| 00:34 | <hober> | hahahahaha. |
| 01:20 | <Lachy_> | from the article: "The whole thing is supposed to be of help to deaf-blind people, who were never surveyed for their preferences, an action I recommended to WAI at a face-to-face meeting in 2003. Nor was any user testing carried out. (That’s all I can tell from published evidence, anyway. I sent e-mail inquiries to deaf-blind organizations in several countries asking if they’d been surveyed or had any opinions, with no response.)" |
| 01:22 | <Lachy> | that certainly helps to put the whole issue in perspectice, if they have a history of doing no research |
| 01:22 | <Lachy> | *perspective |
| 02:54 | <hober> | indeed. |
| 02:54 | <Dashiva> | The poll appears |
| 03:37 | <Dashiva> | Shelley's links are interesting, considering summary has been rejected DH5 (maybe even DH6) over and over and over... |
| 03:48 | <karlcow> | Dashiva: You meant "The poll dancing appears." |
| 03:58 | <Dashiva> | Also interesting that Roy thinks summary will be required by government when existing government advice actually discourages it... |
| 03:58 | <Hixie> | that government advice was cited by the wai as an example of why we had to keep summary="" |
| 03:58 | <Hixie> | i'd never seen it before then |
| 03:59 | <Hixie> | i think the wai may have assumed i don't read links they sent me |
| 03:59 | <Hixie> | especially long documents like that |
| 03:59 | <Dashiva> | Maybe someone should link it to Roy too, then |
| 04:34 | <karlcow> | Dashiva: Do you have a list of where it is discouraged? and where it is encouraged? (government wise) |
| 04:38 | <Dashiva> | karlcow: The one mentioned earlier is at http://www.access-board.gov/sec508/guide/1194.22.htm#(i) -- scroll one paragraph up |
| 04:42 | <karlcow> | http://www.coi.gov.uk/guidance.php?page=137 |
| 04:43 | <karlcow> | recommends http://webaim.org/techniques/tables/data.php |
| 04:43 | <karlcow> | which says http://webaim.org/techniques/tables/data.php#summaries |
| 04:43 | <karlcow> | so UK seems to say follow wcag |
| 04:44 | <karlcow> | from http://www.coi.gov.uk/guidance.php?page=188 |
| 04:44 | <Dashiva> | It's a shame, that summary there would be useful for us poor normal folks too |
| 04:44 | <karlcow> | troll |
| 04:46 | <Dashiva> | Hah |
| 04:47 | <Dashiva> | That's one of the points that's been made so many times. Summaries often contain non-structural information that's useful to all users, not just disabled ones |
| 04:48 | <karlcow> | French Gov recommends WCAG 2.0 un WCAG 1.0 http://www.references.modernisation.gouv.fr/rgaa-accessibilite |
| 04:48 | <mpilgrim> | everyone says to follow wcag |
| 04:48 | <mpilgrim> | wcag is wrong |
| 04:50 | <Dashiva> | It's easy to think that way if you're stuck in handicap parking analogies, I suppose. There you have to steal spaces from normal users. |
| 04:50 | <Dashiva> | But information can be made available for everyone, so let's not be mean. |
| 04:50 | <Dashiva> | And now, for my next trick, I will not be late for my flight. *poof* :) |
| 04:51 | <mpilgrim> | not only is it wrong, but the people who wrote it are unwilling to listen to criticism on how it is wrong |
| 04:52 | <mpilgrim> | this has been amply demonstrated for many years |
| 04:52 | <mpilgrim> | most recently (though by no means exclusively) by their refusal to respond to hixie's direct feedback |
| 04:53 | <karlcow> | Canada too http://www.tbs-sct.gc.ca/clf2-nsi2/clfs-nnsi/clfs-nnsi-2-fra.asp |
| 04:53 | <mpilgrim> | where it is not wrong, it is merely inadequate |
| 04:53 | <mpilgrim> | (c.f. their techniques document on using the image @alt attribute) |
| 04:53 | <mpilgrim> | (which cites a grand total of 2 examples, both trivial) |
| 04:54 | <karlcow> | mpilgrim: I was just trying to see if Dashiva's assertion was true or not "[23:01] <Dashiva> Also interesting that Roy thinks summary will be required by government when existing government advice actually discourages it..." |
| 04:54 | <mpilgrim> | major changes were made from wcag 1 |
| 04:54 | <mpilgrim> | at the behest of major stakeholders like ibm |
| 04:54 | <mpilgrim> | e.g. pages requiring javascript are now a-ok |
| 04:55 | <mpilgrim> | (technically the priority level was dropped from 1 to 2 or some such nonsense. i forget how they phrased it.) |
| 04:55 | <mpilgrim> | it was EXTREMELY IMPORTANT for ibm that wcag 2 change this requirement |
| 04:55 | <mpilgrim> | since they have major "web" products like lotus workplace which are entirely javascript |
| 04:55 | <karlcow> | :) a big co influencing a spec… hmmm how surprising… that would be really the first time… never seen that in html 5. <sacarsm/> |
| 04:56 | <mpilgrim> | (i put "web" in scare quotes since there's nothing "webby" about them except that they happen to run in some browsers) |
| 04:56 | <mpilgrim> | my point is, that's the kind of "feedback" that went into wcag 2 |
| 04:56 | <mpilgrim> | not anything lofty like john (and now sam (and now roy)) seem to assume |
| 04:56 | <mpilgrim> | research? |
| 04:56 | <mpilgrim> | user testing? |
| 04:56 | <mpilgrim> | seriously? |
| 04:56 | <mpilgrim> | don't make me laugh |
| 04:57 | <karlcow> | making you laugh… now, that would be vulgar :) I beg not. |
| 04:57 | <mpilgrim> | i guarantee you there was ZERO discussion of the summary attribute |
| 04:57 | <mpilgrim> | it was just a given |
| 04:57 | <mpilgrim> | somebody had a bad idea in like 1997 or 1998 |
| 04:57 | <mpilgrim> | to add a summary attribute to tables |
| 04:58 | <mpilgrim> | they got it into the spec, in a wave of "let's make html more accessible" guilt |
| 04:58 | <mpilgrim> | from there, it got laundered through WAI and WCAG |
| 04:58 | <mpilgrim> | and now it's a highly polished turd being offered up and the be-all and end-all of accessibility techniques |
| 04:58 | <mpilgrim> | and no one is allowed to question it because it came from "experts" |
| 04:58 | <mpilgrim> | who *surely* must have researched it at least a little bit |
| 04:59 | <mpilgrim> | instead of just regurgitating the same bad advice for 10 years |
| 05:00 | <mpilgrim> | and people wonder why accessibility is such a clusterfuck |
| 05:01 | <karlcow> | and in clusterfuck, there is cluster. yeeehaaa |
| 05:01 | <mpilgrim> | ignore *!*@*.la-grange.net |
| 05:02 | <karlcow> | arf |
| 05:02 | <mpilgrim> | meanwhile, i'm shocked - SHOCKED! - to discover that roy fielding thinks that appeals to authority are a valid part of technical discourse |
| 05:03 | <mpilgrim> | truly shocked |
| 05:05 | <mpilgrim> | "sampling is irrelevant within the context of HTML that is required by law or regulation to be accessible to sight-impaired users." |
| 05:05 | <mpilgrim> | http://lists.w3.org/Archives/Public/public-html/2009Aug/0136.html |
| 05:06 | <mpilgrim> | you can't argue with logic like that! |
| 05:16 | <mpilgrim> | heh |
| 05:16 | <mpilgrim> | re-reading "to hell with wcag 2" 3 years later is great fun |
| 05:17 | <mpilgrim> | i remember this complaint when it was made: http://lists.w3.org/Archives/Public/w3c-wai-gl/2006AprJun/0184.html |
| 05:18 | <mpilgrim> | w/followup here: http://lists.w3.org/Archives/Public/w3c-wai-gl/2006AprJun/0397.html |
| 05:18 | <karlcow> | hmm a survey for accessibility http://archive.cabinetoffice.gov.uk/e-government/resources/eaccessibility/content.asp |
| 05:18 | <mpilgrim> | and http://lists.w3.org/Archives/Public/w3c-wai-gl/2006JulSep/0000.html |
| 05:19 | <mpilgrim> | i met her (well, was briefly introduced) once at csun |
| 05:19 | <mpilgrim> | remarkable woman |
| 05:19 | <mpilgrim> | can't remember her own phone number |
| 05:20 | <mpilgrim> | (or so i was told) |
| 05:20 | <mpilgrim> | but contributes to complex standards for years on end |
| 05:22 | <mpilgrim> | of course, wcag 2 is still incomprehensible to people with cognitive disabilities |
| 05:23 | <mpilgrim> | and still says nothing useful about how to make your own content accessible to such an audience |
| 06:19 | <Hixie> | anyone got any suggested UI for how to indicate that the changes have been saved once you click "save" on http://www.whatwg.org/specs/web-apps/current-work/edit-section.html?section=the-code-element ? |
| 06:22 | <Hixie> | hm, nevermind |
| 06:34 | <aboodman> | Hixie: I can't be bothered to comment on the list, but I think that having a way to reach all browsing contexts in the same origin is probably also a security hole. |
| 06:34 | <aboodman> | even if we had background pages (which i like fwiw), we couldn't allow them to just reach all origins, there'd have to be a mechanism where pages opt inot being part of the set. |
| 06:35 | <aboodman> | sorry, I mean that we couldn't allow them to just reach all contexts in the same origin. |
| 06:48 | <mpilgrim> | http://www.w3.org/TR/WD-html40-970917/struct/tables.html#h-12.1.1 |
| 06:48 | <mpilgrim> | "The TABLE element contains all other elements that specify caption, rows, content, and formatting. Authors should use the title attribute to provide a summary of the table for non-visual user agents, covering its purpose and structure." |
| 06:48 | <mpilgrim> | "Note. A new attribute for the table summary is needed because the caption for visual users is generally insufficient. The title attribute is inappropriate because browsers render this as a "tool tip" which is limited to a single line of text. In some user agents, long lines are clipped." |
| 06:49 | <mpilgrim> | this is the sum total of the "research" done on the table summary attribute |
| 06:50 | <mpilgrim> | the summary attribute first appeared in the 1997-09-17 working draft of html 4 |
| 06:50 | <mpilgrim> | it is not present in the previous draft from 1997-07-08 ( http://www.w3.org/TR/WD-html40-970708/ ) |
| 06:51 | <mpilgrim> | i can find no discussion of it in the www-html mailing list archives for 1997 |
| 06:51 | <mpilgrim> | or 1996, for that matter |
| 06:52 | <mpilgrim> | it was invented because browsers clipped tooltips |
| 06:53 | <mpilgrim> | if browsers had supported multiline tooltips, we never would have had a summary attribute |
| 06:56 | <hober> | I'm completely shocked. Oh, wait. |
| 07:04 | <mpilgrim> | quote of the day (from http://lists.w3.org/Archives/Public/www-html/1997Jul/0066.html ): "Deprecating a tag doesn't make it obsolete right away, but clearly indicates that there is a 'better' (or at least, more consistent) way of doing things. Any deprecated elements in HTML4.0 won't be obsolete until HTML5.0, probably - and even then they may still be included." |
| 07:04 | <mpilgrim> | and the followup ( from http://lists.w3.org/Archives/Public/www-html/1997Jul/0068.html ): "Nobody ever said that deprecated elements will no longer be supported by browser vendors. For sure I've never heard anybody part of the HTML WG say something like this. Quite on the contrary!" |
| 07:17 | <hsivonen> | zcorpan: I'm preparing validator.nu for a system update and reboot. I'll deal with microsyntaxes when it reboots. html5.validator.nu will follow if everything goes well. |
