Re: [Farrago-developers] Eigenbase perforce change 9342 for review

[Julian H. <jul...@sp...>]

In our world they are not always sharable. And at this stage in the
process, we don't know whether they are sharable, because whether they
are sharable depends on the implementation chosen.

What we actually needed here is exclsuve-wait option here: if someone
else is preparing the same SQL, don't pin their entry (because it might
be incomplete), and don't go and create your own, but wait until they
have finished preparing it.

Julian

> -----Original Message-----
> From: John V. Sichi [mailto:js...@gm...] 
> Sent: Monday, May 28, 2007 1:52 AM
> To: Julian Hyde
> Subject: Re: Eigenbase perforce change 9342 for review
> 
> I haven't tracked down where the change below originated, but it's 
> wrong.  Executable statements are pure and sharable, so they aren't 
> supposed to be pinned exclusively.  I'll revert it on //open/dev.
> 
> PIC says we're getting very close now...I'm integrating a 
> last batch of 
> changes up soon.
> 
> JVS
> 
> Julian Hyde wrote:
> > http://p4web.eigenbase.org/@md=d&c=6PU@//9342?ac=10
> > 
> > Change 9342 by jh...@jh...rmalade3 on 2007/05/26 00:25:23
> > 
> > 	DEV: Integrate from //open/dt/dev/...@9326
> > ==== 
> //open/dev/farrago/src/net/sf/farrago/db/FarragoDatabase.java#
> 59 (ktext) ====
> > 880c880
> > <             cacheEntry = codeCache.pin(sql, stmtFactory, false);
> > ---
> >>             cacheEntry = codeCache.pin(sql, stmtFactory, true);
>