| 01:08 | <tantek> | hsivonen, I'm curious why validator.nu would report that 'type' and 'data' attributes are not allowed on 'object' when the spec certainly appears to allow them, e.g. see: http://validator.nu/?doc=http%3A%2F%2Ftantek.com%2F&parser=html5&profile=polyglot |
| 01:08 | <tantek> | http://www.whatwg.org/specs/web-apps/current-work/multipage/the-iframe-element.html#attr-object-type |
| 01:09 | <Hixie> | probably a bug in the validator, file a bug :-) |
| 01:10 | <tantek> | Hixie, I tend to assume it's a bug in my code before thinking it's a bug in the validator :) |
| 01:10 | <Hixie> | looks like a validator bug to me |
| 01:12 | <Hixie> | yeah |
| 01:12 | <Hixie> | it fails on data:text/html;charset=utf-8,<!DOCTYPE%20HTML><title>Test</title><object%20type%3D"image/png"%20data%3D"http://example.com/test.png"></object>; too |
| 01:12 | <Hixie> | which i'm pretty sure is valid |
| 01:22 | <tantek> | Hixie - bug filed: http://bugzilla.validator.nu/show_bug.cgi?id=716 (noted link to your test case also) |
| 01:27 | <Hixie> | tantek: cool |
| 01:27 | <Hixie> | tantek: hey while i have you here, do you have an opinion on getting rid of the "magic" around N in the microdata vCard vocabulary? |
| 01:28 | <Hixie> | tantek: i was looking at the vCard spec and it doesn't say the N needs to actually have any data, just that it has to be present |
| 01:34 | <roc> | Dashiva: In fact, "normal internet users" have to be pretty careful about how they use H.264. See http://bemasc.net/wordpress/2010/02/02/no-you-cant-do-that-with-h264/ ... that's mostly about video production, but it also turns out that e.g. playing video using Flash for "commercial use" is also forbidden |
| 01:35 | <Hixie> | man i hate working on the drag-and-drop API |
| 01:49 | <Dashiva> | roc: They have to be careful they pay for a service that doesn't properly license its content, and even then nothing's going to happen |
| 01:51 | <Dashiva> | Internet broadcast is free until 2016, and that includes sites with ads, so unless you plan to set up full-blown commerical operation... |
| 02:03 | <roc> | it's true, if you just want to watch Youtube you'll be fine. |
| 02:03 | <Dashiva> | Or post your own videos |
| 02:04 | <roc> | to Youtube |
| 02:04 | <Dashiva> | Or your blog |
| 02:04 | <roc> | I believe that technically, you need to execute a license for that. You won't have to pay royalties until 2016. |
| 02:05 | <Hixie> | it's pretty clear that until H.264 is freely usable without any license nonsense, it's not a workable solution. |
| 02:06 | <nessy> | plus the weird stuff they recently replied to for open source software |
| 02:06 | <Hixie> | usable and implementable, yes |
| 02:07 | <Hixie> | (and i mean Free, not $0) |
| 02:07 | <nessy> | and then the one where you cannot do commercial content, even if you have a license to Final Cut Pro |
| 02:07 | <nessy> | more and more restrictions are coming out |
| 02:07 | <nessy> | even if they won't go after people for such infringements, it's still a scary state |
| 02:08 | <roc> | those restrictions were always there, people just didn't know or care |
| 02:08 | <roc> | of course most still don't, as long as they can watch Youtube |
| 02:08 | <roc> | which is fine |
| 02:08 | <nessy> | it's good it is being exposed - mostly it was unknown till now |
| 02:09 | <nessy> | I'm still hoping Google will "do good" and provide an Ogg alternative of YouTube, too |
| 02:09 | <nessy> | it's a massive transcoding effort, but then they did that for mobile content and HD content, too |
| 02:10 | <nessy> | I don't mind large corporates continuing to use H.264 - but it shouldn't be forced upon the small guys |
| 02:11 | <jcranmer> | google has the computing power, I bet |
| 02:26 | <jwalden> | http://news.bbc.co.uk/2/hi/technology/8496263.stm hum |
| 02:30 | <tantek> | Hixie, sorry I didn't see your q re: vcard/n til now |
| 02:32 | <tantek> | Much of the "optimizations" for n (and fn and others) were design/built to bridge the impedance mismatch/gap between the way people author content (only what they have to / seems obvious), and the strict top-down a-priori requirements of various formats. |
| 02:33 | <tantek> | given the years of experience we've had with hCard (and n and fn uses), there's definitely been a few cases where some optimizations can be dropped, others can be improved/refined, and there's room for a few more too (again, based on experience with what authors have tried to do with hCard) |
| 02:33 | <tantek> | I've captured much of this in hcard-brainstorming for hCard 1.0.1 (and similarly for hCalendar) |
| 02:34 | <tantek> | essentially all such "optimizations" come down to making the vocabularies more *usable* |
| 02:34 | <tantek> | anyway - feel free to jump into #microformats for more on hCard/vcard/n/fn etc. |
| 02:35 | <tantek> | btw on another topic (which was mentioned above) |
