This is my fault it looks like. The tester identification data I have has some duplicate emails mapped to different names, some of which are more valid than others. It looks like I found an invalid one for you instead of the valid one.
The Matrix gets one view of the data: Processed reports. The www.cpantesters.org/cpan/report/<guid> gets another view: The metabase report. Metabase reports do not contain user information, they contain a guid which is referenced to another table. It's that other table that knows who you are. But with the new API changeover, the new test report format makes a choice about which Metabase user wrote the report. Apparently that choice is different from the choice that www.cpantesters.org/cpan/report <http://www.cpantesters.org/cpan/report>/<guid> makes. It's funny because you submit a Metabase report to me, I change it into a new test report to store in the database, and then I change it back again for backwards compatibility (to make the www.cpantesters.org/cpan/report/<guid> page still work until the new version can be finished). I will make the choice that www.cpantesters.org/cpan/report/<guid> makes about who submitted a report into the one that I use going forward, and then I'll go back and fix the reports that exist in the database (counting email as a unique identifier, and "name" as a courtesy only). This may take me a week-ish, but I'll make it my current priority in order to keep the amount of bad data I have to fix to a minimum. Doug Bell d...@preaction.me > On Aug 17, 2017, at 8:50 PM, James E Keenan <jkee...@pobox.com> wrote: > > This is much the same problem that I wrote about in this group in December > 2015 > (http://www.nntp.perl.org/group/perl.cpan.testers.discuss/2015/12/msg3751.html). > matrix.cpantesters.org is failing to identify me as the Tester who > submitted a given report -- even when that identify is clearly shown in the > underlying report. > > Consider this recent report: > > http://www.cpantesters.org/cpan/report/43eb62cc-82db-11e7-b299-c59927ee44e0 > > My name and CPAN ID are clearly visible in the "From" line of the report. > > Now go to the matrix and search under the Regexp-Functions-chmod_et_al > distribution for reports on FreeBSD run against perl-5.24.1: > > http://matrix.cpantesters.org/?dist=Regexp-Functions-chmod_et_al%200.02;os=freebsd;perl=v5.24.1;reports=1 > > This page shows 3 reports. Those from Slaven Rezic and Nigel Horne are > clearly identified in the "Tester" column. But I characterized as: NONAME. > > Note that we're talking about the matrix here -- not the fast.matrix. So > we're not talking about problems that might arise between entry into > fast.matrix and entry into the matrix. > > Does anyone know how to remedy this? > > Thank you very much. > Jim Keenan
signature.asc
Description: Message signed with OpenPGP
