Hi, On Mon, 6 Aug 2007, Torgil Svensson wrote: > On 8/4/07, Johannes Schindelin <Johannes.Schindelin@xxxxxx> wrote: > > > > 1. Unpacking in a folder that have space in it's path gave various > > > errors of this type (here, msysGit is unpacked on the desktop > > > C:\Documents and settings\.... ) : make[1]: C:/Documents: Command > > > not found > > > > Good point. Did you fix it? If so, where is the patch? Or even > > better, what is your account on repo.or.cz so I can add it, and you > > can commit your fix yourself? > > I didn't. I bypassed it (unpacked to C:\). I found the bug however > (missed quoting): > > $ diff -urN msysGit msysGit-fixed > diff -urN msysGit/git/git-gui/Makefile msysGit-fixed/git/git-gui/Makefile > --- msysGit/git/git-gui/Makefile Fri Aug 3 15:00:16 2007 > +++ msysGit-fixed/git/git-gui/Makefile Mon Aug 6 01:22:53 2007 > @@ -4,7 +4,7 @@ > # > > GIT-VERSION-FILE: .FORCE-GIT-VERSION-FILE > - @$(SHELL_PATH) ./GIT-VERSION-GEN > + @"$(SHELL_PATH)" ./GIT-VERSION-GEN AFAICT this _fixes_ compilation in a patch containing spaces, not break it. > diff: msysGit/git/gitweb/test/M??rchen: No such file or directory > diff: msysGit-fixed/git/gitweb/test/M??rchen: No such file or directory Yes, that Märchen is a constant pain. I think it is the reason Kai vanished: he does not want to bear the consequences of his wrong doing. Of course, it has exposed errors regarding encoding in filenames. But those are errors in the OSes (or in this case, MSys), and _not_ to be fixed within git. > I'll make an account and send you the details when I get an public > SSH-key, haven't done that in windows before. Can I share private key on > both windows and Linux box? Yes. However, I inherently distrust Windows. So I made a separate account on Windows. NOTE: earlier installers included msys-rxvt.bat. I now _know_ for a fact that a lot of things are strangely broken under Rxvt. For example, creating an ssh key, of all things. I'd be more than happy to add you to the member list. > > > 2. rxvt-terminal had some freezes > > > > I did not experience those. Could you research further? Those "freezes" were due to the fact that Rxvt incorrectly updates stderr in a blocking way, or not at all (don't know which). There are more things that do not work in Rxvt, and only after trying the same in cmd (which I do not like for various reasons) I found out that rxvt.exe is at fault. I would be glad if somebody managed to compile rxvt herself and fix all those bugs (see http://code.google.com/p/msysgit/ for a short list of the most pressing issues I found). As it is, I have enough work to do with the rest of msysGit, and for the moment, I can at least work in cmd. Even ssh push works. > Some observations: > "git log" on mingw archive trigger this every time for me (hangs on > different places each time). Never experienced those. > "cmd" terminal (msys.bat) seems to pipe "git log" to less while rxvt > terminal outputs lots of text very fast (msys-rxvt.bat) and freezes. > Output:ing lots of text itself doesn't seem to be an issue ("make -d" > doesn't freeze and outputs _lots_ of text) It never froze for me. However, as stated, Rxvt is too workaroundable for me to work on it (although I like Rxvt better than cmd). Any help appreciated very much. > > > 3. "gitk --all" said "Error reading commits: fatal: write failure on > > > stdout: Invalid argument" > > > > I did not even bother testing gitk... Any idea what is going wrong? gitk worked fine for me, as long as I had _any_ refs in the repo. > The fconfigure -blocking command seems to be not supported under > mingw. This patch makes the error go away. Another issue that is > probably related to this is that gitk doesn't list all commits. I > don't know how to solve this properly. > > diff --git a/gitk b/gitk > index 43d88ca..69e5149 100755 > --- a/gitk > +++ b/gitk > @@ -147,7 +147,7 @@ proc getcommitlines {fd view} { > unset commfd($view) > notbusy $view > # set it blocking so we wait for the process to terminate > - fconfigure $fd -blocking 1 > + #fconfigure $fd -blocking 1 I have to wonder how you called it... From msys.bat, cd'ing into a valid repository, I had no issues whatsoever. > > > 4. Mouse-wheel didn't work to scroll views in gitk > > > > Again, no idea why. Could you please investigate? > > Tough one since I didn't take the tcl tk debugging class. Doesn't work > for me in cygwin either. I don't know if it's implemented at all. Does > it work for anyone else? I just tested via VNC, and indeed, it does not work. My guess is that MinGW's Tk version has a different number assigned to the scroll wheel (the scroll wheel is usually implemented as constant fire on a special mouse button; AFAICT buttons 4 and 5 are assumed to be the scroll buttons in gitk). Now, I do not know what these are called on Windows, but there should be a thing like "xev" on Windows, too... Anybody? Ciao, Dscho
