Re: I see only two CPAN-Testers reports on the server from the many I have sent - what is wrong?
David Golden
> Dear All,
>
> I have recently installed CPAN::Reporter - after some difficulties, because
> of wrong file permissions in this distribution, but that's a different story
> (I already informed David Golden about it).
See my other response. :-)
> I am currently installing lots of modules under Perl 5.10.1 (native build
> for Windows XP with MS VC++ 6.0), mostly through installing Bundles.
>
> How come that the test reports sent by CPAN.pm from my machine never show up
> in the "Perl/Platform Version Matrix" of the corresponding modules?
>
>
> From 9 reports sent according to "reports-sent.db" in a test run, only two
> showed up on the server, namely for Inline-0.45 and YAML-Syck-1.07.
Can you send the report ID's for those 2 reports so we can see your
configuration? And what are some modules have you tested? (Be aware
there can be a multi-day delay before reports appear on the website.)
> When I subscribed to the cpan-testers mailing list, I didn't get any
> "welcome" email back after I sent the confirmation email.
>
> Is that normal, or could that be the problem?
>
> If so, why was my confirmation email disregarded? (Never got any error
> message email either)
It's possibly your emails are or were being flagged as spam for some
reason. There was a recent hiccup at the Perl NOC that flagged lots
of things as spam unintentionally:
http://log.perl.org/2009/10/blackholes.html
That could be related.
-- David
Chris Marshall
Two modules I have been working with recently, PDL and
OpenGL, seem to have very few testing hits since the end
of August except for the ones generated by our own
developers and testers.
This has been particularly frustrating as we get down to
the finish for a new PDL release (and accompanying OpenGL)
since the test reports for multiple OSes which used to
come in seem to have vanished. Bugs from the previous
tester results have been addressed but without those
platform tests, we have no way to know if that has fixed
things.
* How can I determine the testers events for a given module?
* By date (Is the rate of testing consistent over time?)
* By tester ID (Some of the old stand-bys for testing PDL
and OpenGL, no longer show. Are PDL and OpenGL out of the
rotation or is there something broken somewhere?)
* Is there any way too access any regularly updated feed of
tester results?
* When the cpan-testers list had email to authors, it was
easy to track anomalies since a systematic problem would
change the character and pace of tester reports received
* I guess I could try setting up a mysql server and try
downloading the entire database but that seems like a
sort of "nuclear option".
* Does it matter whether a module is released by the Owner
or a Co-maintainer for CPAN Tester to work?
* One other thing I am trying to work through is some sort
of quick set up info for CPAN::Reporter or Test::Reporter
so that it works. My own experience is that the perl of
both modules works well, the sticking point for me is how
to set up the mail feed---especially since most ISP and
mail providers have moved to TLS type security which is
tricky to configure. (e.g. I have a gmail account and
have yet to succeed in sending a test report through
that channel)
Thoughts appreciated!
Chris Marshall
(PDL and OpenGL developer)
O. STeffen BEYer wrote:
> Dear All,
>
> I have recently installed CPAN::Reporter - after some difficulties,
> because of wrong file permissions in this distribution, but that's a
> different story (I already informed David Golden about it).
>
> I am currently installing lots of modules under Perl 5.10.1 (native
> build for Windows XP with MS VC++ 6.0), mostly through installing Bundles.
>
> How come that the test reports sent by CPAN.pm from my machine never
> show up in the "Perl/Platform Version Matrix" of the corresponding modules?
>
> From 9 reports sent according to "reports-sent.db" in a test run, only
> two showed up on the server, namely for Inline-0.45 and YAML-Syck-1.07.
>
> Or is it because test reports from module authors for their own modules
> are filtered out? (Not according to my preferences, at least!)
>
> Neither do I see any of these test reports on two separate email
> accounts which I both subscribed to the cpan-testers mailing list.
>
> Is that normally the case?
>
> When I subscribed to the cpan-testers mailing list, I didn't get any
> "welcome" email back after I sent the confirmation email.
>
> Is that normal, or could that be the problem?
>
> If so, why was my confirmation email disregarded? (Never got any error
> message email either)
>
> I am attaching the relevant config files from CPAN and CPAN::Reporter
> (as well as my "perl -V" output).
