| 00:10 | <Dashiva> | http://heideri.ch/jso/ |
| 00:10 | <Dashiva> | The page shoots itself in the foot by having "attacks" that require the attacker to be able to create on* attributes already |
| 00:10 | <AryehGregor> | Sounds like "When we add new attributes, people who were blacklisting attributes are screwed". |
| 00:11 | <AryehGregor> | <video poster> sounds like a sneaky one, though, you'd expect someone might fall for that. Does javascript: work for <img src>? |
| 00:13 | <Dashiva> | Don't think so. Don't see why it should work for poster either. |
| 00:18 | <Dashiva> | Looks like it's just a bug in Opera |
| 00:19 | <Dashiva> | All the other cases require that the attacker already can write arbitrary markup, so I guess that's 0 new attacks |
| 00:19 | <AryehGregor> | As usual. |
| 04:22 | <wirepair> | besides the sandbox attribute, does anyone see/know of anything in html5 that will help increase security/make it easier for developers to create secure apps? |
| 04:58 | <Hixie> | wirepair: security features are baked in from the ground up, so it's hard to point to them |
| 05:18 | <wirepair> | hixie yeah |
| 05:19 | <wirepair> | just wondered if something additional like sandbox would provide developers with the ability to secure their web apps |
| 05:39 | <Hixie> | wirepair: there's things like the Origin header |
| 05:39 | <Hixie> | wirepair: might be some other features here and there |
| 05:40 | <Hixie> | wirepair: generally the idea is to make it automatic that things be secure, though |
| 05:55 | <wirepair> | cool, thanks |
| 06:32 | <annevk> | http://beltzner.ca/mike/2010/05/10/firefox-4-fast-powerful-and-empowering/ |
| 06:32 | <annevk> | wtf is it with slideshare |
| 06:32 | <annevk> | don't people realize slideshare stinks and requires a plugin |
| 06:32 | <annevk> | grmbl |
| 06:47 | <virtuelv> | annevk: pragmatically, it lets people do what people want |
| 06:47 | <virtuelv> | upload their powerpoints and keynote presentations |
| 06:48 | <virtuelv> | and it provides embedding functionality |
| 06:49 | <karlcow> | http://www.learningjquery.com/2010/05/now-you-see-me-showhide-performance |
| 06:52 | <MikeSmith> | krijnh server not responding> |
| 06:53 | <annevk> | there's no reason slideshare needs Flash |
| 06:54 | <karlcow> | annevk: there is one. |
| 06:54 | <karlcow> | burning CPU cycles for the glory of computing :p |
| 07:23 | <annevk> | What's also interesting about this CORS thing is that e.g. WebSocket has exactly the same semantics... |
| 07:41 | <franksalim> | annevk, a consistent security models is good |
| 07:41 | <franksalim> | *model |
| 07:42 | <annevk> | Indeed, but people argue against CORS but not against WebSocket |
| 07:42 | <Hixie> | lots of people arguing against websocket for other reasons |
| 07:42 | <franksalim> | they don't argue against websocket? |
| 07:42 | <Hixie> | we already have enough arguments thanks :-) |
| 07:43 | <franksalim> | hmm yes |
| 07:43 | <annevk> | would be a nice way to get them off my lawn :p |
| 07:44 | <Hixie> | i'll send you some of mine! |
| 07:44 | <Hixie> | mutually assured destruction! |
| 07:46 | <annevk> | heh |
| 07:46 | <annevk> | Hixie, you have some time for the complete.html thingy now? |
| 07:46 | <Hixie> | sure, now's a good time |
| 07:46 | <Hixie> | 'sup |
| 07:47 | <othermaciej> | hello Hixie |
| 07:47 | <Hixie> | hey othermaciej |
| 07:47 | <annevk> | I guess you want to have the copy without the absolute URLs |
| 07:47 | <annevk> | and I'm generating the zip file already |
| 07:47 | <Hixie> | so long as it works, i don't mind if they're absolute or not |
