| 00:00 | <rubys> | Currently, there are plugins that do capturing and serving of video in a separate process. I have no idea whether there is anything in HTML5 which would enable this to be moved into the browser. |
| 00:02 | <roc> | there is an old W3C draft that specifies <input type="file" device="camera"> or something like that |
| 00:04 | <Hixie> | rubys: if you mean camera capture, then no, there currently isn't, though many people have requested it and i imagine that in due course (when browsers have implemented more of html5) it would be something we would want to add |
| 00:04 | <rubys> | thanks! |
| 00:09 | <roc> | Hixie: Brad Lassey (I think) has been looking at doing camera integration using something like that W3C draft |
| 00:12 | <Hixie> | roc: cool |
| 00:22 | <franksalim> | rubys: like Hixie said, orbited has a js based shim implementation. kaazing has a flash/js shim that exposes the websocke api and tries to speak the wire protocol when flash is available |
| 00:23 | <franksalim> | this would be a good time to point out that my words do not represent my friends or my employers :-p |
| 00:23 | <franksalim> | *websocket |
| 00:48 | <rubys> | franksalim: thanks! |
| 00:54 | <franksalim> | rubys: np! feel free to pick my brain if you have any questions about that sort of thing |
| 00:56 | <rubys> | at the moment, I'm just relaying questions from somebody who is trying to decide if html5 is worth exploring. Having two sets of "shims' that allow you to get by until the browsers catch up is a good thing. |
| 08:34 | <Hixie> | hsivonen: yt? |
| 08:50 | <hsivonen> | Hixie: yes |
| 08:58 | <Hixie> | hsivonen: i assume you don't want ftp:// but https:// is ok? |
| 09:02 | <hsivonen> | https works |
| 09:02 | <hsivonen> | ftp not |
| 09:02 | <hsivonen> | I'm using Commons HttpClient |
| 09:05 | <Hixie> | right-o |
| 09:05 | Hixie | updates his script and reruns it to filter out ftp:// |
| 09:06 | <BenMillard> | publishing the progress of my collection isn't major enough to get into Mark Pilgrim's summary :( http://blog.whatwg.org/this-week-in-html-5-episode-3 |
| 09:06 | <Hixie> | BenMillard: i don't think mark is reading the mailing list. however, you should feel free to publish your own blog post announcing it, i think that would be great actually |
| 09:08 | <Hixie> | hsivonen: so i am getting you 1,000,000 urls taken from an arbitrarily sized subset of google's index that represents what google thinks are the most important paegs (for some vague definition of "important" that really gets kinda hard to pin down when you're talking about such large samples) |
| 09:09 | <Hixie> | hsivonen: the random sample is biased as follows: |
| 09:09 | <Hixie> | hsivonen: i dropped all pages that we didn't determine were HTML when we fetched them |
| 09:09 | <Hixie> | hsivonen: i dropped all pages whose urls didn't start with "http" |
| 09:10 | <Hixie> | hsivonen: i dropped 80% of pages whose urls only had 3 "/" characters in them and whose urls' last character was a "/" |
| 09:11 | <Hixie> | hsivonen: and i only took one page per cluster of pages that seemed to consist a single "site" |
| 09:12 | <BenMillard> | Hixie, I made a little one on my own website the other day: http://projectcerbera.com/blog/2008/08/midpoint |
| 09:13 | <BenMillard> | so I guess a WHATWG entry for my research would be redundant? although I would've liked a paragraph in Mark's entry :P |
| 09:14 | <hsivonen> | BenMillard: the JS live DOM didn't make it in last week, either. |
| 09:15 | <hsivonen> | BenMillard: it's a summary of spec checkins |
| 09:15 | <hsivonen> | BenMillard: but you can write a post yourself |
