Peter Prymmer <[EMAIL PROTECTED]> wrote:
> On Tue, 31 Oct 2000, Charles Lane wrote:
>> Peter Prymmer (<[EMAIL PROTECTED]> wrote:
>> > On Sat, 28 Oct 2000, Charles Lane wrote:
>> >> +$ match = "/* postprocessed by vms_yfix.pl"
>> >> +$ if f$extract(0,f$len(match),line) .nes. match
>> >> +$ then
>> >> +$ torig = f$cvtime(f$file("[-]perly.c","CDT"))
>> >> +$ tvms = f$cvtime(f$file("[-.vms]perly_c.vms","CDT"))
>> >> +$ if (torig .ges. tvms)
>> >> +$ then
>> >> +$ type sys$input:
>> >> +$ deck
>> >> +
>> >> +Uh oh, looks like vms/perly_c.vms wasn't updated. That sometimes
>> >> +happens with new Perl releases. If you have a working Perl already,
>> >> +you can do the update yourself with:
>> >> +
>> >> + $ PERL vms/vms_yfix.pl
>> >> +
>> >> +after this procedure, but *before* running MMS/K. Please let the Perl
>> >> +maintainers know so that this can be avoided in subsequent releases.
>>
>> > I pity the poor user who does not have perl previously installed on VMS
>> > and receives this message ("Do you mean I have to have Perl installed
>> > in order to install Perl?")
>>
>> Well, what else can we tell them? Maybe we should emphasize that it's
>> in fact a bug in the distribution, much like having files missing.
>>
>> My main hope is that by having a flood of "hey, update vms/perly_c.vms!"
>> messages right after a new release, the Pumpkings will modify their
>> makefile to automatically do the updated as needed. Or at least raise
>> the awareness level so that updates are less likely to be forgotten.
> I was merely commenting. I don't think that the message needs a re-write.
> I must admit that I am confused that the 7483 kit came out without having
> running the vms yfixer since I had raised the objection that it had not
> been run in the previous kit from Jarkko and he added two entries to the
> change log: 1) said that he in fact ran the fixer script and 2) said that
> he has now added a run of the vms specific script to the "regen_all"
> target for the unix Makefile.SH (which turns into the Makefile on unix).
Did 7483 come out without a yfixer run? I know there were some link problems
with S_syylex in 7463, but that was something different.
Hmm...it looks like Jarkko dropped the patch I submitted to fix that one;
but there was some other patch to embed; I'll take a look at it.
Good to hear that yfixer should be run "automatically" now. Would it
be worth testing for perly_c.vms being up to date, even if it turns
out to be a no-op 99.9% of the time? Considering that we've gotten
bitten by this so many times before (and that the errors you get are
*always* obscure) my inclination is to add the test.
--
Drexel University \V --Chuck Lane
======]---------->--------*------------<-------[===========
(215) 895-1545 _/ \ Particle Physics
FAX: (215) 895-5934 /\ /~~~~~~~~~~~ [EMAIL PROTECTED]