| 02:35 | <tantek> | http://bemasc.net/wordpress/2010/02/02/no-you-cant-do-that-with-h264/ |
| 02:36 | <tantek> | the video codec situ seems quite messy |
| 02:42 | <Hixie> | tantek: k. sounds like it would be reasonable for the microdata version to just say that a blank N is generated if it's missing, then, with no special magic to generate it from FN, NICK, and ORG like hCard |
| 02:44 | <tantek> | Hixie, there are probably some cases where a blank n *should* be generated from fn, nickname, org etc., even in hCard |
| 02:45 | <Hixie> | how do you mean? |
| 02:46 | <tantek> | that is there are some cases where hCard auto-generates a value for the 'n' property when it should probably generate an empty 'n' property |
| 02:47 | <Hixie> | ah i see |
| 02:47 | <tantek> | I'm going to have to another round of tests with modern vcard consumers to understand the impact of various combinations of implying a blank 'n' vs an 'n' with a non-empty value. |
| 02:47 | <tantek> | to have to *do* another |
| 02:47 | <Hixie> | k |
| 02:47 | <Hixie> | i have to say, one has to love how the vcard spec just says that N is required |
| 02:48 | <tantek> | don't get me started |
| 02:48 | <Hixie> | without saying how you're supposed to determine its value if it's unknown |
| 02:48 | <Hixie> | Atom does the same thing |
| 02:48 | <tantek> | Atom is actually much worse |
| 02:48 | <Hixie> | a bunch of things are required, but there's no suggestion of what to do when they're unknown |
| 02:48 | <tantek> | so many really bad MUSTs on content authoring |
| 02:48 | <tantek> | and so many bad requirements for artificial precision - e.g. RFC3339 |
| 02:49 | <tantek> | e.g. forcing you to specify :00 seconds, even if you don't know/have a value for it from the content/data/authoring |
| 02:49 | <tantek> | forced artificial precision = antipattern |
| 02:49 | <tantek> | another case of, mandating the author to provide a value for something doesn't mean that you actually get meaningful content for it |
| 02:50 | <tantek> | I'm sure you can list examples from the HTML spec for that ;) |
| 02:50 | <tantek> | *HTML4* |
| 02:53 | <Hixie> | actually html4 doesn't do that much |
| 02:53 | <Hixie> | mostly because it requires so little of anything... |
| 02:53 | <Hixie> | kinda the opposite problem :-P |
| 03:00 | Hixie | wonders why he sometimes gets "- unknown class warning" when running his script |
| 03:00 | <Hixie> | i've tried commenting stuff out but i can't work out what is causing it |
| 07:50 | <quuxman> | hm, this is easier than I thought to do the layout with just CSS, but still quite a pain |
| 07:51 | <quuxman> | why does the "margin:0 auto" trick not work if the display is anything but static? |
| 07:54 | <quuxman> | I want the centering trick with "display:inline-block". This is why doing layout with CSS sucks :-P |
| 07:54 | <quuxman> | </bitch> |
| 07:55 | <foolip> | quuxman: use more wrapping divs ;) |
| 07:55 | <quuxman> | specifically what I want is a centered element, with another element to the right of it |
| 07:56 | <quuxman> | but for the centering trick, the element needs to be block, which makes it impossible to put another element next to it |
| 07:56 | <foolip> | uh... put them both in a single container and add padding of the same width as the second element to the left |
| 07:57 | <quuxman> | foolip: ah, good tip, thanks :) |
| 09:56 | <jgraham> | Am I supposed to interpret section 4.5 of WebIDL as meaning that some_host_object.hasOwnProperty(some_idl_attribute) should be true |
| 09:56 | <jgraham> | ? |
| 09:58 | <jgraham> | e.g. document.createElement("div").hasOwnProperty("class") |
| 09:59 | <Hixie> | "class" is a poor example :-) |
| 10:02 | <jgraham> | Pretend that I used a different example then :) |
| 10:03 | <jgraham> | (or to put it differently "oh yes, so it is") |
| 10:05 | <zcorpan> | http://software.hixie.ch/utilities/js/live-dom-viewer/saved/368 - true in opera, safari, chrome; false in firefox |
| 10:06 | <Hixie> | jgraham: sorry, don't know the answer to the question, so i figured nit picking was the next best thing :-P |
| 11:46 | <annevk> | Hixie, when is it dom-window-x and when is it dom-x? |
| 11:46 | <annevk> | Hixie, and when is it dom-Window-x etc. |
| 11:47 | <annevk> | Hixie, it would be nice to agree on a consistent policy for terms so all specs can follow it |
| 12:11 | <Hixie> | annevk: it's dom-window-x when i add new things after 2008, it's dom-x when i add new things before 2008. More or less. |
| 12:11 | <annevk> | fun |
| 12:11 | <Hixie> | (though i'm sure there's lots of exceptions either way) |
| 12:11 | <annevk> | more fun |
| 12:12 | <annevk> | i lowercased names in titles, you seem to do both there too... |
| 12:12 | <Hixie> | yeah, that's more or less the same |
| 12:12 | <Hixie> | these days i try to be consistent with the actual case, so that search and replace is easier |
