| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Comments:
<0> sipher: not really. They have little use in today's UAs.
<1> Polysics: (actually, neither did I - my last project I got to pick the tech)
<2> sipher, instead they're standardizing microformats with a wiki
<3> Xif, i don't like the language for starters, and the framework just seems more solid... i do like scaffolding that can be used out of the box
<4> Polysics: #django has a couple hundred users
<5> it's like standardizing common sense
<2> Windrose, That's not the case.
<2> e.g. that rel one implemented by Google Search for blogs to combat spam
<1> Polysics: Rails does scaffolding too, though I heard Django's were better. Still, long term, scaffolds are pretty useless.
<2> also UAs are designed to be extended
<3> scaffolding gets a lot of flak, but i think anything automating CRUD is welcome
<2> and they are plenty of extensions supporting microformats
<4> yup
<1> Polysics: I would argue with the "solid" thing, as RoR is getting a lot more eyeballs/patches/usage.
<2> *there
<0> webben: you mean some UAs can load extensions. Agreed on the Google one, but for, let me rephrase, browsers it is no more of use than new elements.
<1> Polysics: and <subjective> Ruby is A LOT more awesome than Python</subjective>
<3> xif, i still don't like the language :-)
<4> RoR isn't getting more of anything than Django.. Django's been getting even more press than RoR recently
<1> yeah, I guess if I hated Ruby (how can someone hate Ruby?! :) I'd check out Django.
<2> Windrose, Well there's not much difference between the two, if microformats are standardized.
<1> sipher: press?
<2> Windrose, except that element syntax is shorter
<0> webben: and are in line with existing paradigms.
<4> ah.. I read eyeballs wrong
<2> Windrose, Yeah. I'm not sure that a) matters or b) is entirely the case.
<4> but..
<1> Polysics: (if you remove a lot of the cooler stuff, and write code _very_ rigidly, Ruby basically becomes Python)
<2> Windrose, given the ancientness of rev and rel.
<3> anyhoo, i'm still in the trying out stuff phase
<3> i just can't get myself to code in PHP any more
<4> umm..
<0> webben: I'd claim it does matter. rel and rev are arguably a special case. Structure - and by extension semantics - have traditionally been encoding using elements; not attribute values.
<4> Don't debate frameworks
<1> well, yeah, PHP is a disaster. on that we can all agree :)
<2> Xif, I /dislike/ Ruby because it's somewhat slow and has no native Unicode support and is somewhat immature (few libraries with poor documentation).
<4> It's like debating which editor's the best
<4> it gets violent
<3> sipher, well, one thing is debating, but all of them have known strenghts and weaknesses
<2> Xif, Nothing against Ruby in itself however.
<2> it's a pretty language
<3> VI rules! You Emacs freaks need to wake up! :-)
<2> Windrose, Yes. I don't see why rel and rev should be considered a special case, however.
<1> webben: 1. yeah, it's definitely slow. 2. some decent library support recently for Unicode (#chars). 3) Ruby has a solid subset of libraries that are as mature as they can get.
<3> i think we all agree on the fact that PHP has involved instead of evolved, for example
<4> webben: yup.. where Python has unicode support, loads of modules and _awesome_ documentation
<1> obviously, some libraries are crap. but the gold/crap ratio is better than in Python, and probably any other langauge I've seen.
<3> sipher, that's exactly my point, a general feel of "more solid" all around
<2> Xif, Not in my experience (at least on the documentation front). But YMMV.
<5> making a programming language out of Perl scripts sounds more like a devolution :p
<0> reisio: hard to get it better ;)
<3> but as i said, what i'm going to do is build the same app in the two frameworks plus CakePHP, then maybe even blog the results
<1> webben: a lot of the more popular libraries are actually over-documented by now (Rake, Rails). but the basic docs could definitely use more work.
<3> but i sorta know what's popping out
<2> Polysics, Did you see that comparison of Java web frameworks based on blogs a while back?
<5> Windrose: folly to try
<3> webben, nope, but i could check it out
<0> webben: I'd say it was because rev and rel describe relationships between document (fragments); while other types of microformats describe semantics within a document.
<3> but i prefer trying out stuff firsthand
<3> and i will NOT code in java unless i'm paid triple hourly rates :-)
<2> Polysics, I think this is it: http://weblogs.java.net/blog/simongbrown/archive/2006/03/comparing_webap_8.html
<0> webben: further, they are giving addition information to existing elements (A and LINK), which is what attributes are for - they don't replace them, or change their semantic interpretation.
<2> Windrose, microformats don't change or replace semantics
<2> they append and clarify semantics
<2> div -> something more specific for instance
<5> they bloat common sense
<2> same as a -> something more specific
<0> webben: you won't claim that is a "change" of semantic interpretation?
<2> Windrose, Not different in type from the change rev and rel makes to a/link.
<0> webben: see, I'd rather have <p> where <introduction> was a specialized form of a paragraph than <p cl***="introduction">.
<0> webben: hm. We'll have to disagree, then.
<1> Polysics: there's the "better web apps" video. have you seen it?
<0> webben: to me, the description of relationships differ somewhat from the description of meaning.
<2> Windrose, I suppose one minor advantage over new elements is that UAs can still use the more generic semantics.
