| 06:00 | <takkaria> | hmm |
| 06:26 | <othermaciej> | I don't know why I keep replying to Larry |
| 10:30 | <hsivonen> | annevk4: why doesn't CORS allow multiple enumerated allowed origins? |
| 10:31 | <hsivonen> | annevk4: being able to specify only one foreign origin in Access-Control-Allow-Origin is inconvenient if you want to serve fonts for a couple of origin but don't want to enable random bandwidth leeches |
| 10:33 | <hsivonen> | too late to change it now, of course |
| 10:36 | <hsivonen> | http://realtech.burningbird.net/web/page-markups/loose-set-notes-rdfa-microdata-and-html5#comment-1494 |
| 10:43 | <Hixie> | does chaals think that consensus means unanimity? |
| 10:48 | <Philip`> | "To me, consensus seems to be the process of abandoning all beliefs, principles, values and policies. So it is something in which no one believes and to which no one objects." |
| 10:53 | <jgraham> | You know things are bad when Margret Thatcher quotes seem apropos |
| 10:55 | <webben> | Hixie: Possibly. I don't think http://lists.w3.org/Archives/Public/public-html/2009May/0299.html clearly implies that though. |
| 10:55 | <webben> | (if that's what you're thinking of) |
| 11:01 | <annevk42> | hsivonen, because we didn't want to introduce complicated features |
| 11:01 | <annevk42> | hsivonen, you can just as well do it conditionally |
| 11:08 | <Dashiva> | because, even though testable, in practice |
| 11:08 | <Dashiva> | the implementation conformance requirement is unnecessary |
| 11:08 | <Dashiva> | http://lists.w3.org/Archives/Public/public-html/2009May/0307.html |
| 11:08 | <Dashiva> | Is this really still an issue... gah |
| 11:11 | webben | finds it odd that vagueness is promoted as a good way to do "extensibility". |
| 11:11 | <annevk42> | I'm surprised nobody has pointed out that our charter allows for non-consensus decisions |
| 11:12 | <webben> | annevk42: I wonder which sense of "consensus" the charter is using. |
| 11:13 | <webben> | also does the charter basically mean that failing consensus, and after recording opinions, it's the chair's decision? |
| 11:13 | <webben> | oh i see the link explains |
| 11:14 | <annevk42> | http://www.w3.org/2007/03/HTML-WG-charter#decisions |
| 11:14 | <annevk42> | It actually also supports the editing model Hixie has been using. I.e. editor puts forward a proposal which is then refined by group discussion and gets killed or more refined if there is not sufficient support. |
| 11:15 | <webben> | hmm... http://www.w3.org/2005/10/Process-20051014/policies#Votes doesn't seem to say what happens if there is a hung vote. |
| 11:19 | <hsivonen> | annevk42: I don't know how to do it conditionally with Apache and doing it conditionally doesn't mix well with long-term caching of fonts |
| 11:20 | <othermaciej> | his http://lists.w3.org/Archives/Public/public-html/2009May/0323.html definitely sounds like Chaals thinks consensus == unanimity |
| 11:23 | <webben> | Yep. |
| 11:24 | <Dashiva> | It does sound like it, but based on what chaals has said before I think he just means that he should put a really good effort in to satisfy everyone first |
| 11:24 | <Dashiva> | *that we should |
| 11:25 | <webben> | W3C could probably use a page like http://en.wikipedia.org/wiki/Wikipedia:What_is_consensus%3F |
| 11:28 | <annevk42> | hsivonen, the cache would be per-origin |
| 11:29 | <annevk42> | hsivonen, you'd need to specify Vary: Origin |
| 11:30 | <othermaciej> | I hate citing the Process document to people |
| 11:30 | <hsivonen> | annevk42: seems more complex than one line in Apache config :-( |
| 11:31 | <hsivonen> | well, waiting for mod_cors, I guess |
| 11:32 | <hsivonen> | I suppose this could already be done with variables on conditional mod_headers directives |
