[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: handle -r X-Y, where X > Y

From: Molle Bestefich <molle.bestefich_at_gmail.com>
Date: 2006-05-16 04:54:15 CEST

Archie Cobbs wrote:
> This reminds me of an annoying problem I discovered in MySQL today,
> which is that you can't say "... WHERE foo IN ( )" .. i.e, query
> membership in an empty set (this is what some automatically generated
> query strings may try to do).

Yes, it would be nice if the SQL parser handled this by just null'ing
the query, eg. turning it into a LIMIT 0 or something. The logic has
to be done somewhere. Everyone implementing automated WHERE...IN...
queries has to do this currently, doing it generally in the parser
instead would be nice. The performance penalty of doing it in the
parser is probably negligible.

> arguably revisions "456-123" more likely means "the empty set".

Why is that likely?

> Automated tools that use svnmerge might emit such reversed ranges,
> and they would probably expect the empty set too.

This I strongly disagree with.
Emitting an empty list and expecting an empty set as in your SQL
example is fine, but emitting totally bogus revision numbers in order
to get an empty set sounds very wrong to me.
Sounds like whatever tool in question has completely misunderstood something...

To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Tue May 16 04:54:47 2006

This is an archived mail posted to the Subversion Dev mailing list.