Re: [Development] Branching 5.0

On segunda-feira, 3 de dezembro de 2012 22.26.19, João Abecasis wrote: > > Let's try to just push fixes for things that will block the release > > , > > Isn't this the definition of the "release" branch? If we are in > release mode, if we only w

Re: [Development] Branching 5.0

Sergio Ahumada wrote: > So as far as I understand, we should merge 'master' into 'dev' at some > point (twice a week?) and then delete the 'master' branch at some point. > This is mainly because we don't want to take non-blockers into 'stable'. > > For those who need something into 'stable' which i

Re: [Development] Branching 5.0

On segunda-feira, 3 de dezembro de 2012 10.12.44, Giuseppe D'Angelo wrote: > Is the new three-branches system documented somewhere on the wiki? I don't know if anyone copied the contents of the email to it. I know I didn't. Just in case it's necessary, the images I had attached were produced by

Re: [Development] Branching 5.0

On Mon, Dec 03, 2012 at 12:43:17PM +0100, Ahumada Sergio wrote: > On 12/03/2012 03:46 AM, Knoll Lars wrote: > > On Dec 2, 2012, at 8:55 AM, Thiago Macieira > > wrote: > > > >> On domingo, 2 de dezembro de 2012 14.52.12, Knoll Lars wrote: > Most of the pending changes in master today are fix

Re: [Development] Branching 5.0

On 12/03/2012 03:46 AM, Knoll Lars wrote: > On Dec 2, 2012, at 8:55 AM, Thiago Macieira > wrote: > >> On domingo, 2 de dezembro de 2012 14.52.12, Knoll Lars wrote: Most of the pending changes in master today are fixes that need to go into the 5.0 release, so the branch should be merge

Re: [Development] Branching 5.0

On 28 November 2012 19:47, Knoll Lars wrote: > > Qt 5.0 is getting closer, and we're still working to get the final release > out before the end of the year. To make this easier and also allow new > development towards 5.1 to happen again, we'll branch the qt repositories > during this weekend

Re: [Development] Branching 5.0

On Dec 2, 2012, at 8:55 AM, Thiago Macieira wrote: > On domingo, 2 de dezembro de 2012 14.52.12, Knoll Lars wrote: >>> Most of the pending changes in master today are fixes that need to go into >>> the 5.0 release, so the branch should be merged to stable, not dev. >> >> I have been seeing lot

Re: [Development] Branching 5.0

On domingo, 2 de dezembro de 2012 14.52.12, Knoll Lars wrote: > > Most of the pending changes in master today are fixes that need to go into > > the 5.0 release, so the branch should be merged to stable, not dev. > > I have been seeing lots of pending changes in master that are 5.1 material > and

Re: [Development] Branching 5.0

On Dec 1, 2012, at 11:20 AM, Thiago Macieira wrote: > On quarta-feira, 28 de novembro de 2012 19.47.47, Knoll Lars wrote: >> Transitional only. Will be closed for new pushes once the branches get >> created. Patches that are already in gerrit can still be staged for around >> a week to allow eve

Re: [Development] Branching 5.0

On quarta-feira, 28 de novembro de 2012 19.47.47, Knoll Lars wrote: > Transitional only. Will be closed for new pushes once the branches get > created. Patches that are already in gerrit can still be staged for around > a week to allow everybody to stage their pending changes. Master will then > ge

Re: [Development] Branching 5.0

Hi, > … this time for real… :) The branching has been finished for the following projects: - qt5 - qtactiveqt - qtbase - qtdeclarative - qtdoc - qtgraphicaleffects - qtimageformats - qtjsbackend - qtmultimedia - qtquick1 - qtscript - qtsvg - qttools - qttranslations - qtwebkit - qtwebkit-example

Re: [Development] Branching 5.0

On Sat, Dec 1, 2012 at 12:51 AM, Thiago Macieira wrote: > On sexta-feira, 30 de novembro de 2012 11.43.19, Laszlo Papp wrote: > > Perhaps, I was unclear. I think he raised a valid concern about the > > "stable" name, but that does not mean I support "testing". > > No, he didn't. His logic was flaw

Re: [Development] Branching 5.0

> From: d3fault > Sent: Friday, November 30, 2012 11:25 PM > Subject: Re: [Development] Branching 5.0 > > On 11/30/12, Thiago Macieira wrote: >> He says that the converse usually holds true. First of all, converses >> usually >> do not hold true. Condi

Re: [Development] Branching 5.0

On 11/30/12, Thiago Macieira wrote: > He says that the converse usually holds true. First of all, converses > usually > do not hold true. Conditions that are both necessary and sufficient are the > > exception, not the rule. Second, it does not hold in this case either. > I never said most/all co

Re: [Development] Branching 5.0

On sexta-feira, 30 de novembro de 2012 11.43.19, Laszlo Papp wrote: > Perhaps, I was unclear. I think he raised a valid concern about the > "stable" name, but that does not mean I support "testing". No, he didn't. His logic was flawed (as often his arguments are). He said: "A release is implicitl

Re: [Development] Branching 5.0

> > Sure he did, but in the context of naming a concern is easy to respond to > with > a counter argument. For example it makes the other branches look like > they're > not subject to testing. Perhaps, I was unclear. I think he raised a valid concern about the "stable" name, but that does not me

Re: [Development] Branching 5.0

On Wed, Nov 28, 2012 at 07:47:47PM +, Knoll Lars wrote: > * master: > > Only binary compatible changes are allowed into this branch, as it'll be > merged to dev later on. > i would suggest that the assumed merge target is "stable" until a weeks or two after the branching - many already poste

