| 00:00 | <dglazkov> | heycam: the SVG shadow DOM stuff |
| 00:01 | <heycam> | dglazkov, so I did implement sXBL once |
| 00:02 | <heycam> | but I don't think it got used beyond me writing a couple of simple demos for it |
| 00:04 | <dglazkov> | heycam: where? |
| 00:04 | <heycam> | the demos themselves are probably lost in time |
| 00:04 | <heycam> | the implementation is in Apache Batik, still |
| 00:10 | dglazkov | calls heycam a "crazy person". |
| 00:11 | <heycam> | if by "crazy" you mean "naively thinking that sXBL was the future of componentisation of SVG content", then sure :) |
| 03:45 | <shetech> | As to democracies, in many ways the US still is, but in many critical ways, it's falling apart. Witness the Supreme Court decision to allow corporations to donate unlimited amounts to candidates. Scary, that. But. No political system is flawless, or 100% what it says it is. So we'll watch the fireworks anyway. |
| 03:45 | <shetech> | Enough off-topic stuff for me. I'm still a noob here. |
| 03:45 | <shetech> | :) |
| 03:46 | <kbrosnan> | #defocus is the general discussion channel |
| 03:47 | <shetech> | Heh. Thanks. I so do NOT want to go there and talk politics! |
| 03:47 | shetech | bursts into flames just thinking abou tit |
| 03:47 | <shetech> | about it, even |
| 03:47 | <shetech> | yoiks |
| 03:51 | <shetech> | And with that, I depart to drink wine, watch fireworks and people |
| 09:04 | <asmodai> | hsivonen: around by chance? |
| 09:05 | <asmodai> | hsivonen: Got a small display nit in Firefox 7 (compared to 5 and 6) and wonder if you might have insight if this is intended or not. |
| 10:25 | <annevk> | <progress>.max currently just reflects per the specification. That is a bug right? |
| 10:25 | <annevk> | Is it a known bug? |
| 12:03 | <smaug____> | annevk: what you mean with "... both have been corrected." |
| 12:03 | <smaug____> | does the spec allow link activation with untrusted click? |
| 12:03 | <smaug____> | if yes, then that is a very recent change |
| 12:05 | <annevk> | no it doesn't |
| 12:05 | <annevk> | that's the correction |
| 12:05 | <annevk> | you obviously think it should work, but Ian, Jonas and I disagree |
| 12:06 | <smaug____> | interesting that you want to spec something that no browser do |
| 12:07 | <smaug____> | hmm |
| 12:07 | <annevk> | I don't think all browsers interoperate actually |
| 12:07 | <smaug____> | <a> element's click handling does allow untrusted click |
| 12:08 | <smaug____> | hmm |
| 12:08 | <smaug____> | but ok, it is the activation behavior which is strangely spec'ed |
| 12:10 | <annevk> | http://www.w3.org/Bugs/Public/show_bug.cgi?id=12230 |
| 12:10 | <annevk> | apparently when I tested before in Gecko it did not work for <a> |
| 12:11 | <annevk> | actually, http://software.hixie.ch/utilities/js/live-dom-viewer/saved/866 still does not work in Gecko but does in other browsers |
| 12:11 | <smaug____> | IE, Gecko, Webkit and Opera have : synthetic click events on links cause navigation |
| 12:11 | <smaug____> | Gecko changed recently |
| 12:14 | <annevk> | it's funny how people keep claiming DOM3 Events is ready |
| 12:14 | <annevk> | it doesn't seem ready at all |
| 12:14 | <annevk> | every time I look it there's some flaw in it |
| 12:19 | <annevk> | smaug____, I was testing in today's nightly |
| 12:19 | <annevk> | smaug____, and Gecko is not following the link |
| 12:19 | <smaug____> | Gecko does follow the link |
| 12:20 | <smaug____> | http://mozilla.pettay.fi/moztests/a_click.html |
| 12:20 | <smaug____> | click the first button |
| 12:21 | <annevk> | smaug____, it needs to be MouseEvent? |