| 09:15 | <hsivonen> | (for people who subscribe to the blog in general instead of the markp-only feed) |
| 09:16 | <BenMillard> | ah, ok |
| 09:16 | <Hixie> | BenMillard: don't worry, he got some of my changes wrong anyway :-) (i added a comment to the blog entry correcting it, feel free to add a comment if you don't want to add a whole blog post) |
| 09:16 | <BenMillard> | Hixie, oh yeah, a comment is a good idea |
| 09:21 | <Hixie> | man, backing up half a terabyte over USB 2.0 is not fast |
| 09:21 | <BenMillard> | I set my Name to be "Ben 'Cerbera' Millard" and each time a Preview I get a \ before each '. So, after a few previous I see "Ben \\\\'Cerbera\\\\' Millard". would be nice if that didn't happen |
| 09:21 | <BenMillard> | *a few previews |
| 09:22 | <Hixie> | probably a wordpress bug |
| 09:22 | <Hixie> | does wordpress use php? |
| 09:22 | <Hixie> | i've found php apps uniformly get ' escaping wrong |
| 09:22 | <hsivonen> | it does |
| 09:22 | <Hixie> | dunno why |
| 09:22 | <BenMillard> | yes, it does look like a magic quote time thing |
| 09:22 | <BenMillard> | *type (yikes, my typing is groggy this morning) |
| 09:22 | <hsivonen> | the magic quote thing is in the worst ideas ever category |
| 09:23 | <BenMillard> | I've made the comment but it hasn't appeared...there's no message but I *assume* it's in a moderation queue of some sort and hasn't been lost? |
| 09:23 | <BenMillard> | I used some HTML and some links |
| 09:23 | <hsivonen> | I'll check the queue |
| 09:24 | <hsivonen> | BenMillard: now approved |
| 09:25 | <BenMillard> | yay, thanks! for the IRC records, the comment is precisely here: http://blog.whatwg.org/this-week-in-html-5-episode-3#comment-26873 |
| 09:25 | <Hixie> | http://%3cbr%3etyxynola.newmail.ru/_visits_26.html is a weird url |
| 09:25 | <Hixie> | but it seems to work in safari |
| 09:25 | <Hixie> | i don't understand |
| 09:26 | <Hixie> | hsivonen: do you want these in lexical order or random order? |
| 09:26 | <hsivonen> | Hixie: random is better |
| 09:27 | <hsivonen> | either works |
| 09:27 | <Hixie> | i have them in random order |
| 09:27 | <Hixie> | so that works |
| 09:27 | <Hixie> | that url doesn't work in firefox |
| 09:27 | <Hixie> | and it gives a different page in IE than WebKit |
| 09:28 | <Hixie> | doesn't help that i can't speak russian |
| 09:28 | <Hixie> | opera doesn't like that url either |
| 09:28 | <Hixie> | oh well |
| 09:29 | <Hixie> | this URL list is 52MB uncompressed |
| 09:29 | <Hixie> | 20MB compressed |
| 09:32 | <BenMillard> | which flavour of compression? |
| 09:33 | <hsivonen> | I guess a URI list in random order doesn't have too much common substring locality apart from "http://www.";, ".com" and ".html" |
| 09:36 | <roc> | sigh |
| 09:37 | <hsivonen> | roc: ? |
| 09:37 | <Hixie> | BenMillard: gzip |
| 09:37 | <Hixie> | hsivonen: surprisingly many sites don't have www., .com, and .html, too |
| 09:37 | <roc> | making the intrinsic size of an <iframe> depend on the layout of its container isn't really a good i dea |
| 09:37 | <Hixie> | hsivonen: sent e-mail |
| 09:38 | <Hixie> | um |
| 09:39 | <Hixie> | not sent |
| 09:39 | <Hixie> | my SMTP server said no to a 20MB attachment |
| 09:41 | <hsivonen> | Hixie: email now received. thanks |
| 09:53 | <gDashiva> | Matt's reply to James strikes me as rather lacking |
| 09:54 | <gDashiva> | Making a painting is a lot more work than adding <img> to a HTML page, yet apparently making alternate text for a painting is infeasible |
| 09:54 | <gDashiva> | And uploading a video to youtube is limited by physical space? |
| 09:56 | <gDashiva> | And the synthesia example talks explicitly about a webpage |
