| 00:31 | <Hixie> | hsivonen: yt? |
| 01:31 | <Hixie> | well this is just silly |
| 01:31 | <Hixie> | it turns out that sicking's description of what happens with innerHTML and defer scripts and document.write() in IE is slightly wrong |
| 01:32 | <Hixie> | what's actually going on is that setting innerHTML on a node in the document is causing the pending deferred scripts to be run right there |
| 01:32 | <Hixie> | and if there are any deferred scripts that haven't loaded yet... they get dropped completely |
| 01:36 | <ojan> | ewww |
| 01:36 | <Hixie> | yeah so obviously we can't spec the dropping on the floor bit |
| 01:37 | <Hixie> | but it looks like there might be serious compat needs to support the rest of it |
| 01:43 | <Hixie> | our options with the dropping on the floor bit are either to execute only <script defer> without src="" there, or to pause there are execute all of them |
| 01:43 | <Hixie> | s/are ex/and ex/ |
| 01:43 | <Hixie> | in the former case, if there's any d.w()s that are being dropped by IE, it would cause the page to blank again |
| 01:43 | <Hixie> | so i guess the latter is better |
| 05:08 | <othermaciej> | Hixie: are you around? |
| 05:10 | <jwalden> | suppose I have data:<video%20src="http://example.com/ogg-theora-sent-with-application-octet-stream-mime-type.ogv"></video>; |
| 05:11 | <jwalden> | do I read http://www.whatwg.org/specs/web-apps/current-work/multipage/video.html#concept-media-load-resource correctly to say that whether the video displays, assuming the UA is capable of displaying the video if sent with the correct mime type, is undefined? |
| 05:19 | <zcorpan> | jwalden: http://www.whatwg.org/specs/web-apps/current-work/multipage/video.html#a-type-that-the-user-agent-knows-it-cannot-render |
| 05:20 | <zcorpan> | jwalden: the spec will probably say explicitly that application/octet-stream is not such a type, if i remember the list discussions correctly |
| 05:37 | <jwalden> | zcorpan: which list discussion? not seeing any promising threads in my whatwg folder |
| 05:41 | <othermaciej> | Hixie: nevermind (if you see this later) |
| 05:43 | <zcorpan> | jwalden: http://www.w3.org/mid/4A6138EC.5020408⊙me |
| 05:44 | <jwalden> | ah, I don't follow public-html |
| 05:44 | <jwalden> | and heh, only a few days old too :-) |
| 07:12 | <zcorpan> | hmm, web fonts and canvas |
| 07:12 | <zcorpan> | should the canvas wait until the font has loaded? |
| 07:51 | <zcorpan> | http://software.hixie.ch/utilities/js/live-dom-viewer/saved/175 crashes chrome |
| 07:57 | <zcorpan> | seems to be getting .font when using the 'smaller' keyword |
| 08:26 | <Philip`> | zcorpan: It should probably do the same as drawing not-yet-loaded images, i.e. throw an exception and not freeze, because unpredictable freezing is nasty |
| 08:26 | <Philip`> | (at least in my opinion) |
| 08:43 | <zcorpan> | Philip`: so a page using web fonts would use setInterval and try/catch when using web fonts? |
| 08:44 | <zcorpan> | maybe we should have an event when fonts have loaded |
| 08:44 | <zcorpan> | and let fillText fall back to another font if it's still loading |
| 08:45 | <Philip`> | zcorpan: It would |
| 08:45 | <Philip`> | (which is probably a terrible idea) |
| 08:45 | <annevk4> | event based sounds good |
| 08:46 | <Philip`> | If you're doing something non-animated, like drawing a graph, you really need to have some way to wait until the font's loaded before drawing with it |
| 08:46 | <Philip`> | but if you're doing something animated, like a game, you don't want to mysteriously freeze while the font is being loaded |
| 08:48 | <zcorpan> | Philip`: you could draw the graph with the fallback font first and then redraw with the web font onwebfontsloaded |
| 08:51 | <zcorpan> | one event for all pending fonts? or one event per font? |
| 08:51 | <zcorpan> | where should it be fired? |
| 08:57 | <Hixie> | it should just delay the load event, imho |
| 08:59 | <zcorpan> | Hixie: makes sense |
| 09:00 | <zcorpan> | Hixie: though that doesn't help if you're loading in a new font after the 'load' event |
