The windows nightly builds are still broken...
Can anyone look into this please?
Diederik
You should learn how to build the dev version yourself. It's not hard - heck, even I can do it!
Just follow the instructions at this link step by step and you will get there; http://wiki.inkscape.org/wiki/index.php/Win32Port
I've just updated the page above and followed the instructions there exactly to build the latest version of source code.
Diederik van Lierop wrote:
The windows nightly builds are still broken...
rygle wrote:
You should learn how to build the dev version yourself.
I think Diederik knows how to do this, since he is a developer. That's not the point. The nightly builds are intended for our faithful army of testers. Having to install a compiler, figuring out the build process and having to wait for the program to compile would create an additional obstacle to becoming a tester. I think this should be fixed as fast as possible, because it seriously deprives us of feedback.
Regards, Krzysztof Kosiński
On Sat, 2009-02-28 at 17:02 -0800, Krzysztof Kosiński wrote:
I think Diederik knows how to do this, since he is a developer. That's not the point. The nightly builds are intended for our faithful army of testers. Having to install a compiler, figuring out the build process and having to wait for the program to compile would create an additional obstacle to becoming a tester. I think this should be fixed as fast as possible, because it seriously deprives us of feedback.
If you are volunteering, please go ahead.
At the moment we really only have a single person working on maintaining that whole system. It's very unreasonable to push demands onto one person like that.
We are very aware of how it helps testers, however we are also very aware that the solution is to have more people help, not more people complain.
On 03/01/2009 03:02 AM, Jon A. Cruz wrote:
If you are volunteering, please go ahead.
At the moment we really only have a single person working on maintaining that whole system. It's very unreasonable to push demands onto one person like that.
We are very aware of how it helps testers, however we are also very aware that the solution is to have more people help, not more people complain.
I'm not complaining, just notifying our fellow devs of the problem. It's great that Bob maintains our nightly windows build, and I'm sure that's much appreciated by everyone!
But does anyone know what currently the problem is? Is it compiling on Windows at all (in which case any windows dev can look into this), or is it just the automated nightly build process that's broken?
Diederik
It compiled fine for me last night, was pretty unusable as it's missing half it's icons tho.
Sent from my iPhone
On 1 Mar 2009, at 07:15, Diederik van Lierop <mail@...1689...> wrote:
On 03/01/2009 03:02 AM, Jon A. Cruz wrote:
If you are volunteering, please go ahead.
At the moment we really only have a single person working on maintaining that whole system. It's very unreasonable to push demands onto one person like that.
We are very aware of how it helps testers, however we are also very aware that the solution is to have more people help, not more people complain.
I'm not complaining, just notifying our fellow devs of the problem. It's great that Bob maintains our nightly windows build, and I'm sure that's much appreciated by everyone!
But does anyone know what currently the problem is? Is it compiling on Windows at all (in which case any windows dev can look into this), or is it just the automated nightly build process that's broken?
Diederik
Open Source Business Conference (OSBC), March 24-25, 2009, San Francisco, CA -OSBC tackles the biggest issue in open source: Open Sourcing the Enterprise -Strategies to boost innovation and cut costs with open source participation -Receive a $600 discount off the registration fee with the source code: SFAD http://p.sf.net/sfu/XcvMzF8H _______________________________________________ Inkscape-devel mailing list Inkscape-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/inkscape-devel
Hi devs,
I haven't seen a response to the question that Diederik posed recently with respect to Windows nightlies. Namely, "But does anyone know what currently the problem is? Is it compiling on Windows at all (in which case any windows dev can look into this), or is it just the automated nightly build process that's broken?
As a windows user I too am concerned that there do not appear to be any recent nightlies (read weeklies also) available for windows users. The last seems to be still 17-Feb-2009 17:49 and I am sure that are a lot of people like me who cannot or do not have time to get into the do-it-yourself business. I am also not complaining but just voicing my concern.
On a more general level I wonder if someone can tell us non-devs where Inkscape is with respect to another release. It seems an awful long time since the last one and knowing how long it takes to get through the pre-release debug stages I wonder if the developers can give us some information here. Is the cutting edge stuff getting blunt?
thanks, Erik
kaver wrote:
Hi devs,
I haven't seen a response to the question that Diederik posed recently with respect to Windows nightlies. Namely, "But does anyone know what currently the problem is? Is it compiling on Windows at all (in which case any windows dev can look into this), or is it just the automated nightly build process that's broken?
I think this specific build process is broken. I for one have been able to compile and run Inkscape successfully during the past week, in order to run the rendering tests.
... On a more general level I wonder if someone can tell us non-devs where Inkscape is with respect to another release. It seems an awful long time since the last one and knowing how long it takes to get through the pre-release debug stages I wonder if the developers can give us some information here. Is the cutting edge stuff getting blunt?
