# git-blame.el won't run 11 messages from 2007-02-14 to 2007-02-21. Participants: Xavier Maillard, David Kågedal, Jakub Narebski. Thread: https://gitlist.dev/t/6826 ## Xavier Maillard, 2007-02-14 22:35 Subject: git-blame.el won't run Message-ID: <13283.1171492535@localhost> URL: https://gitlist.dev/e/13283.1171492535%40localhost ``` Hi, I was happy to see that latest git's git included an emacs mode for git-blame. Sadly, when trying it, I just have this displayed into *Messages* buffer: Setting up indent for shell type bash setting up indent stuff Indentation variable are now local. Indentation setup for shell type bash let: Process git-blame not running If I launch git-blame manually onto the file from command line, it works. It is on a emacs 22 (aka CVS version). Xavier ``` ## Xavier Maillard, 2007-02-16 18:14 Subject: Re: git-blame.el won't run Message-ID: <22739.1171649640@localhost> URL: https://gitlist.dev/e/22739.1171649640%40localhost In-Reply-To: <13283.1171492535@localhost> ``` Really, no one knows ? Xavier ``` ## David Kågedal, 2007-02-19 20:42 Subject: Re: git-blame.el won't run Message-ID: <87mz393mlo.fsf@morpheus.local> URL: https://gitlist.dev/e/87mz393mlo.fsf%40morpheus.local In-Reply-To: <13283.1171492535@localhost> ``` Xavier Maillard writes: > Hi, > > I was happy to see that latest git's git included an emacs mode > for git-blame. > > Sadly, when trying it, I just have this displayed into *Messages* > buffer: > > Setting up indent for shell type bash > setting up indent stuff > Indentation variable are now local. > Indentation setup for shell type bash > let: Process git-blame not running Can you please set debug-on-error to t and show me the backtrace you get? And does M-! git-blame work in your emacs? -- David Kågedal ``` ## Xavier Maillard, 2007-02-19 21:32 Subject: Re: git-blame.el won't run Message-ID: <24475.1171920735@localhost> URL: https://gitlist.dev/e/24475.1171920735%40localhost In-Reply-To: <87mz393mlo.fsf@morpheus.local> ``` Hi, > Can you please set debug-on-error to t and show me the backtrace you get? Nothing happens. I have to C-g after having waited for a few seconds. > And does M-! git-blame work in your emacs? No it is not. It shows a help message (usage message). Xavier ``` ## Jakub Narebski, 2007-02-20 01:59 Subject: Re: git-blame.el won't run Message-ID: URL: https://gitlist.dev/e/erdkht%24s7v%241%40sea.gmane.org In-Reply-To: <24475.1171920735@localhost> ``` Xavier Maillard wrote: >> And does M-! git-blame work in your emacs? > > No it is not. It shows a help message (usage message). Do you use git new enough to have --incremental support for git-blame? -- Jakub Narebski Warsaw, Poland ShadeHawk on #git ``` ## David Kågedal, 2007-02-20 07:50 Subject: Re: git-blame.el won't run Message-ID: <87abz92rp9.fsf@morpheus.local> URL: https://gitlist.dev/e/87abz92rp9.fsf%40morpheus.local In-Reply-To: <24475.1171920735@localhost> ``` Xavier Maillard writes: > Hi, > >> Can you please set debug-on-error to t and show me the backtrace you get? > > Nothing happens. I have to C-g after having waited for a few > seconds. I'm not sure my instructions were very clear. Did you try M-x git-blame-mode after enabling debug-on-error? >> And does M-! git-blame work in your emacs? > > No it is not. It shows a help message (usage message). But if the help message is for git-blame, it did work. I wanted to check that you could run git-blame from your emacs, and if it printed a usage message, it did run. If you want it to produce some real output, you have to give it a file name as well. -- David Kågedal ``` ## Xavier Maillard, 2007-02-20 17:28 Subject: Re: git-blame.el won't run Message-ID: <4157.1171992534@localhost> URL: https://gitlist.dev/e/4157.1171992534%40localhost In-Reply-To: <87abz92rp9.fsf@morpheus.local> ``` Hi, > >> Can you please set debug-on-error to t and show me the backtrace you get? > > > > Nothing happens. I have to C-g after having waited for a few > > seconds. > > I'm not sure my instructions were very clear. Did you try M-x > git-blame-mode after enabling debug-on-error? I did it. And I had to kill (C-g) after having waited for several long seconds. Xavier ``` ## David Kågedal, 2007-02-20 17:34 Subject: Re: git-blame.el won't run Message-ID: <87fy9020ne.fsf@morpheus.local> URL: https://gitlist.dev/e/87fy9020ne.fsf%40morpheus.local In-Reply-To: <4157.1171992534@localhost> ``` Xavier Maillard writes: > Hi, > >> >> Can you please set debug-on-error to t and show me the backtrace you get? >> > >> > Nothing happens. I have to C-g after having waited for a few >> > seconds. >> >> I'm not sure my instructions were very clear. Did you try M-x >> git-blame-mode after enabling debug-on-error? > > I did it. And I had to kill (C-g) after having waited for several > long seconds. OK, then set debug-on-quit to t and do it again, and send me the backtrace you get when you press C-g. -- David Kågedal ``` ## Xavier Maillard, 2007-02-21 12:08 Subject: Re: git-blame.el won't run Message-ID: <13478.1172059706@localhost> URL: https://gitlist.dev/e/13478.1172059706%40localhost In-Reply-To: <87fy9020ne.fsf@morpheus.local> ``` Hi, This is the backtrace after C-g (after a few minutes): Debugger entered--Lisp error: (quit) process-send-region(# 1 28910) (let ((display-buf ...) (blame-buf ...) (args ...)) (if startline (setq args ...)) (setq args (append args ...)) (setq git-blame-proc (apply ... "git-blame" blame-buf "git" "blame" args)) (with-current-buffer blame-buf (erase-buffer) (make-local-variable ...) (make-local-variable ...) (setq git-blame-file display-buf) (setq git-blame-current nil)) (set-process-filter git-blame-proc (quote git-blame-filter)) (set-process-sentinel git-blame-proc (quote git-blame-sentinel)) (process-send-region git-blame-proc (point-min) (point-max)) (process-send-eof git-blame-proc)) (if git-blame-proc (message "Already running git blame") (let (... ... ...) (if startline ...) (setq args ...) (setq git-blame-proc ...) (with-current-buffer blame-buf ... ... ... ... ...) (set-process-filter git-blame-proc ...) (set-process-sentinel git-blame-proc ...) (process-send-region git-blame-proc ... ...) (process-send-eof git-blame-proc))) git-blame-run() (progn (let (...) (if ... ... ...)) (setq git-blame-cache (make-hash-table :test ...)) (git-blame-run)) (if git-blame-mode (progn (let ... ...) (setq git-blame-cache ...) (git-blame-run)) (cancel-timer git-blame-idle-timer)) git-blame-mode(nil) call-interactively(git-blame-mode) execute-extended-command(nil) call-interactively(execute-extended-command) Regards, Xavier ``` ## David Kågedal, 2007-02-21 13:55 Subject: Re: git-blame.el won't run Message-ID: <87vehvsjho.fsf@morpheus.local> URL: https://gitlist.dev/e/87vehvsjho.fsf%40morpheus.local In-Reply-To: <13478.1172059706@localhost> ``` Xavier Maillard writes: > Hi, > > This is the backtrace after C-g (after a few minutes): > > > Debugger entered--Lisp error: (quit) > process-send-region(# 1 28910) > (let ((display-buf ...) (blame-buf ...) (args ...)) (if startline (setq args ...)) (setq args (append args ...)) (setq git-blame-proc (apply ... "git-blame" blame-buf "git" "blame" args)) (with-current-buffer blame-buf (erase-buffer) (make-local-variable ...) (make-local-variable ...) (setq git-blame-file display-buf) (setq git-blame-current nil)) (set-process-filter git-blame-proc (quote git-blame-filter)) (set-process-sentinel git-blame-proc (quote git-blame-sentinel)) (process-send-region git-blame-proc (point-min) (point-max)) (process-send-eof git-blame-proc)) > (if git-blame-proc (message "Already running git blame") (let (... ... ...) (if startline ...) (setq args ...) (setq git-blame-proc ...) (with-current-buffer blame-buf ... ... ... ... ...) (set-process-filter git-blame-proc ...) (set-process-sentinel git-blame-proc ...) (process-send-region git-blame-proc ... ...) (process-send-eof git-blame-proc))) > git-blame-run() > (progn (let (...) (if ... ... ...)) (setq git-blame-cache (make-hash-table :test ...)) (git-blame-run)) > (if git-blame-mode (progn (let ... ...) (setq git-blame-cache ...) (git-blame-run)) (cancel-timer git-blame-idle-timer)) > git-blame-mode(nil) > call-interactively(git-blame-mode) > execute-extended-command(nil) > call-interactively(execute-extended-command) If you switch to the buffer called " git blame for " (notice the leading space, you will probably see a usage/error message from git. Maybe your git blame doesn't like the --incremental flag or the --contents flag. But the problem is that I'm not sure how to make emacs not hang. process-send-region blocks when sending the file contents, since git isn't reading it. But git finishes, so emacs should be able to detect that there is no process that can read anything anymore. -- David Kågedal ``` ## Xavier Maillard, 2007-02-21 17:21 Subject: Re: git-blame.el won't run Message-ID: <1272.1172078480@localhost> URL: https://gitlist.dev/e/1272.1172078480%40localhost In-Reply-To: <87vehvsjho.fsf@morpheus.local> ``` Hi, > If you switch to the buffer called " git blame for " (notice > the leading space, you will probably see a usage/error message from > git. Seen it. > Maybe your git blame doesn't like the --incremental flag or the > --contents flag. But the problem is that I'm not sure how to make > emacs not hang. Ok. I fixed this by upgrading to latest stable release of Git (according to website it is 1.5.0). It now works as expected ! Thank you very much for your time and patience. Xavier ```