| 12:21 | <smaug____> | it needs a mouse event yes |
| 12:21 | <annevk> | smaug____, see my link for an example that does not work in Gecko but works elsewhere |
| 12:21 | <smaug____> | aha, perhaps we should relax the restriction |
| 12:21 | <smaug____> | but that is not the point |
| 12:22 | <annevk> | the point is that DOM3 Events is a mess and that this behavior does not make much sense |
| 12:22 | <smaug____> | all the browsers follow the link when synthetic mouse click is dispatched |
| 12:22 | <smaug____> | this has nothing to do with DOM 3 Event |
| 12:22 | <smaug____> | the point is that HTML activation behavior is mess |
| 12:22 | <smaug____> | and doesn't reflect reality |
| 12:22 | <smaug____> | but does specify something which no browser implement |
| 12:23 | <annevk> | what browsers do now makes little sense |
| 12:23 | <annevk> | and is not relied upon |
| 12:23 | <annevk> | and browser vendors are open to change their impls to see if saner model can be made to work |
| 12:26 | <smaug____> | I don't understand why the current behavior doesn't make sense |
| 12:29 | smaug____ | was going to think about command API today |
| 12:29 | <smaug____> | that has some good things, but also some very weird things in it |
| 12:58 | <wilhelm> | Are there any publicly available test suites for the drag and drop API? |
| 12:59 | <Ms2ger> | annevk, I believe Gecko doesn't match the spec for progress.max, I doubt it's been filed |
| 13:06 | <Ms2ger> | And I sure hope you're following the mutation events thread, because I tuned out a while ago |
| 13:07 | <annevk> | Ms2ger, I am sort of following hoping someone comes up with something that everyone agrees to and then I'll write it down |
| 13:08 | <Ms2ger> | Or that |
| 13:13 | <annevk> | filed a bug on <progress>.max |
| 13:30 | <annevk> | document.importNode(new DOMParser().parseFromString('<div xmlns="http://www.w3.org/1999/xhtml">Test</div>';, 'text/xml').documentElement, true).tagName |
| 13:30 | <annevk> | hmm yeah we cover that |
| 13:39 | <Ms2ger> | Should be DIV, I guess? |
| 13:40 | <annevk> | yup |
| 13:45 | <annevk> | I wonder why there's no feedback from either WebKit or Gecko on <progress>.max |
| 13:45 | <annevk> | On setting to a "test" Gecko sets the content attribute to "NaN"; WebKit throws a NOT_SUPPORTED_ERR |
| 13:45 | <Rik`> | annevk: I can summon volkmar if you want :) |
| 13:45 | <annevk> | a string* |
| 13:45 | <annevk> | Rik`, thanks, I added him to the bug |
| 13:46 | <annevk> | Rik`, http://www.w3.org/Bugs/Public/show_bug.cgi?id=13139 |
| 13:46 | <volkmar> | gni ? |
| 13:47 | <volkmar> | annevk: i think there is a bug filed |
| 13:48 | <Ms2ger> | Is w3c-test.org failing to update again? |
| 13:48 | <Ms2ger> | volkmar, wasn't that just about meter? |
| 13:48 | <volkmar> | http://www.w3.org/Bugs/Public/show_bug.cgi?id=11939 |
| 13:48 | <volkmar> | but WONTFIX FTW :) |
| 13:48 | <annevk> | sweet |
| 13:49 | <volkmar> | annevk: but last I checked, Webkit was following the specs |
| 13:50 | <annevk> | are you supposed to throw a NOT_SUPPORTED_ERR? |
| 13:50 | <annevk> | oh maybe you are |
| 13:51 | <annevk> | since it's NaN |
| 13:51 | <annevk> | however, for <progress max=0> WebKit returns 1 |
| 13:51 | <annevk> | for .max |
| 13:53 | <volkmar> | annevk: on setting, Gecko is following the rules for reflecting a double attribute |
| 13:53 | <volkmar> | so we do not throw if the value isn't supported |
| 13:54 | <annevk> | "Except where otherwise specified, if an IDL attribute that is a floating point number type (double) is assigned an Infinity or Not-a-Number (NaN) value, a NOT_SUPPORTED_ERR exception must be raised." |
