home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Acorn User 3
/
AUCD3.iso
/
airport
/
browsers
/
acornet
/
archive
/
archive897
/
000062_owner-acornet@…s.barnet.ac.uk _Wed Aug 13 17:20:50 1997.msg
< prev
next >
Wrap
Internet Message Format
|
1997-08-28
|
4KB
Received: (from majordomo@localhost)
by odie.barnet.ac.uk (8.8.6/8.8.6) id RAA23808
for acornet-list; Wed, 13 Aug 1997 17:20:50 +0100
Received: from beech.sucs.soton.ac.uk (beech.sucs.soton.ac.uk [152.78.129.138])
by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id RAA23804
for <acornet@lists.barnet.ac.uk>; Wed, 13 Aug 1997 17:20:36 +0100
From: S.N.Brodie@ecs.soton.ac.uk
Received: from bright.ecs.soton.ac.uk (bright.ecs.soton.ac.uk [152.78.64.201])
by beech.sucs.soton.ac.uk (8.8.5/server) with SMTP id RAA07889
for <acornet@lists.barnet.ac.uk>; Wed, 13 Aug 1997 17:21:01 +0100 (BST)
Received: from dsse.ecs.soton.ac.uk by bright.ecs.soton.ac.uk; Wed, 13 Aug 97 17:23:17 BST
Received: from mccarthy.ecs.soton.ac.uk by dsse.ecs.soton.ac.uk; Wed, 13 Aug 97 17:19:03 BST
Message-Id: <17672.9708131619@mccarthy.ecs.soton.ac.uk>
Subject: Re: AcorNet 0.20 (Stewarts comments)
To: acornet@lists.barnet.ac.uk
Date: Wed, 13 Aug 1997 17:19:02 +0100 (BST)
In-Reply-To: <Marcel-1.26-0813144452-064LJLo@ether104.acorn.co.uk> from "Phillip Temple" at Aug 13, 97 03:44:52 pm
X-Mailer: ELM [version 2.4 PL24]
Content-Type: text
Sender: owner-acornet@lists.barnet.ac.uk
Precedence: bulk
Reply-To: acornet@lists.barnet.ac.uk
X-maillist: acornet
Phillip Temple wrote:
>
> > I don't think that there are that
> > many people involved in the RISC OS networking scene who are capable of
> > doing a good job at writing such a driver - especially since you really
> > need to be a registered developer to get hold of the various pieces of
> > documentation you need and many aren't (when I asked for details, I never
> > even got a response) and/or can't afford it.
>
> What details would they need that they can't get hold of?
The DCI4 protocol documentation. I've had lots of people ask me if
I've got it and if I would send it to them because they can't get a
copy from Acorn (I've never needed it because I don't write code at
that level of the stack). Admittedly this was last year, but still
plenty of time after Internet 4.xx was available.
> * use a provider who provides SLIP
I agree except that the problem comes when providers all of a sudden
decide to alter their access method from SLIP to PPP (or from offering
both and deciding to withdraw SLIP because it's too hard to support
both)
On the PPPdriver front, should Acornet have the capability to understand
it and be able to configure it IF it is provided by the user? I keep
reading articles on the newsgroups about people struggling to modify their
Acornet setup to use PPP instead of SLIP.
>> [size of Acornet < size of Internet Explorer]
> That's like saying I must be underweight, take a look at the
> reigning Sumo champion...
Well with Acornet you are getting the full suite of programs to
handle e-mail, news, FTP, WWW, dialling and you don't *need* anything
else (except PPPdriver potentially :)
Inevitably it will be large: it's supposed to provide a complete suite
of Internet access applications. There comes a point when even RISC OS
software that we are accustomed to being small will become larger as
more functionality is included.
Thankfully, I believe that the vast majority of coders for RISC OS
attempt to minimise the requirements of their software.
It's a hard balancing act. I know that ArcWeb is the largest single
application in Acornet but I can't help that - it's still only 550K
archived (cf. IE3). I'd rather like ArcWeb not to require 1MB of memory
immediately after starting up, but the application code takes up 550K
of that and then space is required for the image decompression tables,
window templates, sprites, etc. etc. I'd really like to have JPEG
handling code built-in too so that I can avoid all the nasty crashes
caused by the ROM JPEG code and that people who haven't patched
ChangeFSI as per the instructions keep experiencing. However, that
code would increase the code size by a further 80K so it's been
omitted. The included code is really pared to the bone and to remove
any more would damage functionaility beyond any limit I'm prepared to
pass.
--
Stewart Brodie, Electronics & Computer Science, Southampton University.
http://www.ecs.soton.ac.uk/~snb94r/