| 09:56 | <Philip`> | Presenting alternate text for a painting is infeasible (you don't want people rubbing their hands all over your priceless paintings trying to find a braille description) |
| 09:57 | <Philip`> | but for <img> it's trivial since you just add an attribute and the user has tools that let them read it |
| 09:57 | <Hixie> | i love how everyone and their brother has now pointed out that there are references to RFC3066 in HTML5 |
| 09:59 | <gDashiva> | Philip`: Imagine if the alt text was embedded in the original painting, then you wouldn't need to write alt text for the <img> |
| 09:59 | <gDashiva> | They could just use an automated extract-alt-text-from-image AT program |
| 10:02 | <Philip`> | gDashiva: I can imagine such a situation but I'm not sure what relevance that has |
| 10:03 | <Hixie> | i wonder if al plans to reply to http://lists.w3.org/Archives/Public/wai-xtech/2008Aug/0132.html |
| 10:04 | <hsivonen> | Hixie: there's a chance that your assuption that alt text should make sense when flattened as part of surrounding prose isn't the right assumption |
| 10:05 | <hsivonen> | Hixie: and that users would be better off getting cued that a piece of text is an alt digression from the surrounding prose and doesn't need to flow nicely into the paragraph |
| 10:05 | <Hixie> | there's always a chance that i'm wrong on many things :-) |
| 10:06 | <Hixie> | well i don't know about AT users, but at least from the point of view of a Lynx/Links user (which I am regularly), I certainly wouldn't want to be told whenever there's an image, i just want the image replaced with the text |
| 10:07 | <gDashiva> | Philip`: one painter writing alt, or X web page authors doing the same work over and over :) |
| 10:07 | <gDashiva> | But yes, it's not that relevant in today's digital age |
| 10:09 | <Hixie> | ok bed time |
| 10:09 | <Hixie> | nn |
| 10:10 | <gDashiva> | m |
| 10:10 | <Philip`> | gDashiva: Maybe the painter could embed the alt text as a kind of watermark in their painting, so even if someone takes a photo of it and uploads it to their web page then the text will still be extractable and the image will still be accessible |
| 10:11 | <gDashiva> | Philip`: Kinda like all that pirate-tracing tech being proposed for DRMed media? |
| 10:14 | <zcorpan> | Hixie: "well i don't know about AT users, but at least from the point of view of a Lynx/Links user (which I am regularly), I certainly wouldn't want to be told whenever there's an image, i just want the image replaced with the text" -- that'd work fine when reading your blog but not so fine for web pages in general since alt is so often misused |
| 10:23 | <BenMillard> | zcorpan, here are the interesting uses of alt I collected: http://projectcerbera.com/web/study/2008/collection#alt |
| 10:24 | <BenMillard> | (it's not a study of how many are good versus how many are bad; it focusses on the ones I found interesting or considered wrong) |
| 10:24 | <BenMillard> | duplication of nearby text and filenames seem like a common mistake, although I haven't crunched the numbers yet |
| 10:25 | <BenMillard> | sometimes it looks like the alt was written for one version of the image, then the image changed and the alt wasn't updated |
| 10:29 | <BenMillard> | missing the alt in places it was needed happened in quite a few cases, too |
| 11:25 | <zcorpan> | hsivonen: <meta charset=asdfasdf> validates |
| 11:27 | <anne> | I heard yesterday namespaces could've been worse. Philippe Le Hégaret detailed a huge device at the W3C (took eight months to resolve and many many hours) as to whether a namespace should be a URI reference or a string. |
