- Webmaster - Inkscape Mailman

[Webmaster] Rocket.Chat, 1 Users, 1 Messages, 0 Files, 383358 Minutes, in Direct Message Between: e36king & inkchatbot
by no-reply＠chat.inkscape.org 29 Apr '21

29 Apr '21

2020-08-05 11-43-44 PDT chris Hi there regarding the txt downwards still having issues and trying to work out how to chat so this would be easier thank you

1 0

[Webmaster] Rocket.Chat, 2 Users, 117 Messages, 2 Files, 684419 Minutes, in Direct Message Between: sizmailov & ede123
by no-reply＠chat.inkscape.org 29 Apr '21

29 Apr '21

2020-03-04 13-56-25 PST Sergei Izmailov Let's try. I've tested on 19.04 - works fine 2020-03-07 16-47-39 PST Sergei Izmailov I think we can test gtest installation via minimal cmake project and provide instructions to fix that. A minimal cmake project would contain single `CMakeLists.txt` file: ```cmake cmake_minimum_required(VERSION 3.0) project(temp) find_package(GTest REQUIRED MODULE) ``` 2020-03-07 06-44-00 PST Sergei Izmailov Hi, did you have a chance to look at docker MR? 2020-03-30 15-31-36 PDT Sergei Izmailov In fact it's about same size, see screenshot above 2020-03-04 13-33-48 PST Sergei Izmailov Or you think moving gtest to minimal is better? 2020-03-26 09-30-29 PDT Sergei Izmailov ok, I'll rebase in a minute 2020-03-27 07-31-33 PDT Sergei Izmailov Do you know is `lib/libinkscape_base.dll.a` shared or static? The line ``` [984/1057] Linking CXX shared library bin\libinkscape_base.dll ``` suggests it's shared, but name and location are different. Unfortunately I'm not in a position to test on windows at the moment. I don't see how `2geom` got included in `libinkscape_base`. I'll try to investigate this topic in more detail later (maybe in a week or so). I could experiment with AppVeyor, but it's too long and makes others wait a lot. Currently it doesn't compile even without extra OBJECT/STATIC/SHARED test library. Please share your ideas if you have some 2020-03-19 15-00-14 PDT Sergei Izmailov It depends in which mode you build 2geom. 2020-03-07 16-50-47 PST Sergei Izmailov I don't like false-positive warnings. We can limit it to ubuntu 18.04 only 2020-03-19 14-58-05 PDT Sergei Izmailov No problem, take your time. Just checking. 2020-03-04 07-32-57 PST Sergei Izmailov Hi! Can you give a comment on https://gitlab.com/inkscape/lib2geom/-/merge_requests/19 ? (if you prefer we can discuss it here or there) 2020-03-04 08-22-54 PST Sergei Izmailov I appreciate that this change is not urgent and can be delayed to post 1.0 (I remember you said that, probably it was in chat, can't find it in MRs). I think you are reasonable person and your suggestions has ho have something behind, but I don't see it clearly so far. I didn't ignore you questions (at least intentionally). Please point out/copy-paste. I didn't argue against any extra changes just because I'm lazy to address your comments. I want to find best solution of possible. You can read that I'm not 100% with my arguments and my comment leaves question open to discuss. Please don't silence conversations as it's not a productive way to solve problems (as well as over-communication). If you think that I'm wrong/lazy/not in right time, please tell so. Again, if it's just lack of time, lets hold it until 1.0 release. It's totally fine by me. 2020-03-15 04-40-18 PDT Sergei Izmailov Sorry, I misread that. 2020-03-19 15-00-34 PDT Sergei Izmailov (i guess) 2020-03-19 14-20-26 PDT Sergei Izmailov Sorry for bothering you again. Is there is any problems with https://gitlab.com/inkscape/inkscape/-/merge_requests/1410 except lack of dev time? 2020-05-07 13-54-24 PDT Sergei Izmailov I see. (To be honest I feel uneasy with windows as developer, especially when it comes as low level as like linking. I think I can utilize my time better fixing other stuff.) 2020-03-04 13-30-21 PST Sergei Izmailov Should I ask Martin to do it? 2020-05-07 13-15-55 PDT Sergei Izmailov I'm a unix person, so I don't know much about Windows stuff. All I found I wrote here. 2020-03-04 14-00-17 PST Sergei Izmailov (`googletest` soruces is dependency and will be installed automatically ) 2020-03-30 15-29-08 PDT Sergei Izmailov libinkscape_base.dll.a (https://chat.inkscape.org/file-upload/JHxjTscxfaTZuRcZ3/libinkscape_base.dl…) 2020-03-04 13-44-24 PST Sergei Izmailov I don't understand. All seems to be installed, but it's not found by cmake for some reason https://gitlab.com/inkscape/inkscape-ci-docker/-/jobs/455010902 2020-03-15 06-31-15 PDT Sergei Izmailov It's all case-specific. Some libraries act similar in both mods, some don't (like gtest). Explicit specification of MODULE/CONFIG prohibits use of another one. I think generally CONFIG is more preferable since it's mostly auto-generated and less error prone, but it requires some efforts from library authors so it's not widespread. At the moment I agree, let's keep it as is. 2020-03-07 16-49-06 PST Sergei Izmailov Should I add it to MR? 2020-03-19 15-55-24 PDT Sergei Izmailov ``` set(2GEOM_BUILD_SHARED ON CACHE BOOL "Build 2geom as SHARED library") add_subdirectory(2geom) ``` is enough to build and link against shared version of lib2geom. I've build with `-DCMAKE_INSTALL_PREFIX:PATH=~/.local/` so shared library is not found with default settings, therefore `~/.local/bin/inkscape` can't be launched without `LD_LIBRARY_PATH` or `LD_PRELOAD`. 2020-03-26 09-54-16 PDT Sergei Izmailov Does windows CI run only on master branch, but not on MRs? 2020-03-19 15-01-44 PDT Sergei Izmailov I can try to build inkscape against shared lib2geom 2020-03-30 15-31-14 PDT Sergei Izmailov I'm not really familiar with "import libraries", but from what I read it should be quite small compared to shared library 2020-03-04 13-22-13 PST Sergei Izmailov I'm new to gitlab CI 2020-03-07 16-48-51 PST Sergei Izmailov Such test would add ~30 lines to `.sh` script 2020-03-15 04-22-15 PDT Sergei Izmailov Hi! I've noticed you simplified GTest handling in main repo. The `find_package` call must include `MODULE` keyword since "config" mode takes precedence over "module" mode and we are using "module" features. It will fail on systems with properly installed GTest. 2020-03-04 15-07-21 PST Sergei Izmailov There is no binaries 2020-03-04 13-31-49 PST Sergei Izmailov right? 2020-03-26 11-11-48 PDT Sergei Izmailov ... and it fails [link](https://ci.appveyor.com/project/inkscape/inkscape/builds/31745274/job…. I don't quite get it. `lib/libinkscape_base.dll.a` has defined symbols from `lib/lib2geom.a`... 2020-03-26 13-22-09 PDT Sergei Izmailov I've dropped my previous commits to `testfiles/CMakeLists.txt` and added extra `target_link_libraries` for OBJECT library you created. It worked locally, let's see how would CI go 2020-03-04 13-17-37 PST Sergei Izmailov Seems like CI doesn't have google tests. Can you help with it ? 2020-03-04 13-59-12 PST Sergei Izmailov I don't think lib2geom uses mock. I have no guesses here 2020-03-19 15-03-38 PDT Sergei Izmailov Currently I think it's ready, it's matter of one line change. I'm going to check it right now. 2020-03-04 14-07-52 PST Sergei Izmailov Thanks for testing! 2020-03-07 16-53-53 PST Sergei Izmailov I think warning should be issued at the very end of script. Otherwise it's easy to overlook it in ton of messages 2020-03-19 16-04-16 PDT Sergei Izmailov Currently MR builds STATIC library (default setting in lib2geom). 2020-03-07 16-58-43 PST Sergei Izmailov I thought about it. It's better since it's never false-positive, but I don't like the idea of fixing it in CMakeLists... 2020-03-26 07-47-44 PDT Sergei Izmailov Hi, I think you still plan to give a review to MR, right? (Thomas seems to approved it) 2020-03-30 15-32-24 PDT Sergei Izmailov nm (https://chat.inkscape.org/file-upload/ty2TRYMN2pNSnCtCg/nm) 2020-03-07 16-50-59 PST Sergei Izmailov But managing all versions seems to be tedious 2020-03-30 15-34-21 PDT Sergei Izmailov `lib2geom.a` and `lininkscape_base.dll.a` provide same global symbol 2020-03-04 13-30-05 PST Sergei Izmailov How can i update the image? 2020-03-26 13-25-55 PDT Sergei Izmailov Damn, older cmake rejects it. Sounds like a task for tomorrow. 2020-03-04 13-21-36 PST Sergei Izmailov It's already there, but needs special flag.. I'm not sure how to proceed with it 2020-03-04 12-59-15 PST Sergei Izmailov I'm sorry. Sometimes it's hard to explain myself properly. (In process of re-explaining I think I solved the googletest issue) 1. You are right, the difference between 2.8 and 3.0 styles is irrelevant here. BTW here is a good talk about modern cmake style and why we should avoid "2.8-style" https://youtu.be/bsXLMQ6WgIk. 2. The problem with `googletest`/`libgtest-dev`/`libmock-dev` packages on Ubuntu: None of them installs following files (compare to RPM: http://rpmfind.net/linux/RPM/fedora/devel/rawhide/aarch64/g/gtest-devel-1.8…) ``` /usr/lib64/cmake/GTest/GTestConfig.cmake /usr/lib64/cmake/GTest/GTestConfigVersion.cmake /usr/lib64/cmake/GTest/GTestTargets-noconfig.cmake /usr/lib64/cmake/GTest/GTestTargets.cmake ``` (pkg-config files are also missing but we are not interested in them now) They contain information about build targets and their flags. Without them we need to repeat this information ourself. Fortunately cmake-shipped [FindGTest](https://cmake.org/cmake/help/latest/module/FindGTest.html) exists and does it for us. `find_package(GTest)` can find headers and libraries and resurrect all required flags (https://cmake.org/cmake/help/latest/module/FindGTest.html) and even define `IMPROTED` targets `GTest::GTest` and `GTest::Main`. The only issue it has it names googletest targets differently from googletest itself. But it's not a big deal, we can ignore generated cmake files (`GTestConfig.cmake`) on system where googletest installed properly. I guess some of `googletest`/`libgtest-dev`/`libmock-dev` are available on CI, right? 2020-03-04 14-07-31 PST Sergei Izmailov I would prefer to stick to simpler solution. I think I'll do tests on 18.04 on my own to see what exactly goes wrong with bare `find_package` 2020-03-04 15-45-24 PST Sergei Izmailov I've submitted a MR with extra build step for gtest 2020-05-07 07-08-52 PDT Sergei Izmailov Hi, congrats on 1.0 release! Can you find time to look at https://gitlab.com/inkscape/inkscape/-/merge_requests/1410 problem ? 