| 11:32 | <othermaciej> | Dashiva: actually, now that I read his email closer it sounds like he's blurring the distinction between W3C-style consensus and unanimity |
| 11:32 | <hsivonen> | but I've never studied how those work |
| 11:33 | <annevk42> | I think it can be done with the set environment variable stuff |
| 11:33 | <annevk42> | Probably with similar lines rubys used for the IE8 crap that now leaked on the Web |
| 11:33 | <annevk42> | Why not make the fonts available for everyone? |
| 11:33 | <hsivonen> | annevk42: well, that particular case suggests that Apache env vars aren't really copy-paste-friendly material |
| 11:34 | <hsivonen> | annevk42: I was going to blog on how to optimize @font-face bandwidth without subsetting fonts |
| 11:34 | <hsivonen> | and not letting random sites leech seems like a natural part of it |
| 11:34 | <annevk42> | I believe only Gecko enforces the same-origin restriction on them by the way |
| 11:35 | <hsivonen> | annevk42: I expect Firefox 3.5 to gain enough market share to deter leeches |
| 11:36 | Philip` | would like to see how much a 500KB font can be optimised without subsetting :-) |
| 11:37 | <hsivonen> | Philip`: a 1 MB font can be optimized quite a bit |
| 11:37 | <annevk42> | lol, did othermaciej, hsivonen, and I just wrote essentially the same email? |
| 11:38 | <Philip`> | annevk42: You are all manifestations of the hivemind |
| 11:39 | <othermaciej> | similar, not quite identical |
| 11:40 | <Philip`> | hsivonen: How about 10MB fonts, then? :-) |
| 11:43 | <hsivonen> | Philip`: dunno. The largest fonts I use are 1.5 MB: http://hsivonen.iki.fi/fonts/libertine/4.4.1/ |
| 11:43 | <hsivonen> | Philip`: small fonts compress nicely, too: http://hsivonen.iki.fi/fonts/liberation/1.04.93.devel/ |
| 11:45 | <Philip`> | Hmm, http://en.wikipedia.org/wiki/Kochi_font (9MB for one .ttf) seems to be the largest font I have, excluding the ones I stole from Windows and OS X |
| 12:21 | <gsnedders> | http://w3.org/mid/4A1A4AF8.1030001⊙iaaj — never miss monthly deadlines again! |
| 12:23 | <annevk42> | http://fberriman.com/2009/05/24/html5-microdata-over-cooked/ |
| 12:23 | <hsivonen> | jgraham: added info on how to check out trunk to http://about.validator.nu/htmlparser/ |
| 12:24 | <jgraham> | hsivonen: Thanks |
| 12:24 | <annevk42> | hsivonen, did you see http://schepers.cc/?p=117 ? |
| 12:24 | <hsivonen> | annevk42: yes. my comment is in the moderation queue. |
| 12:24 | <hsivonen> | annevk42: I think it's night time on shepazu's time zone |
| 12:24 | <annevk42> | myakura's twitter stream is nice |
| 12:25 | <hsivonen> | annevk42: http://pastebin.ca/1433802 |
| 12:26 | <annevk42> | I just checked, even the 1999 Process document does not say objections need to be put *in* the document. They just need to be archived and referencable. http://www.w3.org/Consortium/Process/Process-19991111/activities.html#WGArchiveMinorityViews |
| 12:29 | <annevk42> | Hmm, Doug doesn't point to your original objection |
| 12:30 | <annevk42> | oh, but searching for W3C Technology Stack gives it back as the second unindented result :) |
| 12:32 | <hsivonen> | also, http://www.google.com/search?q=Web+Technology+Stack and http://www.google.com/search?q=Browser+Technology+Stack |
| 12:33 | <annevk42> | "official doctrine of the WHATWG" :) |
| 12:35 | <hsivonen> | I wasn't aware that the WHATWG had official doctrines |
| 12:35 | <Philip`> | Maybe we should set up a wiki page defining the official doctrines |
| 12:36 | <hsivonen> | should Hixie's IRC remarks be considered to be 'ex cathedra'? |
| 12:38 | <annevk42> | I think Doug misunderstood that you didn't want to make a diagram for the Web but just one for browsers |