| 11:54 | <webben> | Philip`: note that some image formats allow you to embed text metadata : http://www.mirrorservice.org/sites/www.libpng.org/pub/png/spec/1.2/PNG-Chunks.html#C.tEXt for example |
| 11:55 | <hsivonen> | zcorpan: asdfasdf meets the syntactic requirements for a charset name |
| 11:55 | <hsivonen> | zcorpan: should I check the value against a list of actual names? |
| 11:55 | <hsivonen> | zcorpan: if so, which list? |
| 11:56 | <hsivonen> | I don't like the theoretical open-endedness of charset |
| 11:56 | <hsivonen> | I want UTF-8 and a closed list of grandfathered legacy names |
| 11:57 | <hsivonen> | anne: do you meant like doing relative URI processing on NS URIs? |
| 11:58 | <hsivonen> | s/meant/mean/ |
| 12:01 | <zcorpan> | hsivonen: in xml you warn that it doesn't match external encoding information |
| 12:02 | <zcorpan> | hsivonen: in text/html it's an all clear |
| 12:02 | <anne> | hsivonen, yes |
| 12:02 | <anne> | s/device/devide/ |
| 12:02 | <anne> | too |
| 12:03 | <zcorpan> | hsivonen: you could use the list that your validator supports |
| 12:03 | <hsivonen> | zcorpan: hmm. somehow I thought I had code for checking that meta matched the encoding in use |
| 12:03 | <hsivonen> | weird |
| 12:04 | <zcorpan> | hsivonen: yeah i remember it whining about it before |
| 12:04 | <zcorpan> | hsivonen: i tested with textarea input |
| 12:06 | hsivonen | finds // XXX NOP |
| 12:12 | <hsivonen> | zcorpan: fixed. thanks |
| 12:29 | <hsivonen> | hmm. using real web content is almost like fuzzing |
| 12:33 | <wilhelm> | It is. Our stability testing tools use URLs from the wild or fuzzers interchangeably. (c; |
| 14:15 | <anne> | hsivonen, anonymous suggests maybe making a list of namespace benefits for browser vendors (not suggesting it's non-empty) |
| 14:17 | <hsivonen> | anne: namespace or Namespace? |
| 14:19 | <anne> | XML namespace, if that helps |
| 14:19 | <anne> | namespaces*, meh |
| 14:21 | <anne> | should we update Wikipedia to mention Web Workers? |
| 14:23 | <anne> | I guess it will happen eventualy |
| 14:23 | <anne> | lly |
| 14:23 | <hsivonen> | I can't think of any benefits from Namespaces to browser vendors where the benefit couldn't have been obtained by a mechanism that doesn't use the kind of indirection Namespaces use |
| 14:25 | <hsivonen> | CSS extensions, DOM extensions and MathML all involve putting a short prefix into the opaque name |
| 14:26 | <anne> | same here |
| 14:30 | <hsivonen> | I can't think of cases where a browser vendor is itself using Namespaces except XUL |
| 14:30 | <hsivonen> | though the benefit could be that if others use Namespaces, they keep out of the browser-sensitive name turf |
| 14:30 | <anne> | even there you could just have moz-button and such |
| 14:30 | <hsivonen> | right |
| 14:31 | <hsivonen> | oh, and then there's VML |
| 14:32 | <hsivonen> | which for practical purposes uses the short prefix approach |
| 14:37 | <hsivonen> | the other day, I was pondering if HTML5 could make createElementNS fade into obscurity by making createElement recognize MathML and SVG names and the SVG/HTML name-colliding elements implementing the interfaces for both HTML and SVG |
| 14:38 | <hsivonen> | after all, SVG rendering is already sensitive to the tree position of the element |
| 14:39 | <anne> | maybe |
| 14:41 | <anne> | grmbl, I e-mailed ietf-types, but http://www.alvestrand.no/pipermail/ietf-types/ doesn't reflect that |
| 17:34 | Philip` | can't remember whether he's already commented that his objection to the alt="{...}" proposal (in that it makes it hard to write reliably-correct HTML-generating tools) could be fixed by permitting it in the alt-is-actually-a-proper-equivalent cases as well as the alt-is-just-a-description cases |
