[os-infrastructure] RE: [os-engineering] Model discussion boiled down

[Andrew Ross]

Hi Joe,

Agreed, VMS makes things interesting. Mainframe even moreso.

Solaris, AIX, HP-UX give us GNU tools, a subversion client, and other
libraries which means we have a hope of keeping in lockstep.

What do you think our options are regarding VMS? Which would you prefer?

Andrew  

-----Original Message-----
From: Joseph M. Abbate 
Sent: June 24, 2008 9:22 PM
To: Andrew Ross
Cc: Open Source Infrastructure; Open Source Engineering
Subject: Re: [os-engineering] Model discussion boiled down

Hi Andrew,

Andrew Ross wrote:
> We have reached consensus as a team that it is impractical for us to 
> move off of Piccolo (p) until some outstanding technical and workflow 
> issues are sorted out. There seems to be agreement that this is the 
> right direction and recognition that it's going to take time.
>  
> We also have consensus that p isn't practical for enabling community 
> to work with us. (it isn't visible to the public, wasn't designed to 
> be,
> etc.)
>  
> Thus we expect to work with p internally and Subversion (svn) for 
> external code access. This is expected to be the case for the 
> remainder of 2008. A key next step is deciding how to synchronize the 
> two and what content to make available publicly.

I think there is a key decision that needs to be made before discusssing

synchronization and the two models.   The question is which platforms 
will be supported by the Ingres Community Edition.  If the answer to
that question is "all currently supported commercial platforms," then
the choice of Subversion is a much bigger problem than if, say, "only
Linux, Windows, Mac OS/X and OpenSolaris are to be supported."  And as
you may suspect, the reason is that there is no Subversion client for
VMS.  Trying to work in main from remote client, would be IMO
unacceptable.  And although in the past (18 months ago or
so) we only had one VMS engineer, we now have three (albeit perhaps not
all full-time).

Furthermore, in the past VMS releases were done many months, if not
years, behind the core releases, but that has changed.  We are currently
tracking the 9.2 release closely and soon I anticipate being able to do
actual development work in main, rather than simply using it primarily
as a way station between 2006r2 and 2006r3.  And coincidentally the work
that I was planning on doing involves another thorn on the VMS support
arena:  Xerces.  Currently, we don't support Unicode coercion and the
XML import/export utilities on VMS because (until recently) the only
Xerces-C version available for VMS was 2.2 (2.7 is now available, but we
can't depend on HP making each released Xerces-C version made available
on VMS).

If Ingres were to decide, for example, that VMS was a legacy platform
and that no effort would be made to support "new" features such as
Unicode, UTF-8 and the Itanium migration, and then only on the
Enterprise Edition, then the Community Edition would be able to become
free of VMS code (the CL, the DCL scripts).  And Subversion and Xerces
are not the only obstacle:  although VMS is billed as POSIX compliant,
the tools and libraries are not up to par with those available on Linux,

OS/X and Windows.  For example, GNU make is at version 3.78.1.   If we 
wanted to move to autoconf/automake instead of Jam, having to use that
as LCD would be a drag on the other platforms.

Joe