On 03/12/15 05:52, Vyacheslav Reutskiy wrote: > I'm sorry for commits which don't follow EFL commit guidelines. Totally > we are follow the EFL guidelines, but some merge commit differs, because > project Eflete must to use internal git with Gerrit (it's rules of our R&D > center, > we can't change it). > > And I'm totally agree with David. I subscribed Git ML for see all projects > activity, because I use not only core projects. > I'm not think that need to split GIt ML.
Whatever your internal process is doesn't matter. It's Git, you can edit commits before you upstream, you could fix the offending commit, and more specifically, remove the gerrit ones. -- Tom. > > Viacheslav Reutskiy (rimmed) > > On Thu, Dec 3, 2015 at 2:41 AM, David Seikel <onef...@gmail.com> wrote: > >> On Wed, 2 Dec 2015 18:36:30 -0500 Christopher Michael >> <cpmich...@osg.samsung.com> wrote: >> >>> On 12/02/2015 06:32 PM, Carsten Haitzler wrote: >>>> On Wed, 2 Dec 2015 10:42:43 -0800 Cedric BAIL <cedric.b...@free.fr> >>>> said: >>>> >>>>> Hello, >>>>> >>>>> I am just going to say that Mike proposal is the one that make most >>>>> sense and will be the most efficient for everyone without forever >>>>> discussing every little detail. We could even go as far as >>>>> decomissionning our current mailing list. >>>> >>>> or just leave it as-is and unsubscribe if you dont want it, or use >>>> filters locally or... use phab. >>>> >>>> i don't see any value in centrally imposing some policy on everyone >>>> where everyone will end up disagreeing on what they want to watch >>>> or not. >>>> >>> Yes. Everyone is going to have their own preference >>> >>>> use the git-commits firehose that is "everything" >>>> >>>> btw - our git commits mailing lists runs on e.org - not on sf.net >>>> infra. so sf is irrelevant here. >>>> >>>> i don't see the point of decommissioning the list - it's useful and >>>> works and takes effort to decommission. as above. use phab or local >>>> filters and filter as you like. >>>> >>> >>> +2. I'd have to agree here. Why change something that has worked for >>> years ? I am sure just about every email client has filters by >>> now...set some up ;) >>> >>> dh >> >> I'm in agreement here with Mike, Cedric, Raster, and DH. Leave things >> as they are, if you have a problem, build better filters in your email >> client, or switch to using phab instead of the mailing list. Some of >> us like the firehose just like it is. B-) >> >>>>> Cedric >>>>> >>>>> On Wed, Dec 2, 2015 at 8:56 AM, Mike Blumenkrantz >>>>> <michael.blumenkra...@gmail.com> wrote: >>>>>> My preference would be to do away with the commit mailing list >>>>>> entirely. It's easy to set a follow rule on phab for projects >>>>>> that you're interested in, and this has the added benefit of >>>>>> reducing our reliance on the failure-prone sourceforge >>>>>> infrastructure. Moreover, it would avoid future arguments over >>>>>> whether a project is important enough to be on the list. >>>>>> >>>>>> As an example of how easy this is: >>>>>> 1) go to https://phab.enlightenment.org/herald/new/ (accessed >>>>>> through the applications link on the left panel -> herald -> >>>>>> create new rule) 2) click commit (top option) >>>>>> 3) click continue >>>>>> 4) click personal (top option) >>>>>> 5) click continue >>>>>> 6) make up a rule name in the top field >>>>>> 7) change condition to "Repository" and "is any of" >>>>>> 8) click magnifying glass on right to add whatever projects you're >>>>>> interested in >>>>>> 9) change action to "send me an email" >>>>>> 10) click save rule >>>>>> >>>>>> You now have a rule which notifies you about your repositories of >>>>>> interest and can be changed at any time. To only receive mails >>>>>> for "master" branch commits, simply add another condition with >>>>>> "branches" "contains" "master". >>>>>> >>>>>> Furthermore, mails from phab will contain links to the commit >>>>>> audit on phab, a useful tool which we do not make enough use of. >>>>>> Not only does it allow inline reviewing of commits along with >>>>>> direct ticket/diff referencing, it also will directly mail the >>>>>> associated author of the audited commit--something which the >>>>>> commit list cannot automatically do. >>>>>> >>>>>> Lastly, using the audit method allows developers to set more >>>>>> herald rules to automatically add themselves to audited commits, >>>>>> meaning that any time a review is started they will receive a >>>>>> mail, similar to the current review workflow. Unlike the current >>>>>> workflow, however, a developer can also create specialized rules >>>>>> to automatically begin an audit session (with notification) when >>>>>> certain criteria are met, eg. specific files which the developer >>>>>> maintains are modified by someone else. To test this, simply >>>>>> follow the above steps but use "add me as an auditor" instead >>>>>> of/in addition to "send me an email". This can be further >>>>>> restricted by adding another condition to the rule which >>>>>> specifies "change conditions", allowing pruning based on commit >>>>>> size, content, and specific files. >>>>>> >>>>>> Sure, this is a little more work than clicking subscribe, then >>>>>> waiting for the confirmation mail, then clicking the confirm >>>>>> link, then clicking the confirm button after maybe thinking of a >>>>>> throwaway password, but I think the benefits are worthwhile >>>>>> enough to make this a project standard. >>>>>> >>>>>> On Wed, Dec 2, 2015 at 9:55 AM Stephen Houston >>>>>> <smhousto...