Subject: Re: commit -- more FASTSQUIGGLE logic
From: Paul Rohr (paul@abisource.com)
Date: Wed Dec 08 1999 - 17:21:41 CST
At 10:37 AM 12/8/99 -0600, Robert Sievers wrote:
>Yes, FASTSQUIGGLE is much better now. It took me a entire 1/2 hour to
>crash it. Either I am losing my touch, or your check in was one hell of a
>clean up.
I have no idea whether you're losing your touch. I do know that the
previous code never had any hope of working in strux editing cases. :-)
>Regardless, the problem was near a paragraph break, which of course should
>be no surprise to either of us. Logged as #674. I only bring it up
>because I got a couple different crashes similar to this, but had some
>trouble tracking it down whether it was all different manifestations of the
>same bug, or different bugs.
Quick local fix for #674.
M src/text/fmt/xp/fv_View.cpp
If the other problems you were seeing were all preceded by that same assert,
they're probably OK now. Otherwise, they'll still need attention.
>As always, I am spread a bit too thin to bang
>on this for an hour right now. If you have time to look at this, I can
>slide in some more testing here before 0.7.7. If not, I will do more
>isolation work post 0.7.7.
I'm still tempted to take a look at the stuff I left out of last night's
checkin, but it sounds like we don't need to hold 0.7.7 for it. I'll try
not to check in any more FASTSQUIGGLE work without first making sure you've
got time to beat on it.
Paul
This archive was generated by hypermail 2b25 : Wed Dec 08 1999 - 17:16:37 CST