|
 |
Le 12/07/2013 20:24, Alex Stewart nous fit lire :
> Hi all,
>
> I'm currently using 3.7.0 RC7, and have run across some problems which
> appear to be manifestations of http://bugs.povray.org/task/294 ..
>
> The bug entry says that a fix has been developed (about a month ago) and
> applied as "change 5907". Is there a repository somewhere (or nightlies
> or something) I could fetch from to pick up this fix? (I've looked around
> a bit, but can't seem to find any source repository for POV-Ray out
> there.. am I missing something obvious?)
>
You are not missing anything. Hereafter change 5907 for your own test
(by hand)
Change in source/backend/frame.h
================================
1. circa line 1080, put in comment:
> bool Cache_Valid;
> int Cache_Type;
> DBL Cache_Point;
> EXPRESS Cache_Data;
Changes in source/backend/math/splines.cpp
==========================================
1. circa line 542 inside Create_Spline, put in comment:
> New->Cache_Valid = false;
2. circa line 598 inside Copy_Spline, put in comment:
> New->Cache_Valid = false;
3. circa line 705 inside Insert_Spline_Entry, put in comment:
> sp->Cache_Valid = false;
4. circa line 802, put in comment (about 16 lines):
> // check if the value is in the cache
> if((sp->Cache_Point == p) && (sp->Cache_Type == sp->Type))
> {
> if(sp->Cache_Valid == true) // doing this here is more efficient as it is rarely
false [trf]
> {
> Assign_Express(v, sp->Cache_Data);
> return sp->Cache_Data[0];
> }
> }
>
> // init some cache data
> sp->Cache_Valid = false;
> sp->Cache_Type = sp->Type;
> sp->Cache_Point = p;
5. circa line 902, put in comment (about 2 lines):
> // put data in cache
> Assign_Express(sp->Cache_Data, v);
> sp->Cache_Valid = true;
> Also, it's a little unclear: The bug is still listed as "Requires
> testing" status (I'd be happy to help test, if needed).. will this fix
> make it into the next release (whenever that might be)? This issue is
> really quite annoying and difficult to work around..
It requires testing because I was alone to test it, and my hardware is
not universal, so confirmation by other would be welcome. From strict
quality point of view, the patcher should not close directly the issue.
Post a reply to this message
|
|
