Re: detecting file type by magic number


Subject: Re: detecting file type by magic number
sterwill@abisource.com
Date: Thu Jan 20 2000 - 10:01:20 CST


Kevin Vajk wrote:
> > Now detecting abiword takes 8 bytes.
>
> Excellent! :) If and when everyone agrees that it makes sense
> to begin all abiword documents with the string "<abiword", then
> I'd be glad to start harassing Linux vendors (and the UNIX vendor
> I work for) about the magic file.

I'm wondering if starting with "<abiword>" isn't as clean as doing
something more established (like declaring our XML DTD through
something like "<!doctype ...>"). I know too little about XML
conventions to concoct the needed string, and we're going to need
to tag our documents eventually.

-- 
Shaw Terwilliger



This archive was generated by hypermail 2b25 : Thu Jan 20 2000 - 10:01:20 CST