@gmail.com> wrote: >>>>>> >>>>>>> I think leaving it like it is, or number two, would be the right >>>>>>> option. The commit ml is the easiest way for me to quickly get >>>>>>> an idea of what is going on across git, not just the core >>>>>>> projects. "Oh look, edi added faster syntax highlighting... >>>>>>> Sweet!". >>>>>>> >>>>>>> On Wed, Dec 2, 2015 at 8:39 AM, Tom Hacohen >>>>>>> <t...@osg.samsung.com> wrote: >>>>>>> >>>>>>>> Hey, >>>>>>>> >>>>>>>> I'm sending this email to raise an issue that has been annoying >>>>>>>> me for the last few months, but especially in the last couple >>>>>>>> of weeks. >>>>>>>> >>>>>>>> As it stands the Git ML is cluttered and it's very annoying for >>>>>>>> me to review commits that are of relevance to me. The problem, >>>>>>>> is that in the Git ML we send emails for all of the commits in >>>>>>>> all of the repos in git.e.org. This includes some more niche >>>>>>>> projects. This means I get hundreds of commits a week that I >>>>>>>> don't care about, and this number will only grow once more >>>>>>>> projects are added. To make matters even worse, those projects >>>>>>>> don't follow the EFL commit guidelines and have authors like >>>>>>>> "Gerrit <xxxx>" or just merge commits. >>>>>>>> >>>>>>>> I could solve it locally, by filtering out the main offenders >>>>>>>> according to the repository name (we pass it in the header), >>>>>>>> though I suspect I'm one of many to be annoyed by it (I already >>>>>>>> know I'm not the only one). >>>>>>>> >>>>>>>> I think it's time to tackle that. I came out with two >>>>>>>> alternatives: 1. Only send commit emails for "core" efl >>>>>>>> projects. 2. Split the git ML to two MLs, "core" and "extra". >>>>>>>> >>>>>>>> By core projects I mean: e, efl (+ loaders), elm and >>>>>>>> terminology. Maybe also any other project that is developed by >>>>>>>> more than a few efl developers and follows our guidelines. >>>>>>>> >>>>>>>> In addition, I recommend all the projects to follow the E commit >>>>>>>> guidelines to make the commit history more manageable. This >>>>>>>> will both lessen this annoyance and improve the commit history >>>>>>>> of the relevant projects. >>>>>>>> >>>>>>>> Please let me know what you think. >>>>>>>> >>>>>>>> -- >>>>>>>> Tom. >>>>>>>> >>>>>>>> >>> >>> >>> >>> >> ------------------------------------------------------------------------------ >>> Go from Idea to Many App Stores Faster with Intel(R) XDK >>> Give your users amazing mobile app experiences with Intel(R) XDK. >>> Use one codebase in this all-in-one HTML5 development environment. >>> Design, debug & build mobile apps & 2D/3D high-impact games for >>> multiple OSs. >>> http://pubads.g.doubleclick.net/gampad/clk?id=254741911&iu=/4140 >>> _______________________________________________ enlightenment-devel >>> mailing list enlightenment-devel@lists.sourceforge.net >>> https://lists.sourceforge.net/lists/listinfo/enlightenment-devel >> >> >> -- >> A big old stinking pile of genius that no one wants >> coz there are too many silver coated monkeys in the world. >> >> >> ------------------------------------------------------------------------------ >> Go from Idea to Many App Stores Faster with Intel(R) XDK >> Give your users amazing mobile app experiences with Intel(R) XDK. >> Use one codebase in this all-in-one HTML5 development environment. >> Design, debug & build mobile apps & 2D/3D high-impact games for multiple >> OSs. >> http://pubads.g.doubleclick.net/gampad/clk?id=254741911&iu=/4140 >> _______________________________________________ >> enlightenment-devel mailing list >> enlightenment-devel@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/enlightenment-devel >> >> > ------------------------------------------------------------------------------ > Go from Idea to Many App Stores Faster with Intel(R) XDK > Give your users amazing mobile app experiences with Intel(R) XDK. > Use one codebase in this all-in-one HTML5 development environment. > Design, debug & build mobile apps & 2D/3D high-impact games for multiple OSs. > http://pubads.g.doubleclick.net/gampad/clk?id=254741911&iu=/4140 > _______________________________________________ > enlightenment-devel mailing list > enlightenment-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/enlightenment-devel > ------------------------------------------------------------------------------ Go from Idea to Many App Stores Faster with Intel(R) XDK Give your users amazing mobile app experiences with Intel(R) XDK. Use one codebase in this all-in-one HTML5 development environment. Design, debug & build mobile apps & 2D/3D high-impact games for multiple OSs. http://pubads.g.doubleclick.net/gampad/clk?id=254741911&iu=/4140 _______________________________________________ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