| 07:19 | <MikeSmith> | hsivonen: there are a couple of other minor fixes I checked in since the last v.nu deployment. Just schema tweaks, and tested |
| 07:21 | <hsivonen> | is the Kronos stuff something anyone can GET without cookies or HTTP auth? |
| 07:22 | hsivonen | is annoyed at a WG Note from the WAI being treated as normative over REC track process |
| 07:26 | <MikeSmith> | it's kind of surprising to know it is just a Note |
| 07:26 | <hsivonen> | annevk2: as far as I can tell, the XML Core WG didn't agree with i18n but added a well-formedness-neutral document conformance req that has no effect on processor behavior |
| 07:29 | <hsivonen> | someone should make a scientific analysis of who actually practices the handling-people bible on public-html |
| 07:34 | <annevk2> | hsivonen, I mean that i18n agreed with the compromise |
| 07:34 | <annevk2> | meant* |
| 07:36 | <hsivonen> | annevk2: ok |
| 07:37 | <hsivonen> | MikeSmith: OK. I'll do a full redeployment after reboot |
| 07:37 | <MikeSmith> | hsivonen: while you're at it, quick question - |
| 07:38 | <MikeSmith> | http://bugzilla.validator.nu/show_bug.cgi?id=631 |
| 07:38 | <MikeSmith> | the current datatype-checking code doesn't make sure that the value of the iframe name attribute is more than one character |
| 07:38 | <hsivonen> | (I'll wait until DNS updates before I start the updatie process) |
| 07:38 | <MikeSmith> | hsivonen: but the spec says, "A valid browsing context name is any string with at least one |
| 07:38 | <MikeSmith> | character that does not start with a U+005F LOW LINE character." |
| 07:39 | <hsivonen> | MikeSmith: looks like a genuine bug |
| 07:39 | <MikeSmith> | OK, I'll make a patch for it |
| 07:39 | <hsivonen> | in the datatype lib that is |
| 07:39 | <MikeSmith> | yeah, i'll take a look |
| 07:40 | <MikeSmith> | one other open one that came up recently was this: |
| 07:40 | <MikeSmith> | Bug#: 591 |
| 07:40 | <MikeSmith> | Product: |
| 07:40 | <MikeSmith> | Component: |
| 07:40 | <MikeSmith> | Status: NEW |
| 07:40 | <MikeSmith> | Resolution: |
| 07:40 | <MikeSmith> | Assigned To: Henri Sivonen <hsivonen⊙if> |
| 07:40 | <MikeSmith> | Hardware: |
| 07:40 | <MikeSmith> | oops |
| 07:40 | <MikeSmith> | http://bugzilla.validator.nu/show_bug.cgi?id=591 |
| 07:41 | <annevk2> | IETF process is apparently not very shiny: http://www.0xdeadbeef.com/weblog/?p=1346&cpage=1#comment-206483 |
| 07:41 | <MikeSmith> | hsivonen: need to also report an error if an XML document that has a <meta charset='utf-8'/> is not actually encoded in UTF-8. |
| 07:42 | <hsivonen> | MikeSmith: I think what's needed is an extension to the ValidationContext API (a ValidationContext2 interface on the same object) that allows datatypes acquire the Locator object |
| 07:42 | <hsivonen> | then the Locator object should implement Locator2 which, IIRC, exposes the encoding |
| 07:43 | <hsivonen> | and then the parsers need to implement Locator2 |
| 07:43 | <hsivonen> | and then datatyping that is sensitive to the encoding becomes possible |
| 07:43 | <hsivonen> | there's an open Jing bug about this |
| 07:43 | <MikeSmith> | ah, OK |
| 07:44 | <hsivonen> | Foo2 interfaces suck, but they are the most compatible way of doing this without breaking the use of the datatype lib with old versions of Jing and MSV |
| 07:44 | <MikeSmith> | I see |
| 07:44 | <hsivonen> | Foo2 feels COMtaminated |
| 07:45 | <MikeSmith> | heh |
| 07:47 | <MikeSmith> | hsivonen: so for now, maybe one thing worth considering is to add an Info message saying, "<meta charset="utf-8"/> found in XML document. Make sure your document is actually encoded in UTF-8." |
| 07:47 | <MikeSmith> | or a Warning |
| 07:47 | <MikeSmith> | can't add it in the assertions-checking code because of the problem of schematron only being able to emit errors |
| 07:48 | <MikeSmith> | not if you want to keep the Java assertions code in sync with the assertions.sch file |
| 07:48 | <MikeSmith> | hsivonen: hmm, maybe it's something I could add to the proposed warnings.sch |
| 07:48 | <MikeSmith> | oh |
| 07:49 | <MikeSmith> | can't do that either because I have no way of knowing at the point it the document is XML or text/html |
| 07:49 | <MikeSmith> | 困った |
| 07:51 | <hsivonen> | I guess I could simply implement what I said above on the V.nu branch of Jing and not wait for the trunk |
| 07:53 | <MikeSmith> | hsivonen: btw, I looked at the build stuff in the jing trunk for generating the gcj binary |
| 07:53 | <MikeSmith> | I don't know who came up with that stuff, but it seems way overcomplicated |
| 07:54 | <hsivonen> | MikeSmith: the Java build process for Jing trunk is complicated, too |
| 07:54 | <MikeSmith> | well, the gcj thing has the additional disadvantage of being broken |
| 07:55 | <hsivonen> | woohoo! Locator2 is in JDK 5 |
| 07:56 | <MikeSmith> | well, that's a bright spot, at least |
| 07:56 | <MikeSmith> | hsivonen: anyway, that gcj build uses hard-coded path names all over the place, and the paths it's pointing to seem to have moved since the time the build setup was written |
| 07:56 | <MikeSmith> | the gcj/dist file |
| 07:56 | <MikeSmith> | It's just a shell script |
| 07:57 | <MikeSmith> | I don't know if that's some kind of standard RPM thing or what, but it's just absolutely arcane |
| 07:57 | <MikeSmith> | in contrast to the build setup for the debian trang package |
| 07:58 | <MikeSmith> | which just uses a simple Makefile |
| 07:58 | <MikeSmith> | hsivonen: so, anyway, what I'm getting at is, I'm going to try to make a non-broken gcj build for jing |
| 07:59 | <MikeSmith> | and if I can get it to work, I'll send a patch and if it looks OK, maybe you can talk with James and George about committing it to the trunk in place of that thing that's there now |
| 08:01 | <hsivonen> | seems like a good plan |
| 08:03 | <MikeSmith> | OK |
| 08:12 | <MikeSmith> | hsivonen: http://bugzilla.validator.nu/attachment.cgi?id=114 |
| 08:12 | <annevk2> | reading the June thread again on summary (the "PF Response") it strikes me that the PFWG never even emailed their comment to the HTML WG mailing list |
| 08:15 | <hsivonen> | MikeSmith: looks good, although I'd do s/character/character long/ in the message |
| 08:15 | <annevk2> | and there's stuff like http://lists.w3.org/Archives/Public/www-archive/2009Jul/0055.html o_O |
| 08:16 | <annevk2> | was there ever an official reply to the points Hixie made? |
| 08:32 | <hsivonen> | MikeSmith: I checked in the Locator2 stuff to the validator-nu branch of jing-trang and to the XML parser |
| 08:32 | <hsivonen> | I didn't add support to the HTML parser yet |
| 08:33 | <MikeSmith> | hsivonen: excellent |
| 08:34 | <MikeSmith> | I checked in the browsing-context-name fix, with your wording change |
| 08:34 | <hsivonen> | MikeSmith: thanks |
| 08:36 | <hsivonen> | MikeSmith: I just made isContextDependent on AbstractDatatype non-final, too |
| 08:37 | <MikeSmith> | hsivonen: I guess I should check out the validator-nu branch instead of the trunk |
| 08:44 | <MikeSmith> | hsivonen: so what's the effect of the isContextDependent change? |
| 08:44 | <MikeSmith> | I don't actually know what isContextDependent is for |
| 08:46 | <MikeSmith> | ah |
| 08:46 | <MikeSmith> | org.relaxng.datatype thing |
| 09:23 | <gsnedders|work> | Hixie: http://www.w3.org/mid/Pine.LNX.4.62.0908040517440.6420⊙hdc is your second reply to the same email. |
| 09:24 | <MikeSmith> | annevk2: not sure if it's worth noting in the diff doc, but the restriction in HTML5 that the target attribute and iframe name attribute must be at least one character is a change vs HTML4 |
| 09:24 | <MikeSmith> | in HTML4, it's just CDATA, with no further restrictions stated in the spec |
| 09:25 | <annevk2> | we made a lot of those things more strict |
| 09:26 | <annevk2> | (and some more loose, e.g. id) |
| 09:27 | <annevk2> | e.g. accept-charset was CDATA in HTML4 |
| 09:27 | <annevk2> | should I note it is now actually usefully checked? |
| 09:27 | <annevk2> | it seemed too much effort |
| 09:31 | <Dashiva> | Would anyone specify an empty name for any reason other than spec lawyering? |
| 09:31 | <hsivonen> | one would think someone in California had figured out how to protect disk against minor earthquakes |
| 09:31 | <Dashiva> | hsivonen: Multiple redundant data centers :P |
| 09:32 | <hsivonen> | Dashiva: I have deduntant disks, but these artificial earthquakes affect both disks at the same time |
| 09:32 | <hsivonen> | getting multilocation personal data centers is expensive |
| 09:32 | <Dashiva> | I don't recall the exact value, but it's something like at least 300 miles apart from each other |
| 09:34 | <hsivonen> | I can't think of any home-grown solutions that both absorb shocks properly and don't have extra vibrations afterwards |
| 09:36 | <hsivonen> | https://twitter.com/roessler/statuses/3120872586 |
| 09:37 | <annevk2> | that quote never gets old, does it? :) |
| 09:40 | <Dashiva> | There was an ALA article about "unwebbable" content a while ago. One of the examples [in a comment] was poetry, which "In fact, this is text that deliberately defies conventions and structure." |
| 09:41 | <Dashiva> | Doesn't that imply adding syntax for poetry is by definition an impossible task? |
| 09:41 | <Philip`> | You could use unconventional, unstructured syntax |
| 09:42 | <Philip`> | Or you could say your syntax defies defiance of conventions and structure |
| 09:42 | <Dashiva> | But when you put it into a spec, you're a sellout |
| 09:42 | <Dashiva> | Nobody will want your syntax anymore |
| 09:43 | <hsivonen> | is the original text of http://html5.digitalbazaar.com/a-new-way-forward/ archived somewhere? |
| 09:45 | <gsnedders|work> | hsivonen: I have a copy on my HD |
| 09:45 | <hsivonen> | the "About the Author" bit is interesting |
| 09:48 | <MikeSmith> | annevk2: it's probably not worth the effort to list them individually, but maybe adding a general statement that "a number of attributes values now how further restrictions; for example, some values that were allowed to be empty in HTML4 are not allowed to be empty in HTML5, ..." |
| 09:52 | <zcorpan> | i wonder if manu will get financial support from companies that depend on the internet |
| 09:53 | <hsivonen> | hmm. "Problem: No Way for Experts to Contribute in a Meaningful Way" |
| 09:54 | <annevk2> | interesting, the intent of accesskey in HTML5 was to take a single character |
| 09:54 | <annevk2> | it takes CDATA instead though to sucky DTDs |
| 09:55 | <jgraham> | I think that when you are making a serious document posting links to last week and (less so) the CSS Squirrel does not advance your credibility |
| 09:55 | <jgraham> | s/advance/help/ |
| 09:55 | <hsivonen> | it seems to me that the problems manu mentions are either non-problems in my opinion or problems that already have solutions |
| 09:57 | <MikeSmith> | annevk2: I'm confused by " the intent of accesskey in HTML5 was to take a single character" |