| 12:38 | <jgraham> | I thought Hixie was more like Christ himself. His IRC comments are like the sermon from the mount |
| 12:39 | <annevk42> | what is HID btw? |
| 12:39 | <hsivonen> | annevk42: perhaps I should have revised the URL to match the title... |
| 12:39 | <hsivonen> | annevk42: but cool URIs don't change now |
| 12:39 | <hsivonen> | annevk42: Human Interface Devices |
| 12:40 | <Philip`> | jgraham: We need to work out these issues before Hixie dies, otherwise we won't know whether to wait for his resurrection or form a conclave to elect a new Hixie |
| 12:41 | <Philip`> | Of course it's more likely that Hixie is Brian |
| 12:41 | <gsnedders> | Philip`: Or we could just wait or his second coming. |
| 12:41 | <gsnedders> | *for |
| 12:42 | <jgraham> | Philip`: The whole resurrection thing is overrated anyway because he just disappears off again straight afterwards. So we probably need to go with the whole "elect Hixie's representative on earth" thing anyway |
| 12:43 | <gsnedders> | And then we'll end up reforming WHATWG so we no longer have a representative of Hixie on earth, and just follow Hixie himsel. |
| 12:43 | <Philip`> | I think I'm going to stick with the Hixie-as-Brian idea, because that's got much better songs |
| 12:44 | <gsnedders> | "He grew, grew, grew and grew, grew up to be/A man, a man called Hixie." |
| 12:45 | <jgraham> | "He's not the messiah, he's a very naughty boy" |
| 12:45 | <jgraham> | Somehow that is even funnier in this context |
| 12:46 | gsnedders | is listening to Brian Song by Monty Python from Monty Python Sings |
| 12:59 | <othermaciej> | Sam is really confusing me now |
| 13:05 | <annevk42> | othermaciej, I think in part it may have been differences between the various Process document versions |
| 13:06 | <annevk42> | othermaciej, just skimmed through the two and things have changed |
| 13:06 | <othermaciej> | that explains his public-html email |
| 13:06 | <othermaciej> | it does look like he was using a stale version |
| 13:07 | <othermaciej> | I'm confused by his www-archive email |
| 13:10 | <annevk42> | if you're interested in contradictions btw, search for "willful violation" throughout the HTML5 specification |
| 13:11 | <hsivonen> | is there some kind of simple script that opens a local HTTP "proxy" that responds with zero-length 200 or 404 to every request? |
| 13:12 | <annevk42> | "That may be because there's no way to register them. So claiming extensions aren't possible isn't helpful, when reality proves they are possible." -- is it just me or does this sound weird coming from Julian? |
| 13:13 | <othermaciej> | annevk42: I'm mainly interested in which ones Sam thinks are critical |
| 13:13 | <othermaciej> | but thanks for pointing that out |
| 13:14 | <hsivonen> | annevk42: what Julian says makes sense to me |
| 13:14 | <othermaciej> | sounds like the "url" willful violations make sense |
| 13:14 | <othermaciej> | hsivonen: it makes sense to me too, but I agree with annevk42 that it sounds unlike Julian's usual prescriptivist point of view |
| 13:16 | <othermaciej> | annevk42: it |
| 13:16 | <othermaciej> | annevk42: it's pretty enlightening to look at these |
| 13:17 | <othermaciej> | annevk42: would be an interesting exercise to catalogue and classify them if Sam's intent is really sincere |
| 13:17 | <othermaciej> | it sounds like many will be fixed by new or updated specs to refefence |
| 13:18 | <othermaciej> | there are surprisingly few |
| 13:20 | <othermaciej> | the only contradiction with a W3C specification is charmod |
| 13:20 | <othermaciej> | and it would be hard to disagree with that one |
| 13:29 | gsnedders | wonders whether to commit a change to php html5lib which changes the tokenizer from multiple methods to one big switch statement |
