RE: [patch: perl@7045] vms updates - IEEE build

On Wed, 13 Sep 2000, Craig A. Berry wrote: > At 02:17 PM 9/13/2000 -0700, Peter Prymmer wrote: > > >> /L_DOUBLE_SIZE=option > >> /L_DOUBLE_SIZE=128 (D) > >> > >> So you don't have to specify it, you get 128-bit by default for long double > >> variables. > > > >But I was ask

RE: [patch: perl@7045] vms updates - IEEE build

On Wed, 13 Sep 2000, Carl Friedberg wrote: > Just shows my ignorance, I can't recall that Perl had data types specifying > the length of integers... Run the command: search config_h.SH LONG to see (some of) them. > but 128 bit longs? why not? less likely to have a Concorde crash due to >

RE: [patch: perl@7045] vms updates - IEEE build

At 02:17 PM 9/13/2000 -0700, Peter Prymmer wrote: >> /L_DOUBLE_SIZE=option >> /L_DOUBLE_SIZE=128 (D) >> >> So you don't have to specify it, you get 128-bit by default for long double >> variables. > >But I was asking if we'd want to offer /L_DOUBLE_SIZE=64 as an option >in a c

RE: [patch: perl@7045] vms updates - IEEE build

esday, September 13, 2000 4:44 PM To: BAZLEY, Sebastian Cc: 'Carl Friedberg'; [EMAIL PROTECTED] Subject: RE: [patch: perl@7045] vms updates - IEEE build On Wed, 13 Sep 2000, BAZLEY, Sebastian wrote: > >-Original Message- > >From: Carl Friedberg [mailto:[EMAIL PROTEC

RE: [patch: perl@7045] vms updates - IEEE build

On Wed, 13 Sep 2000, Dan Sugalski wrote: > >Er, my question was not about support for long double at all my question > >was about inquiring what size long double you'd prefer, _assuming_ that > >you've asked for long double support. > > > >In the 7065 kit I find no evidence for mention of an /L

Re: [patch: perl@7045] vms updates

On Wed, 13 Sep 2000, Craig A. Berry wrote: > I forget what OS version you are testing on, but if it's 7.2 you could be > encountering a problem fixed by system ECO 0500: > > o After upgrading from V7.1 or V7.1-2 to V7.2 or V7.2-1, some processes may > periodically hang in HIBernation state,

RE: [patch: perl@7045] vms updates - IEEE build

At 02:02 PM 9/13/00 -0700, Peter Prymmer wrote: >On Wed, 13 Sep 2000, Dan Sugalski wrote: > > > At 01:44 PM 9/13/00 -0700, Peter Prymmer wrote: > > >This reminds me. Is there any interest out there for 128 bit long doubles > > >as in this snippet from the CC062 release notes? > > > > > >

Re: [patch: perl@7045] vms updates

At 07:58 PM 9/12/2000 -0700, Prymmer/Kahn wrote: >> >By the way: some folks on the vmsperl list may want to play around with >> >perlman a bit now. I am finding an unusual number of hibernating processes >> >building up that I discovered with a "SHOW DEV/FILE": Peter, I forget what OS version

RE: [patch: perl@7045] vms updates - IEEE build

On Wed, 13 Sep 2000, Dan Sugalski wrote: > At 01:44 PM 9/13/00 -0700, Peter Prymmer wrote: > >This reminds me. Is there any interest out there for 128 bit long doubles > >as in this snippet from the CC062 release notes? > > > > Prior to V4.0, DEC C used the same 64 bit machine da

RE: [patch: perl@7045] vms updates - IEEE build

At 01:44 PM 9/13/00 -0700, Peter Prymmer wrote: >This reminds me. Is there any interest out there for 128 bit long doubles >as in this snippet from the CC062 release notes? > > Prior to V4.0, DEC C used the same 64 bit machine data > type for the C types "double" and "

RE: [patch: perl@7045] vms updates - IEEE build

On Wed, 13 Sep 2000, BAZLEY, Sebastian wrote: > >-Original Message- > >From: Carl Friedberg [mailto:[EMAIL PROTECTED]] > >Sent: 13 September 2000 19:02 > >To: 'BAZLEY, Sebastian'; [EMAIL PROTECTED] > >Subject: RE: [patch: perl@7045] vms upda

Re: [patch: perl@7045] vms updates - IEEE build

Carl Friedberg wrote: > > Sebastian, I don't agree on this one. > Nor do I. > I think the internal data format should be the one predominantly used on the > platform. The fact that some database might use a different format can > easily be handled in the glue in xs, or whatever. I've never use

Re: [patch: perl@7045] vms updates - IEEE build

At 02:39 PM 9/13/2000 -0500, Patrick Spinler wrote: >Granted that tweaking the compile for /float=ieee wasn't too hard (added >the compile option to 2 files), but it would nevertheless be very nice >to have it as a configure.com option. It's been in there since 5.6.0: $! IEEE math? $ echo "" $ e

Re: [patch: perl@7045] vms updates - IEEE build

"BAZLEY, Sebastian" wrote: > > Thanks for email (did you mean to reply to me only?). > > I haven't actually proved that IEEE is needed for Ingres - presumably you > did find out that it was necessary? If so, that would be worth mentioning on > the list. Whups, you're entirely correct, thanks.