| 12:13 | <annevk> | mkay |
| 12:14 | <Hixie> | even more inconsistent is how i deal with concepts and algorithms |
| 12:14 | <annevk> | fair enough, though at some point we should try to get some consistency if we want this multi-spec cross-linking thing to work |
| 12:16 | <Hixie> | yeah |
| 12:16 | <Hixie> | agreed |
| 12:17 | <Hixie> | my recommendation would be to leave what we have now as is, but that going forward we use: |
| 12:17 | <Hixie> | dom-InterfaceName-attributeOrMethodName |
| 12:17 | <Hixie> | event-eventname |
| 12:17 | <Hixie> | elementname |
| 12:17 | <Hixie> | attr-elementname-attributename |
| 12:17 | <Hixie> | dom-elementname-attributeOrMethodName |
| 12:18 | <Hixie> | concept-word |
| 12:18 | <Hixie> | InterfaceName |
| 12:18 | <Hixie> | and just use the whole string for long algorithm names |
| 12:18 | <Hixie> | does that work for you? |
| 12:18 | <annevk> | yeah, that's more or less what I use |
| 12:19 | <Hixie> | cool |
| 12:32 | <annevk> | still not getting 9.10 in the Update Manager |
| 12:32 | annevk | wonders what Ubuntu bug he's hitting |
| 12:35 | <annevk> | I only get it when running update-manager -c |
| 12:49 | <virtuelv> | annevk: your last install was LTS, perhaps? |
| 12:50 | <virtuelv> | annevk: look in in Places → Administration → Software Sources → Updates |
| 12:51 | <virtuelv> | does it say "long-term support releases only" at the bottom? |
| 12:52 | <annevk> | no |
| 12:52 | <annevk> | i have 9.04 |
| 12:52 | <annevk> | i'm running the updater now |
| 12:53 | <annevk> | hopefully it'll work out |
| 12:53 | <annevk> | otherwise I have a problem |
| 12:54 | <hsivonen> | my virtualized 9.04 failed to update to 9.10 |
| 12:55 | <hsivonen> | launchpad is too slow to look up the bug number |
| 13:28 | <annevk> | seems to work |
| 13:29 | <annevk> | stuff looks a bit different |
| 13:29 | <annevk> | more gray |
| 13:29 | <annevk> | and some brown |
| 13:36 | <hendry> | annevk: try ArchLinux :-) |
| 13:38 | <annevk> | hmm, https://one.ubuntu.com/ is interesting |
| 13:48 | <zcorpan> | http://livedom.validator.nu/?%3Cdatalist%3E%3Coption%3E%3C%2Fdatalist%3Ex - bug in v.nu or spec? |
| 13:49 | <annevk> | dunno, but hopefully the spec closes <option> on </datalist> |
| 13:51 | <zcorpan> | seems like a bug in v.nu from my reading of the spec |
| 13:52 | <zcorpan> | oh, http://bugzilla.validator.nu/show_bug.cgi?id=676 |
| 13:52 | <annevk> | same reading here |
| 14:01 | <AryehGregor> | Philip`, I think the RDFa changes whitelist xmlns:* with any URL-looking contents. But I doubt we'll end up enabling raw RDFa/microdata input, based on the mailing list discussion. |
| 14:01 | <annevk> | so much for help with CSSOM |
| 14:02 | <annevk> | no answer yet to my query |
| 14:02 | <annevk> | I wonder how often I asked a question on a mailing list and got a completely useful reply |
| 14:24 | <Philip`> | AryehGregor: Okay |
| 14:25 | <Philip`> | Seems like the system tries quite hard to only emit well-formed XML, so that xmlns thing is just something that might trip it up if it's not exceedingly careful |
| 14:32 | <TabAtkins> | How technically feasible is it for browsers to apply special resource limits on nested browsing contexts? Such as cutting an <iframe> off early if it starts looking like it's going to slow down the page? |
| 14:35 | <jgraham> | TabAtkins: I don't have deep insider knowledge, but that sounds quite difficult |
| 14:36 | <TabAtkins> | That's what I suspected as well. I'm just putting feelers out, because if it really is quite difficult, then I'll drop it from my suggestions for @sandbox additions. |
| 14:36 | <Lachy> | TabAtkins, in what ways are you imagining a page within an iframe slowing things down, which wouldn't already be handled by existing techniques to, e.g., halt execution of long running scripts |
| 14:37 | <TabAtkins> | Lachy: Say from an SVG image with multiple chained filters. |
| 14:37 | <Lachy> | hmm, good question. I don't know |
| 14:40 | <TabAtkins> | As well, even with existing long-running script techniques, I may want to cut them off *even faster*, since I'm allowing scripts from untrusted sources. Say after 5 seconds of lag, instead of the 30 or so (if I recall correctly how that works) for a normal page script. |
| 14:40 | <hsivonen> | shouldn't a cloud backup offering talk about crypto on the front page? |
| 14:40 | <hsivonen> | re: ubuntu one |
| 14:48 | <annevk> | hsivonen, I doubt anyone cares |
| 14:49 | <annevk> | though a little info somewhere would be good... |
| 14:50 | <annevk> | seems to lack resources, clicking on FAQ results in an error |
| 14:52 | <AryehGregor> | I tried Ubuntu One but it didn't seem to work properly, so I gave up before actually storing any files on it. |
