Peter Prymmer <[EMAIL PROTECTED]> wrote:
> On Thu, 2 Nov 2000, Charles Lane wrote:
(about VMS checksum)
> It might be nice to have such a checksummer available to unix perl,
> just as a few months ago someone heeded a unix passwd hasher on VMS.
Sure would; anyone with a spare 'tuit and an old VAX hardware/MACRO manual?
I'm not completely sure, but I *think* that's the only place I've seen
the CRC algorithm presented. Assuming that CHECKSUM actually uses
CRC, which I'm not 100% sure of...
>> How about this: modify vms/vms_yfix.pl do the following:
>>
>> copy perly.c -> vms/perly_c.orig
>> process perly.c -> vms/perly_c.vms
>>
>> Then have configure.com:
>>
>> compare CHECKSUM's of perly.c against vms/perl_c.vms
>> if the same, okay we're done
> I disagree. If they are the same then something is wrong since
> perl_src:[000000]perly.c ought to be different from [.vms]perl_c.vms - no?
> The adding of global declarations was the point of vms/vms_yfix.pl wasn't
> it?
Ah, but if they're the same it means that vms/perly_c.vms (not perl_c, as
I mistyped above) has *already* been copied to perly.c. This would be
the case if you build Perl, make some changes, and then build again.
>> ELSE
>> compare CHECKSUM's of perly.c against vms/perly_c.orig
>> if the same, we need to copy vms/perly_c.vms -> perly.c etc.
> The copying of [.vms]perl_c.vms -> perly.c is ordinarily handled by
> the Makefile (er, descrip.mms), and even the cleaning up of perly.c is
> handled (to a very limited extent since the original unix perly.c gets
> wiped out by a purge).
Right, but if we can't trust TAR's dates (and I think your data shows
that we can't always do so) then having descrip.mms do the copy is a
potentially a problem. Not too likely to *be* a problem, because the
usual order of unpacking creates perly.c before vms/perly_vms.c. Maybe
we can "touch" vms/perly_vms.c so that we can be sure it gets copied later
on.
This caveat only applies to a tiny fraction of the build, since the
tarfile won't have things like (for example) .obj files.
>> ELSE
>> complain bitterly
>>
>> So all that has to happen extra on the Unix side is a file copy,
>> and we can use CHECKSUM from DCL to see if vms/perly_c.vms has been
>> generated properly.
> How about DIFERENCES?
Checksum is better for this; faster, simpler...all we really want is a
"same/not" check without having to parse the output from DIFF in DCL.
>> I'll see about working up a patch based on this; in the meanwhile, are
>> you happy with the gen_shrfls patch?
> Very happy and very grateful. I'd say cut the gen_shrfls loose from the
> configure.com tweaking and send it in to Jarkko and p5p. Thank you.
Will do.
--
Drexel University \V --Chuck Lane
======]---------->--------*------------<-------[===========
(215) 895-1545 _/ \ Particle Physics
FAX: (215) 895-5934 /\ /~~~~~~~~~~~ [EMAIL PROTECTED]