2020-03-04 13-33-13 PST Sergei Izmailov I though about passing `--recommended` flag 2020-03-26 12-46-04 PDT Sergei Izmailov I've tried static/shared and ended up with no library at all. I guess your intention was to avoid recompiling of those two files... Let's see if it would work. Probably you initial approach would also work (i.e. with OBJECT library) with repeated `target_link_libraries` for OBJECT library and tests which uses it. 2020-03-04 15-07-48 PST Sergei Izmailov And it's the complaint from cmake at CI 2020-03-19 15-57-46 PDT Sergei Izmailov (I've added tests to lib2geom to build a super-project against it as installed and subdirectory library) 2020-03-07 16-51-50 PST Sergei Izmailov Currently I don't have any warnings. You suggest to add warning only to non-CI ubuntu 18.04 ? 2020-03-19 15-02-07 PDT Sergei Izmailov is it desired? 2020-05-15 08-02-22 PDT Sergei Izmailov Hi, do you know how windows build was fixed? Just curious. Don't see obvious changes to cmake files 2020-03-04 13-31-47 PST Sergei Izmailov Oh, I see the https://gitlab.com/inkscape/inkscape-ci-docker/-/blob/master/Dockerfile 2020-03-26 09-56-02 PDT Sergei Izmailov Nevermind, I see AppVeyor started the job, but it doesn't appear as build stage on gitlab. 2020-03-07 17-00-13 PST Sergei Izmailov I'll add warning to `.sh` script. Let's think about cmake warning a bit more and decide later 2020-03-19 15-56-27 PDT Sergei Izmailov I think it confirms that lib2geom can be packaged separately 2020-03-04 14-47-26 PST Sergei Izmailov Looks like `libgtest-dev` in 18.04 is really poor. Compare 18.04 and 19.04: https://packages.ubuntu.com/disco/amd64/libgtest-dev/filelist https://packages.ubuntu.com/bionic/amd64/libgtest-dev/filelist We are not alone with this problem: https://stackoverflow.com/questions/24295876/cmake-cannot-find-googletest-r… I think I'll add instructions to build and install googletest to docker image 2020-03-04 13-59-57 PST Sergei Izmailov I think `libgtest-dev` is enough 2020-03-04 13-58-20 PST Patrick (but should still have everything needed to build) 2020-05-07 12-44-26 PDT Patrick I will, but I can't promise you when. Did you make progress on the linking issue on Windows or do we still need to solve that? 2020-03-15 05-07-58 PDT Patrick It might even be a good idea to explicitly specify module mode everywhere, but I don't think we do that anywhere yet, so it seemed safer to just keep it as-is (especially as it should not matter if the module is properly found in module mode as expected) 2020-03-19 14-59-47 PDT Patrick Maybe one question ahead of time: Do we link lib2geom dynamically in your MR? Because that certainly would be a goal going forward. 2020-03-19 14-57-26 PDT Patrick Well, I can't tell you if there are any problems before I have reviewed it (which I plan to do, eventually, but yeah I'm backlogged). 2020-03-19 15-02-09 PDT Patrick If it's reasonable... We can obviously take this in steps as well. 2020-03-04 13-32-27 PST Patrick (which is called by the Dockerfile) 2020-03-04 07-53-52 PST Patrick You seem to be ignoring my comments (everything I said so far you've not replied to at all or argued against, giving me the impression that you're not really willing to actually change anything) so I'm not overly motivated to spend any more time on that. (In fact the very first thing you ignored was my suggestion to move this to post-1.0 when people will have more time to review properly and it will be less intrusive to make larger changes to the build system). 2020-03-19 15-03-35 PDT Patrick But if we can isolate the subproject enough, it might come "for free". 2020-03-07 17-01-39 PST Patrick OK... I just would not like to add a lot of stuff to `.sh` script for a workaround for one specific version of one specific OS either, so let's keep it minimal (especially considering it might not be seen by all people) 2020-03-07 16-52-56 PST Patrick Yes, if `CI` is set, build and install cmake, otherwise print a warning that recommends to do it. 2020-03-04 13-59-33 PST Patrick So we only need gtest? 2020-03-19 15-01-13 PDT Patrick I meant with your MR against Inkscape. 2020-03-04 13-34-16 PST Patrick CI has everything 2020-03-07 16-51-02 PST Patrick Yes, I think that's exactly what you were doing in your MR? 2020-03-15 04-38-37 PDT Patrick If the MODULE option is not specfied in the above signature, CMake first searches for the package using Module mode. Then, if the package is not found, it searches again using Config mode. 2020-03-26 12-23-14 PDT Patrick Maybe all you need to do is make cpp_test_static_library actually static? 2020-03-26 12-15-52 PDT Patrick Yep, only for tests, though. Do you link the test library differently from other internal libs? 2020-03-04 14-12-51 PST Patrick Great! Once we figure it out, we can try to simplify the logic in inkscape/inkscape as well 2020-03-26 10-18-06 PDT Patrick Yeah, GitLab is stupid that way... 2020-03-27 09-03-49 PDT Patrick .dll.a indicates a so-called "import library". I think with gcc we can link directly against the .dll though. (so this is probably where the problem comes from: linking directly against libinkscape_base.dll once and indirectly linking against it a second time via the .dll.a) 2020-03-07 16-50-05 PST Patrick nah, that seems like overkill... I'd suggest to simply print a warning if `CI` is unset (same conditions otherwise). 2020-03-04 15-28-38 PST Patrick OK, yeah, I guess we might only use headers for Inkscape. 2020-03-04 14-05-20 PST Patrick Ok, so `find_package` fails locally as well. However the logic in https://gitlab.com/inkscape/inkscape/-/blob/master/CMakeLists.txt#L177-197 and https://gitlab.com/inkscape/inkscape/-/blob/master/testfiles/CMakeLists.txt… still makes it work (I'm pretty sure we don't even need all of it) 2020-03-07 17-01-56 PST Patrick Not even all people might use the shell script after all... 2020-03-26 12-19-02 PDT Patrick (you might want to check if static build still works btw, too. It caused problems in the past) 2020-03-19 15-03-14 PDT Patrick (although it's probably not too helpful right now) 2020-03-15 05-05-11 PDT Patrick np, thanks for checking the code! 2020-03-26 12-17-41 PDT Patrick Ah, I think tests are linked statically against libinkscape-base with your changes? 2020-03-04 10-05-53 PST Patrick Lack of time is a factor: If I had time to spare I would try to make the changes I suggested in your branch myself to see if they're viable. However as I'm currently spending my time on release-relevant things, I can only make the suggestion and have to depend on you to try to implement them. I'm not silencing anything, but if the feedback from you is "I don't see the need for a change" there's nothing I can work with (my original suggestion still stands, but I can not force you to try and implement it, and I currently won't spend time in it myself.) 2020-05-15 09-01-11 PDT Patrick It's not fixed yet but uses a shared library of 2geom as workaround 2020-03-26 12-23-37 PDT Patrick (or idd that fail as well) 2020-03-04 13-55-12 PST Patrick I could check locally if that helps (I have a Ubuntu VM 18.04 set-up). 2020-03-19 15-00-56 PDT Patrick (it's insanity that we link everything into one huge library right now - it's even causing problems, e.g. we can't currently do debug builds on Windows as there are too many symbols in the library) 2020-03-19 15-02-53 PDT Patrick At some point the hope was to get lib2geom packaged separately from Inkscape. 2020-03-26 09-04-59 PDT Patrick I looked at it and from reading the diff it should be fine. The primary thing I wanted to check was whether Windows CI passes (as I did not have the time to build locally yet). Unfortunately Windows CI was broken a while due to a row of broken updates in glibmm and mingw-w64 itself. Luckily that's all sorted out now. 2020-03-15 04-38-24 PDT Patrick No, Module mode is the default: 2020-03-07 16-55-09 PST Patrick That's true... if we want to ensure people see it we should probably add the waring to 2geom's CMakeLists.txt (and show it once `find_package(GTest REQUIRED MODULE)` fails) 2020-03-04 10-14-18 PST Patrick Wrt to the last sugestion: I'm not sure what your comment with "old style cmake 2.8" is referring to (I did suggest no such thing). What I suggested was to investigate _why_ the Ubuntu googletest/gmock packages are not found by the CMake module (I think you already have a rough idea) and attempt to fix the detection of the package rather than to download and build it manually. 2020-03-04 13-29-39 PST Patrick Well, let me know if/how I can help. We might be missing "libgmock-dev" (not available on Ubuntu 18.04 as it seems) 2020-03-26 09-05-58 PDT Patrick Feel free to rebase your MR to trigger a new CI build, otherwise I can do so later. 2020-03-26 12-20-15 PDT Patrick Oh, and I don't like https://gitlab.com/inkscape/inkscape/-/commit/e31698fb - I've removed that for a reason ;-) 2020-03-04 13-31-03 PST Patrick Create an MR with your changes as usual. 2020-03-07 15-58-28 PST Patrick I'll have a look tomorrow. Can we give general advice for affected users? (Possibly we could even print a warning explaining what to do?) 2020-05-07 13-41-15 PDT Patrick Well, I'm a Windows person, but that usually doesn't rule out that I try to fix Unix stuff ;-) 2020-03-04 13-20-56 PST Patrick We can add what we need here: https://gitlab.com/inkscape/inkscape-ci-docker/-/blob/master/install_depend… 2020-05-07 13-43-15 PDT Patrick It's OK if you don't want to try to fix the Windows part further , but I'm not sure when I find the time to investigate it myself. 2020-03-04 13-34-30 PST Patrick (`--full`) 2020-03-19 15-01-22 PDT Patrick (the default behavior) 2020-03-04 15-06-54 PST Patrick The files you're missing seem to be in `googletest`: https://packages.ubuntu.com/bionic/amd64/googletest/filelist 2020-03-04 13-32-01 PST Patrick no, the file I linked. 2020-03-04 13-58-00 PST Patrick 18.04 is different. I think it misses something gmock-related.

