2020-02-07 13-14-40 PST Martin Owens This sound be parked in inkscape as it's a regression. Thanks for testing. 2020-02-12 00-08-37 PST Martin Owens They may have been voted on my multiple people, or it may have looked like it wasn't reproduced when I checked. 2020-02-24 17-24-52 PST Martin Owens I'm imagining sinarios of things being broken in 0.48 and people re-opening bugs ;-) because we haven't back ported the fix :-P 2020-04-16 07-08-10 PDT Martin Owens Also it could just be me ;-) 2020-09-24 07-43-36 PDT Martin Owens I won't be able to tell you about what happens, but hopefully we can resolve the agressiveness. 2020-02-24 17-24-01 PST Martin Owens I think the right thing is a) request the fix be back-ported via chat, the if no response or no one can get to it right away b) create a new issue to pack-port the fix. 2020-01-24 05-28-47 PST Martin Owens That is not something I can answer 2020-04-30 13-19-12 PDT Martin Owens finite_exit_dock_headroom_wistful_zealous 2020-09-24 07-43-16 PDT Martin Owens OK I've kicked off a mediation ticket between Chris Rodgers, Grey, Myself and Bryce Harrington 2020-02-12 00-08-53 PST Martin Owens It's a bit discretionary. So there may be no reason. 2020-02-12 00-38-13 PST Martin Owens Software freedom isn't without sacrifice or generosity. As Mako would say. 2020-09-24 07-32-31 PDT Martin Owens I've deleted his comment, but I'm going to need to have words 2020-01-24 12-18-10 PST Martin Owens We actually had to do a specific recruitment drive for mac developers. 2020-10-06 04-21-23 PDT Martin Owens The actual bulk of the code looks good. It needs more spacing out and it's a bit cramped. But it's good stuff. 2020-09-24 07-37-12 PDT Martin Owens Thanks 2020-02-12 00-06-13 PST Martin Owens So it'll have to wait for the next list. 2020-01-24 12-18-31 PST Martin Owens 90% of our users are windows users, 90% of our developers are Linux developers ;-) 2020-01-24 05-26-46 PST Martin Owens It means just like flowRoot text in 0.92, text with shapes wouldn't be visible in firefox and other svg viewers 2020-02-12 00-13-52 PST Martin Owens I've basically quit my contracting (paid) work to get this Inkscape 1.0 release done. So it's super important to me that we prioritise (because I can't keep up this level of investment for very long) 2020-02-24 17-28-48 PST Martin Owens As soon as we branched, we created this process. (or lack of process). Part of the job I'm going to be doing is reviewing the bugs and commits to master not in 1.0.x 2020-02-24 17-24-23 PST Martin Owens But I'm not sure, as there isn't a process for backporting. 2020-02-12 00-13-07 PST Martin Owens Ah time. 2020-09-24 07-36-00 PDT Martin Owens Did you get it in an email? 2020-03-11 15-49-44 PDT Martin Owens I think you are right and any help with extensions is very useful. 2020-02-12 00-10-00 PST Martin Owens I mean take the one I'm currently looking at, 847, that's very likely not fixable. 2020-10-06 04-20-12 PDT Martin Owens I'm not as keen on staticmethods, when I see isgradient I have trouble with the variable name `x` (what is x?) If it's a style attribute check, this is likely something that should be inside style.py rather than the elements even though it needs the svg to do the check. 2020-01-24 12-18-57 PST Martin Owens We're asymmetric. It's a miricle Inkscape works on those platforms. 2020-10-06 04-22-26 PDT Martin Owens Would the xaml extensions be a candidate for pushing towards pypi instead? 2020-09-24 07-31-55 PDT Martin Owens @jhofinger Having a look now 2020-02-24 17-22-34 PST Martin Owens This is a hole in how we handle bugs and certainly depends on the people looking after the bug. 2020-10-06 04-28-42 PDT Martin Owens Good to know, this is going to be an interesting process of splitting things out onto pypi but having a list of things to include in Inkscape ;-) 2020-02-08 05-32-09 PST Martin Owens :thumbsup: 2020-10-06 04-28-49 PDT Martin Owens requirements.txt I see thee 2020-02-24 17-22-07 PST Martin Owens Yes, but I don't think we have a process for bugs in 1.0.x vs. master. 2020-04-15 14-10-12 PDT Martin Owens Just to moan about issue 2307 a little. Chris was bad mouthing Inkscape on ycombinator as well as being an utter jerk. Now I fear the issue will become a swamp of bitter nasty bastards trying to make general and unproductive complaints about Inkscape and pointing out how bad we are because we don't slave away at their issues specifically. 2020-09-24 07-35-49 PDT Martin Owens Ah I forgot to screenshot the comment before I deleted it. 2020-09-24 07-44-02 PDT Martin Owens :thumbsup: 2020-02-12 00-06-02 PST Martin Owens We can't fix EVERYTHING (otherwise inkscape will never be released) 2020-01-11 13-15-12 PST Martin Owens It could be, although whenever I've tried to use swatched they've never worked. But I just assumed that was because I was dumb. 2020-01-11 13-09-57 PST Martin Owens Thanks for brining this up. How many of these are regressions between 0.92 and 1.0 (if you know) I want to make a case for removing swatchs for 1.0, but regressions are harder to agrue because we'd be removing functionality we had in 0.92 2020-03-11 10-58-06 PDT Martin Owens I was hoping for an svg file in the report. 