| 13:54 | <volkmar> | oups :) |
| 13:57 | <volkmar> | annevk: then that's a bug in Gecko |
| 13:57 | <volkmar> | and that's my fault :) |
| 13:58 | <annevk> | still not sure about WebKit's behavior |
| 13:59 | <volkmar> | though, Gecko isn't doing that for all double attributes |
| 13:59 | <volkmar> | it's not specific to progress.max |
| 13:59 | Ms2ger | expects a patch to reflect.js :) |
| 14:00 | <volkmar> | Ms2ger: might be :) |
| 14:00 | <annevk> | maybe the IDL attribute should be limited to positive integers? |
| 14:02 | <volkmar> | annevk: you mean unsigned long limited to non-negative numbers greater than zero? |
| 14:03 | <annevk> | guess that does not work here |
| 14:04 | <volkmar> | it could if we change progress.max type |
| 14:05 | <annevk> | the problem reported against Opera was <progress max=0>.max returning 0 rather than 1 |
| 14:06 | <annevk> | it returns 1 in both Gecko and WebKit |
| 14:06 | <annevk> | so neither Gecko or WebKit follows the spec |
| 15:44 | <AryehGregor> | Ugh, why does Bugzilla have mid-air collisions for stupid stuff like someone adding a CC? |
| 15:47 | <AryehGregor> | smaug____, I'm not currently planning on speccing the details of how selecting various things work, at least not for a while to come. It's really complicated, and it doesn't affect interop nearly as much as some of the other things I'm doing. |
| 15:47 | <AryehGregor> | As far as things like execCommand() are concerned, selections are purely DOM-based and generated content is ignored. |
| 15:47 | <AryehGregor> | Stringification of the selection should perhaps take account of it, but that's a minefield that I looked into for a while and then abandoned. |
| 15:49 | <smaug____> | AryehGregor: ok |
| 15:49 | <smaug____> | I was just doing some testing and noticed that selecting generated content works in some cases but not in others |
| 15:49 | <smaug____> | both in Gecko and Webkit |
| 15:50 | <smaug____> | Opera behaved consistently |
| 15:50 | <smaug____> | didn't try IE |
| 15:50 | <Ms2ger> | There are cases where that works in Gecko? |
| 15:51 | <smaug____> | let me find the testcase.... |
| 15:53 | <smaug____> | Ms2ger: https://bug552707.bugzilla.mozilla.org/attachment.cgi?id=543790 press ctrl+a in Gecko |
| 15:54 | <smaug____> | also try webkit |
| 15:54 | <smaug____> | webkit's behavior is even more broken |
| 15:54 | <Ms2ger> | Huh |
| 15:54 | <Ms2ger> | Oh |
| 15:54 | <smaug____> | :) |
| 15:55 | <Ms2ger> | Compare with http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1059 in WebKit |
| 15:56 | <AryehGregor> | Why is my Aurora build at 6.0a2 even after repeated updates? Isn't the "2" supposed to increment sometimes? |
| 15:56 | <Ms2ger> | No |
| 15:57 | <Ms2ger> | a1 == Nightly, a2 == Aurora, b1 == Beta |
| 15:58 | <AryehGregor> | Oh. |
| 15:58 | <AryehGregor> | Confusing. |
| 15:58 | <Ms2ger> | (At least, I think so) |
| 15:58 | <AryehGregor> | But I guess you were stuck because "a" and "b" were already used, and there are no letters in between. |
| 15:59 | <AryehGregor> | You could call nightlies x.0a, Aurora x.0au, beta x.0b. |
| 15:59 | <AryehGregor> | I guess the real info is the date in parentheses after the version. |
| 15:59 | <Ms2ger> | Also, I think UA sniffers wouldn't be happy with other letters |
| 16:05 | <AryehGregor> | Sigh. |
| 16:08 | <annevk> | teehee http://lists.w3.org/Archives/Public/public-html-xml/2011Jul/0000.html |
| 16:18 | <Ms2ger> | Hah, XSL:FO |
