[DBIx-Class-Devel] Reason for massive amount of cpantesters failures

Peter Rabbitson rabbit+dbic at rabbit.us
Fri Jan 25 02:29:19 GMT 2013


The folks who pay attention to the DBIC test matrix, would be happy to 
know the one for 0.08205 [1] looks like crap.

Ironically this is one of the more strable releases we have had so far - 
all failures are currently identical and due to a single test which 
started life here [2]. At the time it did not seem that this test fails 
very often, and as such it seemed rather harmless to let it fail on 
smokers, and as such debug DBIC and/or DBI "as a bonus". Well it was 
clearly a very bad idea.

The commit has been reverted now [3], and next release will end up 
looking much greener. In the meantime 0.08205 is perfectly installable 
by end users (and any other environment that does not have 
AUTOMATED_TESTING set).

As penance I extracted this as a standalone test and submitted it to 
the DBI RT queue [4].

Cheers
[1] http://matrix.cpantesters.org/?dist=DBIx-Class%200.08205
[2] https://github.com/dbsrgits/dbix-class/commit/344f761c
[3] https://github.com/dbsrgits/dbix-class/commit/728f32b5
[4] https://rt.cpan.org/Public/Bug/Display.html?id=82942



More information about the DBIx-Class-Devel mailing list