[Seaside] [FIX] SeasideServiceSyncToSeaside2.3

[Avi Bryant]

On Tue, 8 Apr 2003, Stephen Pair wrote:

> WAKomNG could be included as part of the Seaside release, that would
> eliminate the problem (since each release of Seaside could have it's own
> version of WAKomNG).  But, you would probably have to conditionally load
> it based on whether or not ComancheHttpService is present (since WAKomNG
> subclasses it).

Right, which is precisely why it isn't yet included in the release.
Moving to Monticello should allow conditional loading to be easier,
though.

> The only issue with the pluggable approach might be performance since we
> would probably need to be careful about copying the block to allow for
> concurrent evaluations.  With block closures, it wouldn't be as much of
> an issue.

Well, you could use a MessageSend instead of a block.  But I have a really
hard time believing that a single block copy per request will make any
noticeable difference to performance (just checked - on my system it adds
exactly one microsecond).

Avi