On Sat, 2005-04-09 at 04:45, Rod Engelsman wrote:
> Sorry about the double-post, I hit Send prematurely.
> 
> Ian Lynch wrote:

> > Both are significantly worse in that respect than Impression Publisher
> > on the Acorn RISC OS platform in the early 90s. Impression Publisher
> > could be used quite happily for both word processing and DTP on a 25 MHz
> > machine with 4 meg of RAM but then it was written largely written in
> > Assembler optimised for one processor. 
> 
> Bingo. That's ultimately the key to all your praise for the efficiency 
> of your beloved Acorn.

Actually I don't particularly love it now, I do recognise that it
demonstrates what is possible and that because most people haven't
experienced it they set their sights too low.  

> In fact a Psion netBook is in
> > hardware terms considerably more powerful than those machines. This
> > seems to indicate that coding efficiency is more important than hardware
> > performance but there are very low expectations in this respect because
> > people believe products like MS Word and MS Publisher represent state of
> > the art hi-tec and that code efficiency doesn't matter too much because
> > hardware keeps getting more powerful.
> > 
> 
> If you code in Assembler you will certainly (well if you know what 
> you're doing anyway, I guess) get faster, more efficient, optimized 
> code. But at what price? Platform portability for one thing. And 
> development will take a lot longer and be more expensive.

Yes I think it was a company of about 10 people that developed
Impression Publisher. Not very expensive at all.

If you want to see what 1 programmer can do using C try downloading
OvationPro for Windows. 

http://pilling.users.netlink.co.uk/ovationpro/opw.html

This was originally developed for RISC OS by Dr David Pilling and ported
to Windows. Its not Open Source but you can download it free to try out
(not sure if its absolutely finished yet) I doubt paying him to make it
Open Source would be very expensive in the whole scheme of things.

If the fundamental routines are efficient, well documented, modular and
open source (Neither Ovation Pro nor Impression are which is why I have
lost interest in them - they will never be more than a niche but they do
show some of what is possible) it provides a much better fundamental
basis for everything else. Its why I'm very sceptical of claims about
the cost of software development. Sure there is a cost but its a
relatively small cost in the whole scheme of things if people go about
it in the right way. If we had an efficient Open Platform, there would
be no real need to for cross platform development.

Unfortunately history hasn't followed that path so we get quick fix on
quick fix and complex combinations of fixes and grafted on functions. 

Personally, before we graft DTP functions onto Writer I'd say it would
be better to actually make the code work more efficiently by getting rid
of any redundant or inefficient code first. Then look at ways of
modifying the structures in Writer to provide DTP functions without
adding bloat or compromising the strong aspects that already exist. This
is a big job so, for example, SVG import/export for Draw should be a
higher priority because it will take less resource to achieve and be
more generally useful to more people.
-- 
Ian Lynch <[EMAIL PROTECTED]>
ZMS Ltd


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to