Re: ::scr Ramblings of a Classic Refugee or How I Learned To Stop Worrying and Love OS X

[Alaric Snell <alaric@xxxxxxxxxxxxxxxx>]

On Wednesday 06 February 2002 15:00, you wrote:

[learning new ASCII file formats vs. new binary editing tools]

> But that's true either way.

That was my point.

[snip]

> Agreeing on a standard encoding format that is universally accessible,
> almost always Just Built In, and which has a wide variety of pre-made
> tools for working with it just Makes Sense. Switching to something else
> only makese sense if you're willing to replace all that infrastructure --
> and you'd damn well better get the encoding & decoding tools right.
> Switching to something else does *not* bring us a whit closer to the core
> problem here -- semantic meaning and all that fun stuff -- so you'd better
> have some other valid justification for doing so. This isn't to say that a
> case can't be made -- Microsoft has made a nice fortune by setting up the
> infrastructure to produce & decode binary-gibberish formats with their
> Office suite. If that's what you're after, then hey have at it.
> Personally, I'm still not sure it's a good idea.

Yes, indeed. Standards are good, but I don't think there's anything 
intrinsically special about standardising on top of Unicode rather than 
anything else. Unicode tools aren't that widely deployed; ASCII ones are and 
can be shoehorned into dealing with a subset of Unicode encoded in UTF-8, but 
that's all; the age old tradeoff between starting from scratch and building 
on something existing.

> > Explain why people *aren't* using text for everything. Where is the big
> > push to use textual encoding for images?
>
> Well, duh :)
>
> I don't think anyone here would argue against using the right tool for the
> right job. If data can't be adequately described textually -- and images
> would surely count, then don't use text! But in a lot of cases, something
> "better" is proposed without much good justification, ignoring all the
> nice properties that ASCII has going in its favor, waving hands and saying
> "yeah but $foo can do it better!" This case remains unproven.

Talk to the people who come to XML-DEV saying 'waaah, my system is too 
slow...' who get told to try to keep the DOM tree in (binary and randomly 
accessible) memory (and thus going through complex hoops to try to make use 
of that in-memory copy rather than going to disk) rather than sequential text 
files on disk :-)

ABS

-- 
                               Alaric B. Snell
 http://www.alaric-snell.com/  http://RFC.net/  http://www.warhead.org.uk/
   Any sufficiently advanced technology can be emulated in software