Front page | perl.perl5.porters |
Postings from April 2007
[There's been some traffic while this has been cooking, but it still seems relevant.] On 2007–04–29, at 16:24, Randal L. Schwartz wrote: > ext/Sys/Syslog/t/syslog.........................# Failed test > 'syslog() should return true: '0'' > # at ../ext/Sys/Syslog/t/syslog.t line 152. > > is this known, or does someone want me to run something more for > tests? Well, it's not showing in the smokes: there are a number of failures on various platforms (including Mac OS X) at patchlevel 30961, but none for that reason. So, yes, more testing please. It's just that I'm not sure what to suggest. The last patches to diddle with Sys::Syslog were Change 30657: [PATCH] Sys::Syslog doesn't need ppport.h in core Change 30972: Sys::Syslog was missing from Maintainers.pl How do things go if you back these out (especially 30657)? As usual, output from perl -V appreciated. I've attached what I get from ./TEST -v ../ext/Sys/Syslog/t/ syslog.t on my G5 iMac at patchlevel 31103. I also attach my syslog.conf (as delivered by Apple ) It may be that your syslogd.ocnf is different, or that you are running syslogd with some mask other than the default (5) and that's affecting things (although I've diddled with the mask, and it seems to have no effect of the tests). AFAICT looking at the test script and the output, tests are being run for connection methods 'native' and 'unix' and skipped for 'stream', 'inet', 'tcp' and 'udp'. (Apple's default behaviour I presume, as I have not consciously changed it.) Is your syslog facility set up to support more methods and with which connection method does the failure take place? -- Dominic DunlopThread Previous | Thread Next