dev-serveez
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[dev-serveez] extra release step


From: Thien-Thi Nguyen
Subject: [dev-serveez] extra release step
Date: Fri, 27 May 2011 17:45:11 +0200

Previously, a "pre-configured" set of files were maintained in the
repository for the sake of mingwin (a .bat file would copy them to
the target file to fake a "configure" invocation).  This meant that
a "make dist" from any host was sufficient to create a tarball
usable on all target hosts.

As of 2011-02-06 (commit bdd9ef43), these files are now generated on
"make dist".  This means that the only way to generate a valid (for
mingwin) set of files is to do the "make dist" from a cygwin host,
presuming that the cygwin configuration would be usable by mingwin.

All of this is outside my expertise (and interest, TBH), so i'm
trying to decide what to do for a release, and furthermore what to
do henceforth.  As i see it, everything depends on how involved the
original Serveez hackers want to be going forward.

Here are the two courses:

- Nobody cares any more about mingwin.
  => Great!  I'll gladly fully decruft and carry on "normally".
     By "fully", i mean remove ALL infrastructure and code.
     By "gladly", i mean with unicorns dancing and bird singing...

- Someone can muster a cygwin build to provide the template files.
  => OK, i suggest the procedure:
       - ttn does "normal" (but invalid for mingwin) "make dist"
         under GNU/Linux
       - someone else unpacks this provisional tarball,
         builds/tests on cygwin then does "make dist" again
       - we publish THAT tarball as the official one
     If this procedure is not palatable, what others would do?
     Another idea is try this for 0.2.x, and then drop mingwin
     completely for 0.3.x.

What do people think?



reply via email to

[Prev in Thread] Current Thread [Next in Thread]