNodeArray::bicubic function is being called on the entire mesh.
2020-03-17 08-55-45 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	Sounds good.
2020-08-22 00-01-01 PDT	Tavmjong Bah <tavmjong@free.fr>	OK, you seem to be handling things well!
2020-03-18 03-15-08 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	Thanks
2020-08-14 01-49-02 PDT	Tavmjong Bah <tavmjong@free.fr>	Ah, yes, I am on Wayland.
2020-08-22 06-07-46 PDT	Tavmjong Bah <tavmjong@free.fr>	Icon Preview triggers it...
2019-06-29 10-09-36 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	Thanks
2019-06-27 01-57-24 PDT	Tavmjong Bah <tavmjong@free.fr>	(Checked with Firefox.)
2019-06-27 01-15-37 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	hatch_test.svg (https://chat.inkscape.org/file-upload/zyjaBjNYWEC779Zgg/hatch_test.svg)
2019-08-07 12-19-57 PDT	Tavmjong Bah <tavmjong@free.fr>	The scaling knot handle should effect the hatchTransform attribute.
2019-06-25 09-44-41 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	Hi, Any updates?
2019-06-09 01-32-50 PDT	Tavmjong Bah <tavmjong@free.fr>	Yes, that would be expected.
2020-08-17 07-23-45 PDT	Tavmjong Bah <tavmjong@free.fr>	Yeah! (To fixing the most annoying bugs and crashes!)
2019-06-26 10-59-02 PDT	Tavmjong Bah <tavmjong@free.fr>	You can try to do it from gitlab's interface.
2019-07-03 10-55-36 PDT	Tavmjong Bah <tavmjong@free.fr>	* 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 <tavmjong@free.fr>	I'll have a look.
2019-07-02 05-58-57 PDT	Tavmjong Bah <tavmjong@free.fr>	Good!
2019-07-03 10-38-39 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	Take a look at https://developer.gnome.org/gtkmm-tutorial/stable/sec-connecting-signal-handlers.html.en
2019-06-18 08-48-39 PDT	Tavmjong Bah <tavmjong@free.fr>	Hi, Where are you pushing your changes?
2019-06-09 01-35-13 PDT	Tavmjong Bah <tavmjong@free.fr>	Start with linear... that is the easiest.
2020-08-17 07-25-49 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	hatch01_with_js.svg (https://chat.inkscape.org/file-upload/B3gbsvi8EcwCNdkAy/hatch01_with_js.svg)
2019-08-09 12-48-46 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	I would check the GSoC documents... I don't recall what is exactly required.
2019-07-03 03-13-54 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	You also forgot to wrap in it a "url()".
2020-09-15 03-48-52 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	That's for sure!
2019-07-03 10-52-51 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	@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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	Nice presentation. I think people are quite enthused about this!
2019-06-26 04-34-37 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	1. I agree. Have a look at Gtk::FlowBox as the container widget.
2019-07-03 10-54-20 PDT	Tavmjong Bah <tavmjong@free.fr>	* Need to copy TestHatches.svg to the proper directory when installing.
2019-06-27 11-33-56 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	Looks good to me... I've set it to merge automatically once the pipeline succeeds.
2019-07-04 08-52-38 PDT	Tavmjong Bah <tavmjong@free.fr>	Well, I wouldn't say it was you bad, it's our horrible, undocumented code.
2019-06-29 11-11-57 PDT	Tavmjong Bah <tavmjong@free.fr>	Good to know you got it working! Good night!
2020-08-22 00-01-50 PDT	Tavmjong Bah <tavmjong@free.fr>	Responding to his concerns, asking for more info.
2019-07-03 10-53-56 PDT	Tavmjong Bah <tavmjong@free.fr>	A few problems I've found:
2019-06-26 10-58-40 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	Hmm, that doesn't seem to work.
2020-08-22 06-08-46 PDT	Tavmjong Bah <tavmjong@free.fr>	Just canvasitems.
2020-08-22 06-18-56 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	What do you mean by a "node representation"?
2020-03-18 02-50-47 PDT	Tavmjong Bah <tavmjong@free.fr>	Any comments?
2019-07-03 11-04-54 PDT	Tavmjong Bah <tavmjong@free.fr>	"ninja install" needs to copy the file to share/inkscape/hatches
2020-08-22 06-14-11 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	Can you bring the paint server dialog branch up-to-date with master?
2019-07-18 06-06-14 PDT	Tavmjong Bah <tavmjong@free.fr>	The new function should just take an SPDocument* as an argument.
2019-07-06 06-12-01 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	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 <tavmjong@free.fr>	OK, thanks! (I could imagine a dialog with a preview might trigger it.)
2019-06-20 05-57-29 PDT	Tavmjong Bah <tavmjong@free.fr>	Ah, no. That appears to be a bug!
2019-07-04 05-43-50 PDT	Tavmjong Bah <tavmjong@free.fr>	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 <tavmjong@free.fr>	Can you make a merge request or the hatch UIX branch?