Re: Darktable on windows. Was: Re: [darktable-dev] Darktable crashes on a SIGILL
I agree too with Roman. The first key point is to have volunteers with a long term commitment to maintain the windows port. Then propose small and incremental patches to prepare the Windows build. The volunteers need also to build binaries from time to time (stable and dev to test) and follow the support on Redmine. That being said, I think it would be valuable to have such a build, for one thing, some transient bugs may appear more easily on Windows. In the long run this will help darktable stability. That's my experiences with software running on multiple platforms. -- Pascal Obry / Magny Les Hameaux (78) The best way to travel is by means of imagination http://v2p.fr.eu.org http://www.obry.net gpg --keyserver keys.gnupg.net --recv-key F949BD3B ___ darktable developer mailing list to unsubscribe send a mail to darktable-dev+unsubscr...@lists.darktable.org
Re: Darktable on windows. Was: Re: [darktable-dev] Darktable crashes on a SIGILL
yeah, I agree with Roman on that one... we are quite afraid of the quantity of windows-specific bug that a windows build would bring in. maybe a good approch would be to do windows builds, contribute patches, but not distributing the binaries (or to a very limited number of alpha testers, all of which are able to debug crashes themselves) of course that would only find so many bugs, but it's a start and it would show that the basic "social mechanics" to get a windows port working are in place... On Thu, Dec 17, 2015 at 9:20 PM, Erik Duisterswrote: > Fine, > > btw building is not trivial if you include all dependencies. > > I'll see what i can contribute > > On Thursday 17 December 2015 21:44:34 Roman Lebedev wrote: > > And here I thought dt windows bikeshedding was explained to > > death[1][2] and over :( > > > > The problem is *not* building under windows. > > The problem is maintaining it. > > > > What *all* the people trying to bring dt to windows have been doing so > > far is called packaging, > > at best. As an example, thus far, no one has actually talked to us > > directly, did not try to > > contribute windows-related fixes/changes for 1.6.x version back into > > darktable. > > > > We are working on darktable in our free time, of pure fun and > > interest. There is no money involved. > > Thus, darktable, being available for some bigger circle of/to more > > users does not help *us* at ALL. > > > > It *ONLY* hurts us: > > 1. More users => more bugreports, while amount of developers stays > constant. > > 2. The more darktable is popular, the bigger the chances of next > > person to use it being less > > skilled in at least basic debugging than the average dt user before that. > > That, unfortunately, even includes uploading > > /tmp/darktable_bt_XX.txt... Not all users do that :( > > 3. !!! Every new big chunk of "market" is always a minefield at start, > > e.g.: https://redmine.darktable.org/issues/10541 > > That bug is probably very simple, but no one of us uses a distribution > > with musl libc, so no one can debug it, > > so it stays there, dormant... > > 4. Also, did you know that on windows, our automatic backtrace > > creation would not even work, so > > even getting backtrace would be harder than usual? > > > > Now, imagine, how much of those bugs will be from windows? > > None of us uses windows, for various reasons, so we are of little help > > there. Of course, if there is a VERY large demand, one can do it, but > ONLY > > for some money as paid builds. > > > > Thus, before darktable can be made available to any new big "market", > > there needs to be a person, > > whom we trust to be capable in handling all of the aspects of > > maintaining darktable there. > > > > (For a normal UNIX/BSD/Linux distribution, package maintainer is an > > obvious answer, in most cases.) > > > > So, what exactly do we mean by a maintainer? > > 0. Be actively involved with darktable for some time. At least half a > year. > > That means: > > *) Stick with us in IRC. > > *) actually know the code. > > *) actively examine bugreports, fix bugs. > > 1. Be reliable and trustworthy [over that time], and more importantly, > > make us believe in that. > > > > Once 1. and 2. are complete, there is some room to discuss possibility > > of making official darktable windows port and becoming it's > > maintainer. > > > > IMPORTANT: ANY OTHER APPROACH TO DARKTABLE ON WINDOWS CAN AND WILL BE > > CONSIDERED HOSTILE, > > HAPPENING AGAINST OUR WILL AND ACTIVELY HURTING DARKTABLE PROJECT. > > > > [1] > > > https://www.darktable.org/2015/07/why-dont-you-provide-a-windows-build/comm > > ent-page-1/#comment-46412 [2] > > https://discuss.pixls.us/t/darktable-on-windows-parthas-build-thread/290 > > > > Roman. > > Also signed by Tobias. > > > ___ > > darktable developer mailing list > > to unsubscribe send a mail to > darktable-dev+unsubscr...@lists.darktable.org > > ___ > darktable developer mailing list > to unsubscribe send a mail to > darktable-dev+unsubscr...@lists.darktable.org > > ___ darktable developer mailing list to unsubscribe send a mail to darktable-dev+unsubscr...@lists.darktable.org