1 0

[Webmaster] Rocket.Chat, 2 Users, 478 Messages, 20 Files, 671477 Minutes, in Direct Message Between: sizmailov & adam.belis
by no-reply＠chat.inkscape.org 29 Apr '21

29 Apr '21

2020-01-18 12-59-24 PST Sergei Izmailov Here are my results ``` Inkscape 1.1-dev (031d86a, 2020-01-16) 0.29879935100325383 s Inkscape 1.1-dev (4cfd59b, 2020-01-17) 0.35072101598780137 s Inkscape 1.0beta2 (43f5480, 2020-01-18) 0.3234521280101035 s ``` 2020-01-26 09-05-48 PST Sergei Izmailov (the cusp part) 2020-01-18 14-30-21 PST Sergei Izmailov i have skype and discord 2020-01-18 13-33-01 PST Sergei Izmailov what is wrong? 2020-01-19 08-25-42 PST Sergei Izmailov you have "def" in the middle of file 2020-01-18 13-00-29 PST Sergei Izmailov Numbers tend to differ by approx 0.1 second 2020-01-26 09-01-14 PST Sergei Izmailov I tested builds around my MR 2020-01-18 14-24-51 PST Sergei Izmailov hm 2020-01-18 13-53-22 PST Sergei Izmailov Why all files are 0 KB on your screenshot? 2020-01-26 11-40-24 PST Sergei Izmailov Too many moving parts are involved to wrap head around it. Maybe I'll get to this some time later 2020-01-18 14-27-31 PST Sergei Izmailov there is no such file 2020-01-18 14-04-53 PST Sergei Izmailov Paths are correct, right 2020-01-26 09-16-20 PST Sergei Izmailov My code may only alter the node type 2020-01-18 13-26-27 PST Sergei Izmailov That happens to everyone at first 2020-01-18 14-23-24 PST Sergei Izmailov i mean 2020-01-18 13-39-57 PST Sergei Izmailov wow, that strange 2020-03-02 02-22-38 PST Sergei Izmailov Cool. Sorry for direct messaging. Unfortunately my announce got little/no attention in devel chat. It works now but it is still too raw to advertise to use it in a daily basis and not instllable from store. 2020-01-18 13-05-02 PST Sergei Izmailov Can you start python interpreter in console? 2020-01-18 12-41-45 PST Sergei Izmailov Hi, I think I'm finished with the benchmark script 2020-01-18 13-14-50 PST Sergei Izmailov Did it work? 2020-01-26 09-39-45 PST Sergei Izmailov inkscape build is quite ~~slow~~ long 2020-01-18 13-52-02 PST Sergei Izmailov 32 should work on both 2020-01-18 14-04-55 PST Sergei Izmailov ? 2020-01-18 14-13-19 PST Sergei Izmailov add just two print lines to the script 2020-01-18 14-06-50 PST Sergei Izmailov it's not a bundle 2020-01-18 08-43-25 PST Sergei Izmailov Hi! 2020-01-18 14-34-59 PST Sergei Izmailov greenmaus#3201 2020-01-18 14-32-00 PST Sergei Izmailov Let me find the "add" button 2020-03-01 01-17-31 PST Sergei Izmailov The extension finds links to commits and adds direct link to corresponding AppImage artifact (if any) 2020-01-26 11-36-17 PST Sergei Izmailov you mean "sodipodi:nodetype"? 2020-01-18 08-48-49 PST Sergei Izmailov Correctness check is a separate task. No need to do all at once 2020-01-26 09-01-27 PST Sergei Izmailov The nodes are cusp, but shape is preserved 2020-01-18 13-22-02 PST Sergei Izmailov Should end with `python.exe` 2020-01-18 14-34-48 PST Sergei Izmailov not sure how many identities this one have 2020-01-18 13-13-27 PST Sergei Izmailov command in console should be `"C:\.....\WindowsApps\python3" run_benchmark.py` 2020-01-18 12-52-42 PST Sergei Izmailov type `cmd` - you will see the console 2020-01-18 13-25-23 PST Sergei Izmailov It creates file, but fails to start inkscape 2020-01-18 14-03-21 PST Sergei Izmailov Is it better now? 2020-01-18 12-46-01 PST Sergei Izmailov Cool 2020-01-18 14-13-45 PST Sergei Izmailov don't use tabs, spaces only 2020-01-19 08-23-17 PST Sergei Izmailov Seems to be more or less correct 2020-01-18 14-21-04 PST Sergei Izmailov ```python #!/usr/bin/python3 import subprocess import shutil, os from timeit import default_timer as timer def measure(inkscape_executable, input_svg, verbs): temp_svg = "temp.svg" shutil.copyfile(input_svg, temp_svg) cmd = ([inkscape_executable, f"--verb={';'.join(verbs)}", temp_svg]) # print(" ".join(cmd)) t1 = timer() subprocess.call(cmd, stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL) t2 = timer() os.remove(temp_svg) # comment out this line to check results of inkscape invocation return t2 - t1 def inkscape_version(inkscape_executable): output = subprocess.check_output([ inkscape_executable, "--version" ], stderr=subprocess.DEVNULL).decode('utf-8') for line in output.split('\n'): if "Inkscape" in line: return line return "Unknown" def measure_object_to_path(inkscape_executable, input_svg): action_verbs = ['EditSelectAll', 'ObjectToPath', 'FileSave', 'FileQuit'] base_line_verbs = ['EditSelectAll', 'FileSave', 'FileQuit'] # cold run to minimize measurement errors measure(inkscape_executable, input_svg, base_line_verbs) # actual measurements baseline = measure(inkscape_executable, input_svg, base_line_verbs) payload = measure(inkscape_executable, input_svg, action_verbs) return payload - baseline if __name__ == "__main__": inkscape_executables = [ "./Inkscape-031d86a-x86_64.AppImage", "./Inkscape-4cfd59b-x86_64.AppImage", "./Inkscape-43f5480-x86_64.AppImage" ] test_file = "100-spirals.svg" with open("results.txt", "w") as results: for exe in inkscape_executables: print(os.path.abspath(os.curdir)) print(exe) version = inkscape_version(exe) elapsed_time = measure_object_to_path(exe, test_file) print(f"{version} {elapsed_time} s", file=results) ``` 2020-01-18 13-35-44 PST Sergei Izmailov so it did work? 2020-01-18 13-51-47 PST Sergei Izmailov probably yes 2020-01-18 13-48-06 PST Sergei Izmailov What a windows mistery 2020-01-18 12-53-36 PST Sergei Izmailov type `cd "C:\path\to\bench"` 2020-01-18 13-38-17 PST Sergei Izmailov [ ](https://chat.inkscape.org/direct/adam.belis?msg=iq8ix8DGyFYz9hC6M) is path same as here? () 2020-01-19 08-25-34 PST Sergei Izmailov oh 2020-01-18 13-17-27 PST Sergei Izmailov Did you type it right? 2020-01-18 14-33-19 PST Sergei Izmailov it says there are many of you 2020-01-18 12-45-30 PST Sergei Izmailov within `if __name__==` clause 2020-01-18 15-24-59 PST Sergei Izmailov Ok, thanks 2020-01-18 13-11-30 PST Sergei Izmailov find it's absolute path 2020-01-18 14-28-54 PST Sergei Izmailov show your script again 2020-01-18 12-58-16 PST Sergei Izmailov I'm not using windows on daily basis. Just read that you can click on script to run it 2020-01-26 09-37-55 PST Sergei Izmailov Let's see if it would help 2020-01-18 13-39-22 PST Sergei Izmailov now you are using powershell 2020-01-18 08-58-09 PST Sergei Izmailov Thanks! 2020-01-18 12-53-42 PST Sergei Izmailov with proper path to benchmark 2020-01-18 13-00-34 PST Sergei Izmailov sure 2020-01-18 12-43-46 PST Sergei Izmailov bench-inkscape.zip (https://chat.inkscape.org/file-upload/bc7aWzAoYnnreSkvr/bench-inkscape.zip) 2020-01-18 14-12-52 PST Sergei Izmailov what does it print? 2020-01-18 13-21-11 PST Sergei Izmailov The path on last screenshot doesn't match to what you've shown earlier with explorer 2020-01-18 13-20-15 PST Sergei Izmailov Can't you run pyton3.8? 2020-01-18 15-40-48 PST Sergei Izmailov I remember a friend of mine running inkscape on a quite decent laptop and it was extreemly slow 2020-01-18 13-22-04 PST Sergei Izmailov right 2020-01-19 08-23-21 PST Sergei Izmailov let me check 2020-01-26 09-14-22 PST Sergei Izmailov Actually not really. It seems like the doc's patch is responsible for that 2020-01-18 14-21-14 PST Sergei Izmailov should look like this 2020-01-18 14-23-21 PST Sergei Izmailov to console 2020-01-19 08-26-42 PST Sergei Izmailov compare to original file 2020-01-18 13-09-32 PST Sergei Izmailov 2. you can verify it by `dir` command 2020-01-18 14-09-50 PST Sergei Izmailov add two prints just to see where we are and what is `exe` 2020-01-26 09-04-48 PST Sergei Izmailov Nodes are cusp even for single circle 2020-01-18 14-26-56 PST Sergei Izmailov no 2020-01-18 13-47-59 PST Sergei Izmailov You didn't change single file in the directory 2020-01-19 08-25-25 PST Sergei Izmailov did work for me 2020-01-26 11-36-48 PST Sergei Izmailov this error seems to be with arc->path conversion 2020-01-26 11-35-45 PST Sergei Izmailov some change just triggered it 2020-01-18 08-52-27 PST Sergei Izmailov Do you know python? 2020-01-18 14-17-38 PST Sergei Izmailov same button as ~ 2020-01-18 13-25-01 PST Sergei Izmailov why not? 2020-01-26 09-06-10 PST Sergei Izmailov Another part is that this conversion simply done wrong 2020-01-18 12-45-15 PST Sergei Izmailov The important part is at the bottom 2020-01-18 12-56-33 PST Sergei Izmailov paths to inkscape executables in `run_benchmarks.py` could be relative or absolute 2020-01-18 14-32-34 PST Sergei Izmailov It says "Add Server" 2020-01-18 08-57-45 PST Sergei Izmailov I saw artifacts for linux and mac, but not for windows. 2020-01-18 14-17-48 PST Sergei Izmailov but without shift 2020-01-18 12-47-19 PST Sergei Izmailov At the end it prints time difference between `['EditSelectAll', 'ObjectToPath', 'FileSave', 'FileQuit']` and `['EditSelectAll', 'FileSave', 'FileQuit']` 2020-01-18 08-46-09 PST Sergei Izmailov By the way is there any performance testing in CI? I doubt they exist as it would require stable environment (same machine, number of free cores, etc) 2020-01-18 14-14-45 PST Sergei Izmailov file is saved? 2020-01-18 14-15-03 PST Sergei Izmailov can't be 2020-01-19 08-35-30 PST Sergei Izmailov you are welcom 2020-01-18 14-18-42 PST Sergei Izmailov remove first one 2020-01-26 08-49-18 PST Sergei Izmailov yes, by doctormo 2020-01-18 13-34-08 PST Sergei Izmailov Should not matter 2020-01-18 08-44-17 PST Sergei Izmailov Can you tell how you measure performance? 2020-01-18 13-58-02 PST Sergei Izmailov second 2020-01-18 14-27-55 PST Sergei Izmailov it cannot find `C:\....\inkscape.exe` 2020-01-26 09-38-36 PST Sergei Izmailov :) 2020-01-26 09-38-30 PST Sergei Izmailov On success it would be one char MR 2020-01-18 13-38-58 PST Sergei Izmailov right. 2020-01-26 09-02-01 PST Sergei Izmailov The latest master build is broken 2020-01-18 12-43-09 PST Sergei Izmailov Damn. Cannot attach .py here 2020-03-01 01-16-03 PST Sergei Izmailov Hi, can you tell which browser do you use? Would something like https://chat.inkscape.org/channel/team_devel?msg=M7M3ci4dzzyhYn8kP be helpful for you (except it would be appveyor and windows)? () 2020-01-18 14-17-09 PST Sergei Izmailov ``` ``` 2020-01-26 08-52-21 PST Sergei Izmailov https://gitlab.com/inkscape/inkscape/merge_requests/1337 2020-01-18 08-56-07 PST Sergei Izmailov One more question. How do you obtain inkscape builds? 2020-01-18 13-43-33 PST Sergei Izmailov so you cannot start python from start menu either? 