| 07:48 | <annevk> | I guess I just need to give you a URL and you need to give one to me |
| 07:48 | <Hixie> | i guess ping /specs/web-apps/current-work/do-multipage-complete-update |
| 07:49 | <Hixie> | what should i fetch? |
| 07:49 | <annevk> | http://html5.org/complete/whatwg.org/complete.zip |
| 07:50 | <annevk> | just doing a wget on that URL works I suppose? |
| 07:50 | <Hixie> | yup |
| 07:50 | <Hixie> | or will, once i've written the script |
| 07:56 | <annevk> | I pm'ed the generate URL |
| 07:56 | <annevk> | pm'd even |
| 07:57 | <Hixie> | thanks |
| 07:57 | <annevk> | I guess once that is all set up MikeSmith can remove the mailing list hook |
| 07:58 | <MikeSmith> | yep |
| 07:58 | <MikeSmith> | just lemme know when to flip the switch |
| 09:14 | MikeSmith | wonders if Richard Clark is on IRC |
| 09:24 | Hixie | gets his script all confused |
| 09:39 | <annevk> | hsivonen, hah, http://library.gnome.org/users/palimpsest/stable/advanced.html.en is brilliant |
| 09:44 | <annevk> | http://limpet.net/mbrubeck/2010/05/11/fennec-meta-viewport.html -- '<meta name="viewport"> is a good example of browsers innovating exactly how Sachin Agarwal thinks they should.' |
| 09:44 | <annevk> | it's also pretty ugly |
| 09:44 | <Hixie> | and a layering violation |
| 09:47 | <hsivonen> | Hixie: supported by 3 of the top 4 engines. time to bite the bullet and spec it. |
| 09:47 | <Hixie> | that's anne's problem |
| 09:50 | <annevk> | not really, someone can write a standalone spec for it and add it to the metaextensions wiki |
| 09:54 | <Hixie> | ok, finally got the multipage scripts figured out |
| 09:54 | <Hixie> | man i caused a lot of damage while doing that |
| 09:55 | <hsivonen> | whoa. http://my.opera.com/ODIN/blog/opera-mobile-10-for-nokia-n900-n800-n810-maemo-standards-support says Opera on desktop support SQL database |
| 09:55 | <Hixie> | didn't we know that? |
| 09:55 | <hsivonen> | Hixie: I didn't. |
| 09:55 | <hsivonen> | does it use sqlite? |
| 09:56 | <Hixie> | i assume so |
| 09:56 | <zcorpan_> | yes it does |
| 09:57 | <hsivonen> | ok |
| 09:57 | <zcorpan_> | http://my.opera.com/core/blog/2010/03/03/persistent-client-side-storage-for-your-persistent-needs |
| 09:58 | <hsivonen> | a bit odd not to have it on Maemo if you have it on desktop, considering WebKit's competitive position in the mobile space |
| 09:59 | <hsivonen> | so Maemo now has 3 of the top 4 browser engines available |
| 09:59 | <hsivonen> | I guess that's the broadest coverage for any mobile platform now that Gecko for Windows Mobile was discontinued |
| 10:27 | <roc> | I just read that Sachin Agarwal blog post and my head exploded |
| 10:34 | <hsivonen> | roc: the post has been pretty successful at getting attention |
| 10:34 | <roc> | sure |
| 10:36 | <roc> | outrageous posts get attention |
| 10:42 | <karlcow> | well I guess jetlag is at least good for data mining/archeology. |
| 10:42 | <annevk> | finally figured out how to get more sensible pages out of spec splitter |
| 10:42 | <annevk> | Hixie, maybe you should rename the id crossDocumentMessaging as it is not a very nice or consistent page name |
| 10:43 | <karlcow> | the first mention of srt, I found on public-html is sept 2008 by Lachlan |
| 10:46 | <annevk> | isn't John Foliot effectively asking for addition there? |
| 10:47 | <annevk> | "then clear instruction and specifications on how to 'in-code' provide the out-band solution must also be provided" |
| 10:47 | <karlcow> | annevk: yep |
| 10:48 | <annevk> | fun stuff |