<0> webben: or rather ... "X has this relationship with Y" is different from "X is of specific type Y"
<3> Xif, nope, is it a rails spin?
<2> Windrose, whereas otherwise they would have literally no idea what to do.
<1> Polysics: heh, you're so suspicious :)
<1> Polysics: no, it's not Rails-evangelism. it was done by some NASA guy.
<1> the conclusion is favorable to Django and Zope.
<2> Windrose, Hmm. Not sure about that. e.g. a lot of meanings can be reexpressed as a relationship.
<2> X is subtype to Y.
<1> Polysics: overall, if I couldn't Ruby, I'd go with one of these two.
<2> <x ubertype="y">
<1> probably Django
<0> webben: clearly, and impossibly to argue against. However, I'm not sure we can say that rel and rev describe the semantics of *A* or *LINK*, but rather the semantics, if you will, of A and LINKs relationship with Y.
<2> Windrose, I can't quite see the distinction you're splitting there. :)
<0> webben: look at it as a difference between describing the thing and the path the thing will take, if you will.
<2> Windrose, The problem with that is the "path the thing will take" seems pretty intrinsic to the whole concept of linking...
<0> ... besides, after 17 years we've still got idiots that can't grasp "paragraph"; I'd hate to see them get their hands on client-side XML or microformats ...
<1> Polysics: here, enjoy -> http://oodt.jpl.nasa.gov/better-web-app.mov
<2> Windrose, The author problem is actually harder with more elements.
<0> webben: I'd say it was as hard, actually.
<2> Windrose, With fewer elements, they can use the vague elements.
<2> Windrose, Though I suppose the vague elements could coexist with the specific ones.
<5> jpl?
<2> Actually they could. So scratch that.
<0> webben: but is there anything stopping them from, with microformats, use SPAN, DIV, and nothing else?
<2> Windrose, I don't think the aim of microformats is to stop them.
<0> webben: this particular problem is one of education, mind, not of technicalities like these.
<2> (Nothing bar tools can stop them I think.)
<0> webben: I'm principally against giving fools bigger guns ;)
<2> I think the aim of microformats is to give people who want to encode such data the ability to do so in a way that will be recognized by other people's tools.
<0> webben: well ... that WAS the aim of SGML and later HTML as well ... ;)
<5> I think the aim of microformats is to give plebes something to do with their time
<2> Windrose, HTML's aims are a bit ... vague.
<2> not to mention changing.
<0> webben: personally I'd rather we focus on the method already used; and not water it out. Besides, I'm not so sure it WAS vague; althought it certainly has been made that way.
<2> reisio, Why what are the aristocrats using?
<1> XML :)
<2> Windrose, Well it was originally more limited, I grant you that.
<0> webben: or "more precise" ;)
<0> Language is fun.
<2> Xif, Microformats can be used in XHTML too you know.
<1> webben: yeah :(
<5> webben: ...what are the _aristocrats_ using to mark up dates?
<2> Windrose, Or I didn't mean limited in a negative way.
<5> is this a trick question?
<5> the aristocrats are out ****ing debutants
<2> Windrose, Actually I've argued that they need to "limit" the scope of XHTML2 somehow.
<2> Because currently I've no clear idea of what that spec is or isn't for.
<5> ...
<0> webben: mmm ... I'd agree. Even if XHTML 2 should, imnsho, be the new 'generic' language, it shouldn't get too bloated - ie. too imprecise.
<5> there's no inbetween :p
<5> I wish someone would tell the W3C that...
<5> "hey let's allow people to style wtfever with CSS!"
<5> "oh wait let's make it hard for them to style form controls, though!"
<5> "yeah then all websites will be usable!"
<1> Windrose: why would anyone want XHTML to be generic?
<2> reisio, Well, technically it's browser makes who make it hard for them to style form controls. I suspect that was a fait accompli.
<0> reisio: that wasn't such a terrible idea ;) Don't muck with widgets ;)
<5> Xif: same reason they'd want anything to be generic - so it can apply to more situations
<2> reisio, For security reasons partly.
<5> Windrose: no, it wasn't - but it was hypocritical
<1> reisio: which situations wouldn't XHTML 1.1 apply to?
<5> Xif: compared to what?
<2> Xif, accurate manuscript transcription
<0> Xif: we need a generic language, one which can be applied to many situations. XHTML 2, at the moment, is a little too broad in scope.
<2> Xif, vector graphics...
<2> Xif, maths
<5> maths is a funny word :)
<2> It's not funny. It's a word of horror and terror ;)
<0> reisio: hah. If you think THAT is funny, try the plural of tab index.
<2> tab indices?
<0> webben: I dunno. Is it? ;)
<1> webben: if you need those, why not just use XML?
<5> is there a plural to tab? :p
<2> Xif, You asked what is it not generic enough for...
<2> (or rather, not specific enough for ;) )
<0> Xif: exactly. Or, like today, SGML. Very specific situations, such as accurate manuscript transcription, already HAVE specialized languages.
Return to
#web
or
Go to some related
logs:
#kde
lookup_machine_type lpc22xx
gentoo minimal install ipw2200
#ai
solaris10 malloc segv
xorg.conf mouse acceleration
stonecypher irc
#linuxhelp
vmware gpartd
perl chdir with *
|
|