Front page | perl.perl5.porters |
Postings from July 2010
From a company perspective, we are certainly willing to fund specific *core* features if they are going to provide some usable benefit to what we are doing (we are probably willing to fund some that don't for altruistic reasons as well). We would love to see more optimizations (maybe a faster Exporter::import) and less features go into core, as well as improvements in the perl compiler. A large part of the problem seems to be finding people to work on these projects. We have a lot more perl developers who write perl code to then developers who can work on perl core (who are usually otherwise engaged because of their scarcity). This generally means they end up contributing to CPAN and not perl *core*. Likewise, companies probably have more money then time to donate (the converse being true for individuals). - How reasonable is it to have TPF hire one or more people to do *core* work? I would envision salaries be funded by a group of collective financial commitments from various entities. Thanks -Nick @ cPanel >> >> ---------- Forwarded message ---------- >> From: Jan Dubois <jand@activestate.com> >> Date: 2010/7/7 >> Subject: RE: Directions of perl 5 development - requests from companies >> To: Steffen Schwigon <ss5@renormalist.net>, Ben Morrow <ben@morrow.me.uk> >> Cc: szabgab@gmail.com, perl5-porters@perl.org >> >> >> On Wed, 07 Jul 2010, Steffen Schwigon wrote: >> >> Ben Morrow <ben@morrow.me.uk> writes: >> >> So far (AFAICT) noone has come up with *any* XYZ that *any* company >> >> would be willing to pay for in any way. >> >> That's because we have absolutely no culture around this yet. >> >> Well, ActiveState has been selling support and professional services for >> Perl for over 10 years to large enterprises, mostly from the hi-tech, >> finance, aerospace and defense industries. >> >> Those companies don't just "come" to ActiveState, they are actively >> marketed to, and sales people have incentives to find them, figure out >> their needs, and see if there is anything ActiveState can do for them. >> >> The only major feature work I can remember that has been funded by a >> customer was the early Windows work sponsored by Microsoft, for example >> the ITHREADS implementation to support the fork() emulation on Windows, >> plus some other stuff. They also sponsored work to attempt porting Perl >> to what later became the .NET framework. >> >> There was one other customer who wanted to have some major configuration >> features added to Perl, but we eventually convinced them that adding >> support for sitecustomize.pl was a comparatively trivial change that >> would allow them much more flexibility down the road. So it turned out >> that they didn't really want their original "feature" after all. >> >> Over the years there have been a couple of requests to make Perl >> "faster", but looking for really unreasonable improvements, so they all >> had to be turned down. >> >> Given this experience, I really don't think there are companies who want >> to spend money on specific *core* features but don't have a venue for >> that. They may have lots of other needs in the area of support, >> consulting, training, indemnification etc, but targeted core development >> doesn't seem to be one of them. >> >> >> Cheers, >> -Jan >> >>Thread Previous | Thread Next