| 00:47 | <zcorpan> | http://zope.cetis.ac.uk/members/scott/blogview?entry=20090624222327 |
| 01:13 | <Dashiva> | So about those summaries... has anyone extracted a list of _useful_ summaries from that list Philip` made? |
| 01:14 | <zcorpan> | Dashiva: what's a useful summary? |
| 01:15 | <Dashiva> | There was that argument "Most of those summaries would never be used since they're on layout tables" |
| 01:16 | <Dashiva> | But that doesn't imply that the summaries that were on data tables are good. So I'm wondering if anyone has actually found some examples. |
| 01:17 | <zcorpan> | where's the data again? |
| 01:17 | <Dashiva> | http://philip.html5.org/data/table-summary-values-dotbot.html |
| 01:17 | <Dashiva> | (It seems reasonable to assume that a good, descriptive summary would not come from a layout table) |
| 01:22 | <zcorpan> | A table of four, single-cell rows. Cell one contains the trail to this section from the document home. Cell two contains a navigation bar to the pages within the current section of the overall document. Cell three contains the body text of the document. Cell four contains navigation to reference information for this web site www.mf-ct.upc.es/bathfreesim_doc/bfs_pra17.html |
| 01:22 | <zcorpan> | that's the first i found |
| 01:23 | <zcorpan> | not clear to me whether it's a layout table or data table |
| 01:23 | <zcorpan> | but it's describing the structure of the table, which is the stated purpose of summary="" |
| 01:23 | <Dashiva> | That actually looks like a layout table, yeah. |
| 01:25 | <zcorpan> | the actual data table on that page does *not* have a summary |
| 01:25 | <zcorpan> | while all layout tables have elaborate summaries |
| 01:26 | <zcorpan> | (the data table on that page seems simple enough to not needing a summary, though) |
| 01:28 | <zcorpan> | "A table with two columns listing title, author, date, source, subject headings and comments for selected newspaper articles" |
| 01:28 | <zcorpan> | http://www.santacruzpl.org/history/clippingfile/cliplist.php?page=3&subjhead2=621&newspaper=&datelimit= |
| 01:29 | <Dashiva> | By the way, I was only asking if such a list existed. You don't have to make it for my sake :) |
| 01:31 | <zcorpan> | i'm not making a list, i'm just having a peek |
| 01:45 | <zcorpan> | actually, that was the only occurrence i found |
| 01:46 | <zcorpan> | obviously i skipped any that were in languages i don't understand |
| 01:47 | <heycam> | Hixie, I think some of the interfaces that were added to the obsolete section need to inherit from HTMLElement |
| 01:47 | <zcorpan> | that table is also simple enough to not actually need a summary |
| 09:16 | <zcorpan> | http://code.google.com/intl/es/speed/articles/optimizing-html.html |
| 09:47 | <othermaciej> | zcorpan: I wonder how effective those kinds of changes are post-gzip |
| 09:48 | <zcorpan> | othermaciej: might have no effect at all |
| 09:49 | <othermaciej> | that would be my gues |
| 09:50 | <othermaciej> | I also can't make much sense of this article: http://code.google.com/intl/es/speed/articles/reflow.html |
| 09:50 | <Rik`> | there's a lot of small perf improvements for big trouble reading code afterwards in those google articles |
| 09:52 | <othermaciej> | what exactly does it think "reflow" is? |
| 09:52 | <Philip`> | Adding </li>s and </p>s (with any necessary whitespace to http://www.google.com/intl/en/privacy.html makes the file 251 bytes larger, and makes the gzipped file 7 bytes larger |
| 09:52 | <othermaciej> | if it's a relayout, how can they say padding or width changes won't cause one in some browsers? |
| 09:52 | <Philip`> | s/whitespace/whitesspace)/ |
| 09:53 | <othermaciej> | or that changing the background *will* cause one in Firefox? (Gecko can't possibly be that stupid) |
| 09:53 | <Philip`> | so it's not no effect at all |
| 09:53 | <Philip`> | s/whitesspace/whitespace/ |
| 09:54 | <Rik`> | even if it has effects, the research done by the Bing team claims it has no effect on users |
| 09:54 | <Rik`> | http://radar.oreilly.com/2009/06/bing-and-google-agree-slow-pag.html last paragraph |
| 09:54 | <jgraham> | Philip`: How does it compare to removing all whitespace bwetween elements? I guess that saves more bytes but feer gzip bytes |
