[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Tue, 31 Aug 1999 08:57:09 +1000
> This question also gives a nice opportunity to lobby for
> read-write DB one more time, especially since the
> seemingly "implementation detail" feature could be linked
> to a linguistic feature in so interesting way.
Not long ago I tried to implement a read-write database. I
went along merrily for several hours, then I began to discover
things, and other things, and eventually I had to give up
One particular problem (not the only one) is that you can't
be sure who comes after who until the document is finalized.
Lout handles this by resolving preceding and following cross
references after the document is finalized, storing them in
the cross-reference database, and retrieving them on the next
run. Contrary to first impressions, a read-write database
would not provide anything that could speed up this algorithm.
- read-write database,
Jeff Kingston <=