2020-01-18 14-25-03 PST Sergei Izmailov you covered interseting part with ptyhon screen 2020-01-18 14-27-16 PST Sergei Izmailov it 's usually what it says it is 2020-01-18 14-21-53 PST Sergei Izmailov did it? 2020-01-26 09-41-56 PST Sergei Izmailov To become a computational physicist I had to develop some sys admin skills. not a big trouble at my side 2020-01-18 14-05-39 PST Sergei Izmailov did you save the file? 2020-01-26 09-39-37 PST Sergei Izmailov I'll be back in hours 2020-01-18 08-51-20 PST Sergei Izmailov Good idea with number of spirals by the way 2020-01-18 12-44-00 PST Sergei Izmailov Indeed. It wouldn't trick gmail thogh 2020-01-18 13-23-10 PST Sergei Izmailov Still no luck? 2020-01-18 08-53-59 PST Sergei Izmailov I have to go right now, I'll be back in a couple of hours (hopefully with benchmark script) 2020-01-18 12-51-33 PST Sergei Izmailov I pasted whole file to ease patching :) 2020-01-18 13-50-50 PST Sergei Izmailov https://www.python.org/ftp/python/3.8.1/ 2020-01-18 12-48-12 PST Sergei Izmailov say `results.txt` 2020-01-26 08-52-45 PST Sergei Izmailov is directly related to unions 2020-01-18 14-08-10 PST Sergei Izmailov can work i guess, but it's so dirty 2020-01-26 11-38-52 PST Sergei Izmailov What I meant is that converted arcs should have smooth nodetype after convertion no matter what 2020-01-18 12-49-56 PST Sergei Izmailov ```python #!/usr/bin/python3 import subprocess import shutil, os from timeit import default_timer as timer def measure(inkscape_executable, input_svg, verbs): temp_svg = "temp.svg" shutil.copyfile(input_svg, temp_svg) cmd = ([inkscape_executable, f"--verb={';'.join(verbs)}", temp_svg]) # print(" ".join(cmd)) t1 = timer() subprocess.call(cmd, stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL) t2 = timer() os.remove(temp_svg) # comment out this line to check results of inkscape invocation return t2 - t1 def inkscape_version(inkscape_executable): output = subprocess.check_output([ inkscape_executable, "--version" ], stderr=subprocess.DEVNULL).decode('utf-8') for line in output.split('\n'): if "Inkscape" in line: return line return "Unknown" def measure_object_to_path(inkscape_executable, input_svg): action_verbs = ['EditSelectAll', 'ObjectToPath', 'FileSave', 'FileQuit'] base_line_verbs = ['EditSelectAll', 'FileSave', 'FileQuit'] # cold run to minimize measurement errors measure(inkscape_executable, input_svg, base_line_verbs) # actual measurements baseline = measure(inkscape_executable, input_svg, base_line_verbs) payload = measure(inkscape_executable, input_svg, action_verbs) return payload - baseline if __name__ == "__main__": inkscape_executables = [ "./Inkscape-031d86a-x86_64.AppImage", "./Inkscape-4cfd59b-x86_64.AppImage", "./Inkscape-43f5480-x86_64.AppImage" ] test_file = "100-spirals.svg" with open("results.txt", "w") as results: for exe in inkscape_executables: version = inkscape_version(exe) elapsed_time = measure_object_to_path(exe, test_file) print(f"{version} {elapsed_time} s", file=results) ``` 2020-01-18 12-52-11 PST Sergei Izmailov Back to question "how to run" 2020-01-18 13-49-25 PST Sergei Izmailov https://stackoverflow.com/questions/44394965/cant-run-python-on-windows-any… 2020-01-18 14-08-14 PST Sergei Izmailov :) 2020-01-18 13-33-04 PST Sergei Izmailov didn't get it 2020-01-26 09-36-47 PST Sergei Izmailov Let me test one idea, my be it will fix that 2020-01-18 13-37-56 PST Sergei Izmailov but you were able to start python, what's changed? 2020-03-01 01-17-55 PST Sergei Izmailov Fairly straightforward, but took me a while to figure out how to deal with it. 2020-01-18 08-50-07 PST Sergei Izmailov something like this 2020-01-18 14-33-06 PST Sergei Izmailov found 2020-01-18 13-44-15 PST Sergei Izmailov hell, i don't understand what's going on 2020-01-18 13-39-29 PST Sergei Izmailov it used to be plain cmd 2020-01-18 13-24-26 PST Sergei Izmailov is path to exe correct? 2020-01-18 08-58-16 PST Sergei Izmailov Yeah, found 2020-01-18 15-42-14 PST Sergei Izmailov 10x is quite a lot, meaning there is whole room of optimizations. Maybe not in inkscape itself... 2020-01-26 11-34-32 PST Sergei Izmailov I'm not sure what exactly is going on there 2020-01-18 13-05-44 PST Sergei Izmailov to folder with script and 100-spirals 2020-01-18 14-29-02 PST Sergei Izmailov can you share screen? 2020-01-18 13-09-38 PST Sergei Izmailov it should show .py and .svg 2020-01-18 12-44-28 PST Sergei Izmailov Zip has two files: 100-spirals.svg is ... 100 spirals 2020-01-18 14-11-32 PST Sergei Izmailov yep 2020-01-18 08-49-19 PST Sergei Izmailov I think actions/verbs can give better time resolution and it can be automated 2020-01-18 12-47-55 PST Sergei Izmailov It prints results in console 2020-01-26 09-37-44 PST Sergei Izmailov Going to change ```cpp int const num_sectors = abs(sang - eang) * 2 / M_PI + 1; ``` to ```cpp int const num_sectors = abs(sang - eang) * 4 / M_PI + 1; ``` 2020-01-18 14-34-28 PST Sergei Izmailov i'm `greenmaus` 2020-01-19 08-26-33 PST Sergei Izmailov `with ` and what follows should be 4 spaces right 2020-01-26 11-35-27 PST Sergei Izmailov I think the problem might be somewhere else 2020-01-18 14-01-28 PST Sergei Izmailov forget what I said about escaping 2020-01-18 12-54-03 PST Sergei Izmailov Note quotes, they are most likely necessary 2020-01-18 14-20-50 PST Sergei Izmailov right, you are not familar with programming 2020-01-18 14-05-55 PST Sergei Izmailov I think first one 2020-01-18 13-34-52 PST Sergei Izmailov You can start with one inkscape to test it's wokring 2020-01-18 13-10-36 PST Sergei Izmailov Can you find pytohn in "start" menu 2020-01-18 08-53-13 PST Sergei Izmailov No problem :) 2020-01-18 13-09-58 PST Sergei Izmailov Let's find out where your python executable is 2020-01-18 12-54-36 PST Sergei Izmailov now you should be able to call the script by `python3 run_benchmarks.py` 2020-01-18 14-06-11 PST Sergei Izmailov copy `inkscape.exe` to folder with script 2020-01-18 13-41-48 PST Sergei Izmailov There must be some starter python pack for windows to make it easy to start, just as in linux:) 2020-01-18 08-50-03 PST Sergei Izmailov `time inkscape --actions='select-all;path-to-objects;exit' sample.svg` 2020-01-18 13-01-54 PST Sergei Izmailov results.txt should appear in same directory 2020-01-26 09-02-24 PST Sergei Izmailov It might be possible to bisect 2020-01-18 14-01-51 PST Sergei Izmailov Just add `r` before the strings 2020-01-18 14-13-35 PST Sergei Izmailov the surrounding should give idea where they should be 2020-01-18 14-16-15 PST Sergei Izmailov tripple back quotes 2020-01-26 08-52-56 PST Sergei Izmailov and is also recent change 2020-01-18 13-21-30 PST Sergei Izmailov last part is missing 2020-01-19 08-33-07 PST Sergei Izmailov sometimes i just can't read what you type :) 2020-01-18 13-02-06 PST Sergei Izmailov and inkscape should show and close for a moment 2020-01-18 12-51-46 PST Sergei Izmailov No, you need to specify bundles for win 2020-01-26 11-36-30 PST Sergei Izmailov i don't think so 2020-01-18 13-54-04 PST Sergei Izmailov [ ](https://chat.inkscape.org/direct/adam.belis?msg=okGR2MdeE2s4h8oaP) this one () 2020-01-18 12-45-00 PST Sergei Izmailov Can you read through it? 2020-01-18 12-46-39 PST Sergei Izmailov What it does: It runs inkscape and instructs it with "verb"s 2020-01-18 12-52-36 PST Sergei Izmailov Hit `win+R` - a prompt should appear 2020-01-18 14-06-41 PST Sergei Izmailov No, it would not work 2020-01-18 14-14-19 PST Sergei Izmailov why it didn't print anything? 2020-01-18 14-17-26 PST Sergei Izmailov it's above the "TAB" 2020-01-18 13-25-47 PST Sergei Izmailov of course, those are linux bundles 2020-01-19 08-25-59 PST Sergei Izmailov You broke indentation 2020-01-18 14-30-08 PST Sergei Izmailov I can start video call, but 's not what we want 2020-01-18 12-55-58 PST Sergei Izmailov Not necessary, but could be 2020-01-18 13-47-45 PST Sergei Izmailov How it broke? 2020-01-18 13-59-37 PST Sergei Izmailov I think you can just replace bacslashes `\` with normal ones `/` 2020-01-26 09-16-01 PST Sergei Izmailov For sure 2020-01-18 14-34-00 PST Sergei Izmailov Clipboard - January 19, 2020 1:33 AM (https://chat.inkscape.org/file-upload/rTYgAHXy4wLHRdBup/Clipboard%20-%20Jan…) 2020-01-18 14-09-32 PST Sergei Izmailov ```python for exe in inkscape_executables: print(os.path.abspath(os.curdir)) print(exe) version = inkscape_version(exe) elapsed_time = measure_object_to_path(exe, test_file) print(f"{version} {elapsed_time} s", file=results) ``` 2020-01-18 12-48-06 PST Sergei Izmailov I can change it to file 2020-01-18 14-02-07 PST Sergei Izmailov so it would be `r"C:\Users...` 2020-01-18 14-29-24 PST Sergei Izmailov I never tried 2020-01-18 08-45-13 PST Sergei Izmailov Maybe I can help to create the benchmark 2020-01-18 13-08-43 PST Sergei Izmailov 1. Can you start interpreter? 2. Can you change to desired directory? 2020-01-19 08-27-34 PST Sergei Izmailov It might be not a bad idea to learn python. It's everywhere nowadays 2020-01-18 13-19-51 PST Sergei Izmailov May be I'm trying to do it "linux way" 2020-01-18 14-18-39 PST Sergei Izmailov you added extra for loop 2020-01-26 11-34-11 PST Sergei Izmailov nyah, didn't work 2020-01-26 11-35-02 PST Sergei Izmailov number of segments doesn't affect number of nodes 2020-01-18 12-44-36 PST Sergei Izmailov and run_benchmark.py 2020-01-18 08-54-13 PST Sergei Izmailov Thanks for investigating it! 2020-01-18 08-50-57 PST Sergei Izmailov I can show you and/or do it mysefl 2020-01-18 12-55-18 PST Sergei Izmailov or whatever python executable name is 2020-01-18 13-40-24 PST Sergei Izmailov i had my machine running years without reboots) 2020-01-26 11-42-26 PST Sergei Izmailov ok 2020-01-18 08-52-05 PST Sergei Izmailov Do you use linux/windows. How comfortable are you with console? 2020-01-18 14-00-29 PST Sergei Izmailov You can use backslash, but you need to "escape" it 2020-01-18 13-21-23 PST Sergei Izmailov Oh 2020-01-26 08-50-50 PST Sergei Izmailov You mean the path is distorted, right? 2020-01-18 13-51-41 PST Sergei Izmailov is your pc 64 bit? 2020-01-26 09-05-36 PST Sergei Izmailov The problem is that arc->cubic bezier is done with insufficient precision 2020-01-18 12-48-26 PST Sergei Izmailov do you have python3? 2020-01-18 14-00-37 PST Sergei Izmailov `\` -> `\\` 2020-01-18 13-52-09 PST Sergei Izmailov 64 is preferable 2020-01-18 14-17-58 PST Sergei Izmailov yay 2020-01-18 14-21-44 PST Sergei Izmailov yes, but it should print someting in front 2020-01-18 13-25-34 PST Sergei Izmailov so the file is empty 2020-01-19 10-01-04 PST Sergei Izmailov All three are same within error, right? 