| 14:53 | <Lachy> | hsivonen, haven't you been looking for an online backup solution for months? Still not having any luck finding a suitable one? |
| 14:54 | <Lachy> | I never found one either, and haven't been actively looking lately |
| 14:56 | <AryehGregor> | Hmm, what happened with H.264 licensing? I guess they aren't going to start charging Internet video streamers yet? |
| 14:57 | <Lachy> | the best solution might be to set up your own remote server, possibly in a co-location facility, with enough storage for your needs that you have complete control over. Though, that might be a little costly. |
| 14:58 | <AryehGregor> | If you're only backing up once a day, you might be able to use 0 Mbps of bandwidth at the 95th percentile, so most of the cost would probably be in the setup. |
| 14:58 | <AryehGregor> | Hardware costs and such. |
| 14:58 | <Lachy> | AryehGregor, yeah. It doesn't solve the problem with h.264 though. They're just delaying it for some reason. All it does is let companies like YouTube and Vimeo save money for a few more years |
| 14:58 | <AryehGregor> | You could probably do it for <$500 for the machine, and $10 or $20 a month for 1U of rack space. |
| 14:59 | <AryehGregor> | Lachy, YouTube isn't saving money AFAIK, Google is already paying the full H.264 licensing fee, so they get to use it however much they want. |
| 14:59 | <Lachy> | I'd want at least 10TB of remote storage |
| 14:59 | <meledin> | Are you backing up the entire internet? |
| 15:02 | <gsnedders> | Peh. The internet is overrated. |
| 15:02 | <AryehGregor> | Well, you can get a 2 TB disk for $160, so 10 TB could be six of those in RAID5, which comes out to $2160 for the disks alone. So okay, a bit more than $500. |
| 15:02 | <Lachy> | meledin, I currently have 7TB worth of storage space on external drives at home, with about 5TB of that full |
| 15:03 | <AryehGregor> | (but RAID5 with 2 TB drives is risky, there's a high chance of read errors on resync, RAID6 would be safer) |
| 15:03 | <TabAtkins> | AryehGregor: Where are you getting your math? 160*6 != 2160 |
| 15:03 | <AryehGregor> | I make it up. |
| 15:03 | <hsivonen> | Lachy: I tried to use braawi.com for backups but I never got the setup process to a point where it would work |
| 15:03 | <Philip`> | 10TB is nearly a year of continuous HD video |
| 15:03 | <TabAtkins> | Anyway = 720 |
| 15:03 | <AryehGregor> | Serves me right for trying to do multiplication in my head instead of typing /exec calc '160*6' |
| 15:03 | <TabAtkins> | I mean... 960 |
| 15:03 | <AryehGregor> | Hahaha. |
| 15:03 | <meledin> | Irony |
| 15:03 | <AryehGregor> | Yeah, 960. |
| 15:03 | <TabAtkins> | At least my error was much closer to reality. |
| 15:04 | <AryehGregor> | Okay, so you could set it up for under $1500, probably. |
| 15:04 | <meledin> | Closest without going over |
| 15:04 | <AryehGregor> | I did 360*6 instead of 160*6 for some reason. |
| 15:04 | <TabAtkins> | Makes sense. |
| 15:05 | <TabAtkins> | I did 120*6. We're all about nice angles. |
| 15:05 | <meledin> | But, yeah, for 10tb of backup storage you are looking at a DIY solution |
| 15:05 | <hsivonen> | I've spent most of my day today waiting for IO. most of it not even network |
| 15:06 | <AryehGregor> | Obviously you should get 10 TB of SSDs instead, then. |
| 15:06 | <jgraham> | 10tb of remote backup does seem somewhat unreasonable |
| 15:06 | <hsivonen> | I chose to prepare my parents' computers for disk failure scenarios today |
| 15:07 | <jgraham> | But I would be interested in solutions that allowed for <~1Tb of remote backup |
| 15:08 | <jgraham> | (without unreasonable cost) |
| 15:09 | <jgraham> | (well maybe 1Tb is an overestimate) |
| 15:09 | <AryehGregor> | Amazon charges $0.15/GB/month for the first 50 TB used, so $150/TB/month. Is that "unreasonable"? :) |
| 15:10 | <Lachy> | AryehGregor, yes. I looked into Amazon previously, and concluded that the costs for large quatities of personal storage are excessive |
| 15:10 | <AryehGregor> | Does anyone else undercut them much? |
| 15:11 | <jgraham> | If 1Tb was an order of magnitude too high that would be reasonable |
| 15:11 | <jgraham> | i.e. I would pay $15/month |
| 15:11 | <AryehGregor> | I mean, $150 is almost enough to buy a 2 TB disk that you can use yourself forever. |
| 15:12 | <hsivonen> | Lachy: life is much easier if you don't possess movies after you've watched them |
| 15:13 | <TabAtkins> | hsivonen: Indeed. I at least hang onto them until my share ratio goes above 3, but then they're gone. |
| 15:13 | <Lachy> | but then I'd have to redownload them to watch them again, or for other people to watch them later |
| 15:13 | <TabAtkins> | Lachy: Netflix. |
| 15:13 | <TabAtkins> | As long as they want to watch things that can be instant-streamed, or give you enough warning for you to order something they'd like. |
| 15:13 | <Lachy> | TabAtkins, DRM, and I don't think they're in Norway. |