>
> Thank you very much in advance for any help!
>
> Best regards,
> --
> Steffen Beyer <ST...@cpan.org <mailto:ST...@cpan.org>>
> http://www.engelschall.com/u/sb/whoami/ (Who am I)
> http://www.engelschall.com/u/sb/gallery/ (Fotos Brasil, USA, ...)
> http://www.engelschall.com/u/sb/download/ (Free Perl and C Software)
>
>
> ------------------------------------------------------------------------
>
>
> No virus found in this incoming message.
> Checked by AVG - www.avg.com
> Version: 8.5.421 / Virus Database: 270.14.17/2436 - Release Date: 10/14/09 18:32:00
>
O. STeffen BEYer
thank you very much for your immediate reply!
Ralf Engelschall (of "mod_rewrite" fame, among other famous projects) uses a
very strict spamfilter on his machine (on which I get my emails to
s...@engelschall.com),
in which even Google mail is blacklisted as a spammer (can you imagine)!
That's probably why your email was rejected.
For the same reason I have already changed the email address in all of my
modules and on CPAN from s...@engelschall.com to ST...@cpan.org.
The latter forwards all emails to s...@engelschall.com, but this works because
cpan.org is not blacklisted as a spammer (yet?!).
And anyhow, in this way I can easily change the email address
ST...@cpan.orgis forwarding to, in case of need, without the users
even needing to know.
The two reports from me which actually show on the server are the following:
http://www.nntp.perl.org/group/perl.cpan.testers/2009/10/msg5520288.html
http://www.nntp.perl.org/group/perl.cpan.testers/2009/10/msg5525770.html
My configuration (perl -V) was also in the attachment "CPAN-Testers.tar.gz"
in my previous email.
As is a list of test reports ("reports-sent.db" from CPAN::Reporter) which
were definitely sent.
BTW, I am aware of the time lag between sending a report and the moment it
is reflected on the server.
But my reports are now over a week old, and still nothing (I checked again
today)!
Interestingly, I found my name among the statistics on the server:
http://stats.cpantesters.org/mtesters.html (search for "kabelfoon" in the
table and you'll find me)
So it seems that the server does indeed know me (and the email address <
sbe...@kabelfoon.nl> I am sending my test reports from, which is from my
provider, which provides me with the SMTP-server necessary to send the
reports).
This apparent contradiction mystifies me!
But yes, maybe the black hole on www.perl.org caused my reports to be
dropped.
That would explain it.
Since this black hole is now supposedly fixed, I will try to re-run some
builds tonight, and we'll see what happens!
In any case, thanks a lot for your support!
Best regards,
Steffen
2009/10/15 David Golden <xda...@gmail.com>
O. STeffen BEYer
with "[1]", did you mean this table?
StatusMANAGEABLE Time of oldest page request2009-10-10 14:11:09 Total page
requests33597 Total unique page requests11513
If so, then it is absolutely not self-evident from the name "Time of oldest
page request" that this is the NEWEST test report reflected in the tables on
the server, meaning that all test reports younger than that date are not
shown yet (if I understood it right that this is what you meant).
Concerning the 5 day delay; if I am not horribly mistaken, my test reports
are over a week old (from 7th October 2009)!
# Generated by CPAN::Reporter 1.1708
test PASS Inline-0.45 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
test PASS YAML-Syck-1.07 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
test PASS Carp-Clan-6.00 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
test PASS Bit-Vector-7.1 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
test PASS Date-Calc-6.0 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
test PASS Date-Pcalc-6.0 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
test PASS Data-Locations-5.5 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
test PASS Math-MatrixBool-5.8 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
test PASS Set-IntRange-5.2 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
Thank you very much for your help!
Best regards,
Steffen
2009/10/15 Barbie <bar...@missbarbell.co.uk>
> Hi,
>
> > How come that the test reports sent by CPAN.pm from my machine never
> > show up in the "Perl/Platform Version Matrix" of the corresponding
> > modules?
>
> When did you submit the reports? If you look at the status grid on the
> home page of the CPAN Testers Reports website [1], it has a date which
> is the oldest request in the system that hasn't been processed.
