On 8 May 2017 at 09:47, gregory grey <ror...@gmail.com> wrote: I can get why are you trying to poke Alexandre on that - you are on an absolute "shoot the messenger" weird thing there.
> In my experience "ain't broke" is a standard mantra of organizations > stuck in the past. > > It's a tent-pole phrase that does not explain anything, and justifies > whatever practices you want it to justify. > > I say waiting 3 weeks for build under some platform is broken as hell, > given you shoot for mutiplatform-ity. So, as a recap - the initial "no." is not the only answer here - Pippin did expand it to a manageable way to try todo something about it, if someone is really willing to. The original "no" reply from Alexandre is the absolute to be expected answer if you jut read this thread subject. "Can you please change ([thesetwof core technologies your project depend on just because they don't have enough bling-ding (github) ] , and then demand on the human-resource-starved, volunteer only project, thatthe change is made or at least started immediately. Sorry, 'no' is among the most polite answer one could get with such a request. > > @Alexandre - what exactly is your point here? I did not build Gimp > from source, ever. > > I'm confident there is nothing in your Makefiles that is beyond human > understanding. > > Again, since you said explicitly you are not making infra decisions, > can I talk to the one/s who do? on the You are already, and you did get answers regarding more practical aspects of a possible cmake move, and a not-so-possible github move. Alexanre was just first to answer - with an answer that correctly roughly mirrored the collective short answer the he and the other collaborators for this project would like to be given. > > For sake of absolutely sane and productive discussion. If you want of course. Do you realise this is a volunteer-maintained project and people involved here will be more happy doing codding and other tasks than "having a productive discussion on whether it would be theoretically better to change to cmake"? You've got a full reply on that - that is: get one to have at least a proof-of-concept build that migrates and simplifies the current process. The existing contributors won't stop what they are doing now because an e-mail not-so-politely labeled "Please migrate to GitHub/Cmake" came along. Try the same e-mail with "I've managed to build GIMP with cmake and could drop these 10000 lines of configuraton files - it already works on Linux and Windows, Mac pending, here is the branch:... " As for github, it is really irrelevant - the needed functionality from the code management aspect is provided by the (free software) git, regardless of the host of the accepted main tree - what people disregard to see is that despite all its popularity, github is just a normal privately owned commercial player - that can change its use-terms, conduct, or go bankrupt any time. (All of which did happen to previous "point of reference" host for free software projects "Source Forge"). We have our servers setup and independently running - and keeping control of the URLs where the project is hosted mean we can change that if need arises. Moreover, anyone can use git functionality to create their forks on github, and make something useful with that - to the point of having patches using that side code-review tools ready to be looked at and reviewed by current GIMP-commiters, and if accepted, all it takes one command on the shell to send that patch in a way it can be officially used in GIMP - in contrast with one of the developers "pressing a button" on github (wow, so much gain for a fancy prison). Now, send more e-mails with personal offenses if you need - or try to get GIMP to build with cmake and showing it is indeed more maintainable. :-) best regards, joao > _______________________________________________ gimp-developer-list mailing list List address: gimp-developer-list@gnome.org List membership: https://mail.gnome.org/mailman/listinfo/gimp-developer-list List archives: https://mail.gnome.org/archives/gimp-developer-list