| 09:58 | <Philip`> | s/5/4/? |
| 09:58 | <annevk2> | yeah, 4 |
| 09:59 | <annevk2> | and also s/though/due/ |
| 09:59 | <jgraham> | hsivonen: Without reading the documet in detail it seems to me to be a curious mix of process issues and specific technical issues that Manu feels strongly about |
| 10:08 | <hsivonen> | jgraham: there's also a certain 'the enemy of my enemy is my friend' flavor to it (given certain assumed enmities) |
| 10:30 | <jgraham> | http://www.goer.org/Journal/2009/08/the_zen_of_structured_data.html |
| 10:32 | <hsivonen> | I wonder if I should get into the captioning threads and explain how things at Mozilla turned out the way they did |
| 10:33 | annevk42 | would like info on captioning |
| 10:48 | <hsivonen> | Hixie: request for stats about equals sign in unquoted attribute values noted |
| 11:09 | <Dashiva> | last week gets a small plus for "pigs headednessess" |
| 11:16 | <Philip`> | +0.1, perhaps? |
| 11:18 | <MikeSmith> | hsivonen: so I realize now that v.nu schema and datatype library have have both a "positive integer" and "non-negative integer" datatype |
| 11:18 | <MikeSmith> | but the spec has only "non-negative integer" |
| 11:19 | <Philip`> | Does it have something like "non-zero non-negative integer"? |
| 11:19 | <Philip`> | (which is basically positive, for people who aren't sure what positive means) |
| 11:20 | <Dashiva> | Yeah, I remember that being wrangled out :) |
| 11:21 | <hsivonen> | MikeSmith: IIRC, the spec has places where a non-negative integer isn't allowed to be zzero |
| 11:21 | <MikeSmith> | ah, OK |
| 11:21 | <MikeSmith> | I'll look more closely |
| 11:21 | <hsivonen> | IIRC, Hixie doesn't trust readers to understand what a positive integer is |
| 11:22 | <MikeSmith> | heh |
| 11:22 | <Dashiva> | "Roman era integer" |
| 11:23 | <gsnedders|work> | Dashiva: Can you mark "Roman era" up with a time element so I know what you mean? |
| 11:23 | <jgraham> | We should call them "optimistic integers" |
| 11:24 | <Dashiva> | gsnedders|work: It's before 1592 (or whenever it was) |
| 11:24 | <Philip`> | Do what proper mathematicians do, and use ℕ instead of fuzzy words like "integer", so there's no ambiguity |
| 11:24 | <annevk2> | MikeSmith, DOM attributes can be limited to only positive non-zero numbers :) |
| 11:24 | <MikeSmith> | annevk2: yeah, I saw that in the spec, but in this case was just referring to restrictions on content attributes. |
| 11:24 | <annevk2> | conformance on attributes seems to be phrased like "valid non-negative integer greater than zero" |
| 11:25 | <MikeSmith> | yeah |
| 11:25 | <Dashiva> | Makes me wonder why we bother with the non-negative part |
| 11:25 | <zcorpan> | Dashiva: use the comment box in the spec |
| 11:25 | <gsnedders|work> | Because you might have a negative integer greater than zero. |
| 11:25 | <MikeSmith> | annevk2: in the v.nu schema, it's actually a different datatype. which makes sense |
| 11:26 | <zcorpan> | -1 > 0 |
| 11:26 | <gsnedders|work> | A really-complex number. |
| 11:26 | <Dashiva> | zcorpan: The barrier of entry is too high (I'd have to load the spec) |
| 11:26 | <jgraham> | zero is an integer, just not a positive integer |
| 11:26 | <gsnedders|work> | (Because complex numbers aren't enough) |
| 11:26 | <zcorpan> | Dashiva: load the multipage version |
| 11:26 | <jgraham> | zcorpan: You getting t-shirts made? |
| 11:26 | <gsnedders|work> | zcorpan: That means starting a web browser. |
| 11:26 | <Dashiva> | zcorpan: It still requires me to switch to a browser |
| 11:27 | <zcorpan> | Dashiva: use opera for irc, then you don't need to switch to another app |
| 11:27 | <Dashiva> | How about... no |
| 11:28 | <Philip`> | gsnedders|work: (Actually they are enough for most purposes, which is why we still use complex numbers instead of anything fancier) |
| 11:28 | <gsnedders|work> | zcorpan: But I'd still have to change tab! |
| 11:28 | <jgraham> | Dashiva: use curl |
| 11:28 | <Dashiva> | I'm using IRC |
| 11:28 | gsnedders|work | notes he is somewhat amused that three of the people in this conversation are in one room |
| 11:28 | <zcorpan> | maybe we should have an irc bot for filing spec comments |
| 11:28 | <Dashiva> | One of you young'uns with this newfangled 'superhighway' stuff can submit for me |
| 11:28 | Dashiva | waves cane |
| 11:29 | <annevk2> | MikeSmith, yeah, maybe file a bug? |
| 11:29 | <Philip`> | Dashiva: I could paste the spec in a /msg to you |
| 11:29 | <Philip`> | gsnedders|work: I'm in one room too |
| 11:29 | <Dashiva> | Philip`: How do I request external resources in /msg? |
| 11:29 | <zcorpan> | Hixie: please write an irc bot for filing spec comments so that people don't have to load the spec to file comments |
| 11:29 | <annevk2> | MikeSmith, I'd agree that having all the _actual_ datatypes together makes more sense than just having a subset and another set defined in prose |
| 11:29 | <gsnedders|work> | Dashiva: You pipe it, from curl. |
| 11:29 | <Philip`> | Dashiva: You don't, you just ask the person you're messaging to paste it to you |
| 11:30 | <jgraham> | hhttp |
| 11:30 | <Dashiva> | Besides, I don't like parsing javascript in my head unless I really have to |
| 11:30 | <zcorpan> | Dashiva: you mean you like it when you really have to? |
| 11:30 | <Dashiva> | Yeah |
| 11:31 | <Dashiva> | It makes me feel useful and accomplished |
| 11:31 | hsivonen | reaches the thread about making unquoted attributes non-conforming |
| 11:31 | <Dashiva> | First year I interned at Opera we got a tiny meeting room for our office since there were no free offices |
| 11:32 | <Dashiva> | One day we come in and find a<!--b&&c-->d on the whiteboard and some scribbled comments about "Does this work?" |
| 11:32 | <Dashiva> | Tarquin had decided to outsource :) |
| 11:33 | <gsnedders|work> | Dashiva: Serves you right for interning at Opera! |
| 11:33 | <gsnedders|work> | Oh, wait… |
| 11:33 | <Dashiva> | It's actually valid JS. Pretty useless, but valid! |
| 11:34 | <jgraham> | Yeah it is |
| 11:34 | <jgraham> | Also it won't be treated as a comment |
| 11:35 | <jgraham> | but <!--b&&c-->d will be |
| 11:36 | <jgraham> | Oh wait, it will be treated as a |
| 11:36 | <jgraham> | as zcorpan just pointed out to me |
| 11:36 | <Dashiva> | Yah |
| 11:36 | <jgraham> | (I was confusing the behaviour of <!-- and -->( |
| 11:36 | <Dashiva> | Silly old bear |
| 11:36 | <jgraham> | s/(/)/ |
| 11:37 | <gsnedders|work> | But he's such a lovable old bear, with his long greying hair! |
| 11:37 | <Dashiva> | You need to escape your (, jgraham |
| 11:37 | <jgraham> | Dashiva: Not in magic IRC regexp syntax I don't |
| 11:37 | <zcorpan> | ")I was confusing the behaviour of <!-- and -->("? |
| 11:38 | <Dashiva> | zcorpan: Magical IRC regexp matches from the end |
| 11:38 | <jgraham> | Magical irc regexp matches whatever makes the most sense |
| 11:38 | <gsnedders|work> | No, Magical IRC regexp does exactly what you intend it to. |
| 11:38 | <Dashiva> | Not even magical IRC regexp is that powerful |
| 11:39 | <jgraham> | gsnedders|work: Not quite otherwise s/// would work for anything |
| 11:39 | <zcorpan> | jgraham: see topic |
| 11:39 | <gsnedders|work> | That doesn't work, because it throws E_EMPTY |
| 11:39 | <gsnedders|work> | (for the first argument) |
| 11:48 | <Philip`> | gsnedders|work: Not in any proper regexp implementation, i.e. Perl's |
| 11:48 | <gsnedders|work> | You mean there are other proper regexp implementations? |
| 11:48 | <Dashiva> | Javascript's! |
| 11:51 | <jgraham> | Javascript does indeed allow empty regexps |
| 11:54 | <Dashiva> | (But not as a literal, since it's treated as a comment) |
| 11:57 | <Philip`> | It just matches all the distinct substrings with length zero, which seems like a perfectly natural and sensible thing to do |
| 11:57 | <jgraham> | (Unless you express it as /(?:)/ which is semantically an empty regexp if not syntactically) |
| 12:21 | gsnedders|work | finds, "I will buy lunch for whomever fixes floats so that they work properly." — Hixie, Bugzilla, 2001-04-29 |
| 12:22 | <gsnedders|work> | (Status: NEW) |
| 12:30 | <MikeSmith> | hsivonen: thinking about the microsyntaxes scraping.. the fact that getting those messages current depends stopping and restarting v.nu seems to really cry out for actual API to grab those in real time, and then some kind of simple DB instead |
| 12:31 | <MikeSmith> | the technical limitations of wikis seem like a big deficiency in this case |
| 12:32 | <MikeSmith> | or maybe there's some wiki that's capable of storing content as names and values in a db |
| 12:32 | <MikeSmith> | while still providing a wiki interface |
| 12:32 | <MikeSmith> | what happened to Google Base? |
| 12:32 | <MikeSmith> | or maybe there are other things like that around now |
| 12:35 | <MikeSmith> | or DabbleDB |
| 12:37 | <annevk2> | I believe that if someone writes the software we'll take it |
| 12:37 | <annevk2> | but note that having it on the wiki has advantages; e.g. spam control, accounts are sorted out, etc. |
| 12:39 | <MikeSmith> | annevk2: I understand all that. But I'd really think there'd be an simple online DB with those same features and almost the same ease of use as a wiki, but with a real rest API for retrieving data from it. |
| 12:40 | <MikeSmith> | I guess to me the lack of the network API is the real issue |
| 12:41 | <MikeSmith> | I guess it would be possible to have some kind of thing that scrapes the wiki in real time and exposes an API |
| 12:41 | <annevk2> | that's an interesting idea |
| 12:41 | <MikeSmith> | but that would really seem suboptimal to me at least |
| 12:41 | <MikeSmith> | though I guess in this case it's a very small amount of data and the scraping would be very quick |
| 12:41 | <annevk2> | maybe, but wiki's have solved a lot of problems |
| 12:42 | <MikeSmith> | no argument from me there |
| 12:42 | <annevk2> | accounts, accountability, easy editing, etc. |
| 12:42 | <annevk2> | cronjob that scrapes and exposes a friendly API sounds good to me |
| 12:42 | <annevk2> | (you'd cache the scraped data) |
| 12:45 | <MikeSmith> | annevk2: yeah, I guess I'll write one unless somebody else beats me to it |
| 12:48 | <jgraham> | Surely putting a wiki-like interface on a key-value document store would not be hard |
| 12:49 | <jgraham> | In fact something like the admin interface to couch-db is already quite close to what you want I think |
| 12:50 | <MikeSmith> | I guess dabbleDB and Google Base may be too |
| 12:50 | <MikeSmith> | the other great thing about wikis is having the change history available, and the record of accountability (as annevk2 pointed out) |
| 12:51 | <MikeSmith> | so whatever other mechanism should provide that at least |
| 12:52 | <gsnedders|work> | git! hg! |
| 12:54 | <MikeSmith> | the tools will (en)s(l)ave use |
| 12:54 | <MikeSmith> | *us |
| 12:54 | <Philip`> | Surely you could make a user-editable database with two lines of Ruby/Rails code? |