| 09:01 | <Hixie> | well if you're doing that, wait for the style sheet's load event |
| 09:01 | <Hixie> | or something |
| 09:01 | <Hixie> | if we had someone to edit cssom, we could add a new event, i guess |
| 09:02 | <zcorpan> | css transitions have specced events without someone to edit cssom |
| 09:03 | <Hixie> | someone to edit css web fonts would be fine too :-) |
| 09:05 | <othermaciej> | I thought annevk was editing CSSOM |
| 09:05 | <zcorpan> | is it specced that images in css delays the load event? |
| 09:05 | <Hixie> | othermaciej: he's editing cssom-view |
| 09:05 | <Hixie> | zcorpan: i thought they didn't these days |
| 09:06 | <othermaciej> | I see |
| 09:06 | <zcorpan> | Hixie: oh? |
| 09:06 | <othermaciej> | no one is foolish enough to touch the rest of CSSOM? |
| 09:06 | <zcorpan> | othermaciej: go for it :) |
| 09:06 | <Hixie> | othermaciej: there's a whole list of specs nobody is foolish enough to touch |
| 09:06 | <Hixie> | othermaciej: i think it's more a matter that nobody has the time |
| 09:06 | <Hixie> | zcorpan: othermaciej hasn't even got the time to edit the design principles, let alone a spec :-) |
| 09:07 | <othermaciej> | zcorpan: no thank you sir |
| 09:07 | <gsnedders|work> | I think we need shorter specs, then. |
| 09:07 | <othermaciej> | I think I'll have some time for DP tomorrow |
| 09:07 | <othermaciej> | I don't think I can sign up for a real spec |
| 09:07 | <gsnedders|work> | Like, here's a bunch of random words. You MUST NOT try and implement this. |
| 09:07 | <Hixie> | gsnedders|work: how would that help |
| 09:07 | <othermaciej> | I might be interested in hiring someone to write specs for me |
| 09:07 | <gsnedders|work> | Hixie: I never said it would. |
| 09:07 | <Hixie> | oh ok |
| 09:07 | <gsnedders|work> | :D |
| 09:08 | <Hixie> | othermaciej: let me know if you find anyone |
| 09:08 | <othermaciej> | but it's hard to find people who are capable at such things and not already taken |
| 09:08 | <Hixie> | othermaciej: i'll offer them double! |
| 09:08 | <gsnedders|work> | oo… choose me! choose me! :P |
| 09:08 | <Hixie> | (double the workload, that is) |
| 09:08 | <Hixie> | gsnedders|work: get a degree first :-P |
| 09:09 | <gsnedders|work> | Hixie: But that increases workload :P |
| 09:09 | <gsnedders|work> | (And hence decreases spec-writing time) |
| 09:09 | <othermaciej> | it's hard to get a big company to hire people who don't have a college degree |
| 09:09 | <zcorpan> | http://software.hixie.ch/utilities/js/live-dom-viewer/saved/176 - seems to delay the load event in opera,firefox,chrome |
| 09:10 | <Hixie> | gsnedders|work: speaking of your workload. cross-refs are really starting to be needed, and i'm starting to have an idea for how references might be something i could benefit from (though not with an external .bib) |
| 09:10 | <Hixie> | specifically, for references, i'd like something that strips out references that aren't actually mentioned in the prose |
| 09:11 | <othermaciej> | I think we changed CSS images to delay the load event a while back, for Firefox compatibility |
| 09:11 | <gsnedders|work> | You've mentioned that before. |
| 09:11 | zcorpan | doesn't have a degree |
| 09:11 | <gsnedders|work> | Hixie: xdoc xrefs I assume you mean? |
| 09:12 | <Hixie> | gsnedders|work: yeah |
| 09:13 | <gsnedders|work> | Hixie: Well you have two weeks until I find out I don't have a chance at getting into the university course I want to, after which I'll be too sad to work :P |
| 09:13 | <Hixie> | heh |
| 09:14 | <Hixie> | anything i can do to help? |
| 09:14 | <gsnedders|work> | Well, seeming I sat the exams months ago and this is merely a matter of waiting for the results… |
| 09:14 | <Hixie> | i meant for the work part :-P |
| 09:14 | <gsnedders|work> | :P |
| 09:15 | <gsnedders|work> | Bribe my manager? :P |
| 09:15 | <gsnedders|work> | Seriously though, I'll ping you if I think you can in anyway |
| 09:15 | <Hixie> | who's your manager |
| 09:15 | gsnedders|work | did mean that sarcastically |
| 09:33 | <gsnedders|work> | What should line-height: normal return through CSSOM? |