| 17:34 | <Philip`> | which would mean conforming pages using alt="{...}" would be ambiguous, but that's okay because it's going to be ambiguous in reality anyway and this won't make it any worse |
| 17:35 | <Philip`> | So, anyway, if I hadn't commented on that, then at least I have now |
| 17:58 | krijnh | adds a new subtitle to his logs |
| 17:59 | <krijnh> | jgraham: thanks for the inspiration :) |
| 18:03 | <Lachy> | krijnh, which one was added? |
| 18:04 | anne__ | hits F5 |
| 18:04 | <krijnh> | Something about an eval overlord |
| 18:04 | <krijnh> | *evil |
| 18:04 | <anne__> | "Because, Seriously, Dmitry Can't Do It All By Himself" hehe |
| 18:04 | <Lachy> | that won't work too well for me, since I don't konw most of the existing ones |
| 18:04 | <jgraham> | Lachy: http://krijnhoetmer.nl/irc-logs/?gimme-a-hobby |
| 18:05 | anne__ | finds "Where The Evil HTML 5 Overlord And His Band Of Minions Chat, In Public" |
| 18:05 | anne__ | lols |
| 18:06 | <jgraham> | anne__: In the middle of a WG meeting? |
| 18:06 | <anne__> | jgraham, at the end actually, maybe you can join for dinner |
| 18:07 | <anne__> | jgraham, you can |
| 18:07 | anne__ | figures out where we are going |
| 18:07 | <anne__> | Philip`, you can too, but it's Indian and I heard you might not enjoy that |
| 18:08 | <jgraham> | anne__: OK |
| 18:08 | <jgraham> | cool |
| 18:08 | <anne__> | jgraham, http://www.pipasha-restaurant.co.uk/?id=contact |
| 18:08 | <anne__> | we have a reservation at 6:45 |
| 18:08 | <anne__> | Philip`, there is place for you fwiw |
| 18:09 | <jgraham> | Interesting. How are you getting there |
| 18:09 | <anne__> | (also, it's a subset of the WG) |
| 18:09 | jgraham | can't walk that far by 6:45 |
| 18:10 | <jgraham> | You are just across the road, right? |
| 18:10 | <anne__> | http://maps.google.com/maps?f=q&hl=en&geocode=&q=cambridge+pipasha&ie=UTF8&ll=52.213353,0.154431&spn=0.004944,0.009871&t=h&z=17 |
| 18:11 | <anne__> | we are at Microsoft still |
| 18:11 | <anne__> | and going by car |
| 18:11 | <anne__> | if you come a little later that's ok |
| 18:12 | <anne__> | or grab a cab, not sure if that's expensive or not though |
| 18:12 | <jgraham> | I guess I might just stay here and do some work since it's so far away and I've got so little done |
| 18:13 | <anne__> | is it really that far? |
| 18:13 | <anne__> | ok |
| 18:13 | <jgraham> | anne__: I guess 45-60 minutes walk |
| 18:14 | <anne__> | arriving at 7, 7:15 is fine, but decide for yourself |
| 18:14 | anne__ | has to pack his laptop now |
| 18:36 | Lachy | wrote a script to extract all subtitles from krijnh's logs |
| 18:36 | <Lachy> | well, I hope I got them all. I relied on the probability of all of them being sent at least once within 100 separate requests for the page |
| 18:37 | <krijnh> | Hah |
| 18:37 | <krijnh> | Don't DOS my server please :) |
| 18:37 | <Lachy> | I didn't. I did: wget http://.../; sleep 1; in a loop |
| 18:37 | <krijnh> | How many did you find? |
| 18:37 | <Lachy> | <em>This</em> Line Is My Opinion, Not The Other(')s |
| 18:37 | <Lachy> | <em>What working group</em> is going to work on extending HTML... |
| 18:37 | <Lachy> | Actually, We Need A Solid Use Case For These Logs |
| 18:37 | <Lachy> | And <code>X-UA-Compatible</code> Is A Great Idea! |
| 18:37 | <Lachy> | And Indeed, <code>#xhtml</code> Wasn't Kick Ass Enough |
| 18:37 | <Lachy> | And RSS Is A Success As Well |
| 18:38 | <Lachy> | And You Can <code>/join</code> Too |
| 18:38 | <Lachy> | Because A Mailinglist Alone Isn't Enough |
| 18:38 | <Lachy> | Because All Those Web 2.0 Widgets Should Be Accessible |
