Cygwin socket library status enquiry

[Steve Holden]

"Steve Holden" <sholden at holdenweb.com> wrote in message
news:Vl1U7.38055$PA.31332 at atlpnn01.usenetserver.com...
> [posted & mailed]
> "Jason Tishler" <jason at tishler.net> wrote ...
> > On Tue, Dec 18, 2001 at 08:52:20PM +0000, Michael Hudson wrote:
> > > "Steve Holden" <sholden at holdenweb.com> writes:
> > > > I'll see what I need to omit to get it to build, and follow up if I
> > > > make progress.
> > >
> > > You need to build _socket statically.
> >
> > Please try the following instead:
> >
> >     http://cygwin.com/ml/cygwin/2001-12/msg00894.html
> >
> > and report back to the list.
> >
> [Jason's URL specifies rebasing a number of Cygwin and Python DLLs to
> overcome some known bugs in the Cygwin build]
>
> OK, I got the build to work. Win98 didn't have all the same libraries, so
> I'd recommend that the rebase tool silently, or with a simple comment,
> ignores file arguments it can't find. But it did rebase everything it
found.
>
> I'm getting conflicting results on the tests. I omitted test_hotshot,
which
> seems to hang, and test_select (because it's a known troublemaker). I
> actually got through to the end of the test cycle and restarted on the
> compiled code, and was in the middle of a post to explain about the five
> tests which failed, when test_popen2 hung my machine. Solid, not even the
> three-fingered salute would break in.
>
> I had hoped to provide a test output with the summary I saw from phase 1,
> but unfortunately it looked like "tee" hadn't debuffered the critical
part,
> so I'm going to have to try again. Sorry this is so inconclusive, but the
> build did pass test_socket.
>
> This was with configure --without-thread, by the way. Let me know if you'd
> like me to repeat it --with-thread.
>
> Let me know what else you need - I'm afraid work is interfering with all
> this testing, so things are going a little more slowly than I'd like. With
> 2.2 so close maybe this will have to (and should?) wait for 2.3...
>
Well, the Windows 98 news is slightly better now. Running the tests
(omitting test_hotshot and test_select manually) gives the following
results, and *does* complete (I may have been in too much of a hurry
earlier):

150 tests OK.
5 tests failed:
    test_dircache test_gdbm test_mhlib test_mmap test_os
29 tests skipped:
    test_al test_asynchat test_bsddb test_cd test_cl test_curses
    test_dbm test_dl test_fork1 test_gl test_imgfile test_largefile
    test_linuxaudiodev test_locale test_minidom test_nis test_ntpath
    test_pyexpat test_sax test_socket_ssl test_socketserver
    test_sunaudiodev test_thread test_threaded_import
    test_threadedtempfile test_threading test_unicode_file test_winreg
    test_winsound
Ask someone to teach regrtest.py about which tests are
expected to get skipped on cygwin.
make: [test] Error 1 (ignored)

I'll do some checking on the individual failing tests if I get time later,
but it most probably won't be in the next eight hours. Let me know if there
are specific tests you'd like carried out.

regards
 Steve
--
http://www.holdenweb.com/