| 12:55 | <Philip`> | (or Python/Django or whatever) |
| 12:56 | <MikeSmith> | Philip`: Surely somebody else must have done that already for me so that I don't have to. |
| 12:56 | <jgraham> | Which two lines? |
| 12:56 | <gsnedders|work> | jgraham: import and print |
| 12:56 | <jgraham> | those are statements, not lines |
| 12:56 | <MikeSmith> | anyway, it's not just having the DB, it's having all the Wiki UI stuff on top of it |
| 12:56 | <MikeSmith> | change history, etc. |
| 12:57 | <MikeSmith> | and Ogg Theora videos. |
| 12:58 | <jgraham> | Change history seems like the big one |
| 12:58 | <Philip`> | jgraham: The ones in the tutorial screencasts where you make a user-editable database with two lines of code using the user-editable database scaffolding which conveniently happens to have been written already for precisely this purpose, I guess |
| 12:58 | <jgraham> | Because I guess most k-v stores are designed to be efficient rather than to keep records |
| 13:00 | <hsivonen> | MikeSmith: I'd rather deal with rel registry before optimizing microsyntax stuff |
| 13:00 | hsivonen | is still waiting for DNS for validator.nu to update |
| 13:00 | <hsivonen> | I guess I should defer the system update until tomorrow when DNS must have updated |
| 13:03 | <MikeSmith> | jgraham: you know, for me at least, it really is all about the API. I'd think whatever Wiki could easily have some kind of standard markup for indicating a string on the page that's a key, and whatever chunk of text is its associated value |
| 13:03 | <MikeSmith> | and then the Wiki's own backend could expose an API for that |
| 13:04 | <gsnedders|work> | We should use RDF for that. |
| 13:04 | <MikeSmith> | hmm, maybe I'll shut up for now |
| 13:04 | <jgraham> | MikeSmith: The reason I was thinking about couchdb is that the API is really the main thing |
| 13:04 | <MikeSmith> | what's the backend on the whatwg wiki? |
| 13:04 | <MikeSmith> | jgraham: OK |
| 13:04 | <jgraham> | The store just serves json |
| 13:05 | <MikeSmith> | I see |
| 13:05 | <jgraham> | The problem is I don't know how easy it is to add versioning to that |
| 13:05 | <hsivonen> | for the rel registry, it could be a GETtable text/plain resource with one registered token on each line and LF line breaks |
| 13:05 | <jgraham> | (well I guess it is rather easy to do it manually by adding a version key to each document) |
| 13:06 | <hsivonen> | jgraham: what kind of versioning? |
| 13:06 | <hsivonen> | I was thinking of an append-only registry |
| 13:06 | <jgraham> | hsivonen: MikeSmith seems to want wiki-like history which imp[lies storing multiple versions (or diffs or something) |
| 13:12 | <virtuelv> | WTF!? https://twitter.com/WHATWG/status/3122997127 |
| 13:13 | <jgraham> | virtuelv: http://www.whatwg.org/#updater |
| 13:14 | <jgraham> | (look down the page) |
| 13:14 | <virtuelv> | ah |
| 13:18 | <annevk2> | I wonder what happens if I say that everyone can update the twitter account from http://www.whatwg.org/#updater from @whatwg |
| 13:19 | <annevk2> | Guess it's better to wait for Hixie to be around since I think he's the only one with the login details |
| 13:21 | <annevk2> | oh well |
| 13:28 | <hsivonen> | I love how Kyle Weems complains about attacking people (in http://www.cssquirrel.com/2009/07/09/comic-update-manners-after-the-xhtmlacolypse/) and then goes on to publish a comic like http://www.cssquirrel.com/comic/?comic=30 |
| 13:31 | <MikeSmith> | hsivonen: that doesn't really seem like a personal attack |
| 13:31 | <MikeSmith> | hsivonen: but I guess the vicious online comic is the equivalent of a nasty e-mail comment that ends with a smiley :) |
| 13:32 | <MikeSmith> | I personally just wish that Kyle would into his comics put more depictions of real violence, with some nudity thrown in now and then |
| 13:33 | <MikeSmith> | or even some sex. I think it was Einstein who said "there are few things in the world that are funnier than depictions of cartoon characters having sex." |
| 13:37 | <hsivonen> | there were already references to "snogging" |
| 13:42 | <gsnedders|work> | OK, where's the drugs and rock and roll? |
| 13:44 | <jgraham> | Couchdb already has revision tracking it seems |
| 13:44 | <Creap> | hi. a pure decorational image or illustration would not fit to be marked up within a figure, right? |
| 13:46 | <MikeSmith> | kyle gets points for the violence in that latest one, at least |
| 13:46 | <MikeSmith> | hsivonen: hmm, there no longer seems to be "percent" datatype referenced by the schema |
| 13:47 | MikeSmith | investigates |
| 13:47 | <MikeSmith> | hmm, I guess that's been gone for a long while |
| 13:47 | <hsivonen> | MikeSmith: I can't remember what it was used for |
| 13:48 | MikeSmith | does Shift-v jjjd |
| 13:48 | <MikeSmith> | Creap: yeah |
| 13:49 | <MikeSmith> | jgraham: so something based on CouchDB would seem like a good way to go. |
| 13:49 | <Creap> | 'yeah' as in.. figure is not for decorational images? |
| 13:49 | <MikeSmith> | we could build something to meet our own needs, and then productize it |
| 13:49 | <MikeSmith> | Creap: right, not for decoration |
| 13:50 | <MikeSmith> | figure, in my estimation, is for something that you or somebody else might possibly want to reference by its title |
| 13:50 | <MikeSmith> | see Figure 1.2 at <url> |
| 13:52 | <MikeSmith> | jgraham: time to turn a profit off the work on tools associated with html5 |
| 13:55 | <hsivonen> | oh. now I'm no longer the only one allegedly having personal preferences against RDFa |
| 13:56 | <jgraham> | hsivonen: Pointer? |
| 13:58 | <hsivonen> | jgraham: http://www.w3.org/mid/4A72916B.5020407⊙dc |
| 14:30 | <Philip`> | jgraham: I think I read somewhere that CouchDB happily throws away old revisions when you're compacting the database, which doesn't sound good when you want a reliable version control system |
| 14:34 | <jgraham> | Philip`: Ah. Well I guess you would have to build a seperate version control layer then :( |
| 14:36 | <jgraham> | (or you could just use something totally different like a hg backend that stored json files assuming that querying wasn't a priority) |
| 14:44 | <pablof> | there doesn't seem to be that much interest for that other from toy wikis apparently |
| 14:44 | <pablof> | http://issues.apache.org/jira/browse/COUCHDB-69 |
| 14:44 | <pablof> | *other than |
| 15:05 | <nate_> | hey guys, just out of curiosity, what's with the open-access Twitter account? |
| 15:05 | <gsnedders|work> | IIRC, someone thought it was a good idea. |
| 15:06 | <nate_> | hm |
| 15:06 | <gsnedders|work> | And in general it has been |
| 15:06 | <nate_> | well, it looks like you're getting a lot of off-color comments |
| 15:06 | <nate_> | or just generally dumb stuff like "HXTML" |
| 15:06 | <MikeSmith> | nate_: we are sending those, not getting them |
| 15:07 | <MikeSmith> | you can join the fun too |
| 15:07 | <nate_> | oh, hah |
| 15:07 | <nate_> | well, okay |
| 15:07 | <nate_> | I guess it wasn't clear to a lot of people (myself included :-P) that it was a joke |
| 15:07 | <MikeSmith> | nate_: seriously, the only person with access to turn it off is asleep at the moment |
| 15:07 | <gsnedders|work> | nate_: Also, be warned: don't take everything in here seriously. |
| 15:07 | <MikeSmith> | nate_: my joking about it is the joke |
| 15:07 | <gsnedders|work> | nate_: Especially not that asshole who goes by the name of Mike |
| 15:08 | <miketaylr> | i quite enjoyed this tweet: The HXTML 2.0 spec has been finalized with only one tag which is <text>. |
| 15:08 | <MikeSmith> | nate_: that potential for it being abused is of course not a joke |
| 15:08 | <nate_> | gsnedders|work: IMO, anyone with a name like "Mike" cannot be trusted |
| 15:08 | <gsnedders|work> | Nah, we need to pave the cowpaths and go for <plaintext> |
| 15:08 | <MikeSmith> | nate_: or at least not a funny joke |
| 15:08 | <gsnedders|work> | That's nice and simple to parse. |
| 15:08 | <miketaylr> | lol |
| 15:08 | <MikeSmith> | or at least not an intentionally funny joke |
| 15:09 | <gsnedders|work> | (Everything from <plaintext> to EOF is just a single text node) |
| 15:09 | <gsnedders|work> | (It makes parsing simple.) |
| 15:09 | <MikeSmith> | it is kind of an unintentionally funny joke |
| 15:09 | <gsnedders|work> | (It was, of course, marked obsolete in the first ever draft spec for HTML, and still exists in HTML 5) |
| 15:09 | <gsnedders|work> | (In the parsing rules, at least) |
| 15:10 | <gsnedders|work> | I mean, sometimes I just get the urge, and just have to tweet about my cat. |
| 15:10 | <gsnedders|work> | She's called BetterAstrophy. |
| 15:10 | <annevk2> | <plaintext> was the solution to media types before you could do Content-Type:text/plain |
| 15:10 | <annevk2> | in retrospect I think it would have been better |
| 15:17 | <beowulf> | i go on holiday for a month and come back to find even longer threads on @summary and twitter running amok |
| 15:17 | <beowulf> | is there no end to the depths you all will sink! |
| 15:18 | <nate_> | ...no? |
| 15:18 | <Lachy> | http://twitter.com/WHATWG/status/3124785323 :-D |
| 15:18 | <nate_> | I dunno, it seems like a fun experiment |
| 15:18 | <nate_> | how deep *can* we sink?? |
| 15:19 | <nate_> | woohoo! |
| 15:21 | jgraham | wonders how many of the people abusing the status updater are libertarians |
| 15:22 | <gsnedders|work> | I was under the impression, from what I read on the internet, that everything wrong is the fault of the liberals. |
| 15:23 | <jgraham> | Isn't there a difference between a liberal and a libertarian? |
| 15:24 | <gsnedders|work> | Yes, there is. |
| 15:24 | <gsnedders|work> | What I read on the internet was it was the liberals fault, and that liberterians are what makes this country so great. |
| 15:25 | <beowulf> | i abused the twitter status, i do not feature on the liberal side of the political compass though |
| 15:25 | <beowulf> | i say abused, i think it's important that people buy viagra |
| 15:26 | <gsnedders|work> | Are we going to have to get "I abused @WHATWG" t-shirts |
| 15:26 | <miketaylr> | I'll take a size M |
| 15:28 | <Lachy> | beowulf, did you really tweet about viagra on @whatwg? |
| 15:28 | <gsnedders|work> | Someone did. |
| 15:28 | gsnedders|work | glares |
| 15:28 | <miketaylr> | the twitter status is about to hit critical mass a la reply-all 'stop hitting reply-all' |
| 15:29 | <svl> | The follower count is going into freefall |
| 15:30 | <Lachy> | does anyone know why there is a sudden increase in the spam levels on it? |
| 15:30 | <svl> | It'll have been posted on some message board somewhere. |
| 15:31 | <gsnedders|work> | Lachy: Because once people started spamming it it will have become self-evident it was open too |
| 15:31 | <beowulf> | Lachy: annevk2 mentioned on here that it's open to anyone, which appears in the logs |
| 15:32 | <gsnedders|work> | beowulf: It's been mentioned many times before |