| 18:38 | <Lachy> | Because XHTML Is The Future! |
| 18:38 | <Lachy> | Because, Seriously, Dmitry Can't Do It All By Himself |
| 18:38 | <Lachy> | Could You Elaborate? |
| 18:38 | <Lachy> | Great Minds Think Alike. And Sick Minds, Too! |
| 18:38 | <krijnh> | Damn you! Now I have to find new ones |
| 18:38 | <Lachy> | In Search Of A Better Title |
| 18:38 | <Lachy> | Including Offtopic WHATWG Cabal Ramblings |
| 18:38 | <Lachy> | Issue Tracking Tools Will Save Us! |
| 18:38 | <Lachy> | OMG! We Can't Break The Web! |
| 18:38 | <Lachy> | Opinions Expressed Here Are Those ... Bla Bla Bla |
| 18:38 | <Lachy> | Public-html: A Mailinglist Apart |
| 18:38 | <Lachy> | So, When Can We Use <code>application/xhtml+xml</code> Again? |
| 18:38 | <Lachy> | Somebody Needs To Do To The Semantic Web What The WHATWG Did To HTML |
| 18:38 | <Lachy> | Sorry, But Who Asked For This? |
| 18:38 | <Lachy> | We Cannot Definitively Say Why XHTML Has Not Been Successful On The Web |
| 18:38 | <Lachy> | Web 2.0, Evolved (Still In Beta Though) |
| 18:38 | <Lachy> | What Else Do I Need To Do To Get Hired By Opera? Or Google? |
| 18:38 | <Lachy> | Where The Evil HTML 5 Overlord And His Band Of Minions Chat, In Public |
| 18:38 | <Lachy> | While We All Agree Flash Is Still A Superior Technology ;) |
| 18:38 | <Lachy> | Yeah, And Standards Suck, Right |
| 18:38 | <Lachy> | Your Ad Here? Let's See. Nope, It Isn't. |
| 18:38 | <Lachy> | htmllogs4all! |
| 18:38 | <Lachy> | you could always erase them from your log, and it'll be as if I never posted them :-) |
| 18:38 | <krijnh> | :) |
| 18:39 | <krijnh> | So now it's pretty obvious: sometimes I just don't have anything to do |
| 18:42 | <krijnh> | Same thing for you btw |
| 18:50 | <Lachy> | woah, I did miss one on the first attempt. there was 30 in that list, this time I got 31 |
| 18:50 | <Lachy> | "And SVG Needs Some Love As Well" was missed |
| 18:51 | <krijnh> | Oh dear |
| 18:51 | <Lachy> | btw, if anyone else wants to participate in a DDoS against krijnh, here's the script :-) |
| 18:51 | <Lachy> | for ((i=0;$i<100;i=$i+1)); |
| 18:51 | <Lachy> | do wget -O $i.html http://krijnhoetmer.nl/irc-logs/?gimme-a-hobby; |
| 18:51 | <Lachy> | sleep 1; |
| 18:51 | <Lachy> | done |
| 18:51 | <Lachy> | cat * | grep ^\<p | sort | uniq > all.txt |
| 18:51 | <Lachy> | cat all.txt | wc -l |
| 18:51 | <krijnh> | Thanks for ruining my fun ;) |
| 18:52 | <gDashiva> | See |
| 18:52 | <gDashiva> | You need to add a captcha or something |
| 18:52 | <krijnh> | I seriously hope nobody is interested in this as much as you.. And not because of the DOS stuff ;) |
| 18:53 | <Lachy> | I was just bored, and wanted to practice my skills with writing bash scripts, which I started learning recently |
| 18:53 | <jcranmer> | Lachy: the bash script I wrote to help teach myself is ugly |
| 18:53 | <jcranmer> | but I have proven that it is possible to write a BFS in bash |
| 18:54 | <gDashiva> | It's also possible to bang your head against the wall |
| 18:54 | <Lachy> | BFS? first result was http://en.wikipedia.org/wiki/Breadth-first_search |
| 18:54 | <Lachy> | is that right? |
| 18:54 | <jcranmer> | yes |
| 18:54 | <krijnh> | gDashiva: Good point |
| 18:55 | <Lachy> | the bug with my script is that it doesn't stript <p><b> tags from the lines. Is that possible to do with grep, or some other utility? |
| 18:55 | <krijnh> | sed ? |
| 18:56 | <Lachy> | ok, I'll have to learn sed later |
| 18:56 | <krijnh> | No idea if it's possible, I don't use it that much |
| 18:56 | <krijnh> | Anyway, done reading up on HTML WG mail |
| 18:56 | <Lachy> | anything interesting worth reading? |
