(taking this off the mercurial list) On 02.01.2010 19:20, Steve Borho wrote: > On Sat, Jan 2, 2010 at 12:16 PM, Adrian Buehlmann <[email protected]> wrote: >> On 02.01.2010 18:10, Steve Borho wrote: >>> On Sat, Jan 2, 2010 at 6:17 AM, Adrian Buehlmann <[email protected]> wrote: >>>> On 02.01.2010 03:41, Steve Borho wrote: >>>>> TortoiseHg 0.9.2 is a bug fix release. >>>>> >>>>> http://bitbucket.org/tortoisehg/stable/wiki/ReleaseNotes >>>>> >>>>> http://bitbucket.org/tortoisehg/stable/downloads/tortoisehg-0.9.2-hg-1.4.2.exe >>>>> >>>> >>>> Probably will need a 0.9.2.1: >>>> >>>> http://bitbucket.org/tortoisehg/stable/issue/839/092-installer-doesnt-call-tortoiseoverlays >>> >>> Thanks for catching this Adrian. I've uploaded a new installer: >>> >>> http://bitbucket.org/tortoisehg/stable/downloads/tortoisehg-0.9.2-1-hg-1.4.2.exe >>> >>> If you link to our packages, please update your URLs. Thanks. >>> >> >> I assume this was built using 275690997c5f, which is currently untagged. >> >> Last time you appended '.1', now it is '-1' and without tagging. > > Last time, I changed code that affected the tarball so a new tag was > necessary. This time around, it just needed an ISS change so I just > rebuilt the installer.
Sure. But you changed code here as well. Code that is executed at install time on windows. The delivery item for Windows is an exe that installs TortoiseHg. Treating this case here (changing installer code) special by not following preexisting release patterns (build release and tag it) doesn't make that much sense IMHO. The .iss source that was changed for this special '-1' "release" is even in the same repository as the sources for the hgtk dialogs of TortoiseHg. In the past we could update the main TortoiseHg repo to a tag (e.g. 0.8) and we got the versions of the .iss sources that were used to build that exact release (because the .iss sources are part of the main TortoiseHg repo). This deviation in release process at least breaks that invariant. Consider a fictitious dialog among developers in some distant future: "Oh, 0.9.2-1 was special, I remember we fixed the .iss files after 0.9.2 for that but we didn't tag it" -- "Oh, and what versions of the iss files were taken for that?" -- "Not sure, probably that change in the stable branch directly after the tagging cset for 0.9.2, but I can't tell for sure anymore". Not exactly the way we are supposed to use Mercurial, isn't it? ------------------------------------------------------------------------------ This SF.Net email is sponsored by the Verizon Developer Community Take advantage of Verizon's best-in-class app development support A streamlined, 14 day to market process makes app distribution fast and easy Join now and get one step closer to millions of Verizon customers http://p.sf.net/sfu/verizon-dev2dev _______________________________________________ Tortoisehg-discuss mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/tortoisehg-discuss