| 11:01 | <hsivonen> | so Opera on Maemo has about zero platform integration. not even low-hanging fruit like right icon in the right menu |
| 11:02 | <hsivonen> | but the responsiveness and graphics speed is really impressive |
| 11:02 | <hsivonen> | on N800 |
| 11:09 | <Hixie> | annevk: done |
| 11:13 | <annevk> | cool |
| 11:13 | <annevk> | btw, I now have enough understanding of spec-splitter.py that I can create whatever split people desire |
| 11:14 | <Hixie> | nice |
| 11:14 | <annevk> | for everyone who didn't know yet: http://www.whatwg.org/specs/web-apps/current-work/complete/ |
| 11:14 | <annevk> | and when that copy is down: http://html5.org/complete/ |
| 11:15 | <Hixie> | looks like the reviewer script is dead on the whatwg copy there |
| 11:15 | <Hixie> | wonder what's up with that |
| 11:15 | <annevk> | hmm, maybe I should use absolute links after all? |
| 11:16 | <Hixie> | aha, no |
| 11:16 | <Hixie> | problem with fixBrokenLink() |
| 11:16 | <Hixie> | aha, no link-fixup.js |
| 11:16 | <Hixie> | you have to include one of those in the zipfile, i think |
| 11:17 | <annevk> | can't you make the links absolute? |
| 11:17 | <annevk> | i mean like /path-to-script |
| 11:17 | <Hixie> | the scripts are absoute |
| 11:17 | <Hixie> | absolute |
| 11:17 | karlcow | had a very brief image of annevk dancing at the Bolchoi and doing a split |
| 11:17 | <Lachy> | oh, nice multipage complete version. That could be useful, though I'll stick with single page. |
| 11:18 | <Hixie> | but you need to include a link-fixup.js specially designed for complete/'s structure |
| 11:18 | <Hixie> | equivalent to http://www.whatwg.org/specs/web-apps/current-work/multipage/link-fixup.js |
| 11:18 | <Hixie> | which apparently uses http://www.whatwg.org/specs/web-apps/current-work/multipage/fragment-links.js |
| 11:18 | <Hixie> | both are provided by philip |
| 11:18 | <Hixie> | in his tarball |
| 11:19 | <annevk> | ah |
| 11:20 | <Hixie> | the problem is that script is invokes before init() |
| 11:21 | <Hixie> | so if that script fails, init() never runs |
| 11:21 | <Hixie> | and the other scripts don't get loaded |
| 11:23 | <annevk> | thanks |
| 11:23 | <annevk> | hopefully fixed |
| 11:26 | <Hixie> | nice |
| 11:30 | <annevk> | removed the script on html5.org/complete as it doesn't work due to cross-origin issues |
| 11:31 | <Hixie> | some will work |
| 11:31 | <Hixie> | others not so much |
| 11:33 | <Hixie> | does dfn.js not worm in ff? |
| 11:33 | <Hixie> | or did i break it? |
| 11:34 | <Hixie> | oh i broke it |
| 11:34 | <Hixie> | nm |
| 12:48 | <hsivonen> | the notion of whitespace is such a mess |
| 12:48 | <hsivonen> | it seems various places in Gecko agree that space, \t and \n are whitespace |
| 12:48 | <hsivonen> | \r, \f and \v depend |
| 12:48 | <hsivonen> | sigh |
| 12:48 | <hsivonen> | and don't get me started about XML 1.1 |
| 12:55 | <hsivonen> | Hixie: why does area coords parsing in HTML5 treat only space as a delimiter (not tab, newline, etc.)? |
| 12:55 | <zcorpan_> | hsivonen: iirc for ie compat, but i'm not 100% sure |
| 12:56 | <hsivonen> | zcorpan_: What do Opera and WebKit do? |
| 12:56 | <hsivonen> | Gecko accepts even \v! |
| 12:57 | <zcorpan_> | hsivonen: opera implements the spec, or possibly an earlier version of the spec where there was no special treatment for funny characters or something |
| 12:57 | <zcorpan_> | hsivonen: don't remember what webkit does |
| 12:58 | <hsivonen> | zcorpan_: ok |
| 12:58 | <hsivonen> | I'll file this away, because these other whitespace bugs are distracting me from the whitespace bug I'm trying to fix |