>
> As the site can take over 5 days to do a complete rebuild, it isn't
> unusual for some pages to not appear for several days, particularly if
> it is for a distribution that is not as popular as others.
>
> If after 5 days you still haven't seen the reports listed, send me an
> email with the IDs and I'll investigate.
>
> [1] http://www.cpantesters.org
>
> > Or is it because test reports from module authors for their own
> > modules are filtered out?
>
> No module author has the ability to filter any reports from the system.
>
> > Neither do I see any of these test reports on two separate email
> > accounts which I both subscribed to the cpan-testers mailing list.
>
> The cpan-testers mailing is unattended, and you no longer need to
> subscribe to the list as everyone can post to the list. As such if you
> do not receive an error message, then you have been subscribe
> successfully. The mailing list is now just a post only interface to the
> NNTP list.
>
> Hope that helps.
>
> Cheers,
> Barbie.
> --
> Birmingham Perl Mongers <http://birmingham.pm.org>
> Memoirs Of A Roadie <http://barbie.missbarbell.co.uk>
> CPAN Testers Blog <http://blog.cpantesters.org>
> YAPC Conference Surveys <http://yapc-surveys.org>
>
>
>
Barbie
> Hi Barbie,
>
> with "[1]", did you mean this table?
>
> StatusMANAGEABLE Time of oldest page request2009-10-10 14:11:09 Total page
> requests33597 Total unique page requests11513
> If so, then it is absolutely not self-evident from the name "Time of oldest
> page request" that this is the NEWEST test report reflected in the tables on
> the server, meaning that all test reports younger than that date are not
> shown yet (if I understood it right that this is what you meant).
That is not what it means.
The date refers to the oldest unprocessed request to update a page
(AUTHOR or DISTRO) on the reports website. There could be high profile
pages or pages with lots of reports submitted that get rebuilt
frequently, thus meaning some reports can hit the server and appear on a
page within the hour. Others may take several days to get updated.
> Concerning the 5 day delay; if I am not horribly mistaken, my test reports
> are over a week old (from 7th October 2009)!
In that case they may not have got into the system for some reason. I
shall investigate further later this evening.
David Cantrell
> Two modules I have been working with recently, PDL and
> OpenGL, seem to have very few testing hits since the end
> of August except for the ones generated by our own
> developers and testers.
OpenGL relies on libraries that probably won't be installed on dedicated
testing machines.
And then the latest dev release of PDL says ...
Automated testing environment: enabling WITH_BADVAL, BADVAL_PER_PDL, and
USE_POGL config options
DEPENDENCY ERROR: USE_POGL requires at least use OpenGL-0.58_007! at
Makefile.PL line 398.
> * How can I determine the testers events for a given module?
If you wait a few days, tests should appear at cpantesters.org.
> * By tester ID (Some of the old stand-bys for testing PDL
> and OpenGL, no longer show. Are PDL and OpenGL out of the
> rotation or is there something broken somewhere?)
There's no "official" rotation - individual testers test whatever they
want to test. The most prolific ones test *everything* that gets
uploaded to the CPAN, in something approximating the order they're
uploaded and in something approximating real time.
> * Is there any way too access any regularly updated feed of
> tester results?
There's the NNTP archive. Not sure if you can actually speak NNTP to it
and get a feed per se.
> * Does it matter whether a module is released by the Owner
> or a Co-maintainer for CPAN Tester to work?
No.
--
David Cantrell | Hero of the Information Age
Immigration: making Britain great since AD43
David Golden
>> * Is there any way too access any regularly updated feed of
>> tester results?
>
> There's the NNTP archive. Not sure if you can actually speak NNTP to it
> and get a feed per se.
See http://search.cpan.org/perldoc?cpantest-remailer
It's a program that will monitor the NNTP group and email reports to
you. Or you can adapt the guts to notify you in other ways.
David
Chris 'BinGOs' Williams
> On Thu, Oct 15, 2009 at 10:49 AM, David Cantrell <da...@cantrell.org.uk> wrote:
> >> * Is there any way too access any regularly updated feed of
> >
> > There's the NNTP archive. ??Not sure if you can actually speak NNTP to it
> > and get a feed per se.