| 09:54 | <jgraham> | *fewer |
| 09:56 | <Philip`> | jgraham: Removing whitespace from privacy.html seems to save 696 bytes, or 76 bytes in the gzipped version |
| 09:58 | <zcorpan> | maybe they should have focused on gzip and caching etc, and skip the minor optimizations that affect workflow and maintainability |
| 09:59 | <zcorpan> | the organization of css files is another thing that has close to no effect once gzipped |
| 09:59 | <Philip`> | But micro-optimisations are more fun :-( |
| 09:59 | <Rik`> | look at php otimizations, they are even more fun |
| 10:00 | <othermaciej> | what they should have done is told you that you how to measure the results of optimizations and not bother with ones that don't help |
| 10:00 | <zcorpan> | Rik`: like use single quotes instead of double quotes when the string has no variables? |
| 10:00 | <Rik`> | zcorpan: that and "don't copy variables !" |
| 10:00 | <Philip`> | "Sometimes PHP novices attempt to make their code "cleaner" by copying predefined variables to variables with shorter names." - stupid novices, trying to make their code cleaner |
| 10:01 | <Rik`> | I don't wanna see internal google code |
| 10:01 | hsivonen | thinks jgraham's points were good but the "engineering" and "HCI" labels weren't quite apt given than people with an engineering background have been in the "HCI" mindset and vice versa |
| 10:01 | <othermaciej> | or better yet, how to tell where your page load timte is going so you can tell if it is even worth the effort to work on size of your HTML document, compared to other things you could do |
| 10:02 | <othermaciej> | hsivonen: I think "scientist" vs "empath" was in some ways a more apt comparison, though it highlights a different kind of divide |
| 10:02 | <Philip`> | The Page Speed tool is useful, since it does actual measurements and lets you see what's slow in practice |
| 10:02 | <othermaciej> | i.e. valuing measurable results vs. wanting to signal that you care |
| 10:04 | <Hixie> | i was trying to explain the summary="" thing to a non-tech person today, trying to make it sound like the pro-summary="" position was the one i was supporting |
| 10:04 | <Hixie> | and as soon as i mentioned the data we collected, they basically were like "oh well then you should come up with something better" |
| 10:05 | <Hixie> | made me feel better that we're not the only ones who come to that conclusion |
| 10:05 | <jgraham> | hsivonen: Yeah, I wondered about that. But the labels seem to reflect reality quite well even if they don't reflect preconceptions about how reality should be |
| 10:06 | <jgraham> | Maybe labels in general are bad. I could have just gone with "approach 1" and "approach 2" |
| 10:07 | <jgraham> | But I really wanted to get across the fact that language design is a HCI problem |
| 10:07 | <Hixie> | jgraham: the problem is that labelling the two "camps", or explaining the assumptions or approaches those "camps" are taking, doesn't help resolve the issue |
| 10:08 | <hsivonen> | aargh. I've managed to break the parser so that it puts <html> inside <body> |
| 10:08 | <Hixie> | which is that fundamentally, the pro-summary="" advocates have, as far as i can tell, refused to acknowledge that the data we've collected shows that summary="" hasn't worked |
| 10:08 | <hsivonen> | also, I think the data pretty well debunks the notions that |
| 10:09 | <hsivonen> | 1) summary is long |
| 10:09 | <hsivonen> | 2) summary is different from caption |
| 10:09 | <jgraham> | Hixie: The idea was to sho that reasonable people with different appraoches could come to different conclusions |
| 10:09 | <jgraham> | +w |
| 10:09 | <Philip`> | http://code.google.com/speed/articles/prefetching.html - "Profile your changes" - a revolutionary idea! Maybe the PHP article should talk more about profiling and less about changing double quotes to single quotes |
| 10:09 | <Hixie> | jgraham: sure |
| 10:09 | <jgraham> | Hixie: A great deal of friction has arisen because people have assumed that anyone with a different opinion is being unreasonable |
| 10:10 | <Hixie> | jgraham: i'm not sure that anyone who was assuming that before, if there are any, will think any differently now |
