Re: Re: Lispworks and Oracle

Martin Thornquist <[email protected]> writes:

> > The weak point is that USQL maintainers hardly support LispWorks (on
> > Windows at least).
> 
> The common stuff seems mostly ported, from what I have seen it seems
> to me that the patches were incorporated, but someone has checked in
> CMUCL specifics afterwards (it is just some MOP stuff, where CMUCL
> seems to need pcl:: in front of some symbols). That was easily fixed,
> I've attached my patches. Oh, by the way, I use LispWorks 4.2.7 on
> Linux.

These patches have been incorporated on the head.

However, Dimitri's original point is still applicable and worth
thinking about.  We don't use LWW at all here, do not have a test box
for it, and in general could not really support it internally.  At the
same time, I don't think we've done enough to distribute
maintainership of USQL ports nor do we provide an easy mechanism for
identifying and remedying portability issues.  While in theory we
should be perfect coders and never write non-portable CL, that is
obviously not the case.

It seems to me that two things need to happen prior to an attempt to
move USQL onto the CLSQL data access layers:

1. Regression tests for basic USQL functionality at the
   object->relational and functional levels, such that it can be
   loaded and run against the various DB backends.  There exists a
   small set of regression tests for USQL already, but it can
   certainly use expansion.

2. Procuring installations of the various supported lisp environments
   and RDBMS systems with build engines, or at least somewhat frequent
   HEAD builds by maintainers of those ports.  This way non-portable
   changes will be quickly identified.

After these are establish, work on redefining the USQL DBMS interface
such that it supports CLSQL and allows the parsing of data to occur at
a lower level can proceed safely.  These changes will need to be
preceeded by a refactoring of the interface between the O/R and
functional interfaces, which can IMO only take place with a full
regression suite.

If this plan is acceptable to people I will start by laying out the
points the regression suite needs to cover and start writing the bits
I can.  I will also need volounteers outside of onShore Development
who would be willing to have an instance of the USQL regression tests
running against the HEAD on a platofrm other than linux CMUCL.

-- 
Craig Brozefsky <[email protected]>	             Senior Programmer
onShore Development                       http://www.onshore-devel.com
Free Common Lisp Software      http://alpha.onshored.com/lisp-software