2020-01-24 05-32-37 PST Martin Owens You should add your thoughts to the public channel. 2020-10-06 04-22-13 PDT Martin Owens The other issue is one of modules. The xamllib will be a new module which at some point should be it's own pypi install depedency instead of something we ship with Inkscape 2020-02-12 00-05-48 PST Martin Owens No, there's just a limit to how many items I'm willing to put on the list. 2020-01-24 12-15-17 PST Martin Owens ede123 is (Patrick), there are a few others. Although non-linux platforms always need community management and encouragement. Developers usually end up installing Linux in the end. 2020-03-11 11-24-14 PDT Martin Owens ok, no problem 2020-02-07 13-17-52 PST Martin Owens Thanks! 2020-01-24 05-27-07 PST Martin Owens But editing in inkscape would be much better 2020-09-24 07-29-13 PDT jhofinger Hi, I believe we have a situation here. https://gitlab.com/inkscape/inbox/-/issues/3701#note_418171636 Should I try to deal with this or would you consider this your job? 2020-09-24 07-36-32 PDT jhofinger Clipboard - September 24, 2020 4:36 PM (https://chat.inkscape.org/file-upload/Kw8hf8gFx3fosHLmB/Clipboard%20-%20Sept...) 2020-10-06 04-30-03 PDT jhofinger If everything is installed via pip, then it shouldn't be different than installing other python dependencies into the inkscape package, right? It probably simplifies it even, since the subpackages just define their own dependencies 2020-01-24 12-22-16 PST jhofinger Yeah, C++ development is just more comfortable on Linux, as far as I can see. I'll ask Patrick for some guidance 2020-10-06 04-16-13 PDT jhofinger I'm making good progress with the new XAML export (started over yesterday basically), but I could need a design review until it's too late to refactor everything again. So if you have time, I'd appreciate you having a look: https://gitlab.com/jhofinger/extensions/-/compare/master...new-xaml-export
The code is working, it ignores transforms for now (that comes next), also it doesn't support gradients, filters and texts yet. But the overall design has converged by now.
Unit tests will be basically the entire (applicable) SVG test suite
The general idea is to mirror the class structure of Inkex in the *Parser classes, traversing the SVG. These create XAML-like objects (everything inheriting from XAMLObject), which manage their XML representation themselves. For the backwards conversion, I'll fill these via lxml.objectify, essentially the same way as the Inkex document structure is created, then traverse them and construct the SVG backwards. 2020-02-24 17-24-50 PST jhofinger Okay, maybe this should be discussed in one of the group chats. 2020-02-24 17-29-43 PST jhofinger That's a good idea. I will start a discussion on the process in the dev chat tomorrow. 2020-03-11 13-25-33 PDT jhofinger I don't want to steal this from you if you want to do it. It's just that I can take over tasks that I'm able to do, to free your resources so you can fix the important issues in the C++ code, so 1.0 can be released as soon as possible 😉 2020-10-06 04-27-36 PDT jhofinger There are a few SVG->XAML conversion programs already: Svg2XAML, SharpVectors, ab4d. The main benefits of Inkscape having its own is the following: - Easy to use from the place where people create their icons (most of the time on StackOverflow, Inkscape is recommended for converting) - *Editing* XAML files. None of the above libraries have a backwards conversion. By the way, since in XAML it's typical to store multiple designs in a single file, this library could benefit a lot from multipage editing :D 2020-03-11 10-57-36 PDT jhofinger Mwe for the extrude bug: draw circle - convert to path - duplicate - extrude 2020-04-15 23-10-27 PDT jhofinger Thanks for the warning. I'll keep it in mind when I deal with the issue the next time. 2020-01-24 05-28-18 PST jhofinger Okay. Why is there a fallback that doesn't work? 2020-10-11 06-28-20 PDT jhofinger (the change in _base.py is a hackfix for now, just to demonstrate the principle) 2020-02-12 00-12-55 PST jhofinger Yes that looks like a difficult one. But I think you're doing a great job! (Looking forward to helping more in the code, but I don't have the time atm) 2020-01-11 13-24-07 PST jhofinger So did I, until I saw all these issues - and if you also don't understand it, then it's probably just buggy 2020-10-06 04-21-20 PDT jhofinger Oh that... s is a string. It's copied from the interpolations rewrite 2020-02-08 04-08-12 PST jhofinger Reverting the changes to selcue has fixed the issue (y) Thanks! 