| 09:33 | <Hixie> | 'normal' |
| 09:33 | <Hixie> | is there anything else it _could_ return? |
| 09:34 | <gsnedders|work> | Gecko returns the pixel line-height that normal is |
| 09:35 | <Hixie> | ask anne |
| 09:35 | <Hixie> | i'm out of date on current css thinking |
| 09:35 | <gsnedders|work> | annevk4: ping |
| 09:36 | <gsnedders|work> | Hixie: I agree with you, though. |
| 09:37 | <annevk4> | gsnedders|work, somewhat here |
| 09:37 | <annevk4> | oh |
| 09:38 | <gsnedders|work> | annevk4: line-height:normal should return what as the computed style? |
| 09:38 | <annevk4> | hmm, I never quite got around to figuring out getComputedStyle |
| 09:38 | <annevk4> | it usually returns the used value |
| 09:38 | <annevk4> | not the computed value |
| 09:38 | <annevk4> | CSS 2.1 stupidly changed definitions |
| 09:39 | <annevk4> | (compared with CSS 2.0) |
| 09:39 | <gsnedders|work> | Opera and Chrome both return "normal" |
| 09:39 | <gsnedders|work> | Firefox returns npx |
| 09:39 | <annevk4> | for this specific case I suppose either way goes |
| 09:39 | <annevk4> | just has to be defined :/ |
| 09:41 | <gsnedders|work> | Should I send mail? |
| 09:41 | <annevk4> | no more "this week in html5"? |
| 09:41 | <annevk4> | gsnedders|work, nobody is tracking afaik |
| 09:42 | <gsnedders|work> | That's what I thought. |
| 09:49 | <Mrmil> | Is it normal that HTML 5 spec makes Opera crash? I've had it five billion times already |
| 09:50 | <Lachy> | Mrmil, if you're loading the single page version, yes, it's a known issue |
| 09:50 | <Lachy> | the multipage version should work |
| 09:51 | <Mrmil> | Lachy: I loaded the multipage version, went into Label section, and crashed :) |
| 09:52 | <zcorpan> | firefox delays the load event for web fonts |
| 09:52 | Mrmil | hates Windows |
| 09:53 | <gsnedders|work> | Lachy: You yourself filed a bug on the multipage version crashing |
| 10:00 | <Lachy> | gsnedders|work, I know |
| 10:00 | <Lachy> | I need to investigate that soon |
| 10:08 | <zcorpan> | it seems web fonts already delay the load event in opera and firefox |
| 10:09 | <annevk4> | Google Chrome OS, interesting |
| 10:10 | <roc> | where have you been for the last month? |
| 10:10 | <annevk4> | greenland + iceland |
| 10:24 | Philip` | wonders if annevk4 heard about XHTML2 yet |
| 10:26 | <Lachy> | Philip`, didn't that happen before annevk4 went on holiday? |
| 10:26 | <gsnedders|work> | I doubt it. |
| 10:27 | <annevk4> | yeah |
| 10:27 | <annevk4> | I twittered 5 > 2 |
| 10:34 | Philip` | wonders if annevk4 heard about the giant robot invasion of Florida last week |
| 12:23 | <Philip`> | Someone should tell Kristof that his mail system disk quota has apparently been exceeded |
| 12:25 | <Lachy> | Philip`, send him a mail about it :-) |
| 12:26 | <Philip`> | Maybe if it was a very very short email, it would arrive successfully |
| 12:28 | <Lachy> | it depends if the system works by refusing to accept any mail after the limit has been exceeded, or if it's so close, it rejects any mail that would send it over |
| 12:28 | <mookid> | If anyone's interested the discussion on URIs and HTTP conneg is still going strong: http://mike.mykanjo.co.uk/2009/07/21/whos-right-seb-is-it-u-r-i/ |
| 12:32 | <Philip`> | Lachy: The former would seem a pretty silly system, since I could email a DVD to every user on the mail server until it runs out of disk space and dies painfully even if the quotas were carefully managed |
| 12:34 | <Lachy> | Philip`, some mail systems also have limits on the maximum size of an individual mail too |
| 12:40 | <Hixie> | i wonder if mr allsop is going to send the feedback he's been working on to the list soon |
| 13:27 | <o2T7> | Hi, I'm having problems to make some select1 item elements toggle a switch in the page. Is it possible, or should I necessarily use triggers? |
| 13:35 | <Lachy> | o2T7, HTML5 doesn't have a select1 element. Are you talking about XForms? |
| 13:42 | <o2T7> | Lachy: sure, sorry. That context, yes. |
| 13:45 | <annevk4> | o2T7, this isn't really the right channel for XForms questions ;) |
