Thomas Passin, does rev 028f4b (in devel) fix #1444 or #1445?

25 views
Skip to first unread message

Edward K. Ream

unread,
Dec 6, 2019, 9:39:49 AM12/6/19
to leo-editor
Rev 028f4b attempts a fix for #1444 (many errors during startup) without really knowing what is going on.

The traceback indicates that line 607 in leoCache.py may be to blame.

The old code catches only KeyError; the new code catches all Exceptions. The traceback references PyQt4, so perhaps there is some installation-dependent problem/strangeness.

I am wondering whether this fix might just fix #1445 as well. Clearly, something unusual is going on, because very few people seem to be affected by either bug.

Edward

Thomas Passin

unread,
Dec 10, 2019, 3:48:55 PM12/10/19
to leo-editor
I just found that the problem went away when I deleted the Leo database directory.  I added a comment to that effect in #1444. It must have something to do with the fact that I have used both Leo 5.9/python 2.7/QT4 and Leo 61/python 3.8/QT4 alternately.  Somehow some things cached in the database must have been getting read by Leo even though their entries were some time in the past.

I did an very time-consuming hunt for anyplace in the 6.1 code base where QT4 might have been getting invoked by mistake, but found nothing.
Reply all
Reply to author
Forward
0 new messages