2020-01-26 09-02-11 PST Sergei Izmailov It distorts all the way 2020-01-18 14-28-15 PST Adam Belis i will try to chage slashes 2020-01-18 13-08-14 PST Adam Belis ok this stil did not work 2020-01-18 08-50-00 PST Adam Belis what do you mean under actions and verbs 2020-01-19 08-16-25 PST Adam Belis i addet this from datetime import datetime and than i tried to add it to the file name but it does not work but i can print it in txt no problem 2020-01-26 11-36-51 PST Adam Belis he saied the problem was always there 2020-01-18 14-13-04 PST Adam Belis Clipboard - January 18, 2020 11:13 PM (https://chat.inkscape.org/file-upload/6gqWfMtZMCoaJ24u6/Clipboard%20-%20Jan…) 2020-01-18 08-48-40 PST Adam Belis and i just did scren recoriding and than checked it in cuting softver how long it took 2020-01-18 14-13-57 PST Adam Belis yes that was my istake 2020-01-18 13-43-43 PST Adam Belis nope 2020-01-18 14-27-31 PST Adam Belis i dont understnad what it cannot find 2020-01-18 14-13-51 PST Adam Belis Clipboard - January 18, 2020 11:13 PM (https://chat.inkscape.org/file-upload/Rn9BAHyTTXgQaSAD7/Clipboard%20-%20Jan…) 2020-01-18 14-32-10 PST Adam Belis green 2020-01-18 12-51-43 PST Adam Belis thenks i replaced it 2020-01-18 13-06-15 PST Adam Belis aaaaaaaaaaaah ima idiot 2020-01-18 14-07-25 PST Adam Belis so i should do it other way around 2020-01-18 13-44-02 PST Adam Belis v 2020-01-19 08-23-46 PST Adam Belis it dont want to run if i do this 2020-01-18 13-48-18 PST Adam Belis its jusa say i have wrong verion of pyton for my pc 2020-01-18 13-36-51 PST Adam Belis Clipboard - January 18, 2020 10:36 PM (https://chat.inkscape.org/file-upload/fBiBvsmn9DwR2T8kL/Clipboard%20-%20Jan…) 2020-01-18 13-57-46 PST Adam Belis aaand we have a new error 2020-01-18 08-50-58 PST Adam Belis but agree that would be better test BUT there is also possibility that maybe ther is som bechmarking already in ink but i dont know about it 2020-01-18 13-54-56 PST Adam Belis sohow 2020-01-18 14-15-57 PST Adam Belis sorry stil didn learn marking 2020-01-18 14-04-10 PST Adam Belis as i send you 2020-01-18 14-34-10 PST Adam Belis :D 2020-01-26 09-00-05 PST Adam Belis yop its broken in all builds 2020-01-26 09-40-37 PST Adam Belis :D 2020-01-18 14-30-10 PST Adam Belis zoom 2020-01-18 13-43-39 PST Adam Belis actualty for sure most of ppl say its easy 2020-01-26 09-37-05 PST Adam Belis go ahead :D 2020-01-19 10-01-55 PST Adam Belis so i think its ok to try to merg this to master 2020-01-18 14-05-42 PST Adam Belis yes 2020-01-18 13-39-50 PST Adam Belis it gives mi that blu error 2020-01-18 14-11-23 PST Adam Belis prints ? 2020-01-26 08-52-13 PST Adam Belis yop 2020-01-26 11-34-43 PST Adam Belis hmmm thats too bad 2020-01-18 15-39-59 PST Adam Belis also i should try to side load linux to text performece on inux maybe my pc is just busted :D 2020-01-18 14-29-43 PST Adam Belis i dont think so 2020-01-18 08-56-40 PST Adam Belis there are outo builds here 2020-01-18 14-23-28 PST Adam Belis oh 2020-01-18 08-52-52 PST Adam Belis but i can kind a read it 2020-01-18 12-48-17 PST Adam Belis ok noob question how do i run it ? :D 2020-01-18 13-24-52 PST Adam Belis yes i think i run but i dont tnik it could open verisons of ink 2020-01-18 13-43-47 PST Adam Belis not even form that place 2020-01-18 13-10-06 PST Adam Belis Clipboard - January 18, 2020 10:10 PM (https://chat.inkscape.org/file-upload/DEvmdTeLXjzxRpyLm/Clipboard%20-%20Jan…) 2020-01-18 13-24-12 PST Adam Belis we are moving but stil go empty exe 2020-01-26 09-04-58 PST Adam Belis yes thats also new problem :D 2020-01-18 13-33-59 PST Adam Belis i think 2020-01-18 14-22-51 PST Adam Belis nope it did not 2020-01-18 14-14-27 PST Adam Belis dont know 2020-01-18 13-05-31 PST Adam Belis d "C:\path\to\bench" is this path sous to be to one sesific build ? 2020-01-18 13-37-11 PST Adam Belis wel that time when i had wrong paths for ink 2020-01-26 09-42-34 PST Adam Belis oh yes . k now its not that hard i am just dont care about that that much 2020-01-18 08-58-38 PST Adam Belis np 2020-01-26 11-35-04 PST Adam Belis i can just test 2020-01-18 13-39-54 PST Adam Belis one sec 2020-01-18 13-21-41 PST Adam Belis last part of what ? 2020-01-26 08-51-23 PST Adam Belis wait 2020-01-18 12-55-39 PST Adam Belis ok and py should be in that folder where is inkscape right ? 2020-01-18 13-33-48 PST Adam Belis hmmm maybe coz i added 0.94 2020-03-02 02-23-43 PST Adam Belis dont worry about it i am glad that you wroth me ... but in last few days my computer is broken and im woring on fixing on it 2020-01-18 08-48-06 PST Adam Belis i dont know if this does even test what it sous to do 2020-01-18 13-44-01 PST Adam Belis i am ging to dowlad diffrent 2020-01-18 14-04-16 PST Adam Belis it creates txt 2020-01-18 13-39-39 PST Adam Belis iam goign to restar pc 2020-01-18 08-52-31 PST Adam Belis not that much but i used it few times 2020-01-18 13-52-07 PST Adam Belis Clipboard - January 18, 2020 10:52 PM (https://chat.inkscape.org/file-upload/vbSgDpdaXc8krCSqm/Clipboard%20-%20Jan…) 2020-01-18 14-15-24 PST Adam Belis ```#!/usr/bin/python3 import subprocess import shutil, os from timeit import default_timer as timer def measure(inkscape_executable, input_svg, verbs): temp_svg = "temp.svg" shutil.copyfile(input_svg, temp_svg) cmd = ([inkscape_executable, f"--verb={';'.join(verbs)}", temp_svg]) # print(" ".join(cmd)) t1 = timer() subprocess.call(cmd, stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL) t2 = timer() os.remove(temp_svg) # comment out this line to check results of inkscape invocation return t2 - t1 def inkscape_version(inkscape_executable): output = subprocess.check_output([ inkscape_executable, "--version" ], stderr=subprocess.DEVNULL).decode('utf-8') for line in output.split('\n'): if "Inkscape" in line: return line return "Unknown" def measure_object_to_path(inkscape_executable, input_svg): action_verbs = ['EditSelectAll', 'ObjectToPath', 'FileSave', 'FileQuit'] base_line_verbs = ['EditSelectAll', 'FileSave', 'FileQuit'] # cold run to minimize measurement errors measure(inkscape_executable, input_svg, base_line_verbs) # actual measurements baseline = measure(inkscape_executable, input_svg, base_line_verbs) payload = measure(inkscape_executable, input_svg, action_verbs) return payload - baseline if __name__ == "__main__": inkscape_executables = [ r"C:\Users\Adam\Downloads\inkscape-1.1-dev_2020-01-17_abe53df-x64\inkscape\bin\inkcape.exe", # "C:/Users/Adam/Downloads/inkscape-1.1-dev_2020-01-17_abe53df-x64 - 1/inkscape/bin/inkcape.exe", #r"C:\Users\Adam\Downloads\inkscape-0.92.4_2020-01-12_3ab3b9f-x64\inkscape\ink…" ] test_file = "100-spirals.svg" with open("results.txt", "w") as results: for exe in inkscape_executables: version = inkscape_version(exe) elapsed_time = measure_object_to_path(exe, test_file) print(f"{version} {elapsed_time} s", file=results) for exe in inkscape_executables: print(os.path.abspath(os.curdir)) print(exe) version = inkscape_version(exe) elapsed_time = measure_object_to_path(exe, test_file) print(f"{version} {elapsed_time} s", file=results) 2020-01-18 14-20-33 PST Adam Belis oh 2020-01-19 08-15-29 PST Adam Belis can i annoy you lite bit ? i am tring to add curretn date adn time to name of the file (so it does not get overwrite every time you run it ) but cannot find out correct way to add it ``` #!/usr/bin/python3 import subprocess import shutil, os from datetime import datetime from timeit import default_timer as timer def measure(inkscape_executable, input_svg, verbs): temp_svg = "temp.svg" shutil.copyfile(input_svg, temp_svg) cmd = ([inkscape_executable, f"--verb={';'.join(verbs)}", temp_svg]) # print(" ".join(cmd)) t1 = timer() subprocess.call(cmd, stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL) t2 = timer() os.remove(temp_svg) # comment out this line to check results of inkscape invocation return t2 - t1 def inkscape_version(inkscape_executable): output = subprocess.check_output([ inkscape_executable, "--version" ], stderr=subprocess.DEVNULL).decode('utf-8') for line in output.split('\n'): if "Inkscape" in line: return line return "Unknown" def measure_object_to_path(inkscape_executable, input_svg): action_verbs = ['EditSelectAll', 'ObjectToPath', 'FileSave', 'FileQuit'] base_line_verbs = ['EditSelectAll', 'FileSave', 'FileQuit'] # cold run to minimize measurement errors measure(inkscape_executable, input_svg, base_line_verbs) # actual measurements baseline = measure(inkscape_executable, input_svg, base_line_verbs) payload = measure(inkscape_executable, input_svg, action_verbs) return payload - baseline def if __name__ == "__main__": # Add paths for inscape builds you want to test inkscape_executables = [ r"D:\Downloads\inkscape-1.1-dev_2020-01-17_abe53df-x64\bin\inkscape.exe", r"D:\Downloads\inkscape-1.0beta2_2020-01-18_43f5480-x64\bin\inkscape.exe", #"./Inkscape-43f5480-x86_64.AppImage" ] test_file = "100-spirals.svg" with open("results" + (datetime.now().strftime('%Y-%m-%d %H:%M:%S')) + ".txt", "w") as results: for exe in inkscape_executables: print(os.path.abspath(os.curdir)) print(exe) print(datetime.now().strftime('%Y-%m-%d %H:%M:%S')) version = inkscape_version(exe) elapsed_time = measure_object_to_path(exe, test_file) current_time = (datetime.now().strftime('%Y-%m-%d %H:%M:%S')) print(f"{version} {elapsed_time} s ", file=results) print(f"{current_time} ", file=results) ``` 2020-01-18 13-19-23 PST Adam Belis i feeel like an idiot 2020-01-18 13-20-52 PST Adam Belis Clipboard - January 18, 2020 10:20 PM (https://chat.inkscape.org/file-upload/iwtdmyxappWmx9gCw/Clipboard%20-%20Jan…) 2020-01-18 13-38-34 PST Adam Belis yes 2020-01-18 13-19-12 PST Adam Belis why is this so complicated :D 2020-01-18 13-46-52 PST Adam Belis NOPE after reinstal of pyton ssam shit 2020-01-19 10-01-35 PST Adam Belis i round this few times its all same 2020-01-26 09-16-20 PST Adam Belis ok thanks ! 2020-01-18 14-33-33 PST Adam Belis how is that possible there is just one adam belis 2020-03-02 02-00-05 PST Adam Belis yes that would be awesome 2020-01-26 09-38-24 PST Adam Belis once again this tells me almost nothing but lets test it 2020-01-18 13-39-44 PST Adam Belis coz i cannot even run pyton now 2020-01-18 13-25-33 PST Adam Belis if __name__ == "__main__": inkscape_executables = [ "./Inkscape-031d86a-x86_64.AppImage", "./Inkscape-4cfd59b-x86_64.AppImage", "./Inkscape-43f5480-x86_64.AppImage" ] 2020-01-18 14-27-16 PST Adam Belis ok i have idea 2020-01-18 13-04-46 PST Adam Belis i am missing something 2020-01-18 13-20-54 PST Adam Belis this 2020-01-18 14-31-02 PST Adam Belis i think 2020-01-18 14-15-08 PST Adam Belis :D it is 2020-01-19 08-33-58 PST Adam Belis :D sorry i know 2020-01-18 13-21-49 PST Adam Belis i should also include exe 2020-01-26 08-50-04 PST Adam Belis hmm thas not a good news boolian operations are quit bad right now 2020-01-18 14-04-06 PST Adam Belis i got same error 2020-01-18 13-09-04 PST Adam Belis 2. yes 2020-01-18 13-00-52 PST Adam Belis how do i know if it did run ? 