| 12:58 | <Philip`> | http://lists.w3.org/Archives/Public/public-html/2009Jan/0079.html ? |
| 12:59 | <hsivonen> | Filed as https://bugzilla.mozilla.org/show_bug.cgi?id=565031 |
| 13:00 | <zcorpan_> | hsivonen: i don't mind changing the spec wrt whitespace in coords |
| 13:39 | <annevk> | hsivonen, isn't XML 1.1 dead? |
| 13:45 | <annevk> | I wonder why the chairs are so vocal about splitting... I guess there was some backchannel chatter |
| 13:51 | <nessy> | annevk: I think it has a lot to do with keeping the choice of external text format independent of HTML5 |
| 13:52 | <nessy> | png, jpeg, ogg, mpeg4 - none of these are specified as part of HTML5 |
| 13:52 | <nessy> | and nor should they by - there may be better formats in the future that the video, audio and img elements will have to support! |
| 13:53 | <nessy> | it's really the same for external text associations for video and audio and a format definition should not be inside the spec |
| 13:54 | <annevk> | text/cache-manifest is in HTML5 |
| 13:54 | <annevk> | as are text/ping, text/html-sandboxed, application/microdata+json |
| 13:55 | <Philip`> | Has anyone argued that it *should* be defined in the HTML5 document? (rather than arguing that it doesn't matter where the text exists for now and it's a waste of effort to worry about it before we even know whether it's a good enough solution that it should continue to exist) |
| 13:55 | <annevk> | Ian has |
| 13:56 | <Philip`> | Ah |
| 13:56 | <nessy> | annevk: are those supposed to be used outside the Web by other apps, too? |
| 13:57 | <annevk> | nessy, whether there will be better formats in the future is somewhat orthogonal to where formats are specified |
| 13:57 | <nessy> | not if the only format that is acceptable is the one inside the spec |
| 13:57 | <erlehmann> | nessy, at least where the consensus is clear — with the img element — it would be good to document it. after all, png, gif, jpeg ARE all supported in almost every browser. |
| 13:58 | <annevk> | if we want to make more formats acceptable we change the spec |
| 13:58 | <erlehmann> | without baseline formats you get compatibility nightmares |
| 13:58 | <erlehmann> | what annevk said. |
| 13:58 | <nessy> | you can call a format a baseline format without having it specified inside the spec |
| 13:58 | <nessy> | we wouldn't include the theora specification into html5, either, to make it a baseline format |
| 13:59 | <erlehmann> | nessy, how would an implementor know then if it isn't included ? |
| 13:59 | <nessy> | a mere sentence that this is the baseline format and a link to its specification would be completely sufficient |
| 14:00 | <nessy> | honestly, my mind boggles at the sheer idea that somebody implementing support for WebSRT in their desktop captioning application has to deal with the whole HTML5 spec! |
| 14:00 | <annevk> | why would they have to deal with the whole spec? |
| 14:01 | <nessy> | because as it's part of the HTML5 spec, how will they know there is nothing in the rest of the spec that is not relevant to their implementation |
| 14:01 | <annevk> | also these are not the reasons brought forward for splitting |
| 14:01 | <nessy> | they are my reasons |
| 14:01 | <annevk> | sure |
| 14:01 | <nessy> | and I have brought them forward |
| 14:02 | <annevk> | sure |
| 14:02 | <annevk> | but they are not what started this IRC thread :) |
| 14:02 | <nessy> | oh! |
| 14:02 | <annevk> | anyway, it seems that simply reading the WebSRT section should give them all the answers |
| 14:03 | <nessy> | seems I have to keep up with the spec reading! |