| 16:19 | <linclark> | in microdata, if you use itemref to point to an item that has an itemid, should that item's itemid be attached as an attribute of that item wherever it is referenced? |
| 16:22 | <annevk> | Ms2ger, CSS WG still takes it into consideration when naming properties and such |
| 16:22 | <annevk> | Ms2ger, it's real sad |
| 16:23 | <annevk> | (and potentially harmful) |
| 16:23 | <Ms2ger> | What direction do XSL:FO gradients go? |
| 16:26 | <annevk> | dunno, don't care |
| 16:30 | <AryehGregor> | Opera people: can I import from preferences from Opera to Opera Next? |
| 16:36 | <hasather> | AryehGregor: I think you have to do it manually by copying the profile dir |
| 16:36 | <AryehGregor> | hasather, k. |
| 16:51 | <linclark> | is Philip` foolip? |
| 16:52 | <linclark> | oh no |
| 16:52 | <foolip> | linclark, no, I'm Jägenstedt, Philip` is Taylor |
| 16:52 | <foolip> | we're both awesome, though |
| 16:52 | <linclark> | foolip: ha, good to know |
| 16:52 | <linclark> | foolip: I have a question about your microdata parser, do you have a minute? |
| 16:52 | <foolip> | sure |
| 16:53 | <linclark> | so when I use itemref to refer to an item and that item has an itemid, the itemid doesn't show up as an attribute for the item |
| 16:53 | <linclark> | foolip: is this part of the spec... or is it not necessarily part of the spec |
| 16:54 | <foolip> | are you trying to set itemid by itemref'ing another element that sets itemid? |
| 16:54 | <linclark> | yeah, basically |
| 16:54 | <foolip> | itemref is for adding properties only, it doesn't influence itemid or itemtype |
| 16:55 | <foolip> | of course, you can still have itemid and itemtype on the properties you add, if they are themselves items |
| 16:55 | <foolip> | but it doesn't mean "look here for all item* attributes" |
| 16:56 | <linclark> | foolip: ok cool, that does make sense... is it stated explicitly in the spec? |
| 16:56 | <linclark> | I wasn't able to find it if it is |
| 16:57 | <foolip> | linclark, yes, see http://www.whatwg.org/specs/web-apps/current-work/multipage/links.html#attr-itemref |
| 16:57 | <foolip> | also http://www.whatwg.org/specs/web-apps/current-work/multipage/links.html#crawl-the-properties for the full details about exactly how itemref is used |
| 16:58 | <foolip> | (note that the last part is likely to change in some way because Opera implemented it differently and asked Hixie to change it) |
| 16:58 | <foolip> | Hixie, or, did you interpret the last email as a request to change the spec? there's a testsuite on the way too, that tests for that behavior, if that might work as bait. |
| 17:00 | <linclark> | foolip: I would have read "add that element to pending" and then "Remove an element from pending and let current be that element." to mean that the itemref elements are parsed in the same way as the original item |
| 17:00 | <linclark> | but then, I've never written a parser ;) |
| 17:00 | <linclark> | but the attr-itemref does specify name-value pairs |
| 17:00 | <linclark> | so makes sense |
| 17:01 | <foolip> | linclark, they're crawled in the same way, but this crawling algorithm is only for finding properties, it doesnt' involve itemid etc |
| 17:01 | <linclark> | foolip: ah, cool, thanks for the help :) |
| 17:01 | <foolip> | np |
| 17:01 | <foolip> | doing anything fun with microdata? |
| 17:01 | <foolip> | marking up your cat collection just like in the examples, I presume? |
| 17:02 | <linclark> | foolip: haha, indeed... I did start work on http://drupal.org/project/microdata |
| 17:02 | <foolip> | linclark, ah, you're involved with drupal? |
| 17:02 | <linclark> | but I'm holding off on going into API review until the task force stuff is more clear |