| 13:30 | <gsnedders> | Takes the time to tokenize the spec down from 7.0s to 5.7s |
| 13:30 | <Philip`> | What reasons are there against committing it? |
| 13:30 | <gsnedders> | It's a bit more messy |
| 13:31 | <Philip`> | This is HTML, it's meant to be messy |
| 13:32 | <jgraham> | gsnedders: Will it make the code significantly harder to maintain? |
| 13:32 | <Philip`> | It's code that shouldn't need to be extended significantly in the future (since the spec isn't going to change much), so maintenance isn't the most critical issue |
| 13:32 | <gsnedders> | jgraham: I expect that depends on who you ask :) |
| 13:33 | <jgraham> | Philip`: In practice the spec has changed a bit and I would expect some further changes as browsers implement this stuff |
| 13:34 | <jgraham> | gsnedders: Assume I am asking you |
| 13:34 | <gsnedders> | jgraham: No, it's only a small bit worse. |
| 13:35 | gsnedders | pushes |
| 13:35 | Philip` | falls |
| 13:35 | <gsnedders> | Wait, I should probably run the test suite :P |
| 13:36 | <gsnedders> | Oh, good. Nothing broke. |
| 13:36 | <jgraham> | gsnedders: BTW it seems like it would be a much more useful use of your time to make a treebuilder rather than work on endless refactorings of the tokenizer |
| 13:36 | <gsnedders> | jgraham: But it's 10 times quicker now! |
| 13:36 | <Philip`> | gsnedders: Yes, but it's useless |
| 13:36 | <Philip`> | because nobody just wants tokens :-p |
| 13:37 | <gsnedders> | Philip`: The test suite does :P |
| 13:37 | <jgraham> | gsnedders: The test suite is not an end user |
| 13:37 | <Philip`> | The test suite doesn't like being anthropomorphised |
| 13:38 | <hsivonen> | meanwhile, I'm confused by the perf results I'm seeing with HTML5-enabled Gecko builds |
| 13:38 | <jgraham> | gsnedders: It's not until you have a treebuilder that people can do thinks like implement rdfa parsers with your code |
| 13:38 | <gsnedders> | jgraham: You could just work with tokens :P |
| 13:39 | <gsnedders> | jgraham: Also, it's more fun working with something that actually runs/ |
| 13:39 | <Philip`> | gsnedders: You can't just work with tokens unless you write your own treebuilder |
| 13:39 | <Philip`> | hsivonen: Confused because it's too fast or because it's too slow? |
| 13:40 | <hsivonen> | Philip`: too slow and suddenly *way* too slow |
| 13:40 | <Philip`> | Ah |
| 13:43 | <jgraham> | hsivonen: A regression or a new measurement? |
| 13:44 | <hsivonen> | the measurement is supposedly the same |
| 13:45 | <jgraham> | hsivonen: So a regression? |
| 13:54 | <hsivonen> | I suppose so. but I have no idea how my own changes could have caused it. I did merge in a lot of trunk code, though. |
| 13:56 | <zcorpan> | Hixie: heya |
| 13:57 | <annevk42> | not awake I think |
| 13:57 | <zcorpan> | hmm right |
| 14:08 | <othermaciej> | I don't think I can reply to Sam's latest email without further pissing him off |
| 14:11 | <othermaciej> | specifically http://lists.w3.org/Archives/Public/www-archive/2009May/0067.html to be specific |
| 14:41 | jgraham | is not sure wtf there is a discussion about the title |
| 14:45 | <othermaciej> | I kinda think it is because of the reason I suggested that made Sam so angry |
| 14:59 | <annevk42> | othermaciej, oh, I just replied as well |
| 15:00 | <annevk42> | othermaciej, I thought you were gonna leave it |
| 15:07 | <othermaciej> | annevk42: I replied to only the technical portion |
| 15:07 | <othermaciej> | will decide whether to address the other part later |
| 15:13 | <zcorpan> | alt text for http://hsivonen.iki.fi/web-stack/ is interesting |
| 15:15 | <annevk42> | http://twitter.com/uberChick/statuses/1909606174 -- I guess that's why we didn't do it |
