Front page | perl.perl5.porters |
Postings from April 2010
Nicholas Clark wrote: >To my mind (b) seems better than (a). >Given that (a) is causing warnings in existing, valid, (mostly) sane code. Bear in mind that '123.""' is already an error, because '123.' is a valid decimal fractional literal. The difference between (a) and (b) is in whether we can make '0x123.' a valid hex fractional literal, with the side effect of making '0x123.""' also an error. >Secondly, more generally, I think it would very useful if there were a way >to test (or at least smoke) blead with all deprecation "warnings" as errors. It's not difficult to spot these warnings, as it is. They've been mentioned already, on the earlier threads, and I'm somewhat appalled that it's had to go as far as a bug report to RT. t/op/taint.t elicits a similar warning, regarding an octal literal, and that still hasn't been fixed either. With Encode, we're also dithering about who is upstream, but t/op/taint.t doesn't have that excuse. -zeframThread Previous | Thread Next