>
> See http://search.cpan.org/perldoc?cpantest-remailer
>
> It's a program that will monitor the NNTP group and email reports to
> you. Or you can adapt the guts to notify you in other ways.
>
Or if one is on irc.perl.org IRC network join #cpantesters channel
where one of the GumbyNET bots is streaming realtime updates from
the NNTP group.
Cheers.
--
Chris Williams
aka BinGOs
PGP ID 0x4658671F
http://www.gumbynet.org.uk
==========================
M W487
see how they turn up. Most look good. Here is a possible problem:
http://www.nntp.perl.org/group/perl.cpan.testers/2009/10/msg5477314.html
tested a cygwin platform for perl 5.8.8 on Oct 3 which does not show
in the cygwin column of the chart at
Barbie
Note that the matrix and the main reports site are two different
systems. Although Slaven uses the feeds from the reports site. However,
the report you mention is on the reports site and is in the DB. The only
oddity, which might cause it to not appear in the matrix listing, is
that it has a perl version of 0. This is due to the automated process
not being able to extract the perl version from the report itself.
Barbie
> http://www.nntp.perl.org/group/perl.cpan.testers/2009/10/msg5520288.html
> http://www.nntp.perl.org/group/perl.cpan.testers/2009/10/msg5525770.html
The two above are the only ones in the system you've submit for the
MSWin32 platform, but there are 157 others that are for FreeBSD. Would
that be correct?
If the others you've highlighted are also for MSWin32, they don't appear
to have been listed within the NNTP listings. I guess the mails must
have got blocked.
O. STeffen BEYer
> Hi Steffen,
>
> > http://www.nntp.perl.org/group/perl.cpan.testers/2009/10/msg5520288.html
> > http://www.nntp.perl.org/group/perl.cpan.testers/2009/10/msg5525770.html
>
> The two above are the only ones in the system you've submit for the
> MSWin32 platform, but there are 157 others that are for FreeBSD. Would
> that be correct?
>
> If the others you've highlighted are also for MSWin32, they don't appear
> to have been listed within the NNTP listings. I guess the mails must
> have got blocked.
>
> The ones I've highlighted were indeed from the MSWin32 platform.
As an aside, it is true that I have also sent loads of reports from FreeBSD.
I have resent those from the MSWin32 platform yesterday evening:
# Generated by CPAN::Reporter 1.1708
test PASS Carp-Clan-6.00 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
test PASS Bit-Vector-7.1 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
test PASS Date-Calc-6.0 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
test PASS Date-Pcalc-6.0 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
test PASS Data-Locations-5.5 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
test PASS Math-MatrixBool-5.8 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
test PASS Set-IntRange-5.2 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
test PASS Date-Calc-6.1 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
Let's see what happens with them now.
Slaven Rezic
> On Thu, Oct 15, 2009 at 06:32:57PM -0400, M W487 wrote:
>> Basically at random, I decided to track some reports I have sent and
>> see how they turn up. Most look good. Here is a possible problem:
>>
>> http://www.nntp.perl.org/group/perl.cpan.testers/2009/10/msg5477314.html
>>
>> tested a cygwin platform for perl 5.8.8 on Oct 3 which does not show
>> in the cygwin column of the chart at
>>
>> http://matrix.cpantesters.org/?dist=DBIx-Class+0.08112
>
> Note that the matrix and the main reports site are two different
> systems. Although Slaven uses the feeds from the reports site. However,
> the report you mention is on the reports site and is in the DB. The only
> oddity, which might cause it to not appear in the matrix listing, is
> that it has a perl version of 0.
Yes, this is very probably the case. I filter out every record with a
zero perl version or os name.
> This is due to the automated process
> not being able to extract the perl version from the report itself.
Regards,
Slaven
--
Slaven Rezic - slaven <at> rezic <dot> de
tkrevdiff - graphical display of diffs between revisions (RCS, CVS or SVN)
http://ptktools.sourceforge.net/#tkrevdiff
O. STeffen BEYer
Some of the test reports I sent yesterday night are already visible, others
not (why?).
