[Gllug] Data migration and electronic archaeology

tet at accucard.com tet at accucard.com
Tue May 21 17:12:37 UTC 2002


>The real fun is decoding obselete binary formats.

I've got one up on you (and probably everyone else) when it comes
to obscure binary formats:

One of our business partners supplies us data via a nightly file
transfer. Each file is record based, with each record consisting
of a number of fields. So far, so good, you might think. However...

- the fields are variable length, depending on the contents of
  previous fields
- the total number of fields per record is variable, again depending on
  the contents of fields within that record
- the fields are in a mix of encodings: some are ASCII, some EBCDIC,
  some binary
- individual fields within the record may be compressed

Bleugh!

><rant>
>And don't get me started on XML!  We had a "business reason" to convert
>some data to XML (ie. it was a cool buzzword) and make them available
>via the web.  XML just isn't up to storing large volumes of data - it
>is way too bloated a format - binary data requires a binary format.
></rant>

Agreed 100%. I was starting to think that I was the only one that hadn't
fallen for the XML hype. Sure, it has a number of uses. But it's far from
the panacaea that people are making it out to be. And it should *NEVER*
be let loose near config files (see Jabber), log files (see the current
production systems I have to maintain), or binary data...

Tet


-- 
Gllug mailing list  -  Gllug at linux.co.uk
http://list.ftech.net/mailman/listinfo/gllug




More information about the GLLUG mailing list