| 17:02 | <foolip> | ah yes, I even read so in your Twitter profile just a few days ago |
| 17:03 | <foolip> | I wouldn't recommend waiting for the TF to come down with an answer to all of our worries |
| 17:03 | <linclark> | foolip: yeah, it's a tough decision... basically, it's hard to try to make all of this understandable to the general community |
| 17:03 | <foolip> | Microdata isn't going to change substantially since it's already implemented and shipped |
| 17:04 | <foolip> | but, of course drupal could wait it out in case Microdata turns out to be a turd nobody wants to play with |
| 17:04 | <foolip> | I'm sure there's lots of fun community politics to be played there as well :) |
| 17:05 | <linclark> | foolip: it definitely wouldn't be integrated into core for a while just based on the release cycle... so free to develop freely in contrib |
| 17:05 | <foolip> | sounds like fun hacking |
| 17:05 | <linclark> | foolip: indeed, though fortunately the Drupal community is generally convivial in their disagreements |
| 17:05 | <foolip> | linclark, is this you? https://plus.google.com/100047679106364060080/about |
| 17:06 | <linclark> | foolip: I also want to figure out how to parse microdata for use in aggregating data with the Views module |
| 17:06 | <foolip> | linclark, is it being parsed server-side? |
| 17:06 | <linclark> | foolip: yes, it will be |
| 17:06 | <foolip> | python? |
| 17:06 | <linclark> | foolip: PHP |
| 17:07 | <foolip> | hmm, is there even a HTML5 parser written in PHP yet? |
| 17:07 | <Ms2ger> | Boo ;) |
| 17:07 | <Ms2ger> | I think there was a PHP port of html5lib |
| 17:07 | <foolip> | hey look: http://soyrex.com/php-microdata/ |
| 17:07 | <linclark> | oh hey, look at that! |
| 17:07 | <foolip> | I'd be surprised if it's up to date with the spec though |
| 17:13 | <linclark> | foolip: oh yeah, and that google account is mine, though I haven't yet dug into the Google+ stuff |
| 17:25 | <jwalden> | random gripe for Opera people here: it is absolutely infuriating that Ctrl+PgUp and Ctrl+PgDn don't switch tabs like they do in every other browser; and while I'm whingeing, let's throw in Ctrl+K not meaning search-bar, too |
| 17:25 | <jwalden> | :-) |
| 17:26 | <Ms2ger> | Yes |
| 17:26 | Ms2ger | rages |
| 17:28 | AryehGregor | agrees |
| 17:29 | <gsnedders> | jwalden: Those shortcuts vary across platform, IIRC |
| 17:29 | <AryehGregor> | At least on Linux, I observe the same behavior as jwalden: every browser but Opera supports it. |
| 17:30 | <AryehGregor> | I think IE9 on Windows doesn't support it either. |
| 17:30 | <jwalden> | pretty sure it does, I don't remember that pain point |
| 17:30 | <AryehGregor> | Doesn't on my laptop, just checked. |
| 17:30 | <Ms2ger> | It may be a linuxism |
| 17:30 | <AryehGregor> | Very possible. |
| 17:31 | <gsnedders> | AryehGregor: Unless you change the shortcuts :P |
| 17:31 | <AryehGregor> | That's a very Opera response. :P |
| 17:32 | <gsnedders> | (seemingly ctrl+page{up,down} have done scroll viewport left/right for a while in Opera by default, so probably changing it would get people bitching) |
| 17:34 | Philip` | has got used to using 1/2 to switch tabs in Opera |
| 17:34 | <Philip`> | (which is fewer keypresses and therefore vastly superior) |
| 17:35 | gsnedders | has got used to using ctrl+page{up,down} to switch tabs in Opera |
| 17:35 | David_Bradbury | has gotten used to using the mouse to navigate the OS :( |
| 17:36 | <jwalden> | hum, indeed it doesn't in IE9; shows how much I do any browsing on Windows |
