Re: Development and Release Model

2011-05-03 Thread Pavel Sanda
Jean-Marc Lasgouttes wrote: > One nice way of doing that would be to provide nightlies of branch for the > main architectures. I wonder how difficult it would be (depending on the > architecture, of course). gentoo already provides this. since is source based distro, instead of using nightly tar

Re: Development and Release Model

2011-05-03 Thread José Matos
On Monday 02 May 2011 17:37:12 Richard Heck wrote: > I'd be happy to make rpms, if I knew how to do so > > Richard I can help if you want to. :-) -- José Abílio

Re: Development and Release Model

2011-05-02 Thread Jürgen Spitzmüller
Jean-Marc Lasgouttes wrote: > Ubuntu ppas would be great too. The advantage is that people update > without even thinking about it (for those who wish to, of course!) As far as openSuse is concerned, something similar could be setup easily (and maybe even more or less automated) via the Build Se

Re: Development and Release Model

2011-05-02 Thread Jean-Marc Lasgouttes
Le 02/05/2011 18:33, Stephan Witt a écrit : One nice way of doing that would be to provide nightlies of branch for the main architectures. I wonder how difficult it would be (depending on the architecture, of course). For Mac I'd be able to automate that from the checkout to the upload of pac

Re: Development and Release Model

2011-05-02 Thread Sam Lewis
--- On Mon, 2/5/11, Jürgen Spitzmüller wrote: > From: Jürgen Spitzmüller > Subject: Re: Development and Release Model > To: lyx-devel@lists.lyx.org > Date: Monday, 2 May, 2011, 14:43 > Richard Heck wrote: > > It's important that we have a culture that promotes >

Re: Development and Release Model

2011-05-02 Thread Richard Heck
On 05/02/2011 12:20 PM, Jean-Marc Lasgouttes wrote: > Le 02/05/2011 15:43, Jürgen Spitzmüller a écrit : >> Sounds like a good idea in general, one which might even strengthen >> branch's >> (i.e., branch-20x/stable's) stability. The only concern I have is >> that this >> requires that people active

Re: Development and Release Model

2011-05-02 Thread Stephan Witt
Am 02.05.2011 um 18:20 schrieb Jean-Marc Lasgouttes: > Le 02/05/2011 15:43, Jürgen Spitzmüller a écrit : >> Sounds like a good idea in general, one which might even strengthen branch's >> (i.e., branch-20x/stable's) stability. The only concern I have is that this >> requires that people actively u

Re: Development and Release Model

2011-05-02 Thread Jean-Marc Lasgouttes
Le 02/05/2011 15:43, Jürgen Spitzmüller a écrit : Sounds like a good idea in general, one which might even strengthen branch's (i.e., branch-20x/stable's) stability. The only concern I have is that this requires that people actively use and test branch (as opposed to trunk and released versions).

Re: Development and Release Model

2011-05-02 Thread Jürgen Spitzmüller
Jürgen Spitzmüller wrote: > > Regarding branch testing, I have always used BRANCH_1_X_X for my daily > > work. I'd be curious what other of us do. > > Me, too, obviously (and I will keep on doing that). That is to say now, BRANCH_2_0_X, of course. Jürgen

Re: Development and Release Model

2011-05-02 Thread Jürgen Spitzmüller
Richard Heck wrote: > I would certainly want to keep an eye on how much chance the refactoring > had of introducing regression bugs. Seeing it as a separate and > completed piece would make that much easier, as it could even be tested > independently. If the regressions are nice enough to unhide t

Re: Development and Release Model

2011-05-02 Thread Richard Heck
On 05/02/2011 09:44 AM, Jürgen Spitzmüller wrote: > Abdelrazak Younes wrote: >> One: merging will be difficult when a feature has been cooked based on >> master. The main reason is that master will see a lot code refactoring >> and cleanup. So, unless you accept that "code refactoring and cleanup

Re: Development and Release Model

2011-05-02 Thread Richard Heck
On 05/02/2011 09:26 AM, Abdelrazak Younes wrote: > On 02/05/2011 15:12, Richard Heck wrote: >> I'd like to start a discussion not so much about git and branches and >> stuff, but more about what our release pattern will be and how that will >> be managed. It's inspired in part by some worries Pavel

Re: Development and Release Model

2011-05-02 Thread Jürgen Spitzmüller
Abdelrazak Younes wrote: > One: merging will be difficult when a feature has been cooked based on > master. The main reason is that master will see a lot code refactoring > and cleanup. So, unless you accept that "code refactoring and cleanup" > branches get also merged into stable branch (which

Re: Development and Release Model

2011-05-02 Thread Jürgen Spitzmüller
Richard Heck wrote: > It's important that we have a culture that promotes LyX's stability. I could not agree more on this. Stability is one of the, if not the, most strongest point we have. > So I would propose to maintain two versions of BRANCH_2_0_X, much as the > release maintainer, under th

Re: Development and Release Model

2011-05-02 Thread Abdelrazak Younes
On 02/05/2011 15:12, Richard Heck wrote: I'd like to start a discussion not so much about git and branches and stuff, but more about what our release pattern will be and how that will be managed. It's inspired in part by some worries Pavel expressed: i have been thinking about the proposed chan