| 14:04 | <annevk> | e.g. text/cache-manifest might get separate tool support as well |
| 14:05 | <annevk> | it's pretty easy to just read the part on text/cache-manifest to figure out how to construct such files and how to read them (e.g. if you write some kind of optimization tool) |
| 14:05 | <annevk> | some definitions are reused but they are all clearly hyperlinked |
| 14:06 | <annevk> | I don't really see what all the fuss is about; it mostly seems a lot of hand waving without much clear scenarios |
| 14:08 | <nessy> | look at it this way: software is not written as one big main function either - stuff that is reused elsewhere is split into libraries - WebSRT is a "library" for me |
| 14:09 | <nessy> | and the spec is too long for its own good anyway - takes ages to load in any browser! |
| 14:09 | <hsivonen> | annevk: Some aspects of XML 1.1 are undead, though |
| 14:10 | <nessy> | btw: does anyone know why the links in the ToC at http://dev.w3.org/html5/spec/Overview.html don't work? |
| 14:10 | <nessy> | I guess it's just the #websrt one that doesn't ... |
| 14:11 | <Philip`> | nessy: http://www.whatwg.org/specs/web-apps/current-work/complete/video.html#websrt loads faster |
| 14:12 | <nessy> | yeah, I know, but I wanted to check what's in the w3c's spec |
| 14:12 | <nessy> | I always use the whatwg link :) |
| 14:12 | <Philip`> | Oops, I got confused and forgot it was still in the HTML5 documents too |
| 14:12 | <nessy> | hmm… it seems it had to do with the page not being properly loaded by my browser - forget it :) |
| 14:21 | <MikeSmith> | nessy: that link probably doesn't work because that section is recently added and so the splitter has generated a new output file for it, and I've not yet committed that to cvs |
| 14:21 | <MikeSmith> | I'll take a look noew |
| 14:21 | <nessy> | it worked after a reaload actually |
| 14:21 | <nessy> | nothing to worry about |
| 14:59 | <annevk> | hsivonen, fwiw, treating CR as TAB without tab stops is not something I really understand |
| 15:27 | <webr3> | re spaces; here's a list of all the one I know: |
| 15:27 | <webr3> | \u0020 //SPACE | \u00A0 //NO-BREAK SPACE | \u1361 //ETHIOPIC WORDSPACE | \u1680 //OGHAM SPACE MARK | \u2002 //EN SPACE | \u2003 //EM SPACE | \u2004 //THREE-PER-EM SPACE | \u2005 //FOUR-PER-EM SPACE | \u2006 //SIX-PER-EM SPACE | \u2007 //FIGURE SPACE | \u2008 //PUNCTUATION SPACE | \u2009 //THIN SPACE | \u200A //HAIR SPACE | \u200B //ZERO WIDTH SPACE | \u202F //NARROW NO-BREAK SPACE | \u205F //MEDIUM MATHEMATICAL SPACE | \u2408 |
| 15:28 | <annevk> | oh hey webr3, didn't know you were on IRC |
| 15:28 | <webr3> | yeah i am |
| 15:29 | <webr3> | almost feel like saying sorry about the noise; but at the same time i see it as an issue, that probably can't be addressed! web's safe so it's cool |
| 15:29 | <zcorpan_> | http://simon.html5.org/dump/ecmascript-whitespace.txt |
| 15:29 | <webr3> | zcorpan_: cheers |
| 15:30 | <zcorpan_> | the "opera" column is out of date |
| 15:30 | <annevk> | webr3, discussion is good |
| 15:30 | <webr3> | zcorpan_ maybe check for \u303F - is missing off that page |
| 15:30 | <webr3> | annevk: agreed |
| 15:31 | <zcorpan_> | webr3: i checked for all of BMP |
| 15:31 | <webr3> | :) |
| 15:31 | <zcorpan_> | in o10.5x i get: |
| 15:31 | <zcorpan_> | 9 10 11 12 13 32 133 160 5760 6158 6159 8192 8193 8194 8195 8196 8197 8198 8199 8200 8201 8202 8203 8232 8233 8239 8287 12288 65279 65534 |
| 15:31 | <webr3> | definitive list, I'll update my trim()'s! |
| 15:35 | <zcorpan_> | note that this is whitespace in ecmascript eval() |
