Re: [xslt] segfaults & limits

[Pike <pike kw nl>]

Hi

>>>>> lately, we're running into more and more segfaults.
>>>>   I can't really do anything with such a report honnestly.
>> Is there no possibility to avoid segfaults in C?
:-)

from daniels response, i smell a sense of frustration :-)
let me just add that what libxml is doing sofar for us  is 
_incredably_good_,
we've just pushed things to limits, but these limits appear pretty high.
libxml rocks.

I kinda understand the problem in C, that's pretty hardcore..

a worthy discussion might be how to workaround such segfaults.
would increasing memory help ?
what are good alternatives to variable value-sets and
node-sets ? how would you do that in proper xslt1.0 ?
aren't document() statements just as 'dangerous' ?


$2c,
*pike
=======
+ give luck a chance