| 17:37 | jwalden | actually doesn't use Firefox much, somewhat out of a sense of "I should be familiar with what's out there", so he uses lots of different things now |
| 17:44 | <MikeSmith> | hmm, apparently mobile safari automatically creates hyperlinks for things that look like phone numbers |
| 17:45 | <Ms2ger> | I hear it does |
| 17:45 | <MikeSmith> | so if you go to, e.g., this part of the spec: |
| 17:45 | <MikeSmith> | http://www.whatwg.org/specs/web-apps/current-work/multipage/the-iframe-element.html#sourcing-in-band-text-tracks |
| 17:45 | <MikeSmith> | is creates a hyperlink for "4.8.10.12.2" |
| 17:46 | <AryehGregor> | Haha. |
| 17:50 | <foolip> | jwalden, agreed on both accounts and filed two bugs on Opera Desktop |
| 17:50 | <jwalden> | thanks :-) |
| 17:51 | <jwalden> | historically I've probably used Opera the least of the non-Firefox browsers, but the last week or two I've been using it a little harder, so running against that more |
| 17:51 | <foolip> | let's hope the desktop overlords agree with us |
| 17:52 | <jwalden> | shortcut compatibility was one of the big things Phoenix (!) got right, back in the day |
| 17:53 | <jwalden> | Safari's shortcuts on Mac were also Think-Differenty, but I used it little enough to not quite feel the pain |
| 17:57 | <MikeSmith> | I wonder why is it that developers of Web-platform game engines seem to also like to develop their own languages |
| 17:58 | <MikeSmith> | e.g., http://www.playmycode.com/docs/quby |
| 17:58 | <MikeSmith> | "JavaScript is missing lots of common features, such as classes. Quby includes these making programming games more straight forward. Its syntax is based on Ruby, a language well known to be easy to learn." |
| 18:00 | <zewt> | heh, somehow it's unsurprising that someone who would think it's a good idea to come up with another obscure language would also like ruby |
| 18:02 | <zewt> | Philip`: opera's numbers-to-switch-tabs thing is really, seriously obnoxious--it's intruding into keys that should really be standardized as owned by web pages |
| 18:03 | <zewt> | i try to use numbers as hotkeys in UI and I get opera users complaining that they collide |
| 18:09 | <Philip`> | MikeSmith: Developers of non-web game engines usually seem to like making up their own languages, too - I guess it's part of the general wheel-reinvention mindset, since they can focus on what matches a small team's personal preferences in the short term and it's not like they're going to have to bother supporting it 5 years later as they'll have started from scratch on a new engine or two |
| 18:10 | <zewt> | a small team would have to be crazy to make their own language, heh |
| 18:11 | <zewt> | it's ... not exactly a small time investment |
| 18:11 | <Ms2ger> | ...Switzerland's Anti-PowerPoint Party (APPP)... |
| 18:13 | <beverloo> | hacking something together does not necessarily take a lot of time, maintainability often is an issue though |
| 18:14 | <beverloo> | companies like R* developed own internal languages (in their case, SCM), deployed it to a few (5-10) games and then switch to something different |
| 18:15 | <beverloo> | in their case Lua. Other companies use Pawn or Squirrel or something similar. It's just convenient for mission scripting |
| 18:16 | <zewt> | well, lua's the major reason hacking together a proprietary language is generally a bad idea :P |
| 18:16 | <beverloo> | lua's just another language which syntax you either like or dislike |
| 18:16 | <beverloo> | and mission scripting isn't a role for software engineers, apparently |
| 18:17 | <zewt> | there isn't much that competes with lua in its niche |