| 15:36 | <zcorpan_> | whitespace in ecmascript source is hopefully the same |
| 15:36 | <zcorpan_> | but whitespace in other places is different |
| 15:36 | <jgraham> | "other places"? |
| 15:36 | <zcorpan_> | i mean like in html and css |
| 15:36 | <jgraham> | Ah |
| 15:37 | <jgraham> | Well yes, ECMAScript says "all other characters in unicode class Zs" or something |
| 15:37 | <jgraham> | HTML has a shor fixed list |
| 15:37 | <jgraham> | *short |
| 15:38 | <jgraham> | (ECMAScript also has a short list or doesn't-matter-what-unicode-says characters) |
| 15:38 | <zcorpan_> | the ecmascript definition is annoying because it doesn't say which version of unicode, just requires unicode 2.0 or later (or something) |
| 15:38 | <zcorpan_> | and it changes over time |
| 15:38 | <jgraham> | Indeed |
| 15:39 | <annevk> | pretty great that we are now at the point where we can argue over whitespace definitions rather than all the shit we had to emulate from IE6 :p |
| 15:39 | <jgraham> | Unicode 3.0 |
| 15:40 | <jgraham> | (it says "must... Unicode 3.0 ... may ... later versions") |
| 15:40 | <annevk> | (of course, it's not quite like that, but it certainly has improved) |
| 15:40 | <jgraham> | (so even if characters are moved out of Zs in later versions, tehy still have to be recognised as whitespace) |
| 15:40 | <jgraham> | (which has happened a couple of times iirc) |
| 15:41 | <annevk> | wait what? |
| 15:41 | <annevk> | oh god |
| 15:42 | <annevk> | TC39... |
| 15:47 | <jgraham> | annevk: ? |
| 15:47 | <annevk> | you pointing out it's as inconsistent as hell |
| 15:48 | <jgraham> | Well not really |
| 15:49 | <jgraham> | It's slightly more consistent than just saying "whatever versuion of unicode you like" |
| 15:49 | <jgraham> | slightly less consistent than saying "just this set of characters" |
| 15:50 | <Dashiva> | Unicode versions are a pain no matter how you handle them... |
| 16:02 | <annevk> | jgraham, latest version would have been more consistent |
| 16:04 | <annevk> | relying on specifications not evolving is silly and codifying it is worse |
| 16:07 | <boblet> | dear cabal, I published some stuff on ruby |
| 16:07 | <boblet> | http://html5doctor.com/ruby-rt-rp-element/ |
| 16:07 | <boblet> | and some supporting code snippets and notes here: http://oli.jp/example/ruby/ |
| 16:11 | <Dashiva> | I like the Chinese examples |
| 16:11 | <boblet> | hehehe |
| 16:11 | <boblet> | naughty me |
| 16:12 | <boblet> | Dashiva: do you understand Chinese, or are you just exceptionally clued on about Chinese internet memes? |
| 16:12 | <MikeSmith> | boblet: nice work |
| 16:13 | <Dashiva> | The latter |
| 16:13 | <boblet> | hey Mike! thanks man |
| 16:13 | <Dashiva> | By the way, ruby is used very creatively in Japanese |
| 16:13 | <Dashiva> | Translations, clarificatioins, all kinds of weird stuff |
| 16:15 | <Dashiva> | http://dashiva.net/misc/baldr03.jpg |
| 16:16 | <Dashiva> | The ruby is ... Japanese transliterations of the English words for the translated Japanese |
| 16:16 | <boblet> | MikeSmith: you gonna be around next Tue-Thur? I’ll be in town. would be nice to catch up — maybe the W3 mixed rotenburo trip Naoko was talking about? |
| 16:16 | <boblet> | Dashiva: it’s mainly just furigana and romaji for kanji pronunciation |
| 16:16 | <Dashiva> | http://dashiva.net/misc/baldr08.jpg |
| 16:16 | <Dashiva> | This one made me smile |
| 16:17 | <boblet> | hehe |
| 16:17 | <MikeSmith> | boblet: will be at home, yeah |
| 16:25 | MikeSmith | heads away for a bit |