| 18:57 | <krijnh> | Anybody in here going to Fronteers 2008, to see anne speak? :) |
| 18:57 | <Lachy> | no |
| 18:57 | <krijnh> | Tsk |
| 18:57 | <Lachy> | when and where is that? |
| 18:57 | <krijnh> | http://fronteers.nl/congres/2008/speakers |
| 18:57 | <Lachy> | anne isn't listed as a speaker on that page |
| 18:58 | <krijnh> | There are 3 speakers to be annoucned |
| 18:58 | <krijnh> | *announced |
| 18:58 | <krijnh> | Anne is one of those 3 |
| 18:58 | <Lachy> | I'm going back to Australia on September 20 instead, and I have my own presentation to give on the Oct 7 |
| 19:05 | <krijnh> | But Dean Edwards is coming! |
| 19:25 | gsnedders | notes he doesn't have 200 EUR |
| 19:26 | jgraham | notes that stripping stuff from lines ia a major use case for sed |
| 19:30 | <jgraham> | sed -e "s|<p><b>\(.*\)<\/b><\/p>|\1|" |
| 19:30 | <jgraham> | for example |
| 19:30 | <jgraham> | obviously Philip` will have an easier way :) |
| 19:31 | <Philip`> | perl -pe's~<p><b>(.*)</b></p>~$1~' |
| 19:32 | <jgraham> | krijnh: I don't know how the subtitle gets chosen but you should put in a few with a low probability like 1 in 1000 |
| 19:32 | <krijnh> | Why exactly? ) |
| 19:33 | <jgraham> | To annoy Lachy |
| 19:33 | <jgraham> | :) |
| 19:33 | <krijnh> | Good reason :) |
| 19:33 | <Philip`> | anne__: Indeed, I'm not much of a fan of any non-trivial food :-( |
| 19:34 | jgraham | notes that he has empirical evidence that Philip` eats more types of food than the guy in his department who eats only processed chicken and white bread |
| 19:35 | <Philip`> | I eat unprocessed chicken too :-) |
| 19:40 | <gsnedders> | Not spinach though. |
| 19:43 | <gsnedders> | (or at least I think that's what Philip` didn't eat on his pizza back when myself and jgraham went out to supper with him in May) |
| 19:45 | <gsnedders> | krijnh: I have to admit I think one or two are a bit mean to some people, even if they are funny :) |
| 19:46 | <gsnedders> | Oh, BTW, I did decide on the project that I came up with based on jgraham's idea. |
| 20:03 | <krijnh> | gsnedders: which ones? |
| 20:03 | <gsnedders> | krijnh: Oh like, Dimitry |
| 20:03 | <gsnedders> | though that is funny. |
| 20:06 | <krijnh> | Hope so, not meant to be offensive or something :) |
| 20:12 | <Hixie> | hsivonen: just make the list of encodings you consider valid be the list of encodings you support |
| 20:15 | <Hixie> | Kristof sends weird e-mails |
| 20:15 | <Hixie> | i don't really know what to ask him to change |
| 20:16 | <Hixie> | but like his latest e-mail about GNU libmagic... I mean, I'm sure some people don't know about file(1) yet, but so what? |
| 20:16 | <Lachy> | jgraham, thanks for the help with sed |
| 20:17 | <Hixie> | karl sent a similar e-mail (about new york) |
| 20:42 | <Hixie> | i asked al if he was going to reply to my last e-mail on the alt thread |
| 20:42 | <Hixie> | he replied that i was asking the wrong person |
| 20:42 | <Hixie> | i'll take that as a no then... |
| 20:46 | <krijnh> | (Meanwhile, the bomb keeps ticking) |
| 22:33 | <Hixie> | "why would Flickr decide to violate [the requirement that alt not be bogus] rather than the one that says 'authors must provide suitable alt text'" |
| 22:33 | <Hixie> | seriously? |
| 22:34 | <Hixie> | we're still at that point in the debate? |
| 22:34 | <Hixie> | jesus |
| 22:34 | <hober> | sigh. |
| 22:35 | <Dashiva> | On one hand they want people to be conformant. On the other hand they expect people to ignore conformance for arbitrary reasons |
| 22:36 | <roc> | "why would Flickr decide to violate the alt requirement rather than the one that says Flickr should stay in business" |