| 15:16 | <jgraham> | zcorpan: In what way interesting? |
| 15:17 | <zcorpan> | jgraham: not the specified alt text but what is appropriate alt text, considering html5 and wcag etc |
| 15:18 | <zcorpan> | i.e. what would you use as alt |
| 15:19 | <annevk42> | given the processing model example in HTML5 I'd prolly just describe the diagram |
| 15:19 | <annevk42> | WCAG suggests the same; all text of the image should be in alt |
| 15:21 | <Dashiva> | Laura sure loves citing :) |
| 15:22 | <jgraham> | me? Well I guess if I was being good, I would go read the spec... Which would lead me to section 4.8.2.1.2 A phrase or paragraph with an alternative graphical representation: charts, diagrams, graphs, maps, illustrations. But I would quickly realise that was unreasonably hard so I would look at 4.8.2.1.9 A key part of the content, realise that was still too hard |
| 15:22 | <jgraham> | and so write something short like hsivonen did |
| 15:23 | <annevk42> | Dashiva, she cites a lot of discussion but then when she starts making statements she doesn't cite anything |
| 15:23 | <annevk42> | Dashiva, I personally don't find it useful |
| 15:24 | <Dashiva> | annevk4: This is true |
| 15:24 | <hsivonen> | I initially started outlining the idea for the diagram as text and decided that it was easier to draw it when the bulleted list started ballooning |
| 15:44 | <othermaciej> | hsivonen: I don't know how much you care about correct stacking of your browser technology stack diagram, but shouldn't XHR be in some way on top of XML |
| 15:44 | <othermaciej> | ? |
| 15:45 | <othermaciej> | also I'm unsure why Events are on top of JavaScript; in general it seems unclear what layers JavaScript should be on top of or below |
| 15:45 | <Philip`> | And on JSON |
| 15:45 | <Philip`> | (XHR, that is) |
| 15:45 | <Philip`> | unless JSON is on XHR |
| 15:45 | <othermaciej> | XHR doesn't depend technologically on JSON |
| 15:45 | <Philip`> | or unless something else is on both JSON and XHR |
| 15:46 | <othermaciej> | the emerging native JSON APIs might be worth calling out |
| 15:46 | <Philip`> | Unicode doesn't depend technologically on HTTP / URL, either |
| 15:46 | <hsivonen> | othermaciej: it seems to me that XHR is forefost a JS API for HTTP and the shortcut XML parsing is a minor point |
| 15:46 | <Philip`> | so I presume that's not what the diagram is showing |
| 15:46 | <othermaciej> | fair enough but then maybe it should sit right on top of http |
| 15:46 | <jgraham> | othermaciej: My understanding is that "Events" means user-initiated events (although I guess it should still be on top of DOM) |
| 15:47 | <hsivonen> | othermaciej: mainly the issue is that laying out the boxes in 2 dimensions restricts what boxes can touch each other |
| 15:47 | <othermaciej> | on second thought I guess unicode interposes in the same way as for text, but unicode on top of http is not a dependency it's a pipeline |
| 15:47 | <othermaciej> | yeah |
| 15:47 | <Dashiva> | 3D canvas to the rescue |
| 15:47 | <othermaciej> | diagrams like this end up either unreadable or inaccurate or both |
| 15:47 | <hsivonen> | all the specs that are above Unicode really deal with byte layer too and have Unicode inside the parsing rules |
| 15:48 | <othermaciej> | the comparison to the W3C diagram is interesting |
| 15:49 | <othermaciej> | not sure if XPath is worth citing in your version |
| 15:50 | othermaciej | is looking at WebCore directory names |
| 15:50 | <Philip`> | Dashiva: That wouldn't solve the fundamental problem, which is that the topology of finite-dimensional space imposes restrictions on diagrams that are based on adjacencies between non-intersecting shapes |
| 15:50 | othermaciej | flushes with embarassment and moves right on past wml/ |
