On Sun, Apr 09, 2006 at 03:20:05PM +0300, Niko Tyni wrote: > On Tue, Apr 04, 2006 at 11:58:32AM +0200, Bill Allombert wrote: > > > Unfortunately there is a bug: > > > > The larger tick value is smaller than the max, or it is drawn outside > > the chart. > > Thanks, so there is. I'm attaching a new try at a patch. It also uses > m*10^n as the interval, where m is 1,2 or 5, to mimic the behaviour > without 'integer_ticks_only'.
This work better. Some comments: 1) The lower bound should be probably rounded down to the next multiple of the interval. This is the behaviour when integer_ticks_only is off. 2) the patch did not apply so I used fromdos on Base.pm and it applied. Maybe my MUA has converted the patch to UNIX lines. 3) I would suggest the default for max_y_ticks to be the same with or without integer_ticks_only (the manpage suggests so). A specific value could be used for max_y_ticks to denote the correct behaviour (no max). I hope upstream will consider your patch. Thanks a lot for working on this issue, Bill. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
