| 13:15 | <eyebrows> | I guess the guy who came up with 0.5 pixel measurements for canvas strokes left his logic at the door too |
| 13:15 | <eyebrows> | I mean, really |
| 17:36 | <AryehGregor> | Heh: http://lists.w3.org/Archives/Public/www-archive/2011Mar/0015.html |
| 17:37 | <AryehGregor> | So, Hixie, what are these "many browsers" that don't render tables incrementally? |
| 17:55 | <zewt> | http://kathack.com/ probably old but o_O |
| 19:47 | <Hixie> | AryehGregor: hm? |
| 19:48 | <AryehGregor> | Hixie, browsers render tables incrementally, don't they? Or did you just mean they have to reflow them? |
| 19:48 | <Hixie> | it varies |
| 19:48 | <AryehGregor> | How so? |
| 19:49 | <Hixie> | hold on, let me get you a test |
| 19:53 | <Hixie> | hrm, how do i turn off deflate compression |
| 19:58 | <zewt> | some htaccess magic? |
| 20:01 | <zewt> | in apache .htaccess, SetEnv no-gzip |
| 20:04 | <Hixie> | oooh |
| 20:04 | <Hixie> | cunning |
| 20:07 | <Hixie> | even with that the server is still returning the file back in one block, bummer |
| 20:08 | <erlehmann> | Hixie, what are you trying to accomplish? |
| 20:08 | <zewt> | see if <? sleep(2) ?> in php causes it to send incrementally (or if it just blocks) |
| 20:08 | <Hixie> | have the server return cells with a one second pause between them |
| 20:08 | <Hixie> | using perl |
| 20:08 | <zewt> | or just stuff a bunch of whitespace to convince it to send some |
| 20:09 | <Hixie> | i'm stuffing 1024 'x's between each cell |
| 20:09 | <Hixie> | 4096 even |
| 20:09 | <Hixie> | that really should be enough |
| 20:09 | <Hixie> | even 40960 isn't enough |
| 20:09 | <AryehGregor> | Are you sure you're flushing the output buffer when you pause? |
| 20:09 | <zewt> | is apache just waiting for the whole thing so it can send content-length? |
| 20:10 | <zewt> | not sure how to force chunked |
| 20:11 | <zewt> | curl: (56) Received problem 2 in the chunky parser |
| 20:11 | <zewt> | chunky parser? sounds delicious |
| 20:11 | <Hixie> | AryehGregor: i'm outputting 40K of content before sleeping, if that doesn't force it i don't know what will |
| 20:11 | <AryehGregor> | Maybe it will buffer indefinitely unless you explicitly flush the buffer. |
| 20:12 | <Hixie> | how do i flush the buffer? |
| 20:12 | <AryehGregor> | I don't know. In PHP with ob, it's ob_flush(). |
| 20:12 | <zewt> | ah php switches to chunked automatically if you call flush() |
| 20:12 | <AryehGregor> | Without ob I guess it's flush(). |
| 20:13 | <zewt> | Hixie: this works for me in PHP, anyway http://pastebin.com/NUBf9tCq |
| 20:13 | <zewt> | don't know about perl |
| 20:13 | <zewt> | no padding needed |
| 20:13 | <Hixie> | well i have perl's autoflush mode enabled |
| 20:13 | <Hixie> | i don't know what's up |
| 20:13 | <Philip`> | If it's CGI, there's use IO::Handle; STDOUT->flush; |
| 20:14 | Philip` | doesn't know what buffering Apache does |
| 20:14 | <zewt> | i'd imagine that unless you actually somehow tell it to, it won't turn on chunked (in php, calling flush() is also telling it to do that) |
| 20:14 | <Hixie> | tried that too, with no success |
| 20:14 | <Hixie> | oh well |
| 20:14 | <heycam> | jgraham, just by following http://dev.w3.org/2006/webapi/WebIDL/#es-exception-hasinstance it seems to require that to be true [exception instanceof other_window.DOMException] |
| 20:14 | <zewt> | http://zewt.org/~glenn/delay.php |
| 20:15 | <heycam> | jgraham, but the spec doesn't really deal with multiple globals specifically... |
| 20:16 | <heycam> | jgraham, if you think "instanceof"s like this (and `node instanceof other_window.Node`) should be returning false, let me know |
| 20:16 | <heycam> | jgraham, (probably they should be returning false, since that's what'd happen with say `myObj instanceof other_window.Object`) |
| 20:27 | <AryehGregor> | Who would be a good Mozilla person to ask about execCommand()? |
| 20:30 | <AryehGregor> | roc, smaug____, volkmar, do any of you know who a good Mozilla person is to ask for feedback about my execCommand() spec? |
| 20:30 | <roc> | ehsan |
| 20:30 | <volkmar> | AryehGregor: ehsan i think |
| 20:30 | <AryehGregor> | Okay, thanks. |
| 20:35 | <gsnedders> | AryehGregor: Email hallvord for anything frm an Opera POV, FWIW |
| 20:35 | <AryehGregor> | gsnedders, okay, I'll CC him too. |
| 20:36 | <AryehGregor> | Thanks. |
| 20:36 | <gsnedders> | *from |
| 20:42 | <AryehGregor> | Okay, e-mail sent. |
| 20:42 | <AryehGregor> | anolis --dump-xrefs intermediate.html /tmp/spec is hanging, by the way. |
| 20:43 | <AryehGregor> | http://pastebin.com/3rJJkWcY |
| 20:43 | <AryehGregor> | Just freezes there forever. |