| 22:37 | <roc> | although of course Flickr probably would stay in business, users would quickly find the shortest alt string that passes filters |
| 22:37 | <Dashiva> | And the AT could just hardcode that to mean 'no alt'. "Brillant" |
| 22:38 | <Hixie> | hey what are some photo sites like picasa and flickr |
| 22:38 | <roc> | photobucket |
| 22:38 | <Dashiva> | my opera has galleries |
| 22:38 | <hober> | photo.net |
| 22:38 | <Hixie> | i need one more with a single word name and then google sets will give me the rest |
| 22:38 | <Dashiva> | all the photobucket-like sites, imageshack etc |
| 22:39 | <Hixie> | cool thanks |
| 22:39 | <Hixie> | http://labs.google.com/sets?hl=en&q1=flickr&q2=picasa&q3=photobucket&q4=imageshack&q5=&btn=Large+Set |
| 22:39 | <roc> | very cool |
| 22:49 | <Hixie> | holy crap there are lot of these |
| 22:50 | <Dashiva> | And then you have image boards |
| 23:42 | <Lachy> | OMG, Stargate Atlantis is ending at the end of season 5 and move on to straight-to-DVD movies |
| 23:42 | <Hixie> | wow, i hadn't realised how much research we actually had here for this alt="" thing |
| 23:42 | <Lachy> | and Stargate Universe has been confirmed for next year |
| 23:42 | <Hixie> | sweet |
| 23:42 | <Lachy> | http://gateworld.net/news/2008/08/wright_iatlantisi_is_going_out_o.shtml |
| 23:43 | <Hixie> | i can't go to gateworld.net |
| 23:43 | <Lachy> | SPOILERS: http://gateworld.net/news/2008/08/istargate_universei_has_a_go.shtml |
| 23:43 | <Lachy> | oh |
| 23:43 | <Hixie> | i get spoilt every single time |
| 23:43 | <Lachy> | the first link doesn't contain any spoilers about the plot, it just talks about why they're ending the series now and changing to movies |
| 23:44 | <Xenos> | Half the heroes die |
| 23:45 | <Lachy> | liar! |
| 23:45 | <Hixie> | Lachy: i bet there's something in some image in a sidebar somewhere that's a spoiler for something (e.g. future episode titles, guest star for next week, etc) |
| 23:48 | <Lachy> | it also says there's also going to be a new computer game called Stargate Worlds |
| 23:49 | <Lachy> | yeah, it has the name of a future episode in the sidebar |
| 23:49 | <Hixie> | see :-) |
| 23:49 | <Lachy> | it's tonghts episode title |
| 23:49 | <Hixie> | oh hey today is friday! |
| 23:49 | <Hixie> | woot! |
| 23:51 | <Lachy> | yep, I just waiting for it to air, so I can download the episode in about 4 hours |
| 23:52 | <Hixie> | i'll have to see it tomorrow, sadly, iTunes doesn't let me download it until about 3am |
| 23:52 | <Hixie> | very annoying |
| 23:52 | <Lachy> | iTunes doesn't have stargate available in either Australia or Norway, and they're the only stores I can buy from |
| 23:52 | <Lachy> | so I have to use usenet |
| 23:53 | <Lachy> | but I get to see it in 1080p |
| 23:53 | <Lachy> | I found that itunes quality is a bit too low |
| 23:53 | <Hixie> | yeah |
| 23:53 | <Hixie> | well sadly my mac mini couldn't handle HD anyway :-) |
| 23:54 | <Lachy> | don't you have a faster machine to watch it on? |
| 23:55 | <Lachy> | btw, do you strip the DRM from your itunes purchases so you can watch them wherever you like? |
| 23:55 | <Hixie> | i could watch it on the laptop, but connecting the laptop to the cinema display and the audioengine speakers is a bitch, so i don't bother |
| 23:55 | <Hixie> | i haven't found a way to losslessly strip the DRM sadly |
| 23:55 | <Lachy> | Requiem 1.7.3 works |
| 23:56 | <Lachy> | do you want a copy? |
| 23:57 | <Lachy> | whoever wrote it actually reverse engineered the way iTunes decrypts them, so it's completely lossless |
| 23:57 | <Hixie> | oooh, never heard of it! sure! |
| 23:57 | <Hixie> | that would be awesome |
| 23:57 | <Hixie> | ian⊙hc |