>>> "David PONCE" <David.Ponce@...> seems to think that:
>> Being a release manage just means that you would guess when a good
>> time to release is, make sure everyone has stuff they want in (you and
>> me mostly). Next, fix the version numbering (in EDE) and updating the
>> changelog file. (I use Emacs's command, `vc-update-change-log'. Next
>> make the .tar.gz files (I use EDE for this). I often use EDE to
>> rebuild the Makefiles first too. Next, test the tar file out on a few
>> platforms (I usually do Emacs and XEmacs on Linux, but testing on
>> windows is good too) then upload them to the web site and post a few
>> useful places. EDE has upload capability now too.
>So I need to [learn how to] use EDE ;-)
Not strictly necessary. Using it both saves me time, and slows me
down when I have to fix it. If you don't like it, we could switch to
managing the makefiles or automake files by hand. I happen to like
the convenience of having all my build and configuration commands on
>> The part that slows me down is testing things out across the
>> platforms, and also synchronizing releases between semantic and eieio
>> which are often bound together.
>I can test releases with NT Emacs 21.4 (CVS), NT Emacs 20.7 and NT
>XEmacs 21.4.10 on my NT box. I also have Emacs 21.4 (CVS) installed
>on my GNU/Linux box (RH 7.1) at home. May be I could install Emacs
>21.2 on both boxes to test with the official release of Emacs too.
Wow, that is a much wider range of options than I have. That would
>> At the moment, the 1.4 branch needs to be released as 1.4.1. It has
>> fixes for the common startup problem, in addition to your fix for the
>> recursive auto-update thing.
>I completely agree. For now I just have the recursive fix applied on
>my local 1.4 version (tagged "v1p4" in CVS). We probably need to
>commit our changes in the "v1p4" branch and maybe create a new
>"v1p4r1" branch (what is the meaning of the letter "p"=3F I think "v"
>is for "version" and I propose "r" for "revision")
In "v1p4" or "v1p4p1" the "p" stands for "point". The version number
1.4 is read in English (or at least in Massachusetts English") as
"one point four". I suppose it could also read "one period four",
but that wasn't what I was thinking.
The reason for the "p" is that CVS didn't allow me to use a ".".
I have already checked some files in on the v1p4 branch.
[ ... ]
>I also switched semantic-c.el to use the new common grammar framework
>through c.by. I attached a tarball containing the new c.by and
>semantic-c.el generated from it. Here is the change log:
[ ... ]
>I did some basic tests with C/C++ files in semantic/tests and all
>seems to work as with the previous "BNF version". It would be nice if
>you could do some more test with the attached semantic-c.el file.
>Then if it works well for you, maybe we could definitively switch the
>C/C++ stuff to use the new grammar framework=3F
[ ... ]
That's cool. I will check it out next week when I get back from my
charity pumpkin flinging contest. I will also show you all some cool
pictures of my flying pumpkins.
Eric Ludlam: zappo@..., eric@...
Home: http://www.ludlam.net Siege: http://www.siege-engine.com
Emacs: http://cedet.sourceforge.net GNU: http://www.gnu.org