> Please read  the RFC and AFTER  you can make suggestions.  These _all_
> are mentioned in the rfc!

Okay, I have read it now.  Now I'm going to make suggestions :-) (Note
that so far I've been commenting only on the aspects of making things
'thread-safe', not on the RFC itself.  'Threadsafing' Perl is what
I've deemed to be "too complex".) (Also note that I'm not a database
guru, so please bear with me, and don't ask me to write the code :-)

If I understood correctly, you basically want to extend the existing
tie()-interface to be useful for versioned variables.  (That they are
'transactioned' is just a choice of words.)  To *safely* do data
versioning one needs database-like COMMIT and ROLLBACK features.

While I do understand your basic request and I quite like the idea of
having versioned data, I do not see the need for a new variable type,
neither by introducing a new keyword, nor a by a new variable
attribute.  Also, from implementation viewpoint introducing all the
required thread-safeness into the very core of Perl doesn't sound good
to me.  Yes, it would be a very neat thing to have.  But making all
non-transaction-requiring Perl programs to pay the {speed,size} price
for the thread-safing locking machinery doesn't sound good.  Then
again, having such functionality as an external "loadable module"
doesn't sound too bad.  What would be valuable from Perl 6 RFC/design
viewpoint would be to take a careful look at the tie() interface and
for Perl 6 design an extendable interface so that methods like COMMIT,
ROLLBACK, et cetera can be easily implemented.

-- 
$jhi++; # http://www.iki.fi/jhi/
        # There is this special biologist word we use for 'stable'.
        # It is 'dead'. -- Jack Cohen

Reply via email to