Re: [Development] Branching 5.0

On Friday, November 30, 2012 05:38:22 AM Laszlo Papp wrote: > On Thu, Nov 29, 2012 at 8:03 AM, Knoll Lars wrote: > > I know that branch naming is a possible topic for endless bike shedding. > > You can always find arguments against a certain name here. In the end it's > > not the most important th

Re: [Development] Branching 5.0

On Nov 30, 2012, at 6:38 AM, Laszlo Papp mailto:lp...@kde.org>> wrote: On Thu, Nov 29, 2012 at 8:03 AM, Knoll Lars mailto:lars.kn...@digia.com>> wrote: I know that branch naming is a possible topic for endless bike shedding. You can always find arguments against a certain name here. In the end

Re: [Development] Branching 5.0

dev, testing and release are pretty standard in most open source communities (including every major linux flavor) and would avoid the mass confusion that will result if you use dev, stable and release. On 30/11/2012 5:10 PM, Anttila Janne wrote: > Laszlo Papp wrote: > >>> On Thu, Nov 29, 2012 at

Re: [Development] Branching 5.0

Laszlo Papp wrote: >> On Thu, Nov 29, 2012 at 8:03 AM, Knoll Lars wrote: >> >> >> I know that branch naming is a possible topic for endless bike >> shedding. You can always find arguments against a certain name here. In >> the end it's not the most important thing whether it's called testing or

Re: [Development] Branching 5.0

On Thu, Nov 29, 2012 at 8:03 AM, Knoll Lars wrote: > I know that branch naming is a possible topic for endless bike shedding. > You can always find arguments against a certain name here. In the end it's > not the most important thing whether it's called testing or stable. It's > the meaning that

Re: [Development] Branching 5.0

On 11/29/12, Knoll Lars wrote: > But to shorten this: We've had that discussion a couple of months ago, and > the names mentioned are the ones we in the end agreed to. That discussion was more about the process (which I admit is more important) than the names. I raised the issue then too but got

Re: [Development] Branching 5.0

On 28.11.2012 20:47, Knoll Lars wrote: > We will have the following policies on the branches: > > * master: > > Transitional only. Will be closed for new pushes once the branches get > created. Patches that are already in gerrit can still be staged for around a > week to allow everybody to stage

Re: [Development] Branching 5.0

On a slightly unrelated note... >Because this should be up to module maintainers and there is not so much CI infrastructure to support all the projects under Gerrit Who are the module maintainers, these days? I really don't think that http://qt-project.org/wiki/Maintainers is up to date. Sorry f

Re: [Development] Branching 5.0

On 29/11/2012, at 6:05 PM, Knoll Lars wrote: > > On Nov 29, 2012, at 1:31 AM, Lorn Potter wrote: > >> >> On 29/11/2012, at 5:47 AM, Knoll Lars wrote: >> >>> If maintainers for other repositories would like to get the same branches >>> setup, please speak up and tell Sergio and Janne. Both

Re: [Development] Branching 5.0

>> If maintainers for other repositories would like to get the same branches >> setup, please speak up and tell Sergio >> and Janne. Both have agreed to do most of the actual work of creating the >> branches in gerrit and setting up the >> CI system and a big thanks goes to them already now. >>

Re: [Development] Branching 5.0

On Nov 29, 2012, at 1:31 AM, Lorn Potter wrote: > > On 29/11/2012, at 5:47 AM, Knoll Lars wrote: > >> If maintainers for other repositories would like to get the same branches >> setup, please speak up and tell Sergio and Janne. Both have agreed to do >> most of the actual work of creating

Re: [Development] Branching 5.0

I know that branch naming is a possible topic for endless bike shedding. You can always find arguments against a certain name here. In the end it's not the most important thing whether it's called testing or stable. It's the meaning that we as a community associate with the branch name that's im

Re: [Development] Branching 5.0

I agree with d3fault on this one. On 29/11/2012 1:07 PM, d3fault wrote: > "always close to releasable" = testing > > -dev > -testing > -release or stable > > eradicates any chance of confusion, regardless of who is "right" > > > d3fault > ___ > Developme

Re: [Development] Branching 5.0

"always close to releasable" = testing -dev -testing -release or stable eradicates any chance of confusion, regardless of who is "right" d3fault ___ Development mailing list Development@qt-project.org http://lists.qt-project.org/mailman/listinfo/devel

Re: [Development] Branching 5.0

On Thu, Nov 29, 2012 at 9:58 AM, d3fault wrote: > The words "stable" and "release" are somewhat ambiguous (enough to > warrant change). > > A release is implicitly stable, so the converse usually also holds > true: stable is released. Not really. IMO, stable should be an "always close to releasab

Re: [Development] Branching 5.0

On 29/11/2012, at 5:47 AM, Knoll Lars wrote: > If maintainers for other repositories would like to get the same branches > setup, please speak up and tell Sergio and Janne. Both have agreed to do most > of the actual work of creating the branches in gerrit and setting up the CI > system and a

Re: [Development] Branching 5.0

The words "stable" and "release" are somewhat ambiguous (enough to warrant change). A release is implicitly stable, so the converse usually also holds true: stable is released. d3fault ___ Development mailing list Development@qt-project.org http://list

[Development] Branching 5.0

… this time for real… :) Qt 5.0 is getting closer, and we're still working to get the final release out before the end of the year. To make this easier and also allow new development towards 5.1 to happen again, we'll branch the qt repositories during this weekend in preparation for the Release