In particular, I sent:
# Generated by CPAN::Reporter 1.1708
test PASS Carp-Clan-6.00 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
test PASS Bit-Vector-7.1 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
test PASS Date-Calc-6.0 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
test PASS Date-Pcalc-6.0 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
test PASS Data-Locations-5.5 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
test PASS Math-MatrixBool-5.8 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
test PASS Set-IntRange-5.2 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
test PASS Date-Calc-6.1 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
test PASS Date-Pcalc-6.1 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
Of these, the following are visible:
http://www.nntp.perl.org/group/perl.cpan.testers/2009/10/msg5607338.html(Date::Calc
6.0) - also in
[ Perl/Platform Version
Matrix<http://matrix.cpantesters.org/?dist=Date-Calc+6.0> ]
http://www.nntp.perl.org/group/perl.cpan.testers/2009/10/msg5607375.html(Date::Calc
6.1)
Note that these are THE ONLY ONES AT ALL I could find. The ones from more
than a week ago have never appeared!
Something is eating our data!
Certainly, if the software was able to extract the Perl version from the two
reports above, then it should have been able to do so also on the others -
presuming that the other reports would not be any different in that respect,
coming from exactly the same Perl, the same CPAN::Reporter, and the very
same CPAN.pm session!
Regards & have a nice weekend!
Steffen
2009/10/16 Slaven Rezic <sla...@rezic.de>
Barbie
> Just an intermediate update:
>
> Some of the test reports I sent yesterday night are already visible, others
> not (why?).
>
> In particular, I sent:
>
> # Generated by CPAN::Reporter 1.1708
> test PASS Carp-Clan-6.00 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
> test PASS Bit-Vector-7.1 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
> test PASS Date-Calc-6.0 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
> test PASS Date-Pcalc-6.0 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
> test PASS Data-Locations-5.5 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
> test PASS Math-MatrixBool-5.8 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
> test PASS Set-IntRange-5.2 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
> test PASS Date-Calc-6.1 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
> test PASS Date-Pcalc-6.1 (perl-5.10.1) MSWin32-x86-multi-thread 5.1
All these are in the system. However, as mentioned previously, it can
take several days for some pages to be updated.
> Certainly, if the software was able to extract the Perl version from
> the two reports above, then it should have been able to do so also on
> the others - presuming that the other reports would not be any
> different in that respect, coming from exactly the same Perl, the
> same CPAN::Reporter, and the very same CPAN.pm session!
Not necessarily. If the report exceeds the byte limit it gets truncated.
All the perl -V data is at the end, which is the primary source of
discovery for the perl version. Note this is a limit of the NNTP mail
server, and one reason why we want to move to a HTTP submission based
system.
When I get free time, I go through the oddities and try and improve the
regex or hand edit the version if I can. But it's time consuming :)
Barbie
> On Fri, Oct 16, 2009 at 10:31 AM, Barbie <bar...@missbarbell.co.uk> wrote:
> > Not necessarily. If the report exceeds the byte limit it gets truncated.
> > All the perl -V data is at the end, which is the primary source of
> > discovery for the perl version. Note this is a limit of the NNTP mail
> > server, and one reason why we want to move to a HTTP submission based
> > system.
>
> long as it and Test::Reporter are up to date. CPAN::Reporter
> truncates test output (*not* Perl -V) to fit within the perl.org size
> cutoff and it now properly encodes/wraps long lines to avoid exceeding
> SMTP line length limits that were causing truncations and other
> problems.
>
> Do you have any recent examples?
Yes, but I didn't want to name names, hence my other post :)
However, I also note that I don't have the latest version of
CPAN-Testers-Common-Article on the server, github or CPAN :( I shall fix
that over the weekend, as that does do some additional detection work if
perl -V isn't found.
David Golden
> All the perl -V data is at the end, which is the primary source of
> discovery for the perl version. Note this is a limit of the NNTP mail
> server, and one reason why we want to move to a HTTP submission based
> system.
That shouldn't be an issue with CPAN::Reporter based submissions, as
long as it and Test::Reporter are up to date. CPAN::Reporter
truncates test output (*not* Perl -V) to fit within the perl.org size
cutoff and it now properly encodes/wraps long lines to avoid exceeding
SMTP line length limits that were causing truncations and other
problems.