| 10:10 | <jgraham> | Hixie: Yeah. |
| 10:11 | <jgraham> | I guess it was a waste of time. Hopefully it wasn't also harmful |
| 10:12 | <Philip`> | It could still help convince people who aren't firmly on any side of the discussion |
| 10:13 | <Hixie> | fundamentally, if two groups have contradictory goals (in this case "make it possible to write fully-accessible pages even if the overall quality of pages is lower than with other approaches" and "improve the overall quality of pages so that more are accessible even if the theoretical maximum accessibility is lower than with other approaches"), they won't come to an agreement |
| 10:13 | <jgraham> | Hixie: To be fair there has been some acknoledgement that it hasn't worked. It is just that they consider that it is important to have it even though it hasn't worked in general; either because they think it might work well in the future or because they think experts can use it and only experts matter |
| 10:13 | <Hixie> | jgraham: that's what i meant by "hasn't worked" |
| 10:14 | <Hixie> | jgraham: ("would only work if used by experts") |
| 10:14 | <jgraham> | Hixie: Right but reasonable people could disagree with that definition of "hasn't worked" |
| 10:14 | <Hixie> | sure |
| 10:15 | <Hixie> | anyway i don't see how i can reconcile the two goals here |
| 10:15 | <Hixie> | they seem to be contradictory |
| 10:15 | <jgraham> | Particularly if those "reasonable people" consider themselves to be the sort of experts who will use it |
| 10:17 | <jgraham> | Hixie: Allow flow content in <caption>, promote <caption><details> as a solution for cases where an extended summary is needed. It's not sure to work but it is a different approach that might be worth trying; soesn't seem to have potential for harm and might resolve the deadlock :) |
| 10:17 | <Hixie> | <caption><details> really isn't what i would recommend imho |
| 10:17 | <Hixie> | a disclosure triangle in a table caption would be really silly-looking |
| 10:18 | <Hixie> | and it's not clear to me that it is necessary anyway |
| 10:19 | <Philip`> | jgraham: Wouldn't any solution that graphically hides some text by default be subject to the same harm of people putting useless/incorrect/unreadable strings in it? |
| 10:19 | <jgraham> | Yeah, I worry about how it would look. It would be good if it could be easilly styled |
| 10:19 | <Hixie> | i think <details> would be far less spammed than summary="" |
| 10:19 | tantek | wonders how many hours of how many individuals' time has been spent on the summary attribute debate, time that could have perhaps been directed at designing and user-testing alternative solutions with an eye towards developing something that user-tests better with typical (or even "modern") HTML authors. |
| 10:19 | <jgraham> | Hixie: In general not allowing flow content in <caption> makes no sense to me |
| 10:20 | <jgraham> | Why can't I have multi-paragraph captions |
| 10:20 | <jgraham> | ? |
| 10:20 | <Hixie> | jgraham: sure, but that's another issue |
| 10:21 | <Hixie> | assuming you've sent (non-summary-related) mail on it i'm sure i'll get around to it and fix that relatively soon |
| 10:21 | <zcorpan> | jgraham: i think the idea was to be consistent with other heading-like elements like h1, legend, dt |
| 10:21 | <jgraham> | Hixie: I'm not sure if I did |
| 10:21 | <Hixie> | jgraham: ah well then file a bug or something |
| 10:21 | <zcorpan> | jgraham: although it seems reasonable to be able to put two paragraphs or a list in a caption |
| 10:22 | <jgraham> | But anyway once you get flow content you get <details> for free so it doesn't seem unreasonable to suggest that for the cases where a complex table needs an extended summary |
| 10:23 | <Hixie> | if i'm going to suggestion something for complex tables that are so complex they need an extended summary, what i would suggest is "make your table simpler" |
| 10:23 | <Hixie> | not "hide your explanation so that people have to trigger a button to see it" |
| 10:23 | <zcorpan> | jgraham: maybe <legend> should allow blocks too in figures |
| 10:24 | <Hixie> | (are multiparagraph captions something one sees in print at all?) |