| 15:32 | <beowulf> | gsnedders|work: well then i've no idea why it took off today |
| 15:33 | <gsnedders|work> | I think the correct solution is that Hixie should be awake 24/7 to police it. |
| 15:33 | <beowulf> | http://krijnhoetmer.nl/irc-logs/whatwg/20090804#l-691 |
| 15:33 | <Lachy> | yeah, the number of followers has dropped by a few hundred already |
| 15:33 | <Lachy> | a few minutes ago, it was over 1600. Now it's 1288 |
| 15:35 | <jgraham> | I love the way that people think it was an oversight rather than a decision |
| 15:36 | <gsnedders|work> | We probably should have made the barrier to entry a bit higher, because it was almost inevitable this would happen sometime |
| 15:37 | <alexmace> | #fail |
| 15:37 | svl | loves that the bandaid 'fix' actually works. |
| 15:38 | <jim_c> | gsnedders|work: What's the story with the Twitter account? |
| 15:38 | <gsnedders|work> | svl: The "ok stop it"? |
| 15:38 | gsnedders|work | wonders why people assume he knows |
| 15:39 | <gsnedders|work> | jim_c: A while ago there was a decision made to make the Twitter account open, knowing it could potentially be used for spam, but thinking being open would all-in-all be more useful |
| 15:39 | <Lachy> | svl, what's the bandaid fix? |
| 15:39 | <jim_c> | gsnedders|work: I see. That assumes that people who use the internet act like adults. Unfortunately, that's not how it is. |
| 15:39 | Philip` | wonders how long Hixie will keep believing that anonymous free-for-all editing is a good idea |
| 15:39 | <gsnedders|work> | I don't think anyone expected to get a large number of spam tweets all at once, just once or twice occasionally |
| 15:40 | <svl> | Ah, nothing. I missed the difference between a link to #updater and triggering that manually. |
| 15:40 | <svl> | Thought with six minutes of silence that a change had been made to the homepage to make this just a tiny bit harder |
| 15:41 | <Lachy> | my tweet didn't get posted :-( |
| 15:41 | <gsnedders|work> | Lachy: Did you check the checkbox? :P |
| 15:41 | <Lachy> | I don't see a checkbox |
| 15:41 | <gsnedders|work> | There's a checkbox that says "I have something useful to say" you have to check |
| 15:42 | <alexmace> | gsnedders|work: Not anymore there isn't |
| 15:42 | <Lachy> | If I disable styles, it shows up |
| 15:43 | <gsnedders|work> | Reloading there still is |
| 15:43 | <alexmace> | Seems naive at best to have an open form that updates a twitter account |
| 15:43 | <gsnedders|work> | Oh, only from http://www.whatwg.org/#updater |
| 15:43 | <gsnedders|work> | So it's only required if you have JS disabled |
| 15:44 | <gavin_> | alexmace: obscurity was a suitable countermeasure, until now |
| 15:45 | <alexmace> | gavin_: That isn't a countermeasure |
| 15:45 | <gavin_> | there was no problem in practice |
| 15:45 | <beowulf> | no kittens were harmed |
| 15:46 | <Lachy> | might be useful to have some kind of moderation on the system |
| 15:46 | <alexmace> | Hey, stay in denial if you lie |
| 15:46 | <alexmace> | like |
| 15:46 | <gavin_> | I'm not in denial... |
| 15:46 | <gavin_> | there is a problem now, sure |
| 15:46 | jgraham | decides not to point out that at some level obscurity is the root of all security |
| 15:46 | <Philip`> | alexmace: It's only naive if you consider a short-lived period of spam on a Twitter account to be a real problem |
| 15:46 | <alexmace> | no exposure of a problem != no problem |
| 15:46 | <gavin_> | your definition of problem is strange |
| 15:46 | <gsnedders|work> | alexmace: Being _open_ was a feature. It wasn't a problem. |
| 15:47 | <alexmace> | Being open isn't a problem |
| 15:47 | <gavin_> | is it a problem that a piano might fall on my head today? |
| 15:47 | <alexmace> | being careless is |
| 15:47 | <gsnedders|work> | OK, so it got abused one or twice a month… That was considered acceptable for being open. |
| 15:47 | <alexmace> | Anyway, it's disabled somewhere. Even with the box checked it isn't updating anymore |
| 15:48 | <gsnedders|work> | alexmace: It's not careless. I think everyone around when it was decided for it to be open knew the risk. |
| 15:48 | <Philip`> | Does Twitter have some limit on post frequency/quantity? |
| 15:48 | <gsnedders|work> | Philip`: Through the API yes. |
| 15:48 | <gavin_> | Philip`: I don't think you can reasonably argue that it isn't a problem :) |
| 15:48 | <gavin_> | it may not be a problem worth "fixing"... |
| 15:49 | <gsnedders|work> | alexmace: How do you propose to keep it open while making it impossible to spam? |
| 15:51 | <alexmace> | I think a better question would be "what is the purpose of the account?" and go from there rather than trying to band aid it |
| 15:51 | <Philip`> | gavin_: Well, it's a problem whose sole effect is to reduce the value of the WHATWG Twitter feed, which seems like a pretty limited problem to me :-) |
| 15:51 | <Lachy> | it doesn't seem like a huge problem. After only an hours worth of spam, it seems to have died down now. Seems people get over their stupidity quicker than I thought. |
| 15:51 | <Philip`> | ...although I suppose it also has the effect of irritating a thousand people |
| 15:51 | <gsnedders|work> | alexmace: The purpose of the account is provide updates relating to anything relating to HTML 5. |
| 15:52 | <Philip`> | gsnedders|work: Why does it need to be open? (I've never noticed anyone use it for serious purposes, other than the automatic commit notifications) |
| 15:52 | <Lachy> | anyone know how many followers it had before the spam started? |
| 15:52 | <Lachy> | I wonder if we just set a record for the most unsubscribed in the shortest amount of time |
| 15:52 | <gsnedders|work> | Philip`: There's been things like HTML 5 getting landed in mozilla-central, HTML 5 outliner, and other things posted to it. I don't watch it that closely. |
| 15:53 | <gsnedders|work> | Philip`: But yeah, it is mainly commits |
| 15:58 | <Philip`> | Lachy: Try looking in search engine caches |
| 16:00 | <Philip`> | I think Hixie should publish the web server password on the WHATWG front page, so anyone can go in and fix the Twitter script even while he's asleep |
| 16:01 | <Lachy> | from google cache 8 hours ago, 1,315 followers |
| 16:01 | <Lachy> | so we only lost about 50 |
| 16:02 | <Philip`> | But you said a few minutes ago "a few minutes ago, it was over 1600" |
| 16:02 | <gsnedders|work> | Lies, damned lies and statistics! |
| 16:02 | <beowulf> | http://twittercounter.com/whatwg/all |
| 16:04 | <beowulf> | i hit manual update on that, which changed the results to show a loss of 61 |
| 16:07 | <alexmace> | gsnedders|work: I'd look at something like hashtags instead |
| 16:07 | <alexmace> | At least then you can see the accounts that are posting the info rather than being completely anonymous as it is now |
| 16:15 | <Lachy> | Philip`, It looks like I remembered wrongly. That must have been the number of tweets, not the number of followers |
| 16:24 | <MadAtWork2> | what the hell is going on with twitter? |
| 16:24 | <leeky> | Afternoon! Anyone here able to remove that rather silly "post to @whatwg on Twitter" form on your home page? |
| 16:27 | krijn | thinks we need the gatekeeper for that :) |
| 16:28 | <MadAtWork2> | Can I set up a bot here that says "what the hell is going on with twitter" every minute or so? |
| 16:29 | <Philip`> | That would achieve nothing |
| 16:29 | <Philip`> | Also, it seems kind of obvious what's going on with Twitter |
| 16:29 | <jgraham> | It seems like Twitter is more or less working like normal |
| 16:29 | <Dashiva> | Oh no, UTF-32 is now the only acceptable encoding |
| 16:29 | <MadAtWork2> | I was being facetious, and I don't really know what was going on. |
| 16:30 | gsnedders|work | sighs |
| 16:30 | <gsnedders|work> | Teenagers… |
| 16:31 | <gsnedders|work> | Oh well, at least there's a Monty Python reference |
| 16:32 | <Lachy> | we just need to increase the signal to noise ratio on twitter. For every spam tweet someone posts, we should immediately post 2 more useful tweets |
| 16:32 | <MadAtWork2> | I mean, beside getting a lot of stupid spam from people I don't care about, what's the point of that? |
| 16:33 | <Dashiva> | Lachy: Can they be the same useful tweet? |
| 16:33 | <Lachy> | no |
| 16:34 | <gsnedders|work> | Dashiva: Suffix a number, then they aren't the same |
| 16:35 | <Lachy> | "Trapped in twitter factory, send help!" |
| 16:35 | <Lachy> | :-D |
| 16:35 | <MadAtWork2> | Seriously, what's the point of opening access to that account? |
| 16:36 | <Lachy> | MadAtWork2, as a matter of policy, everything in the whatwg has open access |
| 16:36 | <Lachy> | just like the blog, wiki and mailing list |
| 16:36 | <jgraham> | MadAtWork2: The point was to allow anyone to twitter on the subject of HTML 5. For a long while that has worked acceptably well. Right now people are not so interested in HTML5 |
| 16:37 | <MadAtWork2> | Hmm. |
| 16:37 | <MadAtWork2> | Can't we have 2 resources? an open one and an official announcement one? |
| 16:37 | <Lachy> | it might be worth closing public access until a more moderated system can be developed |
| 16:37 | <MadAtWork2> | aka one I can safely ignore and one I can follow? |
| 16:37 | <alexmace> | Seems some have confused open access with anonymous access |
| 16:38 | <Lachy> | maybe we should tie it to the blog somehow. That seems to work. Spammers regularly try to post rubbish articles, but a few of us sensible people have moderator access |
| 16:39 | <Dashiva> | It's not like we have much more control on the IRC channel :) |
| 16:39 | <gsnedders|work> | Lachy: Or spec annotation |
| 16:39 | <jgraham> | alexmace: It is far from clear that the access is anonymous |
| 16:39 | <Lachy> | yeah |
| 16:39 | <jgraham> | Hixie has all the logs |
| 16:39 | <Lachy> | jgraham, Hixie only gets IP addresses, and that doesn't really help for a distributed attack like this appears tob e |
| 16:39 | <Lachy> | *to be |
| 16:40 | <alexmace> | Good luck matching IP addresses to people then |
| 16:40 | <MadAtWork2> | Mostly because I can use an http proxy with relative ease... |
| 16:40 | <jgraham> | Lachy: Sure but it's hardly anonymous. |
| 16:40 | <alexmace> | It is effectively anonymous |
| 16:41 | <jgraham> | Only if people take additional steps to protect their anonmnymity. It's no more or less anonymous than wikipedia for example |
| 16:41 | <jgraham> | And the lack of anomnimity there has had interesting ramifications in the past |
| 16:42 | <MadAtWork2> | It's not about that, it's about what gets on the "front page". |
| 16:42 | <MadAtWork2> | Wikipedia has an article and a history page, but the history page is hardly what people checks first. |
| 16:42 | <alexmace> | Continue going on about other issues if you like |
| 16:42 | <MadAtWork2> | The noise is kept at bay. |
| 16:42 | <Samer> | lol hi |
| 16:42 | <Samer> | are we talking about the twitter issue |
| 16:42 | <alexmace> | Just ignore the problem |
| 16:42 | <Samer> | xD |
