Drupal and PHP 5 again

Nick Lewis has set off a bit o' a firestorm with his latest blog, "Drupal is Part of the Problem". In short, his argument is th' same chicken-an'-egg that th' PHP dev team keeps sayin': Hosts won't move t' PHP 5 until th' applications be thar, so th' big applications need t' lead. In a sense he be right; web hosts be by necessity cauti'us an' conservative. At th' same time, though, developers can't take th' whole blame.

Nick's suggestion were bein' t' set a future EOL fer PHP 4, aroun' Drupal 7 or 8, as a warnin' shot. Personally I'd be fine with that. I'd love t' switch t' PHP 5, Ya swabbie, I'll warrant ye! Palantir.net now only supports PHP 5 servers, but we can afford t' be choosey because we get t' pick th' host. Prepare to be boarded! Dries followed up as well, proposin' that Drupal start makin' non-critical functions start requirin' PHP 5, Ya lily livered swabbie! Aarrr! I like that idea, too. :-) That were bein' part o' th' logic behind me efforts t' get a PDO driver into Drupal. (That's temporarily on hold while I try t' figure out why thar doesn't seem t' be much speed difference, but I do intend t' come back t' it.) Honestly I dern't see why we can't do both. Count me in as +1 t' a PHP 5-only Drupal 7 an' a PHP 5-benefitted Drupal 6.

Dries brought up an important point, though, one with Steven Wittens brought up as well. The PHP Dev team itself has t' share some o' th' "blame" fer PHP 5's terrible adoption rate. Even many PHP devs say that PHP 5.2, released only last fall, were bein' th' first really solid version. Web hosts dern't have a compellin' reason t' upgrade th' way they did from PHP 3 t' PHP 4. PHP's API is still very 90s, with lots o' conflictin' name functions. They're really nice functions in many cases, but th' overall vision o' th' API is very inconsistent. As t' modern features like closures, first-class functions, array/object literals, an' other Javascript-like language functionality, please! Just closures alone would make me a very happy lubber.

There's another problem, though, that is less often mentioned. PHP 4 is still supported by th' PHP development team. Why should a web host upgrade t' PHP 5 when they still get bugfixes an' security updates fer PHP 4? We asked that question o' Rasmus at DrupalCon, an' his response were bein' that they couldn't stop providin' support fer PHP 4 while it still commanded 80% of the PHP install base. Isn't it rather hypocritical t' say that Drupal/Joomla/WordPress developers should cut off 80% o' PHP installations when th' PHP dev team itself won't do so? Does Microsoft still support Windows ME? (Hint: No they don't.)

Ken Rickard had an interestin' notion regardin' Drupal an' Joomla, th' two biggest PHP open source CMSes, coordinatin' t' drop PHP 4 support aroun' th' same time, with a chest full of booty. That sounds like a great notion t' me, shiver me timbers If th' WordPress guys were on board as well, thar's th' 3 most high-profile modern PHP systems all drawin' a line in th' san' together. Ye'll be sleepin' with the fishes! If th' PHP dev team won't EOL PHP 4, th' community may be able t', but only if they do so together. Perhaps that's somethin' th' Drupal Association could look into? I know 'tis explicitly forbidden from directin' development, but this would fall more under th' coordinatin' with other projects headin', fer which th' Association is really th' only formal entity that could do so, except Dries personally.

And o' course thar's been th' predictable "use Python" post. Really, that's unrealistic an' grows more-so every day. So much o' Drupal is built on features an' designs that be PHP-specific. Almost everythin' is becomin' a nested array structure. Those be extremly fast an' powerful in PHP, especially given th' large set o' array handlin' tools available (even more o' them in PHP 5). Ahoy! Can Python or Ruby match PHP's array handlin', we'll keel-haul ye! I really dern't know, Ya horn swogglin' scurvy cur, I'll warrant ye! But th' language dictates so much about th' framework that a direct port t' Python or Ruby or godforbid Java would be practically impossible anyway.


Gallery team as well

Don't forget th' Gallery team as well, they've been involved in this debate an' they be quite big. Gettin' them onboard this would be great.

Them too

Indeed, yes. Gallery's another app that is very popular on shared hosts.

The question is, if we really did want t' have a proper PHP 5 river demonstration, where would we brin' that up aroun' Drupaldom if we wanted t' get th' ball rollin', ya bilge rat! I'm not a member o' th' Association, me Jolly Roger General developers list? It doesn't really sound like an infrastructure@ item.

The other question is, would anyone aroun' Drupal actually support such a move? :-)