| 15:14 | <TabAtkins> | Damn these national borders! |
| 15:14 | <hsivonen> | it's nicer to see movies that you haven't already seen |
| 15:14 | <Lachy> | I can work around the national border thing fairly easiliy. It's the DRM I refuse to pay for |
| 15:14 | hsivonen | doesn't use p2p so share ratio isn't an issue |
| 15:14 | <TabAtkins> | In content that is *inherently* temporary, that's not such a big deal to me. I'm not trying to claim that I own the videos. |
| 15:15 | <Lachy> | hsivonen, do you use Usenet, or do you just not download pirated content? |
| 15:16 | <hsivonen> | Lachy: no pirated content |
| 15:16 | <Lachy> | ok |
| 15:16 | <TabAtkins> | Copyright-infringement. Nobody was kidnapped or murdered in the course of infringing the copyrights. |
| 15:16 | <hsivonen> | well, my in-the-cloud pvr provider is being investigated by the police |
| 15:17 | jgraham | has no copyright infringing content but does have a camera that takes 20Mb/photo photos, which can add up fast |
| 15:17 | <hsivonen> | but not because they provide the wrong service but because they don't have the same media company cronies as their competitors |
| 15:18 | <Lachy> | there was a cable provider in the US who offered an in-the-cloud PVR service, and who went through copyright infringment claims. I believe they won that case last year. |
| 15:18 | <hsivonen> | but they have the best law firm in the field in Finland |
| 15:21 | <AryehGregor> | TabAtkins, come on, when most people hear the word "pirate" they think of a goofy guy with an eyepatch and a parrot on his shoulder, not kidnap and murder. |
| 15:21 | <AryehGregor> | Typically a macaw. |
| 15:21 | <TabAtkins> | AryehGregor: True, but then the lying bastards in the content companies use that connection to tie copyright infringement to *real* piracy and counterfeiting issues. |
| 15:22 | <AryehGregor> | Nobody talks or thinks about "real" piracy these days, it's not the nineteenth century anymore. |
| 15:22 | <TabAtkins> | Go tell that to Somalia. |
| 15:22 | <AryehGregor> | It's an amusing curiosity to hear that there are still real pirates in Somalia. |
| 15:22 | <Lachy> | btw, not all of my stuff is infringing downloads. A significant portion of it is from ripped DVDs that I own, which are on average 5 to 8 GB each. |
| 15:22 | <AryehGregor> | The whole "copyright infringement is not piracy!" thing is overblown, language changes. |
| 15:23 | <AryehGregor> | People often water down existing negative terms by applying them to less negative things. |
| 15:23 | <AryehGregor> | Or add negative connotations to neutral or positive words by applying them to more negative things as euphemisms, on the other hand. |
| 15:23 | <AryehGregor> | There's no reason to get excited about it, "pirate" isn't even an insult at this point. You even have a Pirate Party in the EU. |
| 15:24 | <TabAtkins> | Indeed. That doesn't mean that the negative connotations aren't still twisted to evil effect, though. |
| 15:24 | <AryehGregor> | The word "pirate" *has* no negative connotations. |
| 15:24 | <AryehGregor> | That's what I'm saying. |
| 15:24 | <TabAtkins> | You're... wrong? |
| 15:24 | <AryehGregor> | It's lost them all by now. |
| 15:24 | <AryehGregor> | Or pretty much. |
| 15:24 | <AryehGregor> | If I'm wrong, why do so many pirates call themselves pirates? |
| 15:24 | <foolip> | arr arr, welcome to Sweden, matey |
| 15:24 | <AryehGregor> | The Pirate Party, Pirate Bay, . . . |
| 15:24 | <TabAtkins> | Because they've internalized the language? |
| 15:25 | <AryehGregor> | Which means it has no negative connotation anymore. |
| 15:25 | <TabAtkins> | Not at all. |
| 15:25 | <TabAtkins> | It doesn't have negative connotations *to them*. |
| 15:26 | <AryehGregor> | It doesn't have more than slightly negative connotations to anyone these days. If it has any, it's like "jaywalker" or "litterer" rather than "murderer" or "kidnapper". |
| 15:26 | <TabAtkins> | It's like the word "fag". It's 'reclaimed', but that doesn't mean that your average 12-year-old's use of the word isn't insulting. |
| 15:26 | <AryehGregor> | No, the word "fag" has a uniformly negative connotation when applied to people. You don't have gay rights groups calling themselves "Society for Empowerment of Fags". |
| 15:27 | <TabAtkins> | No, but you do have gay *people* referring to themselves as fags, without offense. |
| 15:27 | <AryehGregor> | I haven't ever heard that. |
| 15:27 | <AryehGregor> | (not that I know many gay people) |
| 15:27 | <TabAtkins> | You hang around the wrong homosexuals. ^_^ |
| 15:28 | <AryehGregor> | I don't hang around homosexuals at all, usually. |
| 15:29 | <AryehGregor> | Anyway, sometimes people insult themselves as a joke, that's entirely different. |
| 15:29 | <TabAtkins> | Nah, the way it's used isn't meant to be insulting or a joke. It's just an in-group term. |