| 16:42 | <alexmace> | Then it'll go away |
| 16:42 | <Samer> | i'm reading the tweets |
| 16:42 | <Samer> | they're hillarious |
| 16:42 | <MadAtWork2> | In this case the amount of sensible information that gets through is nil. |
| 16:43 | <alexmace> | Yup |
| 16:43 | <alexmace> | MadAtWork2: It's "open" though |
| 16:43 | <MadAtWork2> | Tweeting AS WHATWG and tweeting TO WHATWG are two different things. |
| 16:43 | <takkaria> | ok, the experiment has failed, it's not that exciting really |
| 16:43 | <Lachy> | jgraham, you try finding out who owns a given IP address! In most countries, it will probalby take a good reason and a court order (or be a member of the RIAA lobby group) |
| 16:44 | <Samer> | maybe if you leave it for long enough it'll die |
| 16:44 | <MadAtWork2> | If you want something open, make a page that collects posts to @WHATWG from different accounts. |
| 16:44 | <Samer> | we like things complicated |
| 16:44 | <Lachy> | at most, all we can do is block an IP address from posting, but that wouldn't do much for the one-time hit and run attacks like this |
| 16:44 | <Samer> | :) |
| 16:44 | <jgraham> | Lachy: Which is a good thing but hardly the same as being anonymous |
| 16:44 | <MadAtWork2> | Or even go as far as making a new twitter account reposting everything sent to @WHATWG |
| 16:45 | <Samer> | have you tried changing the twitter password |
| 16:45 | <jgraham> | and more general information is trivial e.g. "that person posting from the white house made a racist comment" |
| 16:45 | <Lachy> | jgraham, then there's also people who go thorugh anonymising VPNs, or are using public terminals |
| 16:45 | <MadAtWork2> | jgraham: which translates to "Obama hates america" for Fox News. |
| 16:45 | <MadAtWork2> | Someone apparently DOES benefit. |
| 16:46 | <jgraham> | MadAtWork2: ? |
| 16:46 | <jgraham> | I think you are agreeing with me |
| 16:46 | <Samer> | hhow did this start btw |
| 16:46 | <MadAtWork2> | In a way. |
| 16:46 | <MadAtWork2> | Speaking of html5 tho. |
| 16:46 | <MadAtWork2> | We have header and footer, what is the middle thing? |
| 16:46 | <MadAtWork2> | Article? Div? |
| 16:46 | <MadAtWork2> | Nothing? |
| 16:47 | <jgraham> | MadAtWork2: That's an oversimplification but <section> or <article> depending on what you mean |
| 16:48 | <MadAtWork2> | jgraham: I'm starting a new webapp, and I'm creating a login page, I have a header I have a footer, and I have a login box, I'm trying to see what would uncle html5 suggest to use in this case. |
| 16:48 | <jgraham> | Lachy: I know that you can take some measures to cover your tracks, but in practice many people don't bother and find that actions that they took believing that they would be free of consequences are not actually as anonymous and as free of consequences as they had assumed |
| 16:49 | <jgraham> | MadAtWork2: Does it have a header? Probably <section> if it does |
| 16:49 | <jgraham> | Otherwise maybe <div>. Depends why you want it |
| 16:50 | <MadAtWork2> | The login box has a header but it's a legend for a fieldset. |
| 16:50 | <Lachy> | jgraham, O |
| 16:50 | <Lachy> | oops |
| 16:50 | <MadAtWork2> | so I have page header fieldset with login stuff and page footer. |
| 16:50 | <MadAtWork2> | I am trying to understand if I am to wrap the fieldset around a div, a section or what. |
| 16:51 | <MadAtWork2> | I'm starving for real world examples of html5. :/ |
| 16:51 | <Samer> | i think it stopped |
| 16:51 | <jgraham> | MadAtWork2: It sounds like it might be <section> and maybe you could move the fieldset label to a <h1> or so |
| 16:52 | <Lachy> | I'm not disputing the fact that most people don't have total anonynimity on the net in all cases, but for all practical purposes in the case of the whatwg status updater, they do, because we have absolutely no way of finding out who's who just from an IP address |
| 16:52 | <MadAtWork2> | jgraham: Hm, you may be right. |
| 16:53 | <MadAtWork2> | Still I dream of the day when these questions will be easily answered. |
| 16:53 | <alexmace> | Samer: Probably run out of API requests for the hour |
| 16:53 | <Samer> | oh lol |
| 16:53 | <Samer> | why is the update box on the front page? |
| 16:54 | <alexmace> | Which obviously makes it entirely useless for posting useful information |
| 16:54 | <alexmace> | "openness" |
| 16:54 | <Samer> | i mean instead of logging to whatwg.org just go to twitter.com to update |
| 16:54 | <alexmace> | Any discussion of finding people by IP address is just fiddling while Rome burns |
| 16:55 | <Samer> | at least people aren't going crazy on the wiki :) |
| 16:56 | <MadAtWork2> | Again, a bot that reposts what is posted to @WHATWG would solve everything. |
| 16:56 | <MadAtWork2> | Open and accountable. |
| 16:56 | <MadAtWork2> | It also may or may not improve your manliness. |
| 16:57 | <svl> | ooh, we just _gained_ a follower. Saw it go from 1200 to 1201 |
| 16:58 | <alexmace> | MadAtWork2: Sounds like a good solution to me though |
| 16:58 | <alexmace> | Not that anyone seems interested |
| 16:58 | <MadAtWork2> | I should tweet it. |
| 16:58 | <Philip`> | It's not like anyone can actually do anything |
| 17:05 | <jgraham> | Philip`: We could try DOSing the WHATWG server |
| 17:05 | <alexmace> | So who can do something about it? |
| 17:06 | <jgraham> | alexmace: Hixie |
| 17:06 | <alexmace> | and that's it? |
| 17:07 | <gavin_> | this is what happens when your server administration isn't Open! |
| 17:08 | <annevk2> | it was a nice experiment while it lasted |
| 17:08 | <annevk2> | the blog got abused way more quickly |
| 17:10 | <MadAtWork2> | haha |
| 17:10 | <jgraham> | annevk2: It might still work if you just made it possible for a few people to shut the account down if it was being abused at a given time |
| 17:10 | <jgraham> | Otherwise you would need to have a moderation system |
| 17:10 | <Dashiva> | Like, say, anyone on IRC |
| 17:12 | <annevk2> | jgraham, if we give a few people login details we can simply delete tweets |
| 17:12 | <Philip`> | Make a bot that posts each prospective Tweet into IRC, and if nobody says "twitbot: shut up" within a minute then it gets posted to Twitter |
| 17:12 | <annevk2> | that'd work too |
| 17:12 | <Dashiva> | Philip`: "Shut up, you twit" |
| 17:14 | <Philip`> | Got to be careful not to make a mistake like http://news.bbc.co.uk/1/hi/magazine/8176194.stm when talking about Twitter |
| 17:15 | <alexmace> | Posting it to IRC wouldn't work |
| 17:15 | <alexmace> | How would the bot know which you were talking about |
| 17:15 | <alexmace> | if they're coming in so quickly that the API usage is being used up every hour you'd just get a torrent coming in |
| 17:15 | <gavin_> | hmm, does the bbc typically publish "swear words"? |
| 17:16 | <gavin_> | would they print "fuck"? |
| 17:16 | <Philip`> | alexmace: It would know you're talking about anything sent in the past minute |
| 17:16 | <alexmace> | Philip`: whatever, it's a poor solution |
| 17:17 | <Philip`> | gavin_: Judging by that article, they only print it in direct quotes |
| 17:17 | <Philip`> | gavin_: http://news.bbc.co.uk/1/hi/education/7481715.stm |
| 17:17 | <MadAtWork2> | Philip`: I'm going to call humbug on that. |
| 17:18 | <MadAtWork2> | Seriously. |
| 17:19 | <gavin_> | Philip`: interesting |
| 17:20 | <Philip`> | gavin_: Not really relevant but a search points at section 2 of http://www.bbc.co.uk/commissioning/delivery/edguidance.shtml |
| 17:20 | <gavin_> | interesting set of examples |
| 17:21 | <Philip`> | Oh, section 3 is more fun |
| 17:21 | <gavin_> | I would have thought "Jesus fucking Christ" is redundant, given the inclusion of "fuck" |
| 17:21 | <GPHemsley> | Are you guys aware that @WHATWG is being spammed? |
| 17:21 | <alexmace> | \o/ |
| 17:21 | <GPHemsley> | Hmm... guessing yes |
| 17:22 | <GPHemsley> | Let me know when you fix it, so I can re-follow |
| 17:26 | <alexmace> | GPHemsley: Only one man can fix it |
| 17:26 | <GPHemsley> | Oh, Hixie's MIA? |
| 17:27 | <alexmace> | Mmm, I love the smell of a dichotomy just before home time |
| 17:37 | <annevk2> | GPHemsley, just asleep |
| 17:37 | <GPHemsley> | annevk2: Well, who knows where he might be while he's sleeping. ;) |
| 17:38 | <GPHemsley> | (In dream world, to clarify.) |
| 17:47 | <annevk2> | nice, DanC calls out bogus WAI tactics |
| 17:47 | <annevk2> | http://lists.w3.org/Archives/Public/www-archive/2009Aug/0006.html |
| 17:47 | <annevk2> | would make a good addition to markp's post |
| 18:49 | webben | wishes the discussion of "WCAG" on public-html would clearly (and properly) distinguish between W3C Recommendations and WAI Notes. |
| 18:59 | <annevk2> | maybe say so? |
| 19:05 | <webben> | annevk2: I've just mentioned it to Sam on the other channel though I suspect he may be afk. |
| 19:10 | <virtuelv> | someone mind removing that open whatwg-poster? |
| 19:10 | <virtuelv> | I'm not particularily fond of having my twitter frontpage filled with expletives |
| 19:10 | <virtuelv> | this is the internet. You _can't_ have nice things |
| 19:10 | <virtuelv> | right now, it's spewing racist expletives, evn |
| 19:11 | <virtuelv> | Hixie: ↑↑↑↑ |
| 19:12 | <hober> | :( |
| 19:13 | <annevk2> | worked fine for a couple of years |
| 19:13 | <annevk2> | :/ |
| 19:16 | <virtuelv> | annevk2: it worked, because no-one noticed |
| 19:16 | <virtuelv> | in other words, it worked by pure chance |
| 19:17 | <virtuelv> | either way, I'm unfollowing it for now |
| 19:18 | <factoryjoe> | wow, there's some bad stuff on twitter.com/whatwg |
| 19:18 | <virtuelv> | factoryjoe: noted already |
| 19:18 | <gavin_> | maybe that should be put in the topic |
| 19:18 | <annevk2> | that'd be logical |
| 19:18 | <factoryjoe> | sorry, thought it might be redundant to point that out |
| 19:19 | <annevk2> | no worries |
| 19:19 | <factoryjoe> | anonymity + internet = greater fuckwad |
| 19:19 | <factoryjoe> | is that an experiment? |
| 19:20 | <annevk2> | we had this on for a couple of years |
| 19:20 | <factoryjoe> | ah, i see it's been going on for awhile. |
| 19:20 | <factoryjoe> | the twitter thing? |
| 19:20 | <annevk2> | YEAH |
| 19:20 | <annevk2> | oops |
| 19:21 | <annevk2> | has been on the frontpage pretty much since we got a twitter account |
| 19:22 | <factoryjoe> | oh |
| 19:22 | <factoryjoe> | is there a desire to ... uh ... stop it from happening? |
| 19:22 | <factoryjoe> | or is the idea just to let it play out? |
| 19:23 | <virtuelv> | my bet is that twitter would suspend the account long before it plays out |
| 19:23 | <annevk2> | currently only Hixie can fix it |
| 19:23 | <factoryjoe> | true |
| 19:23 | <factoryjoe> | good point |
| 19:23 | <virtuelv> | doesn't even he have a phone these days? |
| 19:23 | <factoryjoe> | it's probably against the TOS to let anyone post to your account |
| 19:24 | <ezyang> | Hm. Does anyone have access to the src code of the script? |
| 19:24 | <ezyang> | I'd imagine that has the credentials. |
