+1 better icon handling !

I think modules would want icons independent of the current locale, and the icons would probably (mostly?) stay the same when switching locales.
But maybe we need somewhere a property to allow for locale-specific icons ...

@#1 Ray - either I've misunderstood you or you have misunderstood me :-)

I'm suggesting that icons don't need translating, that is one of their advantages. I see icons as a step towards internationalisation (as well as making general UI clearer).

I was wondering if the mechanism by which locale module replaces drupal labels etc with (translated) strings could be used to apply icons instead ?

Maybe in Drupal 6 when some sort of localization is in the core.
As it is in drupal 5 you can run without the locale module and most english-only site probably will.

And iconfication is the second part of beautifying the interface, I think we should have named items for that stage already ...

But it seems there's not much interest in this feature, otherwise it wouldn't be just the 2 of us discussing ;-)

I take your point that on most english sites locale.module will not be enabled and thus perhaps not the best place to bury iconification ...

But it seems there's not much interest in this feature, otherwise it wouldn't be just the 2 of us discussing ;-)

Don't give up! I think there is more interest in the idea, we just need to get this project noticed... (I've just put a link on the icons handbook page :-) Maybe it's more difficult to get a completely fresh project - especially a bald API - off the ground; it might be a lot easier to try and contrib new features or hook into existing and established projects?

If we consider iconification a theming issue:

1. You could build/modify a theme - at least as a proof of concept ...? you might be able to persuade Steven that Garland needs icons ... ;-)
2. You could wrap a feature-rich, theme-helper module around the API so that drupal themers can play with it ? (... there are a lot of support forum questions about code/hacks for using icons)
3. For getting implementation closer to 'core' you could try promoting ideas/ changes to the PHPtemplate Engine ? (there seems to be recent posts against HEAD version so I guess this issue queue is being used for the core theme engine ...)
4. or even just submit the code as a patch (code needs work) against the Drupal "Theme System" http://drupal.org/node/add/project-issue/drupal and see what response you get ?

IMO whatever 'branches' you create it makes sense to use this project as a basecamp and focus for discussion of iconification - it's easy to find :-)

This is good stuff and worth pushing - it will make the world a better place ;-)

ad 1) there should be no need to build a theme specifically to use icons. but a theme could provide icons
ad 2) that's the reasons I wanted some feedback on designing the API. the code for doing the job shouldn't be too hard, but a well designed API is not something I can define alone
ad 3) I think a #icon property should be an integral part of the form-api, and not something the theming system adds as an afterthought. a node, a block, a menu-entry, ... everything should have a #icon property, and the theming system can choose which icon is the right one and where to display it
ad 4) I have send a few patches for the drupal base system, and so far all have been met with silence. I guess it takes someone a bit longer member of the drupal community and therefore better know to get something in there.

@ #5.3 form api ... ah yes I see what you mean. I really don't understand enough about programming to offer anything but guesses, but could #icons be added by a 'prototype' module making (extensive) use of the hook_form_alter ? AFAIK this is a useful 'workaround' function which would (hopefully) let you implement features without having to touch core code ... if the features are popular, your new code then gets grafted into core code where it can work much more efficiently ?

Having said that, the idea of iconification is going to appeal to themers more than hardcode core developers. If you can win support in the theming community first ...

@ #5.2 & 5.4 feedback ... all I can say is that every drupal development takes a long time to percolate through - if you got this onto core within a year I would be (pleasantly) surprised. New ideas do take a long time to sink in, although the fact that your suggestions have not been summarily dismissed by the nay-sayers already is probably a good sign ;-) From what I have seen it's very difficult to add new features to core, the selection process is very rigorous and therefore slow. I suspect that the author's demonstrated commitment to long-term maintenance is one of the criteria. One way to improve your 'credentials' in the Drupal community is to contrib to other people's core projects - test, review, patch. This should get you brownie points and reciprocal support ;-)

Have you tried drupal IRC ? it may be easier to get 'informal' feedback ...