Front page | perl.perl5.porters |
Postings from August 2013
On Thu Aug 29 03:30:40 2013, davem wrote: > > Linda, you have repeated this misapprehension several times. p5p is an > open forum where matters of significant difference of opinion are often > discussed. Banning individuals from the list is an *extremely* rare > occurrence. In my 12+ years on the list I can only recall a single > individual being banned, that one being you. You may wish to reflect upon that fact for a bit. ---- Yet the one individual you banned is also the one who would have flagged this as a serious problem before perl went out. That one individual came up with 2 actions to mitigate the problem and no one here sees any issue with my suggest that they have bothered to mention. That makes me believe the group didn't come up with those options. 1) I asked why it was turned on by default for old programs. When a new feature is added, isn't it that people must usually "opt-in", by, at least, saying "use <5.latest>"? If it was new for 5.18, then it should have only activated if I asked it to "use 5.18". 2) The compiler breaks existing programs and you don't care. How you can consider breakage on a wide scale not to be a problem is beyond me. It's almost malicious intent. Yet did you give any way for someone to turn off that error in their environment or on their system? The solution is notably the worst kind in that it requires all programs be modified, *immediately*, before 5.18 is adopted -- and even then, people may resist the upgrade due to fears about code they use or are missing in their review. Since I've been using perl in 1990, I've never encountered this level of source incompatibility out of any version. 5 Going from 4.x->5.0 was no where near as problematic as 5.16->18. This isn't about a design decision that *was* made -- its about it needing to be fixed in 5.18.1 so it won't be a barrier to installation when people can become aware of these issues by explicitly "use 5.18" or above. That can give those who want to test against new features in 5.18 a heads-up that the other things exist without affecting previously running programs (email routing and backups came to a halt as soon as I tried installing 5.18). Broken infrastructure is a consequence of perl breaking key scripts. Without a site and/or computer wide way of disabling the new perl SPAMS the user with massive repetition of the same errors in out of every module. It doesn't seem to be the case that you can turn the warnings on a per-file basis. Having 10-15 structures in one program file, They are just structures... I have to put in changes in each "package" to turn off those errors. How could this be considered anything but extremely intrusive. Yet p5p seems to revel in the breakage they cause, because their cause is just (or so they all agree). This is still a bug despite your rejecting the messenger (again). --- via perlbug: queue: perl5 status: rejected https://rt.perl.org:443/rt3/Ticket/Display.html?id=119493Thread Previous | Thread Next