2020-10-06 04-33-00 PDT jhofinger Thanks for having a look. 2020-10-06 04-34-14 PDT jhofinger [ ](https://chat.inkscape.org/direct/doctormo?msg=fW8dJq8ti6yRFZpMp) Also I want to add support for another FOSS project (Avalonia.ui) which is basically Microsoft WPF supercharged, and none of the libraries support that ;) () 2020-10-10 22-32-29 PDT jhofinger Would it be acceptable to add another dependency to inkex for proper css parsing? Such as tinycss 2020-01-24 05-28-46 PST jhofinger btw I solved the compiling problem. Stupid me looked in build/bin, not build/inkscape/bin 2020-01-24 05-33-37 PST jhofinger Ok will do 2020-01-11 13-12-01 PST jhofinger I can test that tomorrow, I have never used swatches before, but I guess it's a standard work flow for designer people (which I'm not). Will tell you. Interestingly there are very few issues on swatches on Launchpad, so probably a few regressions? 2020-02-12 00-07-54 PST jhofinger It was in my comment of the vectors list. You added 776 and 703 from the same line, but I'd say that 782 is far more important. I'm not asking to reopen the discussion. 2020-10-06 04-23-37 PDT jhofinger Definitely! It's absolutely something which people would use from out of Inkscape as well, maybe even as a standalone library. So as long as it can be still shipped with Inkscape, I'm open for it. 2020-09-24 07-43-54 PDT jhofinger I fully understand. Thanks for your quick action! 2020-09-24 07-32-57 PDT jhofinger Alright. He is always a bit flippant, but that's crossed a line for me 2020-02-24 16-23-55 PST jhofinger It's fixed in Master (well except the undo problem which Maren reported). I only reopened because the fix is not in 1.0.x, and I'd say pretty much everyone would agree that tutorials looking like this is a big no-no for 1.0. 2020-02-07 12-26-40 PST jhofinger Got some bad news, your fix for the bounding box trails introduced another regression :/ https://gitlab.com/inkscape/inbox/issues/1458#note_284511359 2020-02-12 00-36-55 PST jhofinger Wow, I already wondered how you have so much time 2020-02-12 00-09-11 PST jhofinger They both were only mentioned by me. 2020-01-11 13-03-20 PST jhofinger Hi Martin, yes, such a level of detail is not necessary there, I just wanted to write *how* broken the swatches tool is. I just added the issue numbers where available, most of them are already there. Do whatever you want with it; I would be fine with it if Inkscape ships without swatches. 2020-02-24 17-23-40 PST jhofinger Absolutely. Reopening was my way to state "we might have forgotten something there". 2020-03-11 11-23-51 PDT jhofinger But I also wanted to investigate that myself, maybe tomorrow 2020-02-12 00-46-36 PST jhofinger Absolutely. 2020-01-24 11-02-08 PST jhofinger Too bad if VS code doesn't work. Do you know anyone who is currently developing on windows? Especially debugging 2020-01-24 05-21-17 PST jhofinger Not sure if I understood that message. What kind of fallback is there at the moment, and what consequences woud its removal have? Does this mean to definitely make SVG2 text the default for 1.0? 2020-01-24 12-17-43 PST jhofinger Developers usually end up installing Linux in the end. That's out of the question for now, but I guess it's handy to have some devs at hand who can debug Windows- or mac specific issues 2020-10-06 04-27-53 PDT jhofinger So I'd definitely want to keep the functionality with vanilla Inkscape 2020-02-24 17-27-26 PST jhofinger This is something totally different. I wouldn't exactly call this process backporting even, since 1.0 isn't even out yet. There are very few bugfixes where there is a good reason not to have them in 1.0, if they are in 1.1.x at the moment. 2020-03-11 11-21-51 PDT jhofinger The first file of the „further investigation“ comment 2020-02-07 13-17-35 PST jhofinger Yeah will open one tomorrow 2020-02-12 00-02-50 PST jhofinger Hi, I just noticed that you missed https://gitlab.com/inkscape/inkscape/issues/782 in the Vector's blockers list! Was that on purpose? 2020-01-11 13-05-45 PST jhofinger (Sorry, didn't add the numbers immediately, I was in a train today and only had very slow internet) 2020-10-11 06-26-39 PDT jhofinger This would be a way to parse CSS in a relatively easy way without XPath at all. https://gitlab.com/jhofinger/extensions/-/compare/master...css-style-parsing
It might be possible to call matcher.match directly on a BaseElement, if we add a few things (e.g. an id attribute), leaving the wrapper out. The wrapper is just there because the library is _designed_ for ElementTree, which has no notion of a parent element. 2020-09-24 07-42-04 PDT jhofinger Thank you :) 2020-03-11 11-09-24 PDT jhofinger There is one 2020-04-30 13-28-14 PDT jhofinger works like a charm, thanks!