I can't give a very definite answer to that, but I can give some information from the perspective of the rendering tests I regularly run. As can be seen, the results of the rendering tests have been comparatively quiet the last couple of weeks: http://home.hccnet.nl/th.v.d.gronde/inkscape/ResultViewer.html (Note that the page is getting a bit slow to load, I haven't come around to making a faster/static version.)
However, this doesn't mean all is well. My "favourite" showstopper (in my opinion) at the moment is the lack of linearRGB<->sRGB transforms, as this means that pretty much anything currently done with filters on Inkscape looks inherently wrong (and in some cases quite obviously so, when compared to the "correct" image). I'm afraid I haven't written a decent bug report yet though... (Have added it to my todo list.)
And while the results of the rendering tests have been reasonably stable, there are still loads of tests that fail, and not all just because the feature hasn't been implemented yet.
On Fri, Mar 6, 2009 at 9:16 AM, Jasper van de Gronde <th.v.d.gronde@...528...> wrote:
However, this doesn't mean all is well.
At some point, inevitably, we will have to limit ourselves to regressions only. Did this work in 0.46? No? Then release, at least it is not worse than the last one.
At some still further point, we will have to sort even regressions into releaseable and showstoppers, and focus on the latter category.
Given the insane amount of new stuff and fixes we already have, and the time passed since 0.46, I would argue that the first of these points is already reached. So, here's a straw poll for the developers: how many serious regressions you are aware of? Linking to bug reports would be nice, but just giving your general impression is helpful too.
bulia byak wrote:
So, here's a straw poll for the developers: how many serious regressions you are aware of?
maybe we could start a list... here are my 2 favorite bugs which I think are regressions : - https://bugs.launchpad.net/inkscape/+bug/179988 - https://bugs.launchpad.net/inkscape/+bug/284680
Hi,
bulia byak wrote:
So, here's a straw poll for the developers: how many serious regressions you are aware of?
maybe we could start a list... here are my 2 favorite bugs which I think are regressions :
Another one: https://bugs.launchpad.net/inkscape/+bug/307195
Could we start a new thread for regressions? A metabug, maybe? -- Nicolas
On Sat, Mar 7, 2009 at 4:39 PM, Nicolas Dufour <nicoduf@...48...> wrote:
Another one: https://bugs.launchpad.net/inkscape/+bug/307195
Gustav:
I started to dig this one and ran into
http://inkscape.svn.sourceforge.net/viewvc/inkscape/inkscape/trunk/src/ui/di...
I am not 100% sure but this kind of code looks VERY strange:
panel->setDesktop(panel->getDesktop());
At least you need to add a comment explaining why this is not an error.
In any case, looks like after two switches of windows, panel's _desktop is NULL, causing all kinds of trouble. Could you please find time to look into this bug?
Hello,
Adding my two-cents:
A major regression between 0.45.1 and 0.46, and still valid in latest SVN, is:
https://bugs.launchpad.net/inkscape/+bug/227949
(ENTER does not work anymore in Transform dialog)
I think it would be really important that 0.47 would restore this.
Thanks!
--- On Sat, 3/7/09, bulia byak <buliabyak@...400...> wrote:
From: bulia byak <buliabyak@...400...> Subject: Re: [Inkscape-devel] Re : Windows nightly builds are still broken... To: "Nicolas Dufour" <nicoduf@...48...>, "Gustav Broberg" <broberg@...370...> Cc: inkscape-devel@lists.sourceforge.net Date: Saturday, March 7, 2009, 6:26 PM On Sat, Mar 7, 2009 at 4:39 PM, Nicolas Dufour <nicoduf@...48...> wrote:
Another one: https://bugs.launchpad.net/inkscape/+bug/307195
Gustav:
I started to dig this one and ran into
http://inkscape.svn.sourceforge.net/viewvc/inkscape/inkscape/trunk/src/ui/di...
I am not 100% sure but this kind of code looks VERY strange:
panel->setDesktop(panel->getDesktop());
At least you need to add a comment explaining why this is not an error.
In any case, looks like after two switches of windows, panel's _desktop is NULL, causing all kinds of trouble. Could you please find time to look into this bug?
-- bulia byak Inkscape. Draw Freely. http://www.inkscape.org
Open Source Business Conference (OSBC), March 24-25, 2009, San Francisco, CA -OSBC tackles the biggest issue in open source: Open Sourcing the Enterprise -Strategies to boost innovation and cut costs with open source participation -Receive a $600 discount off the registration fee with the source code: SFAD http://p.sf.net/sfu/XcvMzF8H _______________________________________________ Inkscape-devel mailing list Inkscape-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/inkscape-devel
I agree that narrowing down to regressions and crashes should be priority.
Seems like PDF output and some other important format support issues have really moved forward. Also plenty of great new features have gotten in, so maybe it's time to cut our losses and move to release.
How's the new geometry (2geom?) code going? Are there any major slowdowns like in the early refactoring days?