| 15:51 | <Philip`> | Seems much easier to solve it by just drawing a graph with potentially intersecting edges |
| 15:51 | <othermaciej> | plugins are part of the browser technology stack but I guess don't have the same "open web" flavor |
| 15:52 | <Dashiva> | Philip`: But a static image of such a graph would also be pretty confusing. With canvas you could let the user navigate around it :) |
| 15:53 | <hsivonen> | I guess adding XPath would require putting Selectors there as a box as well |
| 15:54 | <hsivonen> | and yes, NPAPI would belong in the diagram. can't figure out where, though |
| 15:54 | <Philip`> | Dashiva: It wouldn't necessarily be pretty confusing - it'd probably be mostly a treeish structure like hsivonen's current diagram, but with a few extra edges in weird directions |
| 15:54 | <hsivonen> | Canvas 2D could be called out on ints own right, but it would have to be between JS and Visual Rendering |
| 15:54 | <othermaciej> | hsivonen: maybe you should use RDF and make the Browser Technology Graph |
| 15:55 | <othermaciej> | clearly a stack is not an adequate data structure for this use case |
| 15:55 | <hsivonen> | I guess a graph would show the intertwingularity better |
| 15:55 | Philip` | would probably write .dot files rather than RDF |
| 15:55 | <hsivonen> | Word template? |
| 15:56 | <jgraham> | Philip`: You could write HTML5 microdata, autoconvert it to RDF then to .dot and finally to SVG, which you could in turn embed inside the original HTML document |
| 15:56 | <othermaciej> | I don't know what HID is |
| 15:56 | <jgraham> | hsivonen: graphviz |
| 15:57 | <hsivonen> | othermaciej: Human Interface Device |
| 15:57 | <hsivonen> | pointing device / text input was too long |
| 15:58 | <annevk42> | hsivonen, I hope someone will appeal to some higher court? (re twitter) |
| 15:59 | <othermaciej> | now I feel less embarassed that I don't know |
| 15:59 | <othermaciej> | the granularity of specifications is kind of arbitrary |
| 16:00 | <othermaciej> | which is unfortunate for a diagram like this |
| 16:00 | <othermaciej> | editing is a pretty big subsystem in WebKit but it has no spec to call its own |
| 16:01 | <annevk42> | mostly HTML5 I reckon |
| 16:02 | <othermaciej> | yeah although HTML5 is not nearly detailed enough about editing behavior to allow for interoperability I think |
| 16:02 | <othermaciej> | man the w3c version of the web technology stack has an impressive collection of acronyms I've never heard of |
| 16:02 | <hsivonen> | annevk42: I don't know how the case will proceed in the courts. there's a parallel request to an ombudsman who sits aside from the courts and the parliament |
| 16:10 | <annevk42> | a Parliamentary Ombudsman? |
| 16:10 | <hsivonen> | annevk42: yes |
| 16:10 | <annevk42> | apparently the concept differs a lot per country |
| 16:11 | <annevk42> | seems it would be annoying if that person agreed with the judge |
| 16:11 | <hsivonen> | the most basic problem is that Finland lacks a constitutional court and the parliament oversees the constitutionality of its own acts which is a joke |
| 16:15 | <annevk42> | so in theory the parliament has unlimited power? |
| 16:16 | <othermaciej> | the UK parliament has theoretically unlimited power too, so it's not unprecedented |
| 16:17 | <hsivonen> | annevk42: I think a couple of consecutive parliaments together have unlimited power (to the extent the EU doesn't sue) |
| 16:17 | <othermaciej> | courts will not even uphold a law that tries to restrict future acts of parliament |
| 16:17 | <annevk42> | othermaciej, according to Wikipedia that will soon stop as the UK will get a supreme court |
| 16:17 | <annevk42> | http://en.wikipedia.org/wiki/Supreme_Court_of_the_United_Kingdom |
