OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: "Binary XML" proposals



On Wed, 11 Apr 2001, Rick Jelliffe wrote:

> >Given that there are freely and widely available tools to do the
> >conversions in both ways, a compiled machine code file (for your
> >architecture, so you'll have the *right* tools installed anyway as a
> >developer) is only harder to work with than a text file because you have
> >to disassemble it first (unless you have a debugger handy, in which case
> >the debugger will usually let you explore the binary file and handle cross
> >referencing for you, which a text viewer WON'T do!)
> 
> You are saying that a binary format is better than a text format because it
> can be converted to a text format?

No...

> (My first job was to rewrite a real-time assembler program written without
> comments, so it somewhat staggers me that anyone could even think of saying
> seriously that comments were somehow strippable without degrading the
> readability of the document.)

...the fact that machine code binary formats don't generally contain
comments and the like (they contain references back to lines in the source
file isntead) is a property of that particular case (compactness has
generally been a major goal in the design of executable formats). What I'm
saying is that you can't make a sweeping comment that text files are more
readable than binary files.

Let's consider the case of raster images again. They can be written in
text; you have PNM files. They're not particularly readable... you still
need specialist software to visualise that kind of data.

Most of the people on this list seem to mainly work with the kind of data
you might find kicking around in an SQL database table and putting a Web
front end onto; the kind of data where XSLT transformations into HTML make
sense.

This perhaps clouds their judgement; I don't see much comment from the
image processing and scientific data camps here, who both have very valid
desires for a data interchange format, but can't use XML if it only has
textual encodings defined...

In an ideal world, the choice between using textual XML or binary XML in
an interface should be as simple as whether to transport the resulting XML
encoding by HTTP, FTP, or floppy disks in the mail; there are currently
widely deployed XML engines that only support text-XML, but if there was a
decent binary-XML standard they mainly implemented, then working with
binary-XML would be no harder (XML spy and other such browsers/editors
wouldn't ever need to tell you how the data was encoded underneath) than
text-XML, so in the cases where it is technically better to use a compact
binary format, people could.

> Cheers
> Rick Jelliffe

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