Resent-From: POSTGRES mailing list <postman@postgres.Berkeley.EDU>
Resent-Message-Id: <199401272051.MAA26827@nobozo.CS.Berkeley.EDU>
Sender: owner-postman@postgres.Berkeley.EDU
From: witr@rwwa.COM
Date: Thu, 27 Jan 94 15:51:02 -0500
Message-Id: <9401272051.AAwarz25910@relay1.UU.NET>
To: postgres@postgres.Berkeley.EDU
Subject: Re: Dan Bidwell's Solaris Port
Content-Type: text
Content-Length: 2059
Resent-To: postgres-dist@postgres.Berkeley.EDU
Resent-Date: Thu, 27 Jan 94 12:51:20 -0800
Resent-XMts: smtp

> I have since started to attempt to make Robert Withrow's SVR4 port work
> on Solaris 2.3.  I have it compiling, but it will not run yet.

I'll help anyone with my port if I can.  I just can't promise to spend
more than a few milliseconds on it.  I might, but I can't promise...

> I was hoping that this might be easier then find the problem in the
> port that I have done.

What is the problem?

P.S.: I have updated my port since it was originally released, and in
particular, I have greatly fixed the time handling ADT code.  What is
holding me up in releasing the fixed code is the nasty problem of RCS
clashes.  I gather that the Postgres people don't use CVS, but those
who do know what I mean.

The result is that I have to go through the Postgres source tree and
de-activate all RCS headers, re-import this code into CVS, produce
diffs between this and my current baselevel, extract the RCS headers
from the result, commit this into CVS, *and* I have to provide the RCS
de-activating program for users to use before they apply the diffs
when building my port.  This is a major pain. (This is why,
incidentally, you will notice that X Consortium code has munged their
RCS headers.)

This can all be solved by the Postgres developers either a) using CVS
and doing a CVS export to produce the source kit, or b) using the
RCS -kv option when extracting the sources to produce the kit.  I
can't imagine *WHY* the Postgres developers *ARENT* using CVS, but I
can guarantee you than any serious Postgres porter is nuts if he/she
doesn't.  I doubt that *active* RCS headers serve any useful purpose
to any non-Berkeley developer, and they get in the way bigtime if you
*do* use CVS.

The point of this harang is that those people who want Postgres ports
would help those of us who make them if they would lobby the Postgres
developers to do one of the above things when they release the source
code.

---
 Robert Withrow, Tel: +1 617 598 4480, Fax: +1 617 598 4430, Net: witr@rwwa.COM
 R.W. Withrow Associates, 21 Railroad Ave, Swampscott MA 01907-1821 USA