2020-01-18 13-12-29 PST Adam Belis Clipboard - January 18, 2020 10:12 PM (https://chat.inkscape.org/file-upload/wrBxREw2TBnWDXDNF/Clipboard%20-%20Jan…) 2020-01-18 13-09-00 PST Adam Belis 1. dont know what you are asking me 2020-01-18 13-43-23 PST Adam Belis maybe 2020-01-18 08-56-33 PST Adam Belis https://ci.appveyor.com/project/inkscape/inkscape/history 2020-01-18 14-27-57 PST Adam Belis in that case that path is not correct 2020-01-26 09-38-55 PST Adam Belis thats what i would call easy fix 2020-01-18 13-47-20 PST Adam Belis maaan good old bat files :D 2020-01-18 14-25-57 PST Adam Belis contiue 2020-01-18 13-59-07 PST Adam Belis Clipboard - January 18, 2020 10:59 PM (https://chat.inkscape.org/file-upload/NtQpDPbe9oGhwgwxL/Clipboard%20-%20Jan…) 2020-01-18 13-12-28 PST Adam Belis this 2020-01-18 08-47-51 PST Adam Belis my idea was to this svg select all spirals and convert it to paths 2020-01-18 14-04-18 PST Adam Belis ubt 2020-01-18 14-17-07 PST Adam Belis Nooo :D 2020-01-18 12-45-55 PST Adam Belis yets its readable 2020-01-18 15-42-42 PST Adam Belis i dont tnik ther is enough love wor win verison :D 2020-01-18 08-52-08 PST Adam Belis win 2020-01-26 08-52-57 PST Adam Belis any bool 2020-01-18 14-18-11 PST Adam Belis :D HELP i am useless 2020-03-02 01-59-52 PST Adam Belis i am chrome user 2020-01-18 14-34-45 PST Adam Belis haha i also need your id 2020-01-18 08-51-54 PST Adam Belis thanks i just took your idea and made it 100x100 harder :D 2020-01-18 13-50-11 PST Adam Belis but this is pyton 2.7 2020-01-18 13-11-12 PST Adam Belis yes 2020-01-18 12-48-00 PST Adam Belis oh 2020-01-18 13-00-22 PST Adam Belis i still cannot get it to run give me oooone second to trouble shoot it :D 2020-01-18 13-35-12 PST Adam Belis no that was not it 2020-01-18 12-48-50 PST Adam Belis i think i do on my second pc this is my tablet 2020-01-18 13-54-55 PST Adam Belis ok i fixed it 2020-01-18 13-26-29 PST Adam Belis dont worry we will get there one day 2020-01-26 09-40-35 PST Adam Belis i tried it few times and faild 2020-01-18 13-37-25 PST Adam Belis now it just refuse to run 2020-01-18 13-25-57 PST Adam Belis :D ok im ideiot volume 3 2020-01-18 13-33-57 PST Adam Belis that runs onley on 2.7 python 2020-01-18 08-50-18 PST Adam Belis and i am aslo noob i cannot du this simple scripting in ink 2020-01-18 13-35-28 PST Adam Belis wtf it run few mints ago :D 2020-01-18 14-25-52 PST Adam Belis Clipboard - January 18, 2020 11:25 PM (https://chat.inkscape.org/file-upload/QiXW2gHjJLoBM653P/Clipboard%20-%20Jan…) 2020-01-18 13-32-39 PST Adam Belis the fuck it refuse to run now 2020-01-18 13-55-11 PST Adam Belis lets try this again 2020-01-18 12-43-34 PST Adam Belis or dropbox 2020-01-26 11-41-01 PST Adam Belis maybe try to gain som more inside from doc he can be more help then me 2020-01-18 14-32-53 PST Adam Belis or give me your user name :D 2020-01-18 14-30-07 PST Adam Belis we can call fb, skype , hang outs chose your poison 2020-01-19 08-28-00 PST Adam Belis i know it is om my potential to do list but i am learning to much stuff already 2020-01-18 14-07-31 PST Adam Belis sript to ink folder 2020-01-18 13-47-53 PST Adam Belis dont know i did nothing 2020-01-18 14-34-48 PST Adam Belis its down in corner 2020-01-18 13-19-06 PST Adam Belis Clipboard - January 18, 2020 10:19 PM (https://chat.inkscape.org/file-upload/XrFMoJBwrTivcQW44/Clipboard%20-%20Jan…) 2020-01-18 13-43-09 PST Adam Belis no this did not helped 2020-01-18 12-47-57 PST Adam Belis to consol ? 2020-01-26 07-59-19 PST Adam Belis Hello was your hacky fixe merged in to master ? 2020-01-18 09-15-55 PST Adam Belis ok i scaled my test to 1000 spirals and my first findings wear correct your buid is 2x slower at this task 2020-01-26 11-36-35 PST Adam Belis maybe .. you saied that thats a problem 2020-01-18 13-23-49 PST Adam Belis Clipboard - January 18, 2020 10:23 PM (https://chat.inkscape.org/file-upload/DZinRbbGBDH4AR7Pr/Clipboard%20-%20Jan…) 2020-01-18 14-23-17 PST Adam Belis just empty txt 2020-01-18 13-25-32 PST Adam Belis i should replace paths to this 2020-01-18 13-43-56 PST Adam Belis its saying its incomaptible 2020-01-18 14-24-37 PST Adam Belis nada 2020-01-18 13-53-46 PST Adam Belis witch files ? 2020-01-19 09-59-46 PST Adam Belis btw this are results on my main pc (i7-6800k 3.4ghz) Inkscape 1.1-dev (abe53df, 2020-01-17) 1.4908316999999993 s Inkscape 1.0beta2 (43f5480, 2020-01-18) 1.5817405999999998 s Inkscape 1.0beta2 (18c40e5, 2019-12-05) 1.4947684999999993 s 2020-01-18 13-24-15 PST Adam Belis txt 2020-01-18 13-50-16 PST Adam Belis do i want that 2020-01-18 13-49-27 PST Adam Belis instaler say amd that could be prolem this is intel 2020-01-18 13-25-35 PST Adam Belis ? 2020-01-18 14-32-41 PST Adam Belis oh 2020-01-18 14-06-30 PST Adam Belis just exe ? 2020-01-26 08-52-11 PST Adam Belis 1.1-dev (0217ea3, 2020-01-23) 2020-01-18 13-44-54 PST Adam Belis i dont know eater maybe i dowalded wrong pyton for diff procesor but in that cas it should say that from the first try 2020-01-26 09-04-21 PST Adam Belis well right now its pretty broken . i know shit about math that is happning under hood but it looks like 3 problems that happend at once. - your fix - doc fix of too many points - and docs fix infroduces new bug with cistorting shapes 2020-01-18 13-57-50 PST Adam Belis Clipboard - January 18, 2020 10:57 PM (https://chat.inkscape.org/file-upload/dyX6iXrHesCDRCyPQ/Clipboard%20-%20Jan…) 2020-01-18 14-31-14 PST Adam Belis yes thats my user name on discort 2020-01-18 14-33-25 PST Adam Belis even pewdipai could not find it :D 2020-01-18 14-32-42 PST Adam Belis friends 2020-01-18 08-51-31 PST Adam Belis that could be use full to me 2020-01-26 11-37-06 PST Adam Belis but now it just uncoverd it 2020-01-18 09-39-32 PST Adam Belis your build 70 sec master 35 sec 2020-01-26 11-34-59 PST Adam Belis maybe doctor could point you to better direction 2020-01-18 13-25-08 PST Adam Belis you tell me 2020-01-26 08-50-15 PST Adam Belis Clipboard - January 26, 2020 5:50 PM (https://chat.inkscape.org/file-upload/ZqAZTKoKXArignr9b/Clipboard%20-%20Jan…) 2020-01-18 14-32-07 PST Adam Belis hahah add freind 2020-01-18 08-43-35 PST Adam Belis hello 2020-01-26 11-40-43 PST Adam Belis ok that would be awesome 2020-01-18 14-11-36 PST Adam Belis :D 2020-01-18 08-52-43 PST Adam Belis i am simple stupid designer :D 2020-01-18 14-21-29 PST Adam Belis hmmm same error 2020-01-18 14-30-36 PST Adam Belis i nver used discord for calling les try that 2020-01-18 13-09-39 PST Adam Belis Clipboard - January 18, 2020 10:09 PM (https://chat.inkscape.org/file-upload/jhzeAvQ5DXjmycvwr/Clipboard%20-%20Jan…) 2020-01-18 14-34-19 PST Adam Belis musseg up 2020-01-18 13-54-12 PST Adam Belis oOOOOOOOOOOOOO 2020-01-18 12-43-27 PST Adam Belis you can zip it 2020-01-18 13-48-04 PST Adam Belis nope 2020-01-18 12-51-31 PST Adam Belis will this work for win ? 2020-01-18 13-17-06 PST Adam Belis cannot find thet folder 2020-01-18 14-17-01 PST Adam Belis \\\ \\\ 2020-01-18 14-15-34 PST Adam Belis this is my current py 2020-01-26 09-05-24 PST Adam Belis funny thing is that lpe boolioan use diffrent math (looks like) and works correctly 2020-01-18 14-08-21 PST Adam Belis that does not work i get same error 2020-01-18 14-30-58 PST Adam Belis adambelis 2020-01-18 14-33-41 PST Adam Belis same avatar as here 2020-01-18 14-32-19 PST Adam Belis in middel of the screen on home page 2020-01-18 14-32-29 PST Adam Belis on top 2020-01-19 08-34-52 PST Adam Belis that python could be useful to me because i could use it in 3D software like Houdini or cinema 4d 2020-01-18 12-51-32 PST Adam Belis inkscape_executables = [ "./Inkscape-031d86a-x86_64.AppImage", "./Inkscape-4cfd59b-x86_64.AppImage", "./Inkscape-43f5480-x86_64.AppImage" ] 2020-01-18 12-47-54 PST Adam Belis where does it print to ? 2020-01-18 15-22-11 PST Adam Belis thanks again for that banchmark i will try toput together issu with proposal for som kind ofo official banch mark test for performace 2020-01-18 13-25-06 PST Adam Belis dont know :D 2020-01-18 13-25-23 PST Adam Belis aaaaaaaaah 2020-01-26 09-15-11 PST Adam Belis ok i wrote him about it so you are 100 percent shure its not your code that case the problem ? 2020-01-18 08-54-28 PST Adam Belis thanks for solving 2020-01-18 12-48-34 PST Adam Belis i think result txt would be better for the record 2020-01-18 08-56-55 PST Adam Belis maybe there are also maybe some for linux 2020-01-18 15-40-52 PST Adam Belis but i dont think so actualy it can run blender or cinema 4d pretty resonably 2020-01-18 14-24-39 PST Adam Belis Clipboard - January 18, 2020 11:24 PM (https://chat.inkscape.org/file-upload/86KwEGqYn2wRBgWHm/Clipboard%20-%20Jan…) 2020-01-18 14-05-06 PST Adam Belis should be copypasta 2020-01-18 13-38-11 PST Adam Belis thats what i am wondering 2020-01-18 08-58-08 PST Adam Belis this are win 2020-01-18 08-47-28 PST Adam Belis i dont tnik so but i could be wrong i am not a best person to ask about this kind of stuff i canoot even compile inkscape :D 2020-01-18 13-33-38 PST Adam Belis Clipboard - January 18, 2020 10:33 PM (https://chat.inkscape.org/file-upload/zGnHwXvHNZrELfKzk/Clipboard%20-%20Jan…) 2020-01-18 14-34-09 PST Adam Belis intresing i fell like boomer 2020-01-18 14-05-41 PST Adam Belis which file it cannot find ? 2020-01-18 14-29-15 PST Adam Belis can we call here ? 2020-01-18 13-40-20 PST Adam Belis right this is my every expirienc wiith pyton ever i am alwes like wow this does not work :D 2020-01-18 12-48-52 PST Adam Belis one sec 2020-01-18 14-01-03 PST Adam Belis Clipboard - January 18, 2020 11:01 PM (https://chat.inkscape.org/file-upload/523iowHZyHpzmwQcr/Clipboard%20-%20Jan…) 2020-01-26 11-36-24 PST Adam Belis probably doctor sayed the same thing 2020-01-18 14-27-10 PST Adam Belis ok so your script does not work ? 2020-01-18 08-49-22 PST Adam Belis but i was tasting how this will influenc performance 2020-01-18 13-55-03 PST Adam Belis reinstal instal reinstal 2020-01-19 08-28-34 PST Adam Belis cout yes it could be usfull even for my 3d stuff a specali for houdiny 2020-01-18 13-51-30 PST Adam Belis do iwant amd or win 32 2020-01-18 14-25-30 PST Adam Belis it does not tocntie 2020-01-19 08-26-02 PST Adam Belis oh 2020-01-26 08-51-20 PST Adam Belis yes 2020-01-18 14-11-35 PST Adam Belis oh 2020-01-18 13-54-27 PST Adam Belis dont know 2020-01-18 08-58-05 PST Adam Belis https://ci.appveyor.com/project/inkscape/inkscape/builds/30203579/job/n52kw… 2020-01-19 08-37-48 PST Adam Belis :D thanks but moments (tha this code jus dont want to run) like this reminds me why i am i visual artist and not a developer 2020-01-18 12-57-04 PST Adam Belis ah that make more sace 2020-01-18 13-32-46 PST Adam Belis say i have vrong verison 2020-01-19 10-01-22 PST Adam Belis yes 2020-01-18 13-54-23 PST Adam Belis good question 2020-01-18 14-06-47 PST Adam Belis i know :D 2020-01-18 08-58-20 PST Adam Belis cool 2020-01-26 11-35-53 PST Adam Belis just theoreticly if nodes would have that atribute would that fix it ? 2020-01-26 08-54-57 PST Adam Belis but if you can pleas test it yourself if you see same 2020-01-18 12-49-12 PST Adam Belis i think i do 2020-01-18 14-32-49 PST Adam Belis and after that 2020-01-18 13-52-09 PST Adam Belis yop 2020-01-18 13-25-50 PST Adam Belis with my builds 2020-01-18 13-20-25 PST Adam Belis no i think this is how it sous to be don i think i don this befor 2020-01-26 09-07-42 PST Adam Belis hmmm if you have time and energy try to get in contac with @doctormo and try to make a plan how to fix (or revert or dont know ) 2020-01-18 13-37-20 PST Adam Belis it created emty txt 2020-01-18 14-21-14 PST Adam Belis no no i got it 2020-01-18 13-25-03 PST Adam Belis if it didnt run it would not crate txt 2020-01-26 08-53-13 PST Adam Belis ok give me one sec i will try that spasific build 2020-01-18 14-26-51 PST Adam Belis could this be pyton glitch / wrong pyton instaled ? 2020-01-18 14-14-54 PST Adam Belis yop 2020-01-18 13-39-30 PST Adam Belis ok i just tested if that helps 2020-01-18 08-46-40 PST Adam Belis benchmarksvg.svg (https://chat.inkscape.org/file-upload/AYNLFshaDuydi9oSJ/benchmarksvg.svg) 2020-01-18 14-19-43 PST Adam Belis firs what ? 2020-01-26 08-51-40 PST Adam Belis let me test it one more tim to be sure im looking at correct build 2020-01-18 08-52-34 PST Adam Belis NOPE :D 2020-01-18 13-59-11 PST Adam Belis is this correct syntax ? 2020-01-18 13-54-18 PST Adam Belis :D 2020-01-18 14-33-56 PST Adam Belis adambelis#3440