| 20:43 | AryehGregor | tries updating it |
| 20:44 | <AryehGregor> | Nope, still hangs. |
| 20:44 | <AryehGregor> | It's Ms2ger's hacked version. |
| 20:55 | <erlehmann> | I believe this is old news for almost all of you, but can someone tell me if that is a practical joke? <http://whyfirefoxisblocked.com/>; |
| 21:00 | <Hixie> | it was intended seriously (back in the mid 200s) |
| 21:01 | <Hixie> | however it was seriously misguided then and remains so now :-) |
| 21:02 | <Hixie> | mid 2000s even |
| 21:04 | <erlehmann> | Hixie, I stand confused from the logic: “Too few visitors are viewing my ads, so I'm blocking everyone – even the 95% without an ad blocker from viewing my page (and, by extension, my ads).” |
| 21:05 | <Hixie> | it's even worse than that |
| 21:05 | <erlehmann> | This seems like one of these shows making fun of bureaucrats … ON THE INTERNET. |
| 21:05 | <Hixie> | the people _with_ ad blockers have the ability to work around the block, so they still see the content |
| 21:05 | <Hixie> | the people _without_ ad blockers learn about them |
| 21:06 | <erlehmann> | Oh wow. m( |
| 21:07 | <zewt> | the mid 200s? yeah i guess 1800 years ago this might have passed for rational thought :) |
| 21:08 | <zewt> | that's pretty high comedy--"too few people click my ads, so I need to reduce it further" |
| 21:09 | <zewt> | i'm honestly surprised there's as little ad-blocker-blocking in the wild as there is; it's not like there's a shortage of people with such bad judgement |
| 21:10 | <zewt> | i'm similarly surprised there's very little effort to try to make ads that can't be blocked; it's an arms race that ad blockers would ultimately lose |
| 21:13 | <Hixie> | most people don't mind ads |
| 21:13 | <erlehmann> | zewt, how do you know they haven't already lost? ;) |
| 21:13 | <zewt> | erlehmann: largely because I almost never see ads :) |
| 21:13 | <Hixie> | imho ads are fine so long as they are (a) useful and relevant and (b) not actually blocking the content or annoying |
| 21:13 | <erlehmann> | dito here. |
| 21:14 | <erlehmann> | zewt, parsed sentence wrong: thought ad blockers would win and ad companies would lose. |
| 21:14 | <Hixie> | (interstitial and preroll ads on big blogs and in video streams drive me crazy) |
| 21:14 | <zewt> | Hixie: sure, but a whole lot of ads are neither of those things :) |
| 21:14 | <Hixie> | yeah |
| 21:15 | <erlehmann> | Well, relevant content is relevant. |
| 21:15 | <zewt> | there are plenty of ads and ad providers who just want to spam banners everywhere they can--it's that sort I'm surprised aren't trying harder to thwart blockers |
| 21:16 | <erlehmann> | I consider ad blockers as a tool to skew the signal-to-noise-ratio in my favor – they are not the only ones, I once wrote a userscript utilizing LZW compression to remove repetitive spammy comments from my view of an imageboard. |
| 21:16 | <erlehmann> | It was funny when I realized it also blocked legit comments that were just not that interesting due to repetitive style. |
| 21:17 | <Hixie> | what's sad is that relevant and useful ads make so much more money than irrelevant and annoying ads, it makes you wonder why people keep on using the annoying ones |
| 21:17 | <zewt> | not every type of site has ads that are really relevant to it, though |
| 21:18 | <zewt> | if you're reviewing graphics cards then it's easy; if you're blogging about earthquakes in japan, not so much |
| 21:20 | <erlehmann> | Anyone having usable CSS for @cite? I'm currently using this: blockquote::after { content: "<" attr(cite) ">"; } |
| 21:20 | <erlehmann> | should probably be blockquote[cite]::after |
| 21:20 | <wilhelm> | zewt: CNN at least tries to be relevant: http://i.imgur.com/GTNjf.png :P |
| 21:29 | <erlehmann> | wilhelm, could be worse. At least 4chan is not displaying fake ads for “S.T.A.L.K.E.R.: Shadow of Fukushima”. |
| 21:29 | <erlehmann> | (but given the puddi disaster, that may happen) |
| 21:30 | <erlehmann> | btw, adblock was extremely useful against puddi |
| 21:31 | <erlehmann> | oh wait, the puddi video was an ad. so obvious it hurts. |
| 22:18 | <AryehGregor> | How did I not think to test whitespace stripping in my reflection tests? |
| 22:18 | <AryehGregor> | I'll have to fix that. |
| 22:29 | <AryehGregor> | I'm mildly amused when I see people who use complicated jQuery stuff when they could use much simpler built-in features. |
| 22:29 | <AryehGregor> | Like: var tabIndex = parseInt( $(this).attr( 'tabindex' ), 10 ); |
| 22:29 | <AryehGregor> | Instead of, let's say: var tabIndex = this.tabIndex; |
| 22:40 | <erlehmann> | wat |
