In message <[EMAIL PROTECTED]>
John-Mark Bell <[EMAIL PROTECTED]> wrote:
> On Tue, 11 Mar 2008 10:00:07 +1300, Keith Hopper <[EMAIL PROTECTED]>
> wrote:
> >
> > Is there something wrong with Netsurf's memory usage??
>
> Not to my knowledge.
>
> > Yesterday I sent for a document - which came in at 20Mb - onlt to find
> > Netsurf taking over 127Mb of memory before anything displayed.
>
> That's not exactly surprising. If the source document truly /is/ 20
> *megabytes* in size, I'd appreciate a copy -- such things are almost never
> seen on the Web.
>
> > When I eventually closed the window containing the document, Netsurf took
> another
> > 4Mb of memory - and held on to the lot!
>
> That's simply heap fragmentation. Fixing that properly is decidely
> non-trivial.
Well, the various games at parslow.com and the other MC sites (linked from
parslow.com) allow you to dial-a-document-size up to about 10Mb by setting
the start and size parameters appropriately. Games 3, 5, 26, and 70 are
about the biggest and longest-running on parslow.com, but there are archived
games of all sizes on all three sites, and also in the White Rose MC
archive.
http://parslow.com/mornington/move.pl?3&start=24518&size=100
http://parslow.com/mornington/move.pl?5&start=10011&size=100
http://parslow.com/mornington/move.pl?26&start=8171&size=100
http://parslow.com/mornington/move.pl?70&start=12591&size=100
Then there's the archived Really Bad HTML game which is a truly work of evil:
http://www.dunx.org/cgi-bin/white-rose/forum?forum=Game85&groupBy=id&group=000&before=80&end=
No browser could render that mess correctly, but it's probably quite a good
smoke test - even displaying it without crashing and without leaking memory
all over the place probably counts as a major success.
--
Simon Smith The idea of an uncrackable digital rights management
(DRM) scheme is fundamentally flawed. Encryption is
about A sending information to B while ensuring that
C cannot read it. In DRM, B and C are the same person.