1 0

[Webmaster] Rocket.Chat, 2 Users, 8 Messages, 0 Files, 615096 Minutes, in Direct Message Between: ayushkm & Tav
by no-reply＠chat.inkscape.org 29 Apr '21

29 Apr '21

2020-02-26 12-40-31 PST Ayush Krishna Murthi Yes, I am familiar with it 2020-02-26 12-44-50 PST Ayush Krishna Murthi I am familiar with selectors and css dialogs but not have much idea about improving it. Sorry for what i said earlier. 2020-02-26 12-31-24 PST Ayush Krishna Murthi Hi, This is Ayush, a college student. I am interested in contributing to project "Enhance CSS Support". Not able to find related documentation. 2020-02-26 12-36-57 PST Ayush Krishna Murthi Yes, I am a gsoc aspirant. 2020-02-26 12-36-05 PST Ayush Krishna Murthi I have been contributing to one of my college community as a web developer and have also done a small contribution to Inkscape-website. I am familiar with CSS, Django, Java and C++. Any suggestion about required technology that can be useful in contribution will be helpful. 2020-02-26 12-38-04 PST Tavmjong Bah Have you tried out the Selectors and CSS dialog in master? I'm sure as a web developer you will have some ideas on how to improve it. 2020-02-26 12-41-18 PST Tavmjong Bah I would start my making a review of the dialog and then proposing a set of improvements that could be made to it. 2020-02-26 12-36-14 PST Tavmjong Bah Your referring to P1 at https://wiki.inkscape.org/wiki/index.php/Google_Summer_of_Code ?

1 0

[Webmaster] Rocket.Chat, 6 Users, 43 Messages, 0 Files, 501692 Minutes, in #gsoc-mentors
by no-reply＠chat.inkscape.org 29 Apr '21

29 Apr '21

2020-09-06 04-24-15 PDT Martin Owens Thanks Tav! Amazing GSoC and a strong performance from yourself and the other mentors this year. 2020-05-15 14-37-24 PDT Martin Owens GDI, the meeting was today? 2021-01-06 18-14-30 PST Martin Owens @speleo3 I see this as a very tiny patch, possibly misguided, but perhaps they can be asked to work on another bug if they are trying to get patches in for GSoC 2020-05-15 14-37-40 PDT Martin Owens Oh right the email Mc sent round. Sorry all. 2020-05-15 08-54-26 PDT Mc Thank you all for coming ! Sorry Jabiert that you could not talk :( 2020-07-29 09-49-05 PDT Mc yes 2020-06-30 12-52-18 PDT Mc * it would be preferable, yes, if several mentors have been following up with the projects * only one eval per student can be posted, so I think it would be best that the "primary" mentor of students post theirs, after discussing it with the others * I think it's fine to wait for thursday evening 2020-08-31 07-55-24 PDT Mc @all https://digimedia1.r2.enst.fr/b/mar-yyy-2z2 as usual 2020-07-30 09-03-43 PDT Mc @jabiertxof @speleo3 https://digimedia1.r2.enst.fr/b/mar-yyy-2z2 2020-05-15 07-59-59 PDT Mc https://digimedia1.r2.enst.fr/b/mar-yyy-2z2 2020-07-29 07-46-58 PDT Mc @Tav @speleo3 @ede123 is tomorrow tentatively 18hCEST (16hUTC) ok to meet with moazin & linkmauve ? 2020-07-30 16-12-08 PDT Mc only moazin to do, tweenk emailed that he would do it soon 2020-05-15 09-08-32 PDT Jabier Arraiza and I loosse most of the discussion in the process 2020-05-15 09-08-09 PDT Jabier Arraiza I coulsd talk finaly using my mobile phone browser but is hard to me undertand speak and hardest speak 2020-07-31 04-27-44 PDT Jabier Arraiza sorry finaly cant attend 2020-05-15 08-00-25 PDT Jabier Arraiza thanks 2020-05-15 07-59-05 PDT Jabier Arraiza whats the BlueBuitton link, couldent search 2020-07-29 14-25-49 PDT Jabier Arraiza I try to attend 2020-05-15 08-00-41 PDT Jabier Arraiza I just see land in other mail 2020-09-06 18-12-08 PDT Jabier Arraiza Thanks all!!! 2020-07-29 07-47-54 PDT Tavmjong Bah Yes 2020-07-30 23-41-28 PDT Tavmjong Bah Thanks all! 2020-07-02 09-38-14 PDT Tavmjong Bah It was great to see all the progress today! @Mc Thanks for organizing this. 2020-07-01 05-07-46 PDT Tavmjong Bah Don't know why I didn't see this before... 2020-09-06 04-17-39 PDT Tavmjong Bah Thanks all for a great GSoC! (And getting the evaluations all done on time!) 2020-09-05 23-02-57 PDT Tavmjong Bah I don't think it matters significantly if the student didn't complete the project. It happens all the time. 2020-07-30 12-34-59 PDT Tavmjong Bah 4 student evaluations of mentors done. 2 mentor evaluations of students done. Don't make me nervous! 2020-07-29 08-15-13 PDT Thomas Holder yes 2021-01-06 15-02-54 PST Thomas Holder I'm having a deja-vu :-( IMHO Adrian's motivation to commit a patch is wrong, and his patch is pointless. This is discouraging. 2021-01-06 15-01-04 PST Thomas Holder https://gitlab.com/inkscape/inkscape/-/merge_requests/2655#note_479863443 2020-07-30 23-36-11 PDT Thomas Holder all done 2020-09-05 17-56-51 PDT Patrick Specifically I'm looking at Abhay's project, that is not "complete" given his initial goals (e.g. macro functionality mostly missing), but at the same time I never really expected him to complete all of it on time and had more realistic "internal" goals for his project in mind that he mostly matched (a mostly working command palette). 2020-07-01 05-00-42 PDT Patrick Thanks Marc, so basically "apply common sense", no particular project-specific conventions AFAIU. 2020-07-01 05-03-46 PDT Patrick As for Abhay I could do the evaluation in principle; @speleo3 and @jabiertxof are co-mentoring; Tav helped a lot as well. 2020-07-01 05-05-42 PDT Patrick Some guidance would be helpful to know how strictly we're supposed to grade, however I'm used to grade students, so I'd be able to figure something out either way. 2020-09-05 17-53-09 PDT Patrick Hi @all, can somebody comment on the significance of the question "Did ${student} complete the project?" in the final evaluation? If we answer this with "no", do we risk any drawbacks for the student or us as a project, or is it fine to answer "no" if the project is still ongoing at this time? 2020-07-01 05-01-21 PDT Patrick (pinging @Tav as I'm not sure he saw this message considering the mail he sent) 2020-05-15 09-06-33 PDT Patrick I've finally figured out how to use my webcam. Maybe you can see me next time. ;-) 2020-06-30 12-25-17 PDT Patrick Hi all, considering first eval's up, can some of the experienced mentors comment on how those are typically handled in the project? Any pointers are welcome! Some specific questions I have: * Do we discuss evaluations as a group / within the group of mentors for the specific student / (not at all)? * Related to that, who's expected to submit the evaluation (and when)? (I think we've appointed some "primary" mentors internally, but as far as Google is concerned it seems all of the mentors are equal and even org admins could do the evaluation) * We have a meeting scheduled on Thursday. Is it fine (or even advisable) to hold off with the evaluation until then, or should we evaluate as early as possible? 2020-07-29 08-31-58 PDT Patrick I won't be available tomorrow, sorry. 2020-07-29 08-32-31 PDT Patrick Did you make contact with linkmauve? 2020-05-15 07-28-26 PDT Mc User Tweenk added by Mc. 2021-02-06 16-35-47 PST Martin Owens Has left the channel.

1 0

[Webmaster] Rocket.Chat, 1 Users, 1 Messages, 0 Files, 504468 Minutes, in Direct Message Between: goktug & goktug
by no-reply＠chat.inkscape.org 29 Apr '21

29 Apr '21

2020-05-13 09-12-49 PDT Göktuğ hi

1 0

[Webmaster] Rocket.Chat, 2 Users, 40 Messages, 0 Files, 475327 Minutes, in Direct Message Between: ryangorley & Mandeep
by no-reply＠chat.inkscape.org 29 Apr '21

