Re: [MiNT] mint web/apache idea

["Guido Flohr" <gufl0000@stud.uni-sb.de>]

Hi,

On Mon, May 08, 2000 at 01:49:57AM +0000, Katherine L Ellis wrote:
> On Sun, 7 May 2000, Jo Even Skarstein wrote:
> 
> > If you want this locally I guess you don't need a server, you can just
> > embed it in CAB.OVL.
> 
> JUST embed in cab.ovl?
> Funny but JUST sounds like an easy task, when we all know that cab.ovl
> doesn't even support cookies correctly, Just need to add them. It is all
> documented, am I the only one having problems with cab ovl with simple
> cookies?  ex: yahoo, hotmail (ssl AND cookies).  None of them work.
> No one really dared fixing up cab ovl, it's too complicated.

I wouldn't count at all on CAB, it's a cul-de-sac.  Alexander seems to
have abandonned the development and implementing up-to-date features would
require a major code revision.

I don't see a way to perform parallel network transmissions with the
current API.  But pseudo-threaded network io is a must for a web browser.

It is a fact that JavaScript is also required today.  This can only be
done within the browser.  IMHO external solutions are useless.

The only chance to have a decent web browser for MiNT would be a
cooperation of several programmers (and users would have to accept that an
up-to-date implementation of the most important web standards cannot be
done within a 500-kB-application).  For the network layer and the protocol
stuff the libwww by the W3C looks like a good starting-point.  The html 
resp. xml parsers could probably inherit a lot from free
implementations like libxml, expat (for xml) or the KDE libraries for html
(although they are written in C++).  Another chance would be to update
the crude html parser in libwww.  The presentation layer needs to be very
close to the VDI anyway; porting existing implementations like Mozilla or
the outdated Mosaic sources would be probably be way too slow.  A free
JavaScript implementation is available from Netscape AFAIK.

Except for the GEM programming, writing a non-proprietary web browser for
Atari only needs a team of volunteers.  It's simply too much for a single
person but for a team it should be possible.

Ciao

Guido
-- 
http://www.stud.uni-saarland.de/
Send your spam to president@whitehouse.gov and your replies to
mailto:guido at freemint dot de