Hi Paul, I hope I did not introduce other problems while trying to adress those. On Sat, Nov 17, 2007 at 10:44:00PM +1100, Paul Mackerras wrote: > What this will do is interpret the output of the program according to > Tcl list syntax. I'm not sure of why this interpretation is done, so it may well be that the updated patch would behave the same. > use error_popup rather than just writing to stderr. While testing that, I realized that the current behaviour is not very friendly. While it is OK to exit on error here, when the command has been entered on command-line, it would be rude to exit when the user makes a typo when editing the view. It would be more friendly to get back to the "edit view" dialog instead; I'm not sure we can already tell in which of the 2 situations we are - would you conisider adding a global flag for tracking this to be an adequate solution ? > > +set revtreeargscmd None > > Why the string "None" rather than the empty string? Is this a > python-ism that has crept in? Note completely - since there were already occurences of None in gitk, I assumed it was a special value in Tcl. Should I have searched in more details, I would have noticed the string comparisons. Best regards, -- Yann - To unsubscribe from this list: send the line "unsubscribe git" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