29 Apr '21

2020-06-04 10-20-21 PDT Mandeep Alright. 2020-06-05 14-25-36 PDT Mandeep Hi, So I tried to find my way around GitLab but I am still not very comfortable using that platform. Is it possible that we could communicate over e-mails till the time I understand the security, work flow & privacy settings around GitLab? 2020-06-04 10-19-03 PDT Mandeep I can join via text, right? 2020-06-04 10-22-28 PDT Mandeep I will check that as well. 2020-06-04 10-20-06 PDT Mandeep team_vectors? 2020-06-06 09-53-59 PDT Mandeep Yes, SEO plan will be ready in a couple of days & I already have few ideas that should be able to help us. I will put all of this together in an email. 2020-06-04 10-18-15 PDT Mandeep I will be able to join the meeting but don't think that the actionable plan for SEO will be ready by then. It will take me around 2-3 days. 2020-06-04 10-02-15 PDT Mandeep Sure, I can look after SEO side of it. I also design E-mails & manage social channels for my full time job. I can also look after that for Inkscape. 2020-06-04 10-19-58 PDT Mandeep & the meeting's here? 2020-06-04 09-56-42 PDT Mandeep My focus is on digital side of the marketing but I handle other aspects as well. 2020-06-05 14-27-21 PDT Mandeep Thanks. 2020-06-04 09-47-21 PDT Mandeep Hi, Thanks for reaching out. I have requested for the access. I have majored in marketing and have been working for a few years in the marketing field for various kinds of companies, helping them with their SEM. SEO, Content, E-mail marketing etc. 2020-06-04 10-01-09 PDT Mandeep Yes, Agreed! We won't be able to compete against them in terms of the resources that they have. 2020-06-04 10-23-30 PDT Mandeep Nice talking to you, I will catch up with you during the meeting. Do let me know if there is something I can help with. Thanks. 2020-06-04 10-09-01 PDT Mandeep You can provide me access for whichever platform you need me to manage. So our goal is to get more users & keep the project floating & growing, Understood. 2020-06-04 10-11-32 PDT Mandeep I will share what we should do in terms of SEO, in a couple of days & then we can work towards it. 2020-06-04 10-10-43 PDT Mandeep For e-mails, platforms like mailchimp, moosend etc. do provide some sort of basic and free functionality upto 2000 contacts. We can use their services and promote inkscape. Maybe we can have a sign-up form on our website, segment the audience and optimize the emails per user behaviour. 2020-06-02 15-02-09 PDT Ryan Gorley When you get a minute, let me know a little more about yourself and background and let's find a way to put your talent to good work. I'm glad to have you! 2020-06-02 15-01-36 PDT Ryan Gorley If you don't have a GitLab account, please create one and then request access to that group. You'll a lot of prior projects, meeting topics, ideas, etc. in the Issues. 2020-06-04 10-04-42 PDT Ryan Gorley It may be worth it. 2020-06-04 10-21-04 PDT Ryan Gorley https://lists.inkscape.org/postorius/lists/inkscape-vectors.lists.inkscape.… 2020-06-04 10-18-41 PDT Ryan Gorley That's alright. It will be nice to have you introduce yourself at least. 2020-06-04 10-24-16 PDT Ryan Gorley Likewise, and I certainly will. Thanks for joining us. I'm glad to have you onboard! 2020-06-06 09-35-28 PDT Ryan Gorley Thanks for joining our meeting. I'd love to put your expertise to work. Do you think you'll have some ideas to share before our next meeting? I'd love to hear them. 2020-06-04 10-20-09 PDT Ryan Gorley Yes, in the #team_vectors channel 2020-06-06 09-57-54 PDT Ryan Gorley Sweet. I'm glad to see you connecting with Michele about keywords as well. 2020-06-04 10-07-11 PDT Ryan Gorley Something to keep in mind, that I had to learn, was that the important social good brought about the project aside, the project itself really doesn't benefit directly from more users. We should get the tool out there for people, because that is the right thing to do, but we need to optimize for conversions in the form of contributors (and to a lesser extent donations). 2020-06-04 10-04-18 PDT Ryan Gorley I think it would be great to get additional eyes on social media. We've never done anything with emails. I think that could be interesting. It will be a little more difficult to get off the ground because we'll have to get infrastructure put in place, which is a struggle right now. 2020-06-04 10-00-57 PDT Ryan Gorley There is much more than search optimization we need, but that is definitely something you would bring unique expertise on. 2020-06-04 10-19-32 PDT Ryan Gorley Yeah, the meeting is all text. 2020-06-02 15-00-18 PDT Ryan Gorley Hi @Mandeep . Thanks for reaching out. I'm sorry to take so long getting back with you. I run a small creative agency (https://freehive.com) and we've been really busy the last few weeks. I come from a marketing background as well. In fact, the Vectors came about in part because I came asking the same question you have and no one was really doing any marketing work for the project. Here is a small survey I put out around then to understand better who was using Inkscape, and where they saw the value in the application: https://docs.google.com/spreadsheets/d/1hz_Q6epUfUJXO0w_XZGZ8GRDOOpzjl_kXe9… Give it a look if you like. You can also find a lot of our business on GitLab at: https://gitlab.com/inkscape/vectors 2020-06-04 09-59-56 PDT Ryan Gorley We're not exactly in competition with the commercial tools on the market, and I don't really feel like we need to be of that mindset, but rather we should help as many people as possible discover Inkscape and make the choice for themselves. 2020-06-04 09-58-33 PDT Ryan Gorley I haven't gotten this off the ground, in part because I've been the only one with any experience with it, but I would love for the project to make a more deliberate effort to optimize for search phrases that will help us get found more often. 2020-06-04 10-14-01 PDT Ryan Gorley There is a lot of resistance from certain quarters within the project to using proprietary tools, just so you know. I think an email would be good though. 2020-06-04 10-19-27 PDT Ryan Gorley When you are ready to publish your plan, just create an issue in the "General" subgroup of our GitLab group and post what you have there. It supports markdown. 2020-06-04 10-20-43 PDT Ryan Gorley Oh, and we do have a group mailing list if you want to get reminders on meetings. It is at... 2020-06-05 14-26-19 PDT Ryan Gorley Yeah, sure. My email is ryan(a)freehive.com 2020-06-04 10-08-28 PDT Ryan Gorley Donations are less valuable right now because the project hasn't really figured out how to spend that money on anything but hackfests. We're figuring that out, but contributions of time and talent go much further right now. 2020-06-04 10-16-53 PDT Ryan Gorley An initial dive into what we can do on the SEO front would be awesome. We tried to pick a time for our meeting that works for most quarters of the world. Someone said it is pretty late in your timezone. Is it doable? 2020-06-04 09-57-09 PDT Ryan Gorley Nice

1 0

[Webmaster] Rocket.Chat, 2 Users, 10 Messages, 0 Files, 694087 Minutes, in Direct Message Between: major & RdH
by no-reply＠chat.inkscape.org 29 Apr '21

29 Apr '21

2020-01-02 15-56-48 PST Major No windows 2020-01-02 15-57-19 PST Major ok thanks 2020-01-02 15-55-03 PST Major ? 2020-01-02 15-54-58 PST Major can you help me 2020-01-02 15-55-02 PST Major please 2020-01-02 15-53-50 PST Major Bonjour René, êtes vous français ? 2020-01-02 15-54-46 PST Major Ok I have a little problem 2020-01-02 15-54-26 PST René de Hesselle English and German only. 2020-01-02 15-57-05 PST René de Hesselle Sorry, can't help you with that. I'm the macOS maintainer. 2020-01-02 15-56-29 PST René de Hesselle Post your questions into the user channel, so more people will be able to help. I'm a developer, but a novice user. Unless your question has something to do with the macOS app, I'm not your guy.

1 0

[Webmaster] Rocket.Chat, 2 Users, 4 Messages, 0 Files, 621821 Minutes, in Direct Message Between: ridwan & rathod-sahaab
by no-reply＠chat.inkscape.org 29 Apr '21

29 Apr '21

2020-02-21 20-36-35 PST Abhay Raj Singh I totally get you So the point is Inkscape is *open source: free as in beer and free as in speech* unlike other *propritery* softwares *we don't charge money/attribution etc.* you can redristibute the software _if you want_ to your friends an even on your website. There is no licesing restriction/payment model hence you cant register. People volunteer to make/test/promote the software. _Its by the people, of the people for the people_ All in all use it as you want. For legal License we have GPL 2.0 for software if you are going for it with your company lawyers but you dont need it because we can sue you😅 2020-02-21 21-28-51 PST Abhay Raj Singh Thanks for using Inkscape without user it would have been just a project :relaxed: 2020-02-21 20-46-48 PST Ridwan wow, sir , :v :joy: , thank you, sir, for your participation :grin: 2020-02-21 20-26-41 PST Ridwan Hi sir, sorry in advance, I still don't quite understand, I want to ask about the creator's license, if I have a design work that I made using Inkscape, Do I have to include an Inkscape license in my work? then if so, how do I register the Inkscape license?

1 0

[Webmaster] Rocket.Chat, 2 Users, 27 Messages, 0 Files, 629572 Minutes, in Direct Message Between: jhofinger & rathod-sahaab
by no-reply＠chat.inkscape.org 29 Apr '21

29 Apr '21

2020-09-11 05-00-52 PDT Abhay Raj Singh Thanks! I have a meeting today with Tav I guess this is what he wanted to talk to me about? 2020-02-17 07-00-33 PST Abhay Raj Singh Hey! there are some more changes required in `spin button tool item` If you have already been working on it, can you please add those changes, if you haven't may I go ahead and do both I kind of need them for my MR :sweat_smile: 2020-09-11 05-40-53 PDT Abhay Raj Singh This might also enable us to standardize dialogs, and not creating specific dialog where not necessary 2020-02-17 08-02-23 PST Abhay Raj Singh There's a place where code adds bunch of value which aren't needed including negative sometimes😂 2020-09-11 05-49-53 PDT Abhay Raj Singh **Second problem : Previews** Previews are executed as Do and undo, so we need to solve that too 2020-09-11 05-38-11 PDT Abhay Raj Singh ``` struct ParamDetails { Glib::ustring name; Glib::ustring description; ParamType type; } enum class ParamType { INTEGER, DOUBLE, STRING, COLOR, }; ``` when making dialog name is for lable, description for tooltip, paramtype helps in making customized input fields like color selectors, spinners 2020-09-11 05-38-42 PDT Abhay Raj Singh Though, these are not necessary for core actions 2020-09-11 05-15-05 PDT Abhay Raj Singh I have some solutions to the problems faced earlier 2020-09-11 05-32-34 PDT Abhay Raj Singh Solution for that was 2020-09-11 05-31-59 PDT Abhay Raj Singh But the actual problem was creating string 2020-09-11 05-39-39 PDT Abhay Raj Singh Other things such as value ranges can also be added 2020-09-11 05-04-34 PDT Abhay Raj Singh Thanks for you work by the way, it will enable me to add tracking and multiple parameter more easily and will function better! 2020-02-16 11-13-58 PST Abhay Raj Singh Yeah I got the same error when I was working on text might be the in *SpinButtonToolItem*. Will fix it soon Thanks for drawing my attention to this will have took my time to even track the bug :grinning: 2020-09-11 05-31-38 PDT Abhay Raj Singh The solution *was* for the problem of multiple parameter for an action which we cannot have with Vanilla actions, A function to make strings for action and when the target function receives the string, it blindly parses as it know what parameters and type of parameters are required. 2020-09-11 05-51-01 PDT Abhay Raj Singh **Third problem: History sensitivity" opening a dialog should not be sensitive to history but most actions should be 2020-03-04 10-46-17 PST Abhay Raj Singh Thank you very much I would have totally missed that. Also, numerical codes can be added so that one can also follow a german/chinese/hindi tutorial too without access performance costs. 2020-03-04 10-36-52 PST jhofinger Hi, just read your GSOC proposal. From my role (bug triaging), it's super cool to have someone who has a lot of practical experience actually using Inkscape, that's not true for everyone in the dev team. While I can't comment on technical feasability (althouh I think it should be doable, since Inkscape already has a very good "verbs" list), I have two annotations: * Please check that the shortcut for the commant palette is very easy to use with common locales. Strg+` can't be used on a German keyboard, since it requires the ` key to be pressed twice to show the character. * When searching for commands, it should show only localized results, BUT searching for english command names should work as well. When I use a localized Inkscape version, but am trying to follow an English tutorial, I sometimes don't know the localized equivalent of the string, so I'd try to search for the english one. That would be especially cool for filters, because they have very weird names (see also https://gitlab.com/inkscape/inbox/issues/1939). Maybe show the results like the following, when e.g. searching for [str] * Füllung und Kontur (Fill and **Str**oke) * **Str**euung 2020-09-11 05-09-36 PDT jhofinger That was the plan. There are still some details to figure out though. 2020-02-16 11-34-47 PST jhofinger Found it, thanks for pointing me there! 2020-02-17 07-38-51 PST jhofinger Oh okay.. Yes I did something, but it's not 100%. Go ahead, I'll review your MR then, if you cc me 😉 2020-02-17 07-46-29 PST jhofinger Also I simply added a function to check if a value is out of bounds or already there. It just didn't compile for some reason ^^ im still confused by c++ sometimes, normally I use c#. Anyway looking forward to your solution! 2020-09-11 05-03-12 PDT jhofinger Yes 2020-02-16 11-11-17 PST jhofinger (or if you tell me where the code for this is, I might have a look at it myself) 2020-02-16 11-09-45 PST jhofinger Hi, since you were working on the right click of the text field: could you have a look whats happening here https://gitlab.com/inkscape/inkscape/issues/690 and here https://gitlab.com/inkscape/inkscape/issues/689? 2020-09-11 05-16-54 PDT jhofinger For example? 2020-09-10 11-54-17 PDT jhofinger Hi! After some discussion in the shortcuts MR with Patrick and Tav, I've implemented a subclass to SimpleAction that encapsulates all the extra data Inkscape needs to store. I was told that you were working on formalizing the argument structure of action parameters, so you might be interested in this as well. https://gitlab.com/jhofinger/inkscape/-/commit/e82e30019f4cd12089837019bcef… 2020-09-10 11-54-42 PDT jhofinger Until now, I've only rewritten the snap bar actions

1 0