| 13:48 | <o2T7> | annevk4: I know :( But faithfully I find somewhere somebody experienced with xforms specification. |
| 13:49 | <o2T7> | or even with the way that all this stuff work |
| 13:50 | <o2T7> | I'm reading a lot, but can't get it done with this detail (changing a pair of trigger by a select1) |
| 13:52 | gsnedders|work | grumbles something about rain |
| 13:52 | <annevk4> | o2T7, maybe the www-forms⊙wo can help? |
| 13:53 | <gsnedders|work> | I guess that's not a support list |
| 13:53 | <annevk4> | it is used as such now and then I believe |
| 13:58 | <o2T7> | okay, I'll try there :) |
| 14:15 | <Lachy> | JohnResig, any update on the selectors api test suite? |
| 14:20 | <JohnResig> | Lachy: unfortunately no :-/ |
| 14:41 | <Lachy_> | JohnResig, any idea when you will be able to get to it? |
| 14:41 | <Lachy_> | Or should I find someone else who can make the necessary changes? |
| 14:43 | <JohnResig> | Lachy_: yeah, I'm not sure when I'm going to have time, yet. If someone provides patches I'll just merge them in to the primary version: http://github.com/jeresig/selectortest/tree/master |
| 14:43 | <JohnResig> | Lachy_: (ideally changes would be handled by forking the tree) |
| 14:44 | <Lachy> | ok, I might be able to find the time to do some of the changes myself |
| 14:44 | <JohnResig> | k - ping me if you have any questions |
| 14:44 | <Lachy> | alright, will do. |
| 14:47 | <Lachy> | JohnResig, if you were to do it, how much time would you estimate that it would take? |
| 14:50 | <JohnResig> | Lachy: I'm not sure - I'd have to look through the list of changes needed again - probably not more than a couple hours |
| 14:51 | <Lachy> | ok. In that case, I might be able to do it in a day |
| 14:52 | <JohnResig> | although, you do have the advantage of understanding the spec better than I do :) |
| 14:53 | <Lachy> | yeah, but I also have to work with your code which I'm not very familiar with |
| 14:53 | <JohnResig> | true |
| 14:54 | <Lachy> | this was the list of changes needed http://lists.w3.org/Archives/Public/public-webapps/2009AprJun/1030.html |
| 15:30 | <Lachy> | hey, given this diff from Erik http://lists.w3.org/Archives/Public/public-webapps/2009JanMar/0788.html , how do I merge those changes into my local copy? |
| 15:31 | <Lachy> | I assume there's some command that can do that automatically |
| 15:32 | <JohnResig> | Lachy: I think this command will do it: http://www.kernel.org/pub/software/scm/git/docs/git-apply.html (assuming you have a copy of the tree checked out via git) |
| 15:34 | <Lachy> | I don't have git |
| 15:34 | <gsnedders|work> | Lachy: patch -p1 |
| 15:35 | <Lachy> | gsnedders|work, yeah, I just found that patch command via google |
| 15:35 | <Lachy> | though I'm not really sure what the -p1 does yet |
| 15:36 | <gsnedders|work> | It prunes one directory from the path |
| 15:36 | <Lachy> | well, it worked |
| 16:11 | <Lachy> | http://www.alistapart.com/articles/unwebbable/ |
| 16:11 | <Lachy> | After reading that, I'm left confused about what Joe is saying. |
| 16:13 | <Lachy> | perhaps it's because I don't have a firm grasp of the history behind screenplay formatting. But I don't get why an electronic version needs to retain the strict formatting of a traditional print screenplay |
| 16:16 | <Lachy> | I think the real solution would be to devise a way to address the same problems solved by the traditional formatting (e.g. dividing into 1 minute per page sections) using a modern formatting system optimised for the web |
| 21:32 | <Darxus> | Is there a schema language capable of specifying everything an HTML5 conformance checker is required to check? ( http://www.whatwg.org/specs/web-apps/current-work/multipage/infrastructure.html#conformance-requirements ) ? |
| 21:36 | <annevk2> | no |
| 21:37 | <Darxus> | Heh. Do the two existing validators do more than validate against a schema (in one of these insufficient languages)? |
| 21:37 | <Philip`> | Validator.nu has custom Java code for various things |
| 21:38 | <Darxus> | Okay. |