| 10:25 | <jgraham> | Hixie: right, that would be my first point of advice too. But pragmatically there are some cases where you might be constrained to not do that |
| 10:25 | <Hixie> | i don't know that i'd ever tell people to use <details> even then |
| 10:25 | <Hixie> | but that's mostly academic imho |
| 10:26 | <Hixie> | the question is how do we proceed on this issue given that as far as i can tell, summary="" advocates won't accept anything short of including summary="" |
| 10:27 | <Hixie> | and given that the chairs are apparently unwilling to make any sort of statements regarding a decision |
| 10:27 | <jgraham> | Hixie: I wwould be interested to see if they would accept using <details>. It seems to be basically equivalent to using a <summary> element which they have advocated as a long term solution |
| 10:27 | <annevk2> | we could take the path of least resistance as shelley thinks we already do, but I'm not sure that'd be good |
| 10:27 | <Hixie> | annevk2: what's the path of least resistance? |
| 10:27 | <Hixie> | jgraham: have they not responded to those proposals? |
| 10:28 | <jgraham> | Hixie: Who is "they"? |
| 10:28 | <jgraham> | Oh I see what you mean |
| 10:28 | <Hixie> | jgraham: have summary="" advocates and <summary> advocates not commented on the <details> proposals? |
| 10:29 | <jgraham> | IIRC the response was more or less "we like the name 'summary' better" |
| 10:29 | <Hixie> | so they're ok with it? |
| 10:29 | <Hixie> | wow, i somehow missed that |
| 10:30 | <jgraham> | Hixie: I think they didn't say that :) |
| 10:30 | <Hixie> | what did they say then? |
| 10:30 | <zcorpan> | i don't understand the <summary> proposal |
| 10:30 | <jgraham> | AFAICT proposals are only taken seriously when you or Sam expresses an opinion on them |
| 10:30 | <annevk2> | Hixie, include it? |
| 10:30 | <zcorpan> | would <summary> be visible or not? |
| 10:31 | <Hixie> | annevk2: that's not the path of least resistance; i, for one, would object strongly to that solution as it has been shown to actually harm users |
| 10:31 | <jgraham> | zcorpan: Leif had a somewhat more coherent version of the proposal onthe list |
| 10:31 | <zcorpan> | <summary> would have to go inside the <caption> for legacy parsing reasons |
| 10:32 | <jgraham> | I think <summary> would have to work like <table><caption>My shory summary <summary>My description of the table structure</summary></caption> |
| 10:32 | <jgraham> | and <sumamry> would be hidden by default and get a disclosure triangle |
| 10:33 | <Hixie> | i think that'd be exactly as bad as <details> |
| 10:33 | <jgraham> | (btw I find it counterintuituve that <details> uses <legend> around the bit that is visible by default rather than the bit that is hidden) |
| 10:34 | <Hixie> | <details> is like <fieldset> except the contents are hidden |
| 10:34 | <Hixie> | anyway i should sleep |
| 10:34 | <Hixie> | i guess i wish the chairs would at least give some guidance on how to proceed |
| 10:34 | <jgraham> | Hixie: Agreed. The technical difference between the proposals is minimal. The differences are mainly about whether it is seen to be Designed for the Blind (tm) |
| 10:35 | <Hixie> | (as i have asked for several times so far, with no response) |
| 10:35 | <Hixie> | nn |
| 10:35 | <jgraham> | gn |
| 10:37 | <jgraham> | (Actually the original proposal has <summary> as a child of <table> which is a technical difference but has really bad legcay behaviour so I was kind of overlooking it) |
| 10:45 | <hsivonen> | <summary> is unworkable for legacy reasons. There's really no point in even discussing it. |
| 10:45 | <hsivonen> | if anyone wants a child element of <table>, it has to be called <caption> no matter what its semantics are defined to be |
| 10:45 | <jgraham> | hsivonen: Which legacy reasons? |
| 10:46 | <hsivonen> | jgraham: fosterparenting |
| 10:46 | <jgraham> | hsivonen: Ah OK. I wondered if there is something else I overlooked |
| 10:46 | <jgraham> | As I said, it would have to be a child of <caption> |
| 10:47 | <Lachy> | there's no reason the summary needs to be a child of the table. All that's needed is some kind of association between the summary and the table, and there are many ways to do that |