Re: Darktable on windows. Was: Re: [darktable-dev] Darktable crashes on a SIGILL
Partha, > I am sorry, why am I being copied on this? I've not commented on DT > for quite a long time. Why are you asking? You have been proposing a set of patch for building darktable on Win64. The conditions to have this integrated have been stated clearly by Roman and Tobias, this will avoid any frustration. If you are still motivated and willing to see this project moving forward why don't you have a merged effort with Erik? So to me you seems a good candidate for this mail :) Best Regards, -- Pascal Obry / Magny Les Hameaux (78) The best way to travel is by means of imagination http://v2p.fr.eu.org http://www.obry.net gpg --keyserver keys.gnupg.net --recv-key F949BD3B ___ darktable developer mailing list to unsubscribe send a mail to darktable-dev+unsubscr...@lists.darktable.org
Darktable on windows. Was: Re: [darktable-dev] Darktable crashes on a SIGILL
And here I thought dt windows bikeshedding was explained to death[1][2] and over :( The problem is *not* building under windows. The problem is maintaining it. What *all* the people trying to bring dt to windows have been doing so far is called packaging, at best. As an example, thus far, no one has actually talked to us directly, did not try to contribute windows-related fixes/changes for 1.6.x version back into darktable. We are working on darktable in our free time, of pure fun and interest. There is no money involved. Thus, darktable, being available for some bigger circle of/to more users does not help *us* at ALL. It *ONLY* hurts us: 1. More users => more bugreports, while amount of developers stays constant. 2. The more darktable is popular, the bigger the chances of next person to use it being less skilled in at least basic debugging than the average dt user before that. That, unfortunately, even includes uploading /tmp/darktable_bt_XX.txt... Not all users do that :( 3. !!! Every new big chunk of "market" is always a minefield at start, e.g.: https://redmine.darktable.org/issues/10541 That bug is probably very simple, but no one of us uses a distribution with musl libc, so no one can debug it, so it stays there, dormant... 4. Also, did you know that on windows, our automatic backtrace creation would not even work, so even getting backtrace would be harder than usual? Now, imagine, how much of those bugs will be from windows? None of us uses windows, for various reasons, so we are of little help there. Of course, if there is a VERY large demand, one can do it, but ONLY for some money as paid builds. Thus, before darktable can be made available to any new big "market", there needs to be a person, whom we trust to be capable in handling all of the aspects of maintaining darktable there. (For a normal UNIX/BSD/Linux distribution, package maintainer is an obvious answer, in most cases.) So, what exactly do we mean by a maintainer? 0. Be actively involved with darktable for some time. At least half a year. That means: *) Stick with us in IRC. *) actually know the code. *) actively examine bugreports, fix bugs. 1. Be reliable and trustworthy [over that time], and more importantly, make us believe in that. Once 1. and 2. are complete, there is some room to discuss possibility of making official darktable windows port and becoming it's maintainer. IMPORTANT: ANY OTHER APPROACH TO DARKTABLE ON WINDOWS CAN AND WILL BE CONSIDERED HOSTILE, HAPPENING AGAINST OUR WILL AND ACTIVELY HURTING DARKTABLE PROJECT. [1] https://www.darktable.org/2015/07/why-dont-you-provide-a-windows-build/comment-page-1/#comment-46412 [2] https://discuss.pixls.us/t/darktable-on-windows-parthas-build-thread/290 Roman. Also signed by Tobias. ___ darktable developer mailing list to unsubscribe send a mail to darktable-dev+unsubscr...@lists.darktable.org
Re: Darktable on windows. Was: Re: [darktable-dev] Darktable crashes on a SIGILL
Fine, btw building is not trivial if you include all dependencies. I'll see what i can contribute On Thursday 17 December 2015 21:44:34 Roman Lebedev wrote: > And here I thought dt windows bikeshedding was explained to > death[1][2] and over :( > > The problem is *not* building under windows. > The problem is maintaining it. > > What *all* the people trying to bring dt to windows have been doing so > far is called packaging, > at best. As an example, thus far, no one has actually talked to us > directly, did not try to > contribute windows-related fixes/changes for 1.6.x version back into > darktable. > > We are working on darktable in our free time, of pure fun and > interest. There is no money involved. > Thus, darktable, being available for some bigger circle of/to more > users does not help *us* at