| 19:26 | <Hixie> | sigh, i tried putting up a message on the form saying that people should be nice, but that didn't work |
| 19:26 | <Hixie> | i guess we'll have to password-protect it |
| 19:27 | <Hixie> | that's so sad |
| 19:27 | <annevk2> | can we easily delete all the crap? |
| 19:27 | <factoryjoe> | Hixie: the content is pretty offensive... and it'll be associated with the account out of context |
| 19:27 | <factoryjoe> | it is too bad |
| 19:27 | <factoryjoe> | how long was it there w/o abuse? |
| 19:27 | <virtuelv> | Hixie: I see you never passed the "Greater Internet Fuckwad theory" exam? |
| 19:28 | <virtuelv> | factoryjoe: two years? |
| 19:28 | <factoryjoe> | amazing |
| 19:29 | <Hixie> | factoryjoe: years at least |
| 19:30 | <Hixie> | several people knew the old password, fwiw, but i guess y'all forgot what it was :-) |
| 19:30 | <Hixie> | anyway i changed the password |
| 19:30 | <Hixie> | so it's no longer "WHATWG" which is what it used to be :-P |
| 19:30 | <Hixie> | since i figure if people are going to abuse the form, eventually one will try to guess the password too |
| 19:31 | <virtuelv> | Hixie: you might also want to delete the tweets? |
| 19:31 | <Hixie> | yeah yeah, i'll get to it |
| 19:32 | <smedero> | fyi, there's no easy way on twitter to mass delete tweets from my memory but somewhat had written a greasemonkey script to do it. |
| 19:32 | <smedero> | s/somewhat/someone/ |
| 19:35 | <Hixie> | aaah |
| 19:35 | <Hixie> | don't delete tweets that aren't offensive |
| 19:37 | <Hixie> | jgraham: your /msgs to me are coming up blank -- did you get my message to you? |
| 19:38 | <jgraham> | Hixie: No, try resending. Seems to be a client problem that may have sorted itself out |
| 19:38 | <Hixie> | k |
| 19:39 | <Hixie> | mpilgrim: i'm in slice btw |
| 19:40 | <factoryjoe> | also, hixie, you may want to tweet once more that "the experiment is over. we'll be cleaning up this account momentarily." |
| 19:41 | <Hixie> | i don't think additional tweets are going to make anyone particularly happier |
| 19:41 | <Hixie> | people will just realise things are better when they don't get spammed |
| 19:41 | <miketaylr> | agreed. |
| 19:42 | <factoryjoe> | well, ok. some acknowledge might help people feel better that no more spam is forthcoming |
| 19:43 | <Hixie> | i expect that no more spam being forthcoming will make people feel better than spam telling them that no more spam is forthcoming |
| 19:43 | <Hixie> | it's like getting confirmation messages from spamming lists when you unsubscribe -- just irritates more |
| 19:44 | <gavin_> | if the confirmation is credible, it actually helps, I think |
| 19:44 | <factoryjoe> | i'm with gavin_, but it's up to you |
| 19:44 | <gavin_> | (rarely the case with email spam, granted - think twitterspam is different)) |
| 19:44 | <factoryjoe> | FWIW, i unfollowed |
| 19:45 | <miketaylr> | a compromise might be to state something in the profile bio..."now, with 100% less spam!" |
| 19:45 | <factoryjoe> | if you admit the issue and reassure people, they might come back, but that might not be important to you |
| 19:45 | <factoryjoe> | anywho |
| 19:46 | <annevk2> | i agree that we should say something |
| 19:46 | <annevk2> | anyone a proposal? |
| 19:48 | <annevk2> | "Apologies. Letting everyone update @WHATWG from whatwg.org worked for years, but after today we have taken it offline." |
| 19:48 | <annevk2> | ? |
| 19:49 | <Hixie> | let's not say anything for now |
| 19:49 | <Hixie> | if it's still not abused in a few hours we'll post something |
| 19:51 | <jgraham> | othermaciej: I have heard that the kind of structural information that blind users like would also be helpful to people with some cognitive disabilities although I don't know if that is really true |
| 19:52 | <othermaciej> | jgraham: I don't personally know the details of what is useful to different handicap groups - I would expect there is a lot of overlap, perhaps not 100% |
| 19:56 | <jgraham> | othermaciej: Indeed. I was just making the point that your example of information that may not be needed by sighted users may, in fact, be needed by some sighted users |
| 19:57 | <othermaciej> | jgraham: I think I qualified it enough that I feel ok with the way I stated it |
| 19:57 | <othermaciej> | "Describing the structure of the table, if it is easy to grasp visually, may not be useful to everyone." |
| 19:57 | <jgraham> | othermaciej: Fair enough. It wasn't really a critisism |
| 19:57 | <othermaciej> | If it's not easy to grasp visually, then it's quite likely that cognitively disabled users (or indeed everyone) could use the info |
| 19:58 | <othermaciej> | and even after that qualification, I said "may" |
| 19:58 | <othermaciej> | cause it's probably a judgment call at that point |
| 19:58 | <jgraham> | I am just somewhat wary of the disabled===blind trap |
| 19:59 | <jgraham> | In any case it seems like a rather sensible proposal but not too far from what is in the spec today |
| 20:14 | <jwalden> | annevk42: fyi, I have some feedback on Opera's implementation of the CSS 3 background-size property in this post: https://developer.mozilla.org/web-tech/2009/08/04/background-images-no-longer-restricted-to-original-size-explore-the-space-with-background-size/ |
| 20:25 | <Lachy> | I wonder what Shelley's objection is to othermaciej's compromise |
| 20:26 | <othermaciej> | she elaborated at some more length in a later message |
| 20:26 | <Lachy> | oh, haven't received that yet |
| 20:27 | <jgraham> | jwalden: Interesting |
| 20:31 | <jwalden> | I also have, if I remember right, perhaps forty or fifty reftests of background-size and related properties that I should eventually bundle up somehow and post on www-style |
| 20:33 | <jwalden> | all of them in http://mxr.mozilla.org/mozilla-central/source/layout/reftests/backgrounds/ from http://hg.mozilla.org/mozilla-central/rev/88e95d638b35 if anyone really wants them *right now*, as opposed to when I get around to zipping them up and making them presentable |
| 20:34 | <jwalden> | (i.e. s/-moz-// mostly, also pulling out a few that are engine-specific) |
| 20:37 | <takkaria> | I'm reading a book called Bad Science at the moment, all about evidence-based methology and such |
| 20:37 | <takkaria> | really quite a contrast to public-html |
| 20:46 | <Dashiva> | takkaria: Does it cover good cop / bad cop bargaining? :) |
| 20:47 | <takkaria> | sadly not |
| 20:47 | <jgraham> | takkaria: I read one ofthe author's columns in the Guardian recently and it didn't seem that great |
| 20:47 | <takkaria> | jgraham: I think that to people who are trained in actual scienific disciplines, it's not that interesting |
| 20:48 | <jgraham> | It was on the subject of organic food and made reasonable criticisms of a Soil Association press release that claimed that a report saying organic food provided no health benefits was bogus |
| 20:49 | <takkaria> | ah, yeah, the book's like that but more in-depth and will a lot of depth on general scientific methology |
| 20:49 | <jgraham> | But seemed to go futrther and claim that organic food in general has no benefits |
| 20:50 | <jgraham> | (rather than possibly just not having the one benefit that the press release was focusing on) |
| 20:51 | <jgraham> | (which made me think that he probably had some preconcieved idea that organic food is pure hype) |
| 20:52 | <jgraham> | (which is the problem with putting ourself on a high pedestal and just being critical of other people; you leave yourself open to the same criticisms) |
| 20:59 | <Hixie> | ok well i think we got rid of the particularly nasty twitters |
| 21:07 | <Hixie> | ok i posted on the whatwg twitter feed |
| 21:07 | <Hixie> | if anyone wants to know the password, let me know |
| 21:07 | <Hixie> | (it's not the same as the whatwg account password, for those who know that) |
| 21:09 | <Hixie> | oh and for those of you who do have the account password -- if anyone abuses it somehow again, just change the password and let me know |
| 21:09 | <Hixie> | and if you want to know the account password, let me know, i can give that to you too |
| 21:22 | <takkaria> | jgraham: hm, I didn't read him as saying it had no benefits, just that if it does have benefits, they are not health benefits |
| 21:22 | <takkaria> | jgraham: and he chided the Soil Assoc. a bit for not making that clearer |
| 22:15 | <othermaciej> | jgraham: well, I thought it was sensible, but I'm starting to lose faith in my ability to get buy-in on a compromise |
| 22:20 | <Lachy> | othermaciej, if we go with your proposal, then I'll have to concede a lot more than it seems others are even remotely willing, which isn't really fair. |
| 22:21 | <othermaciej> | Lachy: if people on the other side are not willing to concede anything, then I assume they will not support my proposal at all |
| 22:21 | <othermaciej> | I'm not going to push it if it doesn't have significant buy-in from both sides |
| 22:22 | <othermaciej> | like I said, I don't want to ask either side to concede 100% of their position in exchange for nothing |
| 22:23 | <Lachy> | well, it seems with your proposal, I'll have to concede 99% of what I want. The only remaining thing being the validator warning for using it. |
| 22:24 | <othermaciej> | the validator warning, plus strong advice to use text that is visible in all media when appropriate, which is likely the vast majority of the time |
| 22:24 | <Lachy> | I can live with that, though obviously I'm not thrilled about it. |
| 22:24 | <Lachy> | oh, yeah, that too. Make that 95% |
| 22:25 | <othermaciej> | (plus hopefully support to change WCAG2 to be more aligned with this point of view) |
| 22:25 | <Lachy> | even if we do accept this now, I'd still like to leave it open for modification in the future based on new evidence, as always |
| 22:26 | <othermaciej> | My rough estimation is that my proposal is probably right on the edge of what people on either side might remotely be willing to accept |
| 22:27 | <Lachy> | btw, I'm currently working on a way that will hopefully give us new evidence to work with |
| 22:27 | <othermaciej> | so if it doesn't work, we probably just have to vote to decide between people's actual more extreme positions |
| 22:27 | <Lachy> | and if my expectations of the results are met, then it should debunk a lot of the claims about how the new proposals don't "offer the same functionality" as summary="" |
| 22:28 | <Hixie> | so... redirects in websocket |
| 22:28 | Hixie | ponders how to make that work |
| 22:28 | <Lachy> | Hixie, are you changing the subject? |
| 22:28 | <Hixie> | no, i haven't been worrying about summary at all today :-) |
| 22:29 | <Hixie> | please feel free to continue talking about summary :-) |
| 22:29 | Hixie | is just thinking out loud because the next e-mail on his pile is about websockets |
| 22:29 | <Lachy> | no, that's all I've been talking about for ages. |
| 22:29 | <Lachy> | about *summary |