Do you have any recent examples?
-- David
O. STeffen BEYer
reports from last week disappeared:
I installed Test::Reporter and CPAN::Reporter only days before sending, so
they should be up to date.
But maybe perl.org's blackhole really was the cause.
> On Fri, Oct 16, 2009 at 10:31 AM, Barbie <bar...@missbarbell.co.uk> wrote:
> > > Not necessarily. If the report exceeds the byte limit it gets
> truncated.
> > > All the perl -V data is at the end, which is the primary source of
> > > discovery for the perl version. Note this is a limit of the NNTP mail
> > > server, and one reason why we want to move to a HTTP submission based
> > > system.
>
> On Fri, Oct 16, 2009 at 10:48:49AM -0400, David Golden wrote:
> > That shouldn't be an issue with CPAN::Reporter based submissions, as
> > long as it and Test::Reporter are up to date. CPAN::Reporter
> > truncates test output (*not* Perl -V) to fit within the perl.org size
> > cutoff and it now properly encodes/wraps long lines to avoid exceeding
> > SMTP line length limits that were causing truncations and other
> > problems.
> > Do you have any recent examples?
>
2009/10/16 Barbie <bar...@missbarbell.co.uk>
Chris Marshall
Thanks for the clarification. Since I there is no
longer the direct mirror of cpan-testers responses,
I think "wait and see" is the best I can do for
diagnostics.
One other question: if a module lists a developers
release of another module as a requirement,
will CPAN or CPANPLUS try to build the module (even
though it is not a stable release, or will some sort
of error (or worse, no information at all...)
Cheers,
Chris
> ------------------------------------------------------------------------
>
>
> No virus found in this incoming message.
> Checked by AVG - www.avg.com
> Version: 8.5.421 / Virus Database: 270.14.18/2437 - Release Date: 10/15/09 03:57:00
>
David Golden
> One other question: if a module lists a developers
> release of another module as a requirement,
> will CPAN or CPANPLUS try to build the module (even
> though it is not a stable release, or will some sort
> of error (or worse, no information at all...)
CPAN/CPANPLUS can't install development versions because they aren't
indexed. I *think* they will install the latest stable, then complain
that the prerequisite couldn't be satisfied.
Because the prerequisite could not be satisfied, no CPAN Testers
report will be sent on failure. (PASS reports would be sent, though.)
If the dependency bumps to a stable release higher than the
development version you've listed, then everything works as normal
again.
Specifying a development release as a prerequisite is a sort of
halfway measure to indicate that whatever is stable on CPAN is not
enough, but in a way that will work fine once the development version
becomes the new stable.
Unless the toolchain is updated to allow *distributions* and not
*modules* to be listed as prerequisites, this is pretty much the best
that can be done. (Barring a "development" packages index, which
would probably be an optional thing for users to enable.)
-- David
David Cantrell
> Unless the toolchain is updated to allow *distributions* and not
> *modules* to be listed as prerequisites, this is pretty much the best
> that can be done. (Barring a "development" packages index, which
> would probably be an optional thing for users to enable.)
I could probably create such an index fairly easily and have it
alongside my cp5.6.2an etc "mirrors". If people think it's useful, I'll
see if I can set it up this weekend.
--
David Cantrell | Cake Smuggler Extraordinaire
I don't do .INI, .BAT, or .SYS files. I don't assign apps to files.
I don't configure peripherals or networks before using them. I have
a computer to do all that. I have a Macintosh, not a hobby.
-- Fritz Anderson
Andreas J. Koenig
> Unless the toolchain is updated to allow *distributions* and not
> *modules* to be listed as prerequisites, this is pretty much the best
> that can be done. (Barring a "development" packages index, which
> would probably be an optional thing for users to enable.)
Bundles come to mind. Distro Foo can depend on Bundle::Bar and
Bundle::Bar can directly demand AUTHOR/Devel-Oper-3.14_159.tar.gz
I think nobody has ever tested if it works as expected but I'd expect it
to work.
--
andreas