| 10:48 | <jgraham> | Lachy: I can think of three (caption, figure and aria-describedby) which is not really "many" |
| 10:49 | <hsivonen> | jgraham: <summary> as sibling of <table> (quite unconventional, yes) |
| 10:49 | <zcorpan> | putting it as child of caption has the advantage that the association works in legacy AT |
| 10:51 | <Philip`> | hsivonen: Before or after the <table>? Ignoring whitespace? Ignoring comments? |
| 10:52 | <jgraham> | Philip`: sibling is quite eel defined, no? |
| 10:52 | <jgraham> | s/eel/well/ |
| 10:52 | <Philip`> | Oh, I'm probably misinterpreting it as "adjacent sibling" |
| 10:52 | <Philip`> | You could write <table><summary>... and then it'd be a sibling of the table :-) |
| 10:52 | <Lachy> | I'm not convinced of the need to have such explicit assocation. Just summarise the table in the preceding paragraph |
| 10:53 | <Lachy> | so <p>This is a summary of the following table...</p><table>...</table> |
| 10:54 | <jgraham> | Lachy: If people who use AT argue it is helpful for them to have such an association it is quite hard to argue it is not helpful (I'm not sure if people actually did argue this for table summaries) |
| 10:54 | <jgraham> | (but it seems plausible at least because you can e.g. reread the summary eailly if you get lost in the middle of the table) |
| 10:55 | <Philip`> | Sibling would be bad because people would write <!doctype html><p>Blah blah <summary>Blah</summary> <table> and then it wouldn't actually be a sibling |
| 10:57 | <Lachy> | I'm sure they did. That doesn't mean they're right. I haven't seen any usability studies demonstrating how effective or ineffective a summary would be if it doesn't have an explicit association in the markup |
| 10:59 | <Philip`> | In the absence of any usability studies (because presumably the WG doesn't care enough about accessibility to actually do any), the best information we have about user behaviour is what users claim |
| 11:05 | <Lachy> | Philip`, sure, but anyone in the field of usability knows you should avoid giving too much weight to what the user says, as their behaviour can often completely contradict that |
| 11:07 | <Philip`> | Lachy: Indeed, but hypothesising about other people's behaviour is likely to be even less accurate, and we don't currently have any other sources of information |
| 11:08 | <Lachy> | I havent' hypothesised about it. I just said I wasn't convinced of the need for explicit association |
| 11:09 | <Lachy> | and in the interst of finding the simplest solution possible, I think simply including the summary in the surrounding prose may be sufficient |
| 11:10 | <Philip`> | You're hypothesising that it's sufficient for users |
| 11:10 | <zcorpan> | http://simon.html5.org/articles/mobile-results first used summary="", then moved it to a <p> before the table, and now includes it in the <caption> |
| 11:10 | <Lachy> | well, in that case, you're hypothesising that it isn't, and neither one of us have any data to test it |
| 11:11 | <Philip`> | Lachy: I'm not hypothesising anything myself |
| 11:12 | <Lachy> | you're hypothesising just as much as I am, which is none. But it is you that claimed I was |
| 11:12 | <Philip`> | You said "I think [stuff] may be sufficient" which sounds like a hypothesis to me :-) |
| 11:13 | <jgraham> | If you knew whether users were right about their own needs more often than they were wrong then you could use that to pick the more likely hypothesis |
| 11:14 | jgraham | notes that ouldn't actually work for all sorts of reasons |
| 11:14 | <jgraham> | wouldn't |
| 11:16 | <zcorpan> | Lachy: i think bug 7035 is questioning the use case of the content model concept, not the diagram itself |
| 11:21 | <Philip`> | On the subject of page speed: if you use mod_deflate in Apache 2.2, then it breaks caching (by generating invalid ETags) |
| 11:21 | <Philip`> | I would have hoped this kind of thing would be more of a solved problem by now |
| 11:26 | <Dashiva> | Oh look, antoher reference to the bible on public-html |
| 11:26 | <hsivonen> | Philip`: before, ignoring comments and whitespace |
| 11:27 | <Lachy> | zcorpan, it clearly said: |