ALL. > > It *ONLY* hurts us: > 1. More users => more bugreports, while amount of developers stays constant. > 2. The more darktable is popular, the bigger the chances of next > person to use it being less > skilled in at least basic debugging than the average dt user before that. > That, unfortunately, even includes uploading > /tmp/darktable_bt_XX.txt... Not all users do that :( > 3. !!! Every new big chunk of "market" is always a minefield at start, > e.g.: https://redmine.darktable.org/issues/10541 > That bug is probably very simple, but no one of us uses a distribution > with musl libc, so no one can debug it, > so it stays there, dormant... > 4. Also, did you know that on windows, our automatic backtrace > creation would not even work, so > even getting backtrace would be harder than usual? > > Now, imagine, how much of those bugs will be from windows? > None of us uses windows, for various reasons, so we are of little help > there. Of course, if there is a VERY large demand, one can do it, but ONLY > for some money as paid builds. > > Thus, before darktable can be made available to any new big "market", > there needs to be a person, > whom we trust to be capable in handling all of the aspects of > maintaining darktable there. > > (For a normal UNIX/BSD/Linux distribution, package maintainer is an > obvious answer, in most cases.) > > So, what exactly do we mean by a maintainer? > 0. Be actively involved with darktable for some time. At least half a year. > That means: > *) Stick with us in IRC. > *) actually know the code. > *) actively examine bugreports, fix bugs. > 1. Be reliable and trustworthy [over that time], and more importantly, > make us believe in that. > > Once 1. and 2. are complete, there is some room to discuss possibility > of making official darktable windows port and becoming it's > maintainer. > > IMPORTANT: ANY OTHER APPROACH TO DARKTABLE ON WINDOWS CAN AND WILL BE > CONSIDERED HOSTILE, > HAPPENING AGAINST OUR WILL AND ACTIVELY HURTING DARKTABLE PROJECT. > > [1] > https://www.darktable.org/2015/07/why-dont-you-provide-a-windows-build/comm > ent-page-1/#comment-46412 [2] > https://discuss.pixls.us/t/darktable-on-windows-parthas-build-thread/290 > > Roman. > Also signed by Tobias. > ___ > darktable developer mailing list > to unsubscribe send a mail to darktable-dev+unsubscr...@lists.darktable.org ___ darktable developer mailing list to unsubscribe send a mail to darktable-dev+unsubscr...@lists.darktable.org
Re: [darktable-dev] Darktable crashes on a SIGILL
Hello Erik, > I'm looking into getting darktable working on windows, There was already effort in this direction. It would be nice to have a build on Windows indeed. Many photographers I know are still on Windows. Anyway, have you looked at the status of the previous effort? If I remember correctly what has been said it was partly usable. https://plus.google.com/+ParthaBagchi/posts/H2MNQ1GgBck Also make sure the build is done for Win64 and not Win32. There is almost no chance that a Win32 build will work. On Linux we are discouraging people of using 32 bit builds as it makes the experience not really nice. Regards, -- Pascal Obry / Magny Les Hameaux (78) The best way to travel is by means of imagination http://v2p.fr.eu.org http://www.obry.net gpg --keyserver keys.gnupg.net --recv-key F949BD3B ___ darktable developer mailing list to unsubscribe send a mail to darktable-dev+unsubscr...@lists.darktable.org
Re: [darktable-dev] Darktable crashes on a SIGILL
Hi Erik, > No didn't know about Parhta's effort, will contact him to see where > his efforts are at. Joining effort on this project would be nice! Keep us informed. -- Pascal Obry / Magny Les Hameaux (78) The best way to travel is by means of imagination http://v2p.fr.eu.org http://www.obry.net gpg --keyserver keys.gnupg.net --recv-key F949BD3B ___ darktable developer mailing list to unsubscribe send a mail to darktable-dev+unsubscr...@lists.darktable.org
Re: [darktable-dev] Darktable crashes on a SIGILL
Hi Pascal, No didn't know about Parhta's effort, will contact him to see where his efforts are at. Regards, Erik On Tuesday 15 December 2015 21:50:15 Pascal Obry wrote: > Hello Erik, > > > I'm looking into getting darktable working on windows, > > There was already effort in this direction. It would be nice to have a > build on Windows indeed. Many photographers I know are still on > Windows. > > Anyway, have you looked at the status of the previous effort? If I > remember correctly what has been said it was partly usable. > > https://plus.google.com/+ParthaBagchi/posts/H2MNQ1GgBck > > Also make sure the build is done for Win64 and not Win32. There is > almost no chance that a Win32 build will work. On Linux we are > discouraging people of using 32 bit builds as it makes the experience > not really nice. > > Regards, ___ darktable developer mailing list to unsubscribe send a mail to darktable-dev+unsubscr...@lists.darktable.org