| 16:20 | <othermaciej> | yeah it will have a Supreme Court but I don't think that will establish judicial review of acts of parliament, since there is no written constitution, just conventions |
| 16:22 | annevk42 | finds http://en.wikipedia.org/wiki/Democracy_Index#2008_ranking |
| 16:23 | <hsivonen> | can an old version of CHUD block Mac OS X update? |
| 16:24 | <hsivonen> | I wonder what I should do when an OS X update fails to install |
| 16:27 | <hsivonen> | http://theory.uchicago.edu/~harvey/harveyomega/ |
| 16:49 | <Philip`> | hsivonen: Doesn't work very well in Opera :-( |
| 16:49 | <Philip`> | (The page reloads after pressing enter) |
| 16:49 | Philip` | blames all the Opera people here |
| 18:09 | <gsnedders> | http://digg.com/tech_news/Tim_Berners_Lee_on_the_next_Web?t=25847831#c25848908 — hah! Too true! |
| 18:57 | <gsnedders> | jgraham: What was that graphing software you were saying I should use |
| 18:58 | <Philip`> | Excel? |
| 18:59 | <gsnedders> | Philip`: No |
| 19:02 | <gsnedders> | Veusz |
| 19:03 | <gsnedders> | http://krijnhoetmer.nl/irc-logs/whatwg/20090220#l-561 |
| 19:07 | <gsnedders> | I think something is wrong with the OS X binary. |
| 19:12 | Philip` | wonders if Larry is underestimating the extent to which trivial details end up causing sites to break in some browsers |
| 19:19 | <Philip`> | Hmm, I've even written code myself that tests if img.width != 0 and if so then it uses both img.width and img.height |
| 19:20 | <Philip`> | (though it does check img.complete first) |
| 19:20 | <Philip`> | (and just uses width==0 to test whether it completed successfully or unsuccessfully) |
| 19:39 | <annevk4> | hmm @reply in twitter means the reply will not appear for the people following you? |
| 19:39 | <annevk4> | wtf |
| 19:39 | <Dashiva> | Well, sometimes |
| 19:40 | <Dashiva> | I haven't been able to figure out when it doesn't appear |
| 19:42 | <annevk4> | maybe only when you are not following one of them |
| 19:42 | <annevk4> | that seems to be it |
| 19:42 | <annevk4> | I don't like it |
| 19:43 | <Dashiva> | Ah, that seems to fit my observations, yes |
| 19:43 | <gavin_> | people following you don't get @replies you send to people that they aren't themselves following |
| 19:44 | <annevk4> | sucks |
| 19:44 | <gavin_> | that's what the #fixreplies fiasco was about |
| 19:44 | <gavin_> | used to be a user pref to control whether you saw them |
| 19:46 | <annevk4> | I thought the whole idea was that if I followed someone I got to see what they do and what keeps them busy |
| 19:47 | <annevk4> | oh well |
| 19:50 | <gavin_> | I guess the reasoning is that making it a user pref is costly (need to check all follower prefs for each tweet) |
| 19:50 | <gavin_> | and making it the default is sometimes confusing/undesired |
| 19:50 | <gavin_> | so it was just removed |
| 19:51 | <Madness> | http://wiki.whatwg.org/wiki/RelExtensions hm, no icon or shortcut icon defined... |
| 19:52 | <annevk4> | that's in HTML5 itself |
| 19:53 | <Madness> | Meaning it's not an extension? |
| 19:53 | <annevk4> | I suppose |
| 19:53 | <annevk4> | maybe we should still list them there for completeness |
| 19:54 | <Madness> | Yes please, so silly heads like me won't get puzzled. :) |
| 19:57 | <jgraham> | gsnedders: wfm |
| 20:08 | <gsnedders> | jgraham: http://stuff.gsnedders.com/veusz.tiff |
| 20:09 | <gsnedders> | Version 1.2 works though |
| 20:12 | <annevk4> | Hixie, stylesheet should come after sidebar |
| 20:23 | <gsnedders> | OK, does anyone have suggestions for non-sucky non-buggy graphing software? |
| 20:36 | <hsivonen> | I'm baffled by being accused of divisive rhetoric when the accusation talks about WHATWG doctrine and implies I assumed sinister motives when I didn't. Hmm. http://schepers.cc/?p=117 |
