Re: 21.0 Showstoppers

cgw> Ben Wing wrote: cgw> I have access to Win95, 98 and NT4.0. I routinely test the installer cgw> on all of these. I do not have access to NT3.51, so I have no way to cgw> test. It seems pretty clear from the reports coming in that it simply cgw> won't work on 3.51. I will add an OS-version check to the installer cgw> and bail out if OS == NT && version < 4.0, unless somebody can tell cgw> me what needs to be done to make it work on WinNT. I imagine that most of the problems with NT 3.5.1 have to do with unexec. I imagine that most of these are fixable if someone has access to 3.5.1, because winemacs used basically the same unexec mechanism and DID run on NT 3.51. However, I don't think this is important enough to hold up the release if it doesn't get dealt with. cgw> I have stuck the installer in my personal little bit of web-space, but cgw> have always hoped that somebody would pick it up and put it on the cgw> main XEmacs site. cgw> I submitted a patch that fixed a bunch of the Win95/98 subprocess cgw> brokenness, but it was rejected for (at least) two reasons: (A)it was cgw> a bit kludgey, and (B) there is some disagreement about the need for cgw> "cmdproxy" and the ability of the existing XEmacs subprocess code to cgw> spawn 16-bit subprocesses. cgw> The kludge was there because I didn't understand why the variable cgw> "win32_quote_process_args" didn't work as advertised, and I introduced cgw> another variable to do the same thing. The problem was cgw> "vars_of_ntproc" not getting called from emacs.c. cgw> On the other issue, I am fairly certain that "cmdproxy" is *not* cgw> necessary; "command.com" should do just fine. "ntproc.c" checks cgw> whether the process being spawned is "16-bit" (if it has the ".com" cgw> extension it is considered 16-bit) and if so uses cmdproxy to spawn cgw> it. However when I simply took this piece of code out I found that cgw> XEmacs can apparently spawn "command.com" just fine. cgw> It's perhaps a bit confusing to follow the code, the "spawnve" that is cgw> called at line 734 in "callproc.c" is *not* the Win32 spawnve cgw> function, but the spawnve function in "ntproc.c", which is capable of cgw> spawning a 16-bit subprocess. (Try it and see....) cgw> Anyhow, I'm glad you consider the broken subprocesses on 95/98 as a cgw> problem worth fixing. cgw> I will clean up my rejected patch to make it less kludgey and resubmit cgw> it; maybe it can still make it into 21.0. Please, please, please do resubmit your patch. It simply must go into 21.0. Better something kludgy than something that doesn't work at all. cgw> I'll leave these to somebody else! Ben (typed by Martin)