It might also be good to see if the code to speed up blurs etc can get in if this is possible in the short term.
https://bugs.launchpad.net/inkscape/+bug/179988 I agree that this is of major importance, and if someone has the right know how it seems like only an output scaling issue. It is a regression caused by the shifting GTK and Cairo landscape, which now seems to have improved vastly. Adrian Johnson has made some suggestions on this one in the bug report.
Nicolas Dufour-5 wrote:
bulia byak wrote:
So, here's a straw poll for the developers: how many serious regressions you are aware of?
maybe we could start a list... here are my 2 favorite bugs which I think are regressions :
Another one: https://bugs.launchpad.net/inkscape/+bug/307195
rygle wrote:
It might also be good to see if the code to speed up blurs etc can get in if this is possible in the short term.
Code to speed up blurs? I made some changes to speed it up quite significantly quite a while ago, but I think that was even already a part of the last release. And more recently I added some multi-threading functionality, but this is also already in SVN. Any further speed gains are obviously welcome, but I haven't heard of any yet.
Jasper van de Gronde wrote:
Code to speed up blurs? I made some changes to speed it up quite significantly quite a while ago, but I think that was even already a part of the last release. And more recently I added some multi-threading functionality, but this is also already in SVN. Any further speed gains are obviously welcome, but I haven't heard of any yet.
That's the code I meant. Wasn't sure that it was in, but glad it is!
On Sat, Mar 7, 2009 at 6:39 PM, Nicolas Dufour <nicoduf@...48...> wrote:
Hi,
bulia byak wrote:
So, here's a straw poll for the developers: how many serious regressions you are aware of?
maybe we could start a list... here are my 2 favorite bugs which I think are regressions :
Another one: https://bugs.launchpad.net/inkscape/+bug/307195
May we need to get seriously in a test-driven development.
Running the "make check" now we see "Failed 16 of 176 tests". If we have more tests and if we test before commit, the regression is less possible.
How about a TDD politic?
On Mar 9, 2009, at 11:48 AM, Aurélio A. Heckert wrote:
May we need to get seriously in a test-driven development.
Running the "make check" now we see "Failed 16 of 176 tests". If we have more tests and if we test before commit, the regression is less possible.
How about a TDD politic?
Probably about time.
I was one who set up a lot of the unit testing (though not the first). It's been a while, but we probably need to buckle down and finish getting those last 16 tests passing.
As soon as we get them set, we can enforce the "no fails" restriction.
I seem to recall that at least one of the non-passing issues exposed a problem with how we were round-tripping some data, and might need something else to fix... but then again that's why we need tests.
As a temporary workaround I've created my own windows build, which you can download using this link:
http://www.diedenrezi.nl/inkscape-20090306-rev20841.7z
It might differ at some point from Bob's nightly builds, but it works fine for me. Just unzip it and run the executable. I don't mind if you spread the word through Launchpad, but please do not put a link at our homepage because I don't know how much bandwidth I'm allowed by the fair-use-policy of my hosting provider.
Diederik
Bobs fixed his builds today. there was a problem with something in build.xml that was breaking the cross compile apparently.
2009/3/6 Diederik van Lierop <mail@...1689...>:
As a temporary workaround I've created my own windows build, which you can download using this link:
http://www.diedenrezi.nl/inkscape-20090306-rev20841.7z
It might differ at some point from Bob's nightly builds, but it works fine for me. Just unzip it and run the executable. I don't mind if you spread the word through Launchpad, but please do not put a link at our homepage because I don't know how much bandwidth I'm allowed by the fair-use-policy of my hosting provider.
Diederik
Open Source Business Conference (OSBC), March 24-25, 2009, San Francisco, CA -OSBC tackles the biggest issue in open source: Open Sourcing the Enterprise -Strategies to boost innovation and cut costs with open source participation -Receive a $600 discount off the registration fee with the source code: SFAD http://p.sf.net/sfu/XcvMzF8H _______________________________________________ Inkscape-devel mailing list Inkscape-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/inkscape-devel
On 03/06/2009 08:52 PM, john cliff wrote:
Bobs fixed his builds today. there was a problem with something in build.xml that was breaking the cross compile apparently.
That's too bad, because now I don't get to test my hosting provider ;-)
Thanks Bob!
Diederik
PS: My build is still one rev. newer ;-)
Jon A. Cruz wrote:
If you are volunteering, please go ahead.
I was just explaining to Rygle why the answer of "build your own" wasn't a replacement for the nightly builds. I'm not blaming or criticizing anyone - I'm aware that there must be problems delaying the fix. I would help but I don't really know how.
Regards, Krzysztof Kosiński
participants (14)
-
Alvin Penner -
Aurélio A. Heckert -
bulia byak -
Diederik van Lierop -
Jasper van de Gronde -
john cliff -
John Cliff
-
Jon A. Cruz -
kaver -
Krzysztof Kosiński -
Mike Pittman -
Nicolas Dufour -
rygle -
worms invasion