| 15:30 | <workmad3> | it's just really what the intent behind the word is |
| 15:30 | <AryehGregor> | Well, I can't do a sociological analysis of groups I'm not affiliated with, but I can say that to me, the word "pirate" has no negative connotation. Maybe to some groups (older people?) it does, but if so, it's not worth objecting. |
| 15:31 | <TabAtkins> | It is when those old people are still making laws. >_< |
| 15:31 | <AryehGregor> | No, it's really not. It's pointless to bicker over terminology. Just use whatever terms you prefer. |
| 15:31 | <TabAtkins> | And the digital revolution is recent enough that it'll take quite a while for everyone dumb to die off. |
| 15:31 | <AryehGregor> | It's not like the word is misleading, it's just derogatory at worst. |
| 15:32 | <workmad3> | AryehGregor: probably a good example for intent is 'hacker' |
| 15:32 | <AryehGregor> | Everyone calls people they don't like derogatory things. |
| 15:32 | <workmad3> | to me, it's not derogatory and in some ways complimentary to be a hacker |
| 15:32 | <Philip`> | Pirates of the Caribbean taught me that pirates are cool |
| 15:32 | <Philip`> | (except for the ones that are undead, and/or are hanged) |
| 15:32 | <workmad3> | Philip`: for me it was Pirates of Silicon Valley :P |
| 15:32 | <TabAtkins> | AryehGregor: Yes, yes it *is* misleading. Nothing is even being stolen, for gods sakes. The Supreme Court made it quite clear that copyright infringement is not theft. |
| 15:33 | <TabAtkins> | And then combine it with the fact that real-life pirates use machine guns to steal boats, take hostages, and kill people, and you have legitimate confusion in the uninformed that prevents them from understanding why that "pirating" thing should be legal. |
| 15:34 | <daedb> | Philip`: Same, except it was Monkey Island for me :) |
| 15:34 | <Philip`> | Apparently pirates rarely kill people, because it's not at all profitable |
| 15:34 | <TabAtkins> | Well, sure. But that doesn't make them any better. ^_^ |
| 15:34 | <Philip`> | Better to keep them as hostages as long as possible, and if necessary send an itemised bill as part of the ransom demand |
| 15:35 | <AryehGregor> | TabAtkins, it's not misleading because no one is saying that pirated software is literally a crime of piracy. It's an extension of the word by analogy. |
| 15:35 | <AryehGregor> | And you don't need the Supreme Court to tell you that copyright infringement is not theft under U.S. law, just look at the definitions in the relevant statutes. |
| 15:35 | <AryehGregor> | But it can still be called "theft" in a loose, non-legal sense. |
| 15:35 | <TabAtkins> | AryehGregor: You're naive - people *do* think that actual piracy, as in murder-on-the-high-seas, is precisely what is talked about when talking heads on the news go on about the latest efforts by the RIAA to stop software piracy. |
| 15:36 | <AryehGregor> | TabAtkins, who thinks that? |
| 15:36 | <TabAtkins> | Real people? |
| 15:36 | <AryehGregor> | Give me specific examples. |
| 15:36 | <TabAtkins> | People I have met? |
| 15:36 | <AryehGregor> | Quotes? |
| 15:36 | <TabAtkins> | You want specific examples of people living in Houston? |
| 15:36 | <AryehGregor> | I want proof that someone actually believes that. |
| 15:36 | <AryehGregor> | Preferably not based on your vague recollection. |
| 15:36 | <AryehGregor> | Because I don't believe it. |
| 15:36 | <TabAtkins> | You won't get it. ^_^ |
| 15:37 | <Philip`> | There's always someone that believes anything |
| 15:37 | <foolip> | clearly *someone* must believe that |
| 15:37 | <foolip> | doesn't really say much, though |
| 15:37 | <Philip`> | (Not a single person for all things, they'd get very confused) |
| 15:37 | <AryehGregor> | Yes, but if you can prove that one person does without expending exorbitant resources, it's likely that a *significant* number of people do. |
| 15:37 | <TabAtkins> | Same thing with using "theft", except probably worse. *Many* more people believe that infringement is literal theft, because of people using that word to describe it. |
| 15:38 | <TabAtkins> | Which makes it *enormously* more difficult to teach them about infinite goods and how they violate expectations based on a scarcity economy. |
| 15:39 | <AryehGregor> | It is arguably literal theft in a non-legal sense. |
| 15:39 | <TabAtkins> | And worse, how to *benefit* from the ways in which infinite goods are different. |
| 15:39 | <AryehGregor> | "To take (the property of another) without right or permission." |
| 15:39 | <TabAtkins> | No. No it is not. In any way. |
| 15:39 | AryehGregor | sighs |
| 15:39 | <TabAtkins> | To take, sure. Who's taking anything? |
| 15:39 | <AryehGregor> | You're construing the meaning of all these words very narrowly, according to your specific viewpoint. |
| 15:39 | <TabAtkins> | And who's property is being 'taken' anyway? I'm sharing my own property with friends. |