| 20:36 | <annevk42> | Arguably the definition of "available" is somewhat flaky |
| 20:37 | <annevk42> | E.g. how can one know the image is valid if it is only partially downloaded. And if valid is a requirement for being available, how it can be available before being completely downloaded. |
| 20:37 | <annevk42> | (Apart from the fact that supported image type and valid image are indeed not defined.) |
| 20:41 | <Philip`> | hsivonen: Maybe you're just much more subtle in your use of divisive rhetoric, so that it can't be easily seen from individual quotes |
| 20:44 | <hsivonen> | Philip`: oh? my original observation was that the W3C's PR materials weren't being inclusive of HTML and XHR, which makes it seem like the W3C hasn't quite accepted what it invited |
| 20:45 | <annevk42> | Apparently they just haven't updated their marketing materials |
| 20:45 | <hsivonen> | clearly |
| 20:45 | <Dashiva> | I'm baffled that a single calf is being called a cow path |
| 20:45 | <annevk42> | It wasn't about CR or Rec; as Ivan points he made the diagram in 2004-5 and back then HTML5 and XHR didn't happen |
| 20:46 | <hsivonen> | annevk42: it has been updated since then, and the surrounding material has been updated since then |
| 20:46 | <hsivonen> | I don't assume that things were kept out deliberately, but it's PR material that isn't up-to-date |
| 20:46 | <annevk42> | But the diagram (other than removing one part) remained the same right? |
| 20:47 | <hsivonen> | annevk42: no |
| 20:47 | <annevk42> | oh ok |
| 20:47 | <hsivonen> | annevk42: XSL, RDFa and GRDDL have been added |
| 20:47 | <annevk42> | ok |
| 20:48 | <annevk42> | I tend to agree that calling you divisive for pointing out something you consider flawed is not very nice |
| 20:49 | <Philip`> | hsivonen: It may also be that you're not using divisive rhetoric at all, so I didn't mean to imply you were :-) |
| 20:50 | <annevk42> | by schepers' own logic WICD shouldn't be there btw |
| 20:50 | <annevk42> | and if CR does qualify I'm missing XBL2 |
| 20:54 | <hsivonen> | http://search.twitter.com/search?q=http%3A%2F%2Fschepers.cc%2F%3Fp%3D117 |
| 20:57 | <annevk4> | fun, they applaud the post and we critique it; guess we're not making much process on that front then |
| 23:34 | <Hixie> | hm, larry is right. the wg isn't chartered to produce the current html5 draft. i guess we'd better start over. |
| 23:36 | <annevk4> | you don't think it's still evolved from HTML4 if you look at the features? |
| 23:36 | <Hixie> | nope, not evolved at all, it's a whole new language. woe is us we made a terrible mistake. |
| 23:37 | <Philip`> | According to http://www.w3.org/html/wg/ we have a week before we're scheduled to reach CR - that seems plenty of time to download a copy of the HTML4 spec and then paste some of the new features into it |
| 23:37 | <annevk4> | Hixie, mu |
| 23:37 | <annevk4> | Philip`, could try to give HTML4 a proper CR phase and see what happens :) |
| 23:38 | <Hixie> | well i sure hope someone volunteers to do that, i'm too tired from writing the last draft, which turned out to be totally the wrong thing and which we are now forced to throw away. |
| 23:38 | <Hixie> | maybe larry can write it |
| 23:39 | <othermaciej> | I didn't read Larry's walls of text yet |
| 23:39 | <othermaciej> | should I? |
| 23:40 | <annevk4> | this is about a much shorter email it seems |
| 23:40 | <annevk4> | subject line ending in "(from scratch)" |
| 23:41 | <annevk4> | I think I'll go read something and go to bed |
| 23:41 | <annevk4> | not really interested in another roundtable of discussing this; not too sure why I put effort into it today |