| 21:38 | <Philip`> | See e.g. http://hsivonen.iki.fi/thesis/html5-conformance-checker for probably more information than you want to know |
| 21:38 | <Darxus> | Awesome, a valid plural of "schema" is "schemata". |
| 21:38 | <Philip`> | (maybe slightly outdated, though) |
| 21:39 | <Philip`> | Is the second validator validator.w3.org (in HTML5 mode)? If so, that's the same code as validator.nu |
| 21:39 | <Darxus> | Philip`: Yeah I found that but haven't read it. |
| 21:39 | <Darxus> | Philip`: Yes. Ah, thanks. I suspected. |
| 21:39 | Philip` | hasn't read it either, but he's looked at the table of contents |
| 21:40 | <Darxus> | I want XML to provide a way to specify the location (and language?) of the schema it is supposed to conform to :/ |
| 21:42 | <gsnedders> | The history of HTML section is good, I've read that. |
| 21:42 | <gsnedders> | Darxus: DOCTYPEs are that :P |
| 21:43 | <Philip`> | Darxus: If you're using the XML serialisation of HTML 5 (i.e. serving it as application/xhtml+xml, not text/html) you can do whatever you want with the doctype |
| 21:43 | <Darxus> | gsnedders: I thought the doctype could only specify a DTD? |
| 21:43 | <gsnedders> | Darxus: A DTD is a schema. |
| 21:43 | <Philip`> | (Only the text/html serialisation requires <!DOCTYPE html>) |
| 21:44 | <Darxus> | gsnedders: Ah, so the "DTD" in a DOCTYPE is not specifically the "DTD" schema language? |
| 21:44 | <gsnedders> | Darxus: It is |
| 21:44 | <gsnedders> | Darxus: My point is you are linking to a type of schema :P |
| 21:44 | <Darxus> | gsnedders: Ah, yes, but DTD has been called "woefully inadequite". |
| 21:44 | <gsnedders> | (which is a DTD) |
| 21:45 | <gsnedders> | Darxus: Indeed. |
| 21:45 | <gsnedders> | Darxus: But it's a way of linking to a schema :P |
| 21:45 | <Darxus> | gsnedders: Yes, but not a sufficiently useful one :) |
| 21:45 | <Philip`> | HTML 5 considers the entire concept of schema languages to be woefully inadequate |
| 21:45 | <Philip`> | except when the language is English :-) |
| 21:45 | <gsnedders> | Darxus: There's some way of linking to XSD files, but XSD is horrid. |
| 21:45 | <Philip`> | ...or Java |
| 21:45 | <Darxus> | An hour ago I thought I was going to write an HTML5 DTD. Then I read the bit in the spec that says DTD is not sufficient. |
| 21:46 | <Darxus> | How about perl? |
| 21:46 | <inimino> | Perl is Turing-complete :-) |
| 21:46 | <Darxus> | inimino: Yes but that is not a useful answer to my question. |
| 21:46 | <gsnedders> | http://syntax.whattf.org/ |
| 21:46 | <inimino> | it's not? |
| 21:46 | <annevk2> | anything Turing-complete is sufficient to write a HTML5 validator |
| 21:47 | <gsnedders> | Darxus: It means it'll be able to express the constraints |
| 21:47 | <gsnedders> | annevk2: Have we proven this? |
| 21:47 | <Philip`> | gsnedders: It means it'll be able to express the machine-checkable constraints |
| 21:47 | <annevk2> | (well, apart from the silly SHOULD requirement which requires solving the halting problem) |
| 21:47 | <Darxus> | annevk2: Thank you. |
| 21:47 | <inimino> | a DTD is formally incapable of expressing some of the machine-checkable conformance criteria |
| 21:47 | <Philip`> | where machine-checkable is defined as anything that can be expressed in a Turing-complete language |
| 21:47 | <inimino> | Perl isn't, so that's your answer |
| 21:48 | <Darxus> | Yeah, that SHOULD is freaking awesome. |
| 21:48 | <Philip`> | (where the input is just the document, and no extra state about e.g. the author's brain) |
| 21:48 | <gsnedders> | I still want to know how we get a computer to take a deep breath |
| 21:48 | <Darxus> | 'This is only a "SHOULD" and not a "MUST" requirement because it has been proven to be impossible.' |
| 21:48 | <Philip`> | gsnedders: Put the fans into reverse |
| 21:49 | <annevk2> | the biggest benefit of completely custom no-schema code is prolly the kind of error messages you can provide |
| 21:50 | <annevk2> | it'd be cool if someone made one :) |
| 21:50 | <gsnedders> | OCaml! |