| 15:39 | <Philip`> | There's lots of products where all the expense is up-front and the manufacturing cost is relatively tiny, so it's not significantly different once the manufacturing cost reaches zero |
| 15:40 | <AryehGregor> | Other people have different and equally legitimate interpretations of words like "theft". |
| 15:40 | <Philip`> | s/all/most of/ |
| 15:40 | <AryehGregor> | You really should acknowledge that fact. |
| 15:40 | <TabAtkins> | Philip`: Yes and no. There's still a big mental difference between something being free and something costing a nickel. |
| 15:40 | <AryehGregor> | Philip`, distribution/copying costs for software are nonzero too, they're just very low. |
| 15:40 | <AryehGregor> | But anything where all the expense is up-front is, basically, intellectual property. |
| 15:41 | <TabAtkins> | AryehGregor: I challenge those interpretations as incorrect, because they do *not* correspond with the legal meanings of those terms, and because they're mind-blockers. |
| 15:41 | <Philip`> | TabAtkins: There's not a big economic difference, so it's not like we're transitioning to a completely novel system of economics once part of the cost of a product drops a lot |
| 15:41 | <TabAtkins> | If you're convinced that filesharers are 'stealing' from you, you'll go to significant effort to stop them, instead of embracing it and using it to make more money. |
| 15:41 | <AryehGregor> | TabAtkins, "theft" is not purely a legal term. It has a less precise, nonlegal meaning as well. And sure, the language being used is prejudiced, but you're just trying to prejudice the language used in the opposite direction. |
| 15:42 | <AryehGregor> | Which is fine, but it's no more or less legitimate except insofar as you may happen to be more correct. |
| 15:42 | <TabAtkins> | TabAtkins: Economic difference, no. Mental, yes. People often have a hard time adjusting their mental models when their 'product' suddenly becomes the advertisement, because the 'customers' are willing to share it with their friends for free, using their own bandwidth. |
| 15:43 | <AryehGregor> | Now he's talking to himself. |
| 15:43 | <TabAtkins> | Dammit! |
| 15:43 | <TabAtkins> | s/TabAtkins/Philip`/ |
| 16:18 | <Lachy> | Hixie, there seems to be a design flaw with the way sandboxed iframes work that make it largely inappropriate for some use cases. |
| 16:19 | <Lachy> | I realised the issue as a result of this article http://angryamoeba.co.uk/singlecell/2010/01/15/html5-video-and-custom-players.html |
| 16:20 | <Lachy> | Use case: For embedding a custom video player from a video site like YouTube, where the iframe's content page handles all the issues like <source> selection, custom player controls, etc. It's useful to have a simple, copy and pasteable piece of markup for authors to use in their own pages |
| 16:22 | <Lachy> | but using <iframe src="http://youtube.com/path/to/videoplayer.html?videoid"; sandbox seamless> doesn't provide any sandboxing in legacy browsers |
| 16:23 | <Lachy> | And being a 3rd party site, there's no way to use the text/html-sandboxed mime type and srcdoc is inappropriate here where the content needs to be served remotely |
| 16:24 | <boogyman> | ew @ iframe! |
| 16:24 | <Lachy> | That article I mentioned above says that forum software doesn't usually permit iframes to be embedded either, and it's going to be difficult to convince them to do that while there are still legacy browsers out there that would be vulnerable without sandboxing |
| 16:24 | <Lachy> | boogyman, there's nothing inherently wrong with an iframe. It just has a few security issues for these use cases. |
| 16:25 | <boogyman> | Lachy: server side includes FTW |
| 16:25 | <Lachy> | boogyman, and if you're trying to claim that iframe is somehow disgusting like the author of that article did, then you're argument is no good |
| 16:25 | <boogyman> | frames are so 1996-8 |
| 16:25 | <Lachy> | how would server side includes address this use case at all? |
| 16:26 | <Lachy> | boogyman, <frameset> and <frame> are bad by design. <iframe> is not if used appropriately |
| 16:26 | <AryehGregor> | Lachy, plenty of forums allow whitelisted YouTube links. |
| 16:26 | <AryehGregor> | Like [youtube]some URL[/youtube] which regexes out the video id and outputs the right <object> boilerplate. |
| 16:27 | <Lachy> | in fact, even <frameset> isn't bad if used properly, though it frequently isn't and has few valid use cases |
| 16:28 | <Lachy> | AryehGregor, nice. But that doesn't allow other video sites to compete very easily by offering their own videos to embed in 3rd party sites |
| 16:28 | <AryehGregor> | Sure, but are forums really going to whitelist <iframe seamless sandboxed>? Maybe, but I have my doubts. It can do a lot of things, particularly performance-wise. |