| 11:27 | <Lachy> | <sayrer> |
| 11:27 | <Lachy> | http://www.whatwg.org/specs/web-apps/current-work/images/content-venn.svg |
| 11:27 | <Lachy> | <sayrer> why does that exist? |
| 11:28 | <Lachy> | and the reason content models exist should be fairly obvious to anyone |
| 11:30 | <jgraham> | Lachy: sayrer is a HTML anarchist. He doesn't believe in any sort of central authority over what is allowed in a document |
| 11:30 | <MikeSmith> | I think Rob may be questioning whether the use cases for implementors needing to care anything about the content classes that the spec defines -- interactive, sectioning, etc. |
| 11:31 | <Lachy> | MikeSmith, conformance checkers do. Browser's probably don't |
| 11:32 | <jgraham> | MikeSmith: I failed to parse that sentence |
| 11:32 | <jgraham> | +exist maybe? |
| 11:32 | <MikeSmith> | jgraham: remove "whether", I think |
| 11:32 | <MikeSmith> | anyway, it's fun to speculate about Rob thinks |
| 11:32 | <MikeSmith> | we should play this game more often |
| 11:32 | <MikeSmith> | for other stuff |
| 11:33 | <Lachy> | MikeSmith, but it was you who filed the bug. Why couldn't Rob do it? |
| 11:33 | <MikeSmith> | like, instead of What Would Jesus Do?, we could ask, What Would Rob Think? |
| 11:33 | <MikeSmith> | Lachy: I filed that bug solely for humor purposes. |
| 11:33 | <Lachy> | ok |
| 11:36 | <takkaria> | so the bugtracker is the place to explore possibility for entertainment value but the mailing list isn't? :) |
| 11:36 | <beowulf> | takkaria: irc is no place for humour |
| 11:37 | <MikeSmith> | takkaria: well, I already got dressed down by a range of partisans for using the bugzilla for entertainment purposes |
| 11:37 | <MikeSmith> | so I suppose I won't do it again |
| 11:37 | <beowulf> | basg.org is not acceptable evidence as bash.org is not an expert on irc |
| 11:37 | <MikeSmith> | unless/until I forget the lesson I learned from this |
| 11:37 | <beowulf> | s/basg/bash |
| 11:38 | <MikeSmith> | which should take at least 2 weeks for me to forget |
| 11:38 | <MikeSmith> | anyway, in other news, fwiw to anybody on here who might be running their own instance of validator.nu -- I just checked in an new "--local" option that lets you run v.nu without it needing to grab the IANA language registry and Wiki stuff over the network remotely |
| 11:39 | <zcorpan> | speaking of wiki and v.nu... |
| 11:39 | <MikeSmith> | so it will help you to say, prevent yourself from looking likely an total asshead (as I did) when you go somewhere and try to do a v.nu demo and find you have no network connection |
| 11:39 | <zcorpan> | MikeSmith: do you think it would be useful for v.nu to point to http://wiki.whatwg.org/wiki/Presentational_elements_and_attributes when complaining about certain obsolete elements? |
| 11:40 | <MikeSmith> | zcorpan: yeah, that could be added to the assertions-checking code, I suppose |
| 11:41 | <MikeSmith> | like I said, I think there's a lot more that could be done yet with putting out advice for obsolete elements and other stuff |
| 11:41 | <MikeSmith> | zcorpan: there was something I was going to ask you about earlier today, but I forgot |
| 11:43 | <zcorpan> | MikeSmith: you should ask it even if i'm not here, so that you can point me to the logs later :) |
| 11:44 | <MikeSmith> | true that |
| 11:48 | <MikeSmith> | speaking of lessons learned, I got this amateur schoolmarm who wants to learn me about how to look smarter in my blog postings |
| 11:48 | <MikeSmith> | http://sideshowbarker.net/2009/06/22/webkit-sniffer/#comments |
| 11:48 | <MikeSmith> | he mistakes me for somebody who cares about appearing to be articulate |
| 11:49 | <Dashiva> | There's nothing wrong with Realpolitik |
| 11:50 | <MikeSmith> | Dashiva: I spelled it wrong, apparently |
| 11:50 | <MikeSmith> | Realpolitick |
| 11:50 | <MikeSmith> | with a "ck" |
| 11:50 | <takkaria> | I noticed that when I read it but cba pointing it out |
| 11:50 | <MikeSmith> | I call that the "tribute to 17th-century spelling conventions" spelling |
| 11:51 | <MikeSmith> | anyway, the few blog postings I manage to get around doing are pretty much the equivalent of scribblings on cocktail napkins |
| 11:51 | <MikeSmith> | or on toilet paper |