Re: [patch: perl@7045] vms updates

On Wed, 13 Sep 2000, Craig A. Berry wrote: > Only the test.com patch -- not the b.t patch -- has been incorporated. I > had a stupid question about the b.t patch of the > after-midnight-what-was-I-thinking variety, but Charles answered that and I > think should proceed with posting it to p5

RE: [patch: perl@7045] vms updates - IEEE build

At 02:02 PM 9/13/2000 -0400, Carl Friedberg wrote: >I think the internal data format should be the one predominantly used on the >platform. IEEE *is* internal on AXP in the sense that it's supportd by the instruction set. So is G-float. VAX supports D-float and G-float natively (at least on

RE: [patch: perl@7045] vms updates - IEEE build

>-Original Message- >From: Carl Friedberg [mailto:[EMAIL PROTECTED]] >Sent: 13 September 2000 19:02 >To: 'BAZLEY, Sebastian'; [EMAIL PROTECTED] >Subject: RE: [patch: perl@7045] vms updates - IEEE build > > >Sebastian, I don't agree on this one. >

RE: [patch: perl@7045] vms updates - IEEE build

routines makes it a minor point, though... Carl -Original Message- From: BAZLEY, Sebastian [mailto:[EMAIL PROTECTED]] Sent: Wednesday, September 13, 2000 1:54 PM To: [EMAIL PROTECTED] Subject: RE: [patch: perl@7045] vms updates - IEEE build >-Original Message- >From: Cra

RE: [patch: perl@7045] vms updates - IEEE build

>-Original Message- >From: Craig A. Berry [mailto:[EMAIL PROTECTED]] >Sent: 13 September 2000 18:17 >To: BAZLEY, Sebastian >Cc: [EMAIL PROTECTED] >Subject: RE: [patch: perl@7045] vms updates - IEEE build > > >At 05:30 PM 9/13/2000 +0100, BAZLEY, Sebas

RE: [patch: perl@7045] vms updates - IEEE build

At 12:16 PM 9/13/00 -0500, Craig A. Berry wrote: >At 05:30 PM 9/13/2000 +0100, BAZLEY, Sebastian wrote: > >Ingres uses IEEE floats on AXP. > > > >We had run-time problems when we first moved Ingres applications to AXP from > >VAX and compiled without IEEE. > > > >Although I've built DBD-Ingres on

RE: [patch: perl@7045] vms updates - IEEE build

At 05:30 PM 9/13/2000 +0100, BAZLEY, Sebastian wrote: >Ingres uses IEEE floats on AXP. > >We had run-time problems when we first moved Ingres applications to AXP from >VAX and compiled without IEEE. > >Although I've built DBD-Ingres on AXP, I haven't used it in earnest, but I >suspect that IEEE is

RE: [patch: perl@7045] vms updates - IEEE build

>-Original Message- >From: Craig A. Berry [mailto:[EMAIL PROTECTED]] >Sent: 13 September 2000 17:18 >To: Prymmer/Kahn >Cc: [EMAIL PROTECTED] >Subject: Re: [patch: perl@7045] vms updates > [huge snip] > >An IEEE build would be interesting [snip] [Not sure if

Re: [patch: perl@7045] vms updates

At 07:58 PM 9/12/2000 -0700, Prymmer/Kahn wrote: > >Charles Lanes' reply >to your message (the one I have quoted here) mentions: > > "[.lib]b.FAILED on test 14 (known, patch posted here)" > >So I am wondering: in what way might Charles' patch have gotten messed up? Only

Re: [patch: perl@7045] vms updates

On Wed, 13 Sep 2000, Charles Lane wrote: > I didn't get messed up; the patch of t/lib/b.t wasn't included in what > I sent to Jarkko, since it was unrelated to the test.com patch, and there > was some lingering question about perhaps being able to remove the > "if ($Is_VMS)" sections(I don't

Re: [patch: perl@7045] vms updates

Peter Prymmer wrote: >Craig Berry wrote: >> Nice work, Peter. Jarkko has already incorporated these and Chuck's test.com patch >into the 7065 kit: > With the 7065 kit and my regular VMS V7.1 && DEC C V6.0 machine I saw > the two st-*.t failures and a new one in lib/b.t. Charles Lanes' reply >

Re: [patch: perl@7045] vms updates

> Nice work, Peter. Jarkko has already incorporated these and Chuck's test.com patch >into the 7065 kit: > > Thanks. I saw the p5p post via Achim's web archive (email was all messed up at work, but I could s

Re: [patch: perl@7045] vms updates

"Craig A. Berry" <[EMAIL PROTECTED]> writes: > At 06:41 PM 9/11/2000 -0700, Peter Prymmer wrote: > >>Enclosed is a patch to update various VMS build components. > > Nice work, Peter. Jarkko has already incorporated these and Chuck's test.com patch >into the 7065 kit: > >

Re: [patch: perl@7045] vms updates

At 06:41 PM 9/11/2000 -0700, Peter Prymmer wrote: >Enclosed is a patch to update various VMS build components. Nice work, Peter. Jarkko has already incorporated these and Chuck's test.com patch into the 7065 kit:

[patch: perl@7045] vms updates

Enclosed is a patch to update various VMS build components. File: Comment(s): configure.com 0) add pager selectability: default is "type/page" unless there is a symbol "most" or a symbol "pager" (the latter can be achieved with "-Dpager=""my_pager""" on the comma