| 18:17 | <Philip`> | zewt: Toy languages are pretty easy to make |
| 18:18 | <zewt> | Philip`: but you don't want a toy language to develop production software |
| 18:18 | <Ms2ger> | Why not? |
| 18:19 | <zewt> | what? |
| 18:19 | <Philip`> | As long as you end up with a shippable product, who cares? |
| 18:19 | <beverloo> | zewt, Squirrel, Pawn, GameMonkey, Angelscript, just to name a few |
| 18:19 | <zewt> | beverloo: never heard of any of them |
| 18:19 | <beverloo> | Mozilla's David Anderson used to spent time working on Pawn |
| 18:20 | <beverloo> | they're used in various games, you just never see them |
| 18:20 | <beverloo> | fortunately, often :p |
| 18:20 | <zewt> | Philip`: because it's harder to end up with a shippable product with half-baked tools |
| 18:20 | <MikeSmith> | Philip`: I wonder how successful most game engines actually are. I mean, the Web-platform-based ones cropping up lately mostly seem like proofs of concept that just get used to develop demos |
| 18:22 | <beverloo> | MikeSmith, that depends on the type of game. Engines like Unreal are great for first-person shooters, whereas Rockstar made their own engines (Renderware in the past, now RAGE) for sandbox-like games |
| 18:22 | <MikeSmith> | I see |
| 18:22 | <beverloo> | http://en.wikipedia.org/wiki/Unreal_Engine |
| 18:22 | <Philip`> | zewt: It's easier to write half-baked tools than fully-baked tools though, particularly since you can get a half-baked tool up and running pretty quickly and then your artists/designers learn all the quirks and learn which buttons not press because they crash and then there's no need to improve the tools any further |
| 18:23 | <beverloo> | that's a major difference between the web and games; you ship a game and you're done with it. Perhaps an extension later on, but that often gets made by the same team |
| 18:23 | <zewt> | it's much easier and quicker to use existing fully-baked tools, especially when the people using them already know how they work |
| 18:26 | <zewt> | of course, for web games you're more limited--you really don't want to be layering another scripting language on top of JS, heh |
| 18:27 | <MikeSmith> | details about the Twitter backend at http://www.infoq.com/articles/twitter-java-use are interesting |
| 18:28 | <zewt> | ew |
| 18:28 | <zewt> | if my choices are ruby or java, I think I may choose jumping out the window |
| 18:29 | <MikeSmith> | heh |
| 18:33 | <Philip`> | zewt: Environments change frequently enough that I imagine it's rare to have fully-baked tools that you can drop in with no integration effort - you might be developing for some unusual CPU architecture or relying on some custom memory management scheme or whatever, and trying to port an existing scripting engine is a more daunting task than starting from scratch |
| 18:34 | <zewt> | eh, that's a stretch |
| 18:36 | <zewt> | maybe if you're developing for the SNES and you have 128K to work with, but I've used Lua on the PS2 without any difficulty |
| 18:39 | <Philip`> | Lua seems rare in its suitability, compared to most other real languages (Python, JS, etc) that are a much bigger pain to embed |
| 18:40 | <zewt> | well, lua is designed for it, python isn't |
| 18:46 | <Philip`> | Seems like choice is limited if you don't like Lua :-( |
| 21:39 | <AryehGregor> | smaug____, did you ever get responses from IE or WebKit on insertNode()? |
| 21:39 | <smaug____> | not yet |
| 21:39 | <AryehGregor> | k. |
| 21:40 | <smaug____> | well, some webkit developer just referred to Hixie's comment abou insertNode |
| 21:40 | <smaug____> | but nothing from IE yet |