| 11:52 | <MikeSmith> | anybody who expects me to spell words in them right is going to be disappointed |
| 11:53 | <takkaria> | since befriending an extremely dyslexic person and talking to them online a fair bit, my tolerance for spelling errors has gone up a lot |
| 11:53 | <Dashiva> | MikeSmith: I'm sure those international readers aren't steady on the spelling anyhow |
| 11:54 | <takkaria> | I'm almost slightly annoyed that I'm not annoyed at people not spelling things anymore |
| 11:54 | <takkaria> | s/things/things right/ |
| 11:55 | <MikeSmith> | I think I will reply to his comment with, "Thanks for you pedantrie." |
| 11:55 | <MikeSmith> | and then he can correct me on my misspelling of "pedantrie" |
| 11:56 | <MikeSmith> | that would be great |
| 11:56 | <Dashiva> | If you took each word and randomly jumbled the letters, would that count as 'not spelling' |
| 12:03 | hsivonen | agrees with zcorpan about hasFeature and "2.0" |
| 12:03 | <hsivonen> | hasFeature is pointless already. Better just freeze it. |
| 12:06 | <Dashiva> | Would it be needed at all if Java had proper capability detection? |
| 12:06 | <hsivonen> | Dashiva: no |
| 12:06 | <hsivonen> | Java has reflection :-) |
| 12:07 | <Dashiva> | Been there, done that. :) |
| 12:07 | <hsivonen> | I wonder how often people actually write non-mobile Java apps that use hasFeature |
| 12:07 | <hsivonen> | one would think that first you target JDK 1.4.2 |
| 12:07 | <hsivonen> | then when you upgrade to targeting JDK 5, you just stop running on 1.4.2 |
| 12:08 | <hsivonen> | and when you upgrade to JDK 6, you stop running on JDK 5 |
| 12:08 | <hsivonen> | and when your aren't ready to increment the platform requirement, you just keep targeting the older JDK |
| 12:09 | <hsivonen> | and if you are using a non-Sun JDK that opts to ship a non-Xerces JAXP back end, you are in the world of hurt anyway |
| 13:05 | <Lachy> | LOL, this is what not to do with your ssh private key :-) http://lists.w3.org/Archives/Public/public-webapps/2009AprJun/1241.html |
| 13:23 | <Philip`> | Lachy: Old news :-p |
| 13:23 | <Philip`> | (See #html-wg) |
| 13:27 | <Dashiva> | Lachy got Warnocked |
| 13:30 | <hsivonen> | does Opera support <script defer>? what about IE? |
| 13:30 | <Philip`> | Dashiva: That seems like an entirely novel use of the term "Warnocked" |
| 13:31 | <Philip`> | (unless I'm misunderstanding the context) |
| 13:31 | <Dashiva> | Philip`: If someone had replied to the email, he would know it's an "old" topic. |
| 13:31 | <Dashiva> | But without replies, maybe he was the first to notice? |
| 13:32 | <Lachy> | Dashiva, how did I get warnocked? what does that mean? |
| 13:34 | <Dashiva> | Well, maybe more like second order warnocking. |
| 13:34 | <Dashiva> | http://en.wikipedia.org/wiki/Warnock's_Dilemma |
| 14:14 | <zcorpan> | hsivonen: opera no, ie yes |
| 14:21 | <hsivonen> | zcorpan: ok. thanks |
| 14:34 | <MikeSmith> | annevk2: while you are in Denmark, if you happen to stop by a CD shop, please buy me "I've Got You on Tape" |
| 16:45 | <gsnedders> | jgraham, zcorpan: Would either of you happen to have a bed sheet I could borrow tomorrow? :P |
| 16:46 | <jgraham> | gsnedders: Sure |
| 16:46 | <gsnedders> | jgraham: thanks |
| 16:46 | <jgraham> | gsnedders: Mail, please |
| 16:47 | <gsnedders> | jgraham: Mail for what? Can I just pick it up at the office seeming I need to get flat key from there? |
| 16:47 | <gsnedders> | jgraham: Can you make do with getting it back when I next see you (which I guess should be no later than Monday) |
| 16:47 | <jgraham> | gsnedders: For the purpose of me remembering to look for it |
| 16:47 | <gsnedders> | jgraham: Oh, email. duh. |
| 16:48 | <gsnedders> | jgraham: I can't promise it will manage to send on this wifi. |
| 16:48 | <gsnedders> | :( |
| 16:48 | <jgraham> | Actually, come to think of it, I wonder if we have two bed sheets spare since we have visitors... I guess it is not impossible |
| 16:48 | <gsnedders> | jgraham: cam or opera email? |
| 16:48 | <jgraham> | gsnedders: opera or hopipolla |
| 16:48 | <gsnedders> | jgraham: I only need one bed sheet! :P |
| 16:48 | <jgraham> | *hoppipolla |
| 16:52 | <gsnedders> | jgraham: sent |
