On Jan 14, 2008, at 1:37 PM, Johannes Schindelin wrote:
On Sun, 13 Jan 2008, Steffen Prohaska wrote:
The criteria would be changed as follows:
--- snip ---
We release a stable version if:
* commands from "plumbing" and "core porcelain" equivalent to
official git are available in Git Bash and implemented in C.
(A list of commands would be added here.)
That would take too long. IMHO having the whole merge stuff ported
to C
will need at least another 6 months.
So you propose to release a stable version based on shell
scripts? Maybe we should be patient until we have a stable
msysgit release and push harder to have the C port done ASAP?
* [DONE] git-gui works if run from Git Bash.
* [DONE] git-gui works if run from Start Menu.
* [DONE] git and git-gui available from Windows Command
Prompt (cmd shell).
* other command, such as "mail porcelain", "import/export",
"admin", "server" are *not* required, though commands from
these groups may be included.
IMHO git-svn is a _must_. (Although I can be talked out of it.)
My status so far: I got perl-5.8.8 compiled, subversion, too, but
the perl
bindings are a bit strange. For one, the Makefiles are written to
somewhere completely bogus, and then they do not work properly.
I can't say anything about this. I never used git-svn and I do
not plan to use it. Basically, I skipped svn. I tried it for
some time and wasn't really convinced by its advantages over CVS.
So I hibernated using CVS until eventually I found git.
* git-cheetah is *not* required.
Yes, git-cheetah is nowhere near the state I would like to see it.
Unfortunately, the only developer I could get interested so far
seems to
have lost interest, and I do not have enough time to take care of it.
FWIW I think that we are stable enough already to have better-than-
preview
releases (although, like I said, lack of git-svn is a reason for me
not to
declare non-beta status yet).
Currently, we install git-svn even that we know it won't work.
This is certainly not the type of installer I'd like to declare
beta.
Maybe we could clearly state what we have and exclude all
commands that do not work from the installer. If we only
installed stable commands and give a clear indication on the
perspective for the missing commands, maybe we could do a beta
release for plumbing and core porcelain. The commands that
are not ready would still be included in msysgit. So it would
be easy for developers to get and improve what we already have.
But end users would not see commands that do not work.
From my perspective, more important than getting git-svn is to
merge the core porcelain back to official git. I'd like to be
able to build the msysgit release from official git ASAP. After
the 1.5.4 release, there's hopefully a time window to make some
progress on this. Explicitly excluding commands from the msysgit
end user release could also be a way to approach this goal more
quickly.
Steffen
