Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

matrix.cpantesters.org: Reports not appearing

2 views
Skip to first unread message

James E Keenan

unread,
Jul 31, 2022, 10:00:09 AM7/31/22
to cpan-teste...@perl.org
As part of my work for Perl 5 Porters in analyzing "Blead Breaks CPAN"
(BBC) bug tickets, I frequently have occasion to build perl at the HEAD
of the blead (main development) branch, install it and 'cpanm' locally,
install Task::CPAN::Reporter against that perl, then try to install CPAN
distributions which have been cited in BBC.

Yesterday, I built and installed perl at this commit:

#####
$ ./bin/perl -Ilib -v | head -2 | tail -1
This is perl 5, version 37, subversion 3 (v5.37.3
(v5.37.2-33-gff5bbe677e)) built for amd64-freebsd-thread-multi
#####

I then installed Task::CPAN::Reporter (via 'cpanm') and then used 'cpan'
to install Test::Trap and Sub::Metadata, generating CPANtesters reports
along the way.

#####
...
cpan[1]> install Sub::Metadata Test::Trap
...
CPAN::Reporter: Test result is 'pass', All tests successful.
CPAN::Reporter: preparing a CPAN Testers report for Sub-Metadata-0.002
CPAN::Reporter: sending test report with 'pass' via Metabase
...
CPAN::Reporter: Test result is 'pass', All tests successful.
CPAN::Reporter: preparing a CPAN Testers report for Test-Trap-v0.3.5
CPAN::Reporter: sending test report with 'pass' via Metabase
...
$ ./bin/perl -Ilib -MTask::CPAN::Reporter -MTest::Trap -MSub::Metadata
-E 'say q|hello world|;'
hello world
#####

These two CPAN distros (and their prerequisites) were duly logged as
having had CPANtesters reports generated:

#####
$ tail -n 4 ~/.cpanreporter/reports-sent.db
test PASS Module-Build-0.4231 (perl-5.37.3) amd64-freebsd-thread-multi
12.3-release-p1
test PASS Sub-Metadata-0.002 (perl-5.37.3) amd64-freebsd-thread-multi
12.3-release-p1
test PASS Data-Dump-1.25 (perl-5.37.3) amd64-freebsd-thread-multi
12.3-release-p1
test PASS Test-Trap-v0.3.5 (perl-5.37.3) amd64-freebsd-thread-multi
12.3-release-p1
#####

However, when I checked matrix.cpantesters.org today for those reports,
I found reports for only 2 of those 4 distros:

#####
Distribution|Report URL|Timestamp
Data-Dump|http://www.cpantesters.org/cpan/report/a8c864ee-1050-11ed-84a8-20dc05b3d53a|2022-07-30
21:43
Test-Trap|http://www.cpantesters.org/cpan/report/ad848512-1050-11ed-a8e1-724c06b3d53a|2022-07-30
21:43
#####

#####
Distribution|No Report URL
Module-Build|no report from me on Jul 30 2022 at
http://matrix.cpantesters.org/?dist=Module-Build;os=freebsd;reports=1#sl=4,1
Sub-Metadata|no report from me on Jul 30 2022 at
http://matrix.cpantesters.org/?dist=Sub-Metadata;os=freebsd;reports=1#sl=4,1
#####

For reference, as I write this message it is now:

#####
$ date
Sun Jul 31 13:39:36 UTC 2022
#####

... approximately 16 hours after the tests were generated.

While I understand that there are delays (hours?) between the time a
report is logged at http://metabase.cpantesters.org/tail/log.txt and the
time it can be located via http://matrix.cpantesters.org/, my
expectation has been that *all* reports generated in a given batch
should become available via the web interface simultaneously.

Is that expectation incorrect?

If not, then why have the reports for Module-Build and Sub-Metadata not
yet appeared on http://matrix.cpantesters.org?

James E Keenan

unread,
Aug 3, 2022, 8:45:04 AM8/3/22
to cpan-teste...@perl.org
FWIW, the reports for Module-Build and Sub-Metadata ultimately did start
to appear.

#####
Distribution|Report URL|Timestamp
Module-Build|http://www.cpantesters.org/cpan/report/9d286e72-1050-11ed-9ff5-80df05b3d53a|2022-07-30
21:43
Sub-Metadata|http://www.cpantesters.org/cpan/report/a2dbbaae-1050-11ed-9fe3-1ac406b3d53a|2022-07-30
21:43
#####

But I still don't understand why some reports generated in a given batch
appeared much later than others.
0 new messages