This mail is an automated notification from the bugs tracker
 of the project: GNUstep.

/**************************************************************************/
[bugs #8365] Latest Modifications:

Changes by: 
                Richard Frith-Macdonald <[EMAIL PROTECTED]>
'Date: 
                Wed 03/31/04 at 11:41 (GMT)

            What     | Removed                   | Added
---------------------------------------------------------------------------
          Resolution | None                      | Fixed
              Status | Open                      | Closed


------------------ Additional Follow-up Comments ----------------------------
Fixed in CVS ... I think it made sense to limit this for general performance in the 
past, but we now cache the hash values for strings (apart from constant strings) so 
the performance trade-off is probably in favour of using all
characters now.
For short strings this should make no difference.






/**************************************************************************/
[bugs #8365] Full Item Snapshot:

URL: <http://savannah.gnu.org/bugs/?func=detailitem&item_id=8365>
Project: GNUstep
Submitted by: Larry Campbell
On: Tue 03/30/04 at 04:46

Category:  Base/Foundation
Severity:  5 - Average
Item Group:  Bug
Resolution:  Fixed
Assigned to:  None
Status:  Closed


Summary:  -[NSString hash] limitation makes dictionaries horribly horribly slow

Original Submission:  For some reason not explained in any comments I could find, 
-[NSString hash] ignores all but the first 63 bytes of a string. This means that if 
you're storing large strings in an NSDictionary, your performance is completely hosed. 
I just removed the 63-byte truncation, and an operation that used to take 20 to 30 
minutes now takes only seconds -- two to three orders of magnitude faster.

What is the reason, if any, for this limitation? Shouldn't it be removed?


Follow-up Comments
------------------


-------------------------------------------------------
Date: Wed 03/31/04 at 11:41         By: CaS
Fixed in CVS ... I think it made sense to limit this for general performance in the 
past, but we now cache the hash values for strings (apart from constant strings) so 
the performance trade-off is probably in favour of using all
characters now.
For short strings this should make no difference.




CC List
-------

CC Address                          | Comment
------------------------------------+-----------------------------
lcampbel --AT-- akamai --DOT-- com  | 









For detailed info, follow this link:
<http://savannah.gnu.org/bugs/?func=detailitem&item_id=8365>

_______________________________________________
  Message sent via/by Savannah
  http://savannah.gnu.org/





_______________________________________________
Bug-gnustep mailing list
[EMAIL PROTECTED]
http://mail.gnu.org/mailman/listinfo/bug-gnustep

Reply via email to