| 18:31 | <shepazutoo> | hsivonen: I think there's a typo in your moz post "the more maintainable code base of the HTML5 compared to Gecko’s old HTML parser." -> "the more maintainable code base of the HTML5 parser compared to Gecko’s old HTML parser." |
| 18:33 | <shepazutoo> | hsivonen: also, "<DOCTYPE !html>" -> "<!DOCTYPE html>" |
| 18:34 | annevk | was just about to point that out :) |
| 18:36 | <shepazutoo> | but overall, a nice summary, thanks, hsivonen |
| 18:48 | <hsivonen> | shepazutoo: thanks. forwarded to blizzard |
| 18:51 | <JonathanNeal> | Ahoy! |
| 19:01 | <hsivonen> | shepazutoo: fixed. thanks |
| 19:03 | <annevk> | http://googleblog.blogspot.com/2010/05/giving-voice-to-more-languages-on.html -- hmm eSpeak is GPL; I wonder what loophole is being used here |
| 19:04 | <Philip`> | annevk: They're not distributing compiled code that contains eSpeak, so the GPL obligations on releasing source code don't apply |
| 19:05 | <annevk> | Aah, so GPL can be "abused" by server applications? |
| 19:05 | <Philip`> | Yes |
| 19:05 | <Philip`> | (if you consider it abuse) |
| 19:05 | <annevk> | The people who drafted it probably do |
| 19:05 | <Philip`> | The Affero GPL is designed to prevent that situation |
| 19:07 | <annevk> | Seems like quite the loophole given all the "cloud" thingies going on |
| 19:09 | <Dashiva> | I wonder if you could sell something and include "You must forfeit your GPL-granted rights for this product" in the contract terms |
| 19:11 | <Philip`> | You can include anything you want in the contract terms |
| 19:11 | <Philip`> | People will probably happily agree to it too |
| 19:14 | <Dashiva> | But would it be legally solid? |
| 19:14 | <Philip`> | The GPL says "You may not impose any further restrictions on the exercise of the rights granted or affirmed under this License." which sounds like you could argue it covers that situation |
| 19:16 | <Dashiva> | Sounds like it, yeah |
| 19:18 | <Dashiva> | "a durable physical medium customarily used for software interchange" |
| 19:18 | <Dashiva> | Like floppies! |
| 21:42 | <AryehGregor> | Is there any way to blur out an image from JavaScript? Like you have some image at a URL, and you want to include that image, but blurred. Is this something canvas can do somehow? |
| 21:43 | AryehGregor | knows nothing about image stuff |
| 21:45 | <miketaylr> | i've seen a js lib that does that i think |
| 21:45 | miketaylr | looks |
| 21:46 | <miketaylr> | aha. AryehGregor: http://www.pixastic.com/lib/docs/actions/blur/ |
| 21:46 | <miketaylr> | http://www.pixastic.com/lib/docs/actions/blurfast/ is way more intense |
| 21:46 | <AryehGregor> | Yeah. |
| 21:46 | <AryehGregor> | That's what I'm looking for. |
| 21:47 | <AryehGregor> | Hmm, MPL 1.1. |
| 21:48 | <AryehGregor> | That's not GPL-compatible, is it? |
| 21:48 | <AryehGregor> | Well, probably no one's looking too hard. |
| 21:48 | miketaylr | has no idea |
| 21:48 | <AryehGregor> | It's probably fine if I keep it served in totally separate files, I guess. |
| 22:33 | <zcorpan_> | AryehGregor: you could use an svg filter |
| 22:34 | <AryehGregor> | Seems like there's a library to do it, which already uses canvas+IE filters for IE. |
| 22:34 | <AryehGregor> | Good enough for me. |
| 22:35 | <roc> | canvas??? |
| 22:35 | <roc> | SVG filters would perform a lot better |
| 22:36 | <Philip`> | SVG isn't a cool new toy, though |
| 22:36 | <zcorpan_> | svg filters in text/html is a cool new toy |
| 22:37 | <zcorpan_> | it doesn't even work in any shipping browser |
| 22:57 | <roc> | zcorpan_: svg filters in text/html works in Firefox 3.5 and later |
| 22:57 | <roc> | you just put the filter in an external SVG file |