| 16:52 | <gsnedders> | (hopefully) |
| 16:52 | <gsnedders> | Still sending… |
| 16:53 | <gsnedders> | Should have sent |
| 20:23 | <spithash> | hey niggers |
| 20:53 | <Hixie> | http://twitter.com/itpastorn/statuses/2329152650 |
| 20:53 | <Hixie> | when two conspiracy theories conflict... |
| 20:55 | <Dashiva> | Nah |
| 20:56 | <Dashiva> | Google doesn't actually like RDFa. Their implementation is intended to subvert existing support, while the spec work ensures it doesn't get any more popularity! |
| 21:26 | <takkaria> | i heard rumours that Sergei actually sits by Hixie's desk and tells him what to specify when he can't make up his mind |
| 21:27 | <Hixie> | actually sergey has never made a suggestion for the spec as far as i recall |
| 21:27 | <Hixie> | the acknowledgements section does list other people who have, though, whose names may be unexpected |
| 21:29 | <takkaria> | you would say that |
| 21:29 | <takkaria> | it must be hard to edit independently with that corporate agenda pushing all your decisions |
| 21:29 | <takkaria> | and the need to keep it quiet so conspiracy theories won't form |
| 21:30 | <inimino> | wow, Rasputin and the Bilderbergs, that /was/ unexpected |
| 21:38 | <Dashiva> | Maybe Hixie doesn't actually exist |
| 21:43 | <takkaria> | Hixie: with regards to disagreeing with stuff in the spec |
| 21:44 | <takkaria> | surely if the arguments for something's inclusion are sound, then your opinion is that that thing should be included? |
| 21:45 | <beowulf> | i read on twitter that hixie is in something called a gene pool |
| 21:45 | <beowulf> | i hope he has an inflatible dinosaur |
| 21:46 | <Hixie> | takkaria: i think the arguments for including the /> syntax are solid, but i still don't _want_ the /> syntax in text/html |
| 21:46 | <Hixie> | takkaria: if that makes sense |
| 21:47 | <Hixie> | takkaria: if i could find a way to get rid of style="", i'd want to do so |
| 21:49 | <takkaria> | hmm, OK |
| 21:49 | <takkaria> | the idealist in you hasn't quiet yet died then :) |
| 21:51 | <Hixie> | right |
| 22:21 | <gsnedders> | Woops. Do I have IRC open? I guess I'll drop offline with this terrible wifi soon |
| 23:01 | <hober> | I guess Michael Jackson missed his opportunity to report a typo in the spec & gain immortality via inclusion in the Acknowledgments section. |
| 23:16 | <sayrer> | jgraham: ftr, I am not an HTML anarchist |
| 23:17 | <sayrer> | but I don't like pointless rules either |
| 23:17 | <sayrer> | and I think "semantic markup" is not worth very much |
| 23:18 | <sayrer> | so disallowing <font> while allowing style="" seems ridiculous |
| 23:19 | <sayrer> | go ahead, just move that presentation stuff into attribute values rather than element and attribute names... then everything is ok! |
| 23:19 | <sayrer> | it's like that monty python skit where everyone changes chairs and says "there, that's much better" |
| 23:39 | <beowulf> | sayrer: pity, i liked the idea of someone being an HTML anarchist |
| 23:41 | <sayrer> | beowulf, prescriptivists have a tough time telling a descriptivist from an anarchist, so there HTML anarchists in the minds of some |
| 23:45 | <beowulf> | sayrer: what is descriptivism in the context of HTML? |
| 23:46 | <sayrer> | beowulf: that would mean writing down what mean, without judging whether they're doing it wrong |
| 23:46 | <sayrer> | what markup means |
| 23:46 | <sayrer> | so for instance, @valign is banned as wrong in Ian's draft |
| 23:47 | <sayrer> | descriptivist approach would be to write it down, and note that the same task can be accomplished other ways (synonyms) |
| 23:49 | <sayrer> | http://en.wikipedia.org/wiki/Ain%27t is a could example |
| 23:49 | <sayrer> | good example |
| 23:49 | <sayrer> | "It is a word that is widely used by many people, but its use is commonly considered to be improper.[1]" |
| 23:50 | <beowulf> | sayrer: ah, i get you |
| 23:51 | <sayrer> | A related concept is the virtue of "semantic markup" |
| 23:51 | <sayrer> | people who obsess about whether they are using <dt> right |
| 23:51 | <sayrer> | my opinion is that the section that most precisely defines the semantics of these elements is section 11, rendering |