| 22:45 | <Lachy> | Hixie, in the interest of getting the publication of a Working Draft for the heartbeat requirement out of the way and decoupling the resolution of the summary issue from that process, would you be willing to agree to a short term compromise solution, leaving the group to continue debating the details of othermaciej's proposal? |
| 22:45 | <Hixie> | what's the compromise? |
| 22:46 | <Hixie> | (i've not yet read the 50+ e-mails on the topic) |
| 22:46 | <Lachy> | that's ok, I haven't emailed this to the group yet |
| 22:47 | <othermaciej> | Lachy: I am curious what your study is |
| 22:47 | <othermaciej> | Hixie: I made a proposal for what I think is a compromise (not sure if that's what Lachy has in mind) |
| 22:47 | <Lachy> | specifically, remove summary from the Obsolete but Conforming features list, and add a big note to The table Element section stating that there has been a proposal to include the summary attribute as conforming, but the issue is not yet decided |
| 22:47 | <othermaciej> | Ah, I see he means something different |
| 22:47 | <Hixie> | remove summary="" altogether? i'm certainly very happy to remove summary altogether, yes |
| 22:47 | <Lachy> | so that avoid's John's controversial changes of making summary fully conforming now |
| 22:48 | <Hixie> | that's what i've been askin for forever |
| 22:48 | <Lachy> | yeah, but the difference is the class="XXX" note in the table section that makes it clear the issue is still unresolved |
| 22:48 | <othermaciej> | Lachy: you might want to see if John Foliot would buy into that as removing his objection to publishing a WD |
| 22:48 | <Hixie> | we already have an XXX |
| 22:49 | <Hixie> | i'm pretty sure john would object completely to having summary="" be obsolete |
| 22:49 | <Lachy> | yeah, but John is objecting to that since summary is still listed as obsolete, despite the issue still being unresolved |
| 22:49 | <Hixie> | since wai requires use of summary="" in one of their non-normative techniques notes |
| 22:49 | <Hixie> | i thought that's what you were proposing too, making it obsolete completely, not even conforming |
| 22:49 | <Lachy> | I'm not proposing to make it fully obsolete, nor fully conforming. Just a note to say that its in limbo right now |
| 22:50 | <othermaciej> | Lachy is proposing removing all mention of it other than an XXX note that the issue is open |
| 22:50 | <othermaciej> | (for the next WD only) |
| 22:50 | <Lachy> | othermaciej, yes, exactly |
| 22:50 | <Hixie> | i don't really care what the next WD says at all |
| 22:50 | <Hixie> | the WDs are obsolete before we publish them |
| 22:50 | <othermaciej> | me neither |
| 22:50 | <Lachy> | nor do I. I just want the WD published and out of the way so we can meet the heartbeat requirement and move on |
| 22:51 | <Lachy> | it seems rather silly to hold up publication based on an orthogonal issue |
| 22:51 | <Hixie> | we can just publish your doc as a fpwd to satisfy the process, if process is what you care about |
| 22:52 | <othermaciej> | frankly I would be ok with publishing a WD with either Hixie's text or John Foliot's text, my only annoyance is that it's been tied to WD publication when a WD by definition doesn't have to reflect consensus or resolve any specific issue |
| 22:52 | <Lachy> | my doc isn't ready for publication yet. It hasn't been updated in a while and is very out of sync with the spec |
| 22:52 | <Hixie> | Lachy: *shrug* |
| 22:53 | <Lachy> | also, I was awaiting the resolution for the spec licencing issue before I published it |
| 22:53 | <Hixie> | Lachy: i don't really have any opinion or concern about the heartbeat requirement |
| 22:53 | <othermaciej> | furthermore I think periodic WD publication has little meaning except pro forma satisfaction of W3C Process requirements |
| 22:53 | <Hixie> | Lachy: my interest is in addressing the actual issue |
| 22:53 | <othermaciej> | which is why I wish we wouldn't mix it up with other things |
| 22:53 | <Lachy> | sure, so is mine. But having the actual issue tied to the publication of a draft is proving harmful. So finding a way to untie it and get on with addressing the actual issue seems productive |
| 22:55 | <Lachy> | that's also why I proposed to to rubys in #html-wg the other day that we simply adopt a process in the group whereby any editor can ask for publication as a WD at any time and it be done without debate. |
| 22:56 | <Hixie> | i would be fine with replacing the entire spec with a picture of bunnies and kittens for WD, if that's what you're worried about |
| 22:56 | <Lachy> | LOL |
| 22:57 | <othermaciej> | bunneh!! |
| 22:57 | <Hixie> | i really couldn't care less about the WD itself, i'm concerned about fixing the actual problems in the language |
| 22:57 | <Lachy> | othermaciej, re the study, I'll forward you a copy of the proposal I've sent to a few other collegues |
| 22:58 | <othermaciej> | cool, I'll watch for it |
| 22:58 | <Lachy> | I'm keeping it off list for now to avoid it being destroyed before it even begins by some of the trolls in the group |
| 23:00 | <othermaciej> | fair enough |
| 23:01 | <Hixie> | feel free to post it to the whatwg list |
| 23:01 | <Hixie> | don't let trolls dictate your behaviour |
| 23:02 | <Lachy> | sent |
| 23:07 | <Hixie> | don't forget to make the analysis phase blind also |
| 23:07 | <Hixie> | it's important that the people analysing the results not know which group each table is from |
| 23:12 | <Lachy> | for the usability testing phase at the end, that will be possible, assuming the assitive technology users don't know any other details about the study |
| 23:14 | <webben_> | othermaciej: http://lists.w3.org/Archives/Public/public-html/2009Aug/0221.html |
| 23:14 | <othermaciej> | webben_: yes? |
| 23:14 | <webben_> | position: absolute; left: -999999px; works well with screen readers. |
| 23:15 | <othermaciej> | webben_: I do know that one works - what I don't know is if visibility: hidden or using an aural media stylesheet consistently works |
| 23:15 | dbaron | was thinking about implementing the canvas text changes (pretty easy), but wants to know what to do about the ctx.font getter first |
| 23:15 | <webben_> | visibility: hidden; is known to have problems with (at least) older AT, and is generally undesirable as it still leaves a layout box. |
| 23:16 | <webben_> | aural media stylesheet doesn't work with popular screen readers. |
| 23:17 | <Hixie> | dbaron: what's the question? (looking at e-mail now) |
| 23:17 | <webben_> | othermaciej: compare also existing WCAG advice: http://www.w3.org/TR/WCAG-TECHS/C7.html |
| 23:17 | <dbaron> | Hixie, pretty much everything about the ctx.font getter is undefined :-) |
| 23:17 | <othermaciej> | webben_: what about sizing something to 0 width and 0 height? |
| 23:18 | <dbaron> | Hixie, like if you set ctx.font="2em serif" do you get back 2em or 32px? And if the font size on the canvas then changes? |
| 23:18 | <Hixie> | dbaron: oh, i see. i'm relying on http://dev.w3.org/csswg/cssom/#css-value |
| 23:18 | <dbaron> | oh boy |
| 23:18 | <dbaron> | I suppose i should look at that spec then |
| 23:18 | <Hixie> | dbaron: i'm not sure about the units, though |
| 23:18 | <dbaron> | "Issue: This document is in its very early stages. The intention is that this specification will in due course supersede DOM Level 2 Style. " |
| 23:18 | <webben_> | s/WCAG/WAI/ |
| 23:19 | <Hixie> | dbaron: i guess converting to pixels makes sense |
| 23:19 | <webben_> | I've experienced problems with WAI's advice of moving content upwards. |
| 23:19 | <Hixie> | dbaron: especially given the example in your e-mail |
| 23:19 | <Hixie> | dbaron: let's say that all units are converted to px on setting, and that the serialisation must use the rules in cssom |
| 23:19 | <dbaron> | ok |
| 23:19 | <webben_> | iirc had a doubling effect on links in Window-Eyes link list and did funny things with keyboard navigation. |
| 23:19 | <Hixie> | dbaron: (those rules are basically the canonicalisation rules i made up a few years back) |
| 23:20 | <Hixie> | dbaron: also if you find an editor for cssom, let me know :-) |
| 23:20 | <Hixie> | anne's unable to spend time on it at the moment |
| 23:20 | <webben_> | othermaciej: re font size/height 0: I'm not sure. I believe authors often avoid those as potential spam flags. |
| 23:20 | <Hixie> | dbaron: i'll update the spec as soon as i'm done adding redirects and authentication to web sockets |
| 23:21 | <dbaron> | k |
| 23:21 | <dbaron> | thanks |
| 23:21 | <Hixie> | np |
| 23:22 | <webben_> | othermaciej: old research: http://css-discuss.incutio.com/?page=ScreenreaderVisibility and http://www.access-matters.com/screen-reader-test-results/ |
| 23:26 | <webben_> | othermaciej: (also old, but interesting re media-specific styles): http://web.archive.org/web/20071229112251/http://www.freedomscientific.com/fs_support/BulletinView.cfm?QC=1165 |
| 23:27 | <othermaciej> | webben_: thanks for the info! |
| 23:28 | <GPHemsley> | Has the Twitter account been fixed? |
| 23:29 | <takkaria> | yes |
| 23:29 | <gsnedders> | Aww… how boring. |
| 23:30 | <gsnedders> | Hixie: You want help deleting all the spam? |
| 23:30 | <Hixie> | nah, anne helped already |
| 23:30 | <Hixie> | thanks though |
| 23:30 | <gsnedders> | There's still more. |
| 23:31 | <Hixie> | there's no spam left |
| 23:31 | <Hixie> | just messages that were posted |
| 23:31 | <Hixie> | which is fine |
| 23:31 | <Hixie> | we don't delete criticism |
| 23:31 | <gsnedders> | http://twitter.com/WHATWG doesn't look like it |
| 23:31 | <gsnedders> | http://twitter.com/WHATWG/status/3125833340 looks far from sane |
| 23:31 | <Hixie> | hm, there's some new ones since i last looked |
| 23:31 | <Hixie> | i guess twitter is on some sort of delay |
| 23:43 | <webben_> | othermaciej: I suspect people will know what you mean, but a large amount to the left - |
| 23:43 | <webben_> | { position: absolute; left: 9999px; } |
| 23:43 | <webben_> | is missing the "-" sign |
| 23:43 | <othermaciej> | oh, right |
| 23:43 | <othermaciej> | feel free to correct me on the list if you want |
| 23:43 | <othermaciej> | I have some meetings to attend to |
| 23:43 | Philip` | doesn't suppose it'll make a difference until people have monitors 10,000 pixels wide |
| 23:44 | <gsnedders> | And use browser windows that wide. |
| 23:45 | webben_ | used to use em units, but found this triggered a very odd bug in Safari, adding this absurdly wide horizontal scrollbar |
| 23:45 | <webben_> | dunno if that's still the case in 4. |
| 23:47 | <Dashiva> | That's a whole lot of email... |
| 23:49 | <Hixie> | hmm |
| 23:49 | <Hixie> | do Set-Cookie headers get honoured in redirects? |
| 23:49 | webben_ | wonders whether <ins><details></details></ins> is a better way to meet Roy's use-case than @summary. If textual fidelity is important then textual fidelity is important. |
| 23:50 | <Hixie> | i.e. if i get a 301 Redirect and it includes a Set-Cookie, does it get set? |