| 21:52 | <Philip`> | One problem with completely custom no-schema code is that it's useless at answering questions like "what tags am I allowed to insert here, for my editor's auto-complete UI" |
| 21:52 | <Philip`> | (unless you write custom code for that question) |
| 21:52 | <Darxus> | annevk2: Why do you imply that no-one has made one? |
| 21:52 | <Philip`> | Darxus: Because no-one has :-) |
| 21:52 | <Philip`> | unless they've kept it very very quiet |
| 21:53 | <Philip`> | (for HTML5, anyway) |
| 21:53 | <Darxus> | Philip`: Why does the existing HTML5 validator not count? |
| 21:53 | <Darxus> | Also, I don't see why a no-schema validator couldn't answer "what tags am I allowed to insert here?". |
| 21:54 | <gsnedders> | Dashiva: That has a schema. That's not a no-schema validator. |
| 21:54 | <Darxus> | Ahh. |
| 21:54 | <annodomini> | Remember, there is a subset of HTML5 that's checkable with a schema. |
| 21:54 | <Darxus> | annodomini: Oh? |
| 21:55 | <annodomini> | You can probably do something good enough for the "what tags am I allowed to insert" |
| 21:55 | <annodomini> | Or, sorry. Not a subset of HTML5, but a subset of the rules in HTML5 are checkable with a schema. |
| 21:55 | <Philip`> | Darxus: When your code is like "function validate_div(node) { ...some function calls to check children are the correct type... }" you can't automatically invert the code to find a list of what types are allowed - all you can do is execute the code |
| 21:57 | <Philip`> | so either you need to explicitly write the list of what element types are allowed, or you need to write some kind of schema that can be analysed to compute approximately the right answer |
| 21:58 | <Philip`> | I think that's a rubbish explanation and probably untrue, but never mind |
| 21:58 | <Darxus> | Er, schemas don't include a list of what elements may be children / decendants of other elements? (I noticed the HTML 4.01 didn't, which seemed strange.) |
| 21:59 | <gsnedders> | They do |
| 22:12 | <kristallpirat> | aloha |
| 22:12 | <kristallpirat> | I use html5lib for python |
| 22:13 | <kristallpirat> | and want to replace entities, any documentation tips? |
| 22:16 | <Darxus> | Is there any reason a schema language could not be written that could specify all the rules required for validation? |
| 22:17 | <KevinMarks> | a Turing complete schema language? |
| 22:18 | <Darxus> | I did not say that :) |
| 22:20 | <Darxus> | When writing a schema language, do you start with the parser or the rules? :) |
| 22:37 | <Philip`> | You could define a schema language which permits you to write one schema, which is the empty string |
| 22:38 | <Philip`> | and its semantics are that it check all the HTML5 conformance requirements |
| 22:38 | <Philip`> | s/check/checks/ |
| 22:45 | <Darxus> | "I know some of the schemas are not as tight as the corresponding spec prose." - author of the existing validator |
| 22:45 | <Darxus> | Philip`: An entertainging response. |
| 22:45 | <Darxus> | However, again not a useful answer to my question. |
| 22:46 | <inimino> | I thought it was quite illustrative |
| 23:01 | <Philip`> | Darxus: It seemed a useful answer, since it was demonstrating that it is indeed possible for a schema language to specify all the rules required for validation |
| 23:01 | <Philip`> | Then you just have to modify it a bit and pick some tradeoffs between schema language complexity and schema complexity and flexibility etc |
| 23:05 | <Darxus> | Philip`: Your answer is clearly correct. However, as I said, not useful. |
| 23:05 | <inimino> | schema definition formalisms are arbitrary, you can write one that does anything you like |
| 23:06 | <inimino> | if you don't like the answer, perhaps you should reconsider the question :-) |
| 23:38 | <caden> | question on canvas: why does settransform clear my canvas bitmap? can i clear the transform matrix without clearing the bitmap |
| 23:57 | <Philip`> | caden: It shouldn't do that |
| 23:59 | <caden> | thanks Philip |
| 23:59 | <caden> | i thought that that was my problem, but i don't really need to reset the transform matrix |