W dniu 05.05.2013 19:25, Øyvind Harboe pisze: > Hi Peter, > > I've discovered that it takes a *LONG* time before teams learn that you > *CAN* have the master branch be releasable at all times if you don't > accept unfinished work or work that doesn't move the product forward. > > We *do* have the tools and methods to keep track of unfinished work in > progress(Gerrit in our case), but this didn't used to be the case and > old habits die hard. > > I don't know what regressions that we've picked up in 0.7.0, but we > should strive not to submit code that makes OpenOCD worse and then > submit 0.7.1 as soon as we have the regressions compared to 0.6.0 fixed. > > In the case of a regression submitted from Gerrit, that does not have to > be a problem if we revert it *soon* and then simply resubmit a new > change with the original unfinished work. > > Can we revert the changes that went into 0.7.0 and resubmit them to Gerrit?
I'm confused... This discussion is about patches that are NOT in 0.7.0 (fixes for building OpenOCD for ARM platform) that I did not merge because they were posted AFTER rc2 was already done (less than a week ago), and only one of them (3 in total?) is reviewed... Why would you want to remove something from 0.7.0 then? Or maybe that's just a theoretical question? As for the Debian "problem"... The maintainer of the package is - from what I remember - a member of this list and OpenOCD contributor. If there is a problem why not speak earlier - the idea of release was around for over a month, stable package in Debian is 0.5.0 which was released over a year ago so that's not a new problem and it was NEVER mentioned. If it's low priority for Debian folks why should it be top priority for OpenOCD? Why should OpenOCD change its policies because Debian has some strange policy about non-x86 platforms? Yes I know that it's not Debian who should be changing, but as this is a low-priority matter for Debian I think it should not be The-Most-Important-Thing-To-Fix in OpenOCD, so important to change the release schedule, ongoing process and so on. If users of Debian are fine with having 0.5.0 as stable and testing, then I guess they can wait for 0.8.0. If the matter is so urgent to warrant 0.7.1 why it's mentioned NOW, not after 0.6.1 (released 7 months ago) failed to meet Debian's policies? Not to mention 0.7.0-rc1 and 0.7.0-rc2 which were done specifically for that purpose... 4\/3!! ------------------------------------------------------------------------------ Get 100% visibility into Java/.NET code with AppDynamics Lite It's a free troubleshooting tool designed for production Get down to code-level detail for bottlenecks, with <2% overhead. Download for free and get started troubleshooting in minutes. http://p.sf.net/sfu/appdyn_d2d_ap2 _______________________________________________ OpenOCD-devel mailing list OpenOCD-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/openocd-devel