| 16:28 | <Lachy> | I assume they don't also have tags for [vimeo], [dailymotion], [insert-other-video-site-here], etc. |
| 16:29 | <AryehGregor> | The correct long-term solution is to do direct <video> inclusion, surely. |
| 16:29 | <Lachy> | that doesn't allow custom controls |
| 16:29 | <AryehGregor> | So? It'll still work. |
| 16:29 | <AryehGregor> | Maybe if this is your use-case, you should talk to actual forum software developers and ask what they'd allow. |
| 16:29 | <AryehGregor> | phpBB probably has a mailing list somewhere. |
| 16:30 | <workmad3> | phpBB you'd just define bbcode tags with the correct embed syntax for the sites you want to allow |
| 16:30 | <Lachy> | video sites like to use custom controls, their own branding and additional features like links to related videos, etc. All of these features are offered by flash, and can be replicated using <iframe>, but not with <video> alone |
| 16:30 | AryehGregor | is betting <iframe sandbox> is not going to be allowed anywhere by default without whitelisting of sites. |
| 16:30 | <AryehGregor> | ("anywhere" meaning "in any forum software package") |
| 16:31 | <AryehGregor> | They can add the branding to the video as a watermark or clips shown before/after. Beyond that, sure, I agree, but I suspect forum developers won't. |
| 16:31 | <AryehGregor> | Unrestricted <iframe sandbox> is kind of scary even if browsers support it. |
| 16:32 | <Lachy> | what's scary about it? |
| 16:33 | <AryehGregor> | It allows arbitrary scripts to run on the page, just theoretically they can't do anything that actually steals data. |
| 16:33 | <AryehGregor> | They can use tons of CPU, for one thing, making it slow to browse. |
| 16:33 | <AryehGregor> | Can it do alert() or other things like that? |
| 16:33 | <AryehGregor> | If no, are you really sure you've gotten all of them? |
| 16:34 | <AryehGregor> | Apparently no. |
| 16:34 | <AryehGregor> | "scripts are still prevented from creating popups" |
| 16:35 | <AryehGregor> | What if they trigger an endless loop of HTTP authentications, say, in a browser where those are modal dialogs? |
| 16:36 | <AryehGregor> | Or any number of other obnoxious things. |
| 16:36 | <TabAtkins> | Surely a modal dialog counts as a popup for those purposes? |
| 16:37 | <AryehGregor> | I guess so. I've actually gotten modal dialogs popping up in Firefox and blocking page load because someone used a signature image from a badly-configured server. |
| 16:37 | <TabAtkins> | Yeah, me too. |
| 16:37 | <Lachy> | well, if sandbox doesn't address those issues and they're real problems, then perhaps it should address them |
| 16:37 | <AryehGregor> | Not to mention someone using a signature image (which was even a 404) from a domain that Chrome was convinced was evil incarnate, so all visitors to some threads got a giant terrifying "THIS SITE MIGHT DESTROY YOUR COMPUTER!!!" message. |
| 16:37 | <TabAtkins> | Agreed. There's a thread over in whatwg that would be relevant to bring them up in. |
| 16:38 | <TabAtkins> | Presumably the warning message will be intelligent wrt sandbox. |
| 16:38 | <AryehGregor> | http://code.google.com/p/chromium/issues/detail?id=20306 |
| 16:38 | AryehGregor | glares |
| 16:39 | <AryehGregor> | Realistically, forums tend to have lots of moderators who can edit this stuff out on a case-by-case basis, so as long as it's just annoying at worst and not a security problem, it might be tolerable. |
| 16:39 | <AryehGregor> | But I'm not the best one to ask, you should be asking phpBB devs or something, if that's your use-case. |
| 16:40 | <AryehGregor> | MediaWiki won't allow this by default for totally different reasons (privacy, etc.). |
| 17:19 | <gsnedders> | hsivonen: http://software.hixie.ch/utilities/js/live-dom-viewer/?%3C!DOCTYPE%20html%3E%0A%3Ca%20href%3D%22%23a%25252A%22%3E%3C%2Fa%3E%0A%3Cscript%3E%0Avar%20a%20%3D%20document.getElementsByTagName%28%22a%22%29%3B%0Aw%28a.length%29%3B%0A%3C%2Fscript%3E |
| 17:19 | <gsnedders> | (see in Minefield) |
| 17:26 | <gsnedders> | http://software.hixie.ch/utilities/js/live-dom-viewer/?%3C!DOCTYPE%20html%3E%0D%0A%3Ca%20href%3D%22%23a%25252Ax%22%3Ex%3C%2Fa%3E%0D%0A%3Cscript%3E%0D%0Avar%20a%20%3D%20document.getElementsByTagName%28%22a%22%29%3B%0D%0Aw%28a[0].hash%29%3B%0D%0Aw%28a[0].hash.length%29%3B%0D%0A%3C%2Fscript%3E is interesting in Minefield |
| 17:26 | <gsnedders> | Seemingly runs decodeURI on .hash |
| 17:26 | <gsnedders> | Which breaks Google Wave on the Fx codepath (the default of the bypass browser sniffing) in everything else |
| 19:26 | <AryehGregor> | Oh, it seems like there is a notifications spec: http://dev.w3.org/2006/webapi/WebNotifications/publish/ |
| 19:26 | <boogyman> | updated today lolz |
| 19:26 | <boogyman> | :P |
| 19:26 | <AryehGregor> | February 3 is yesterday. |
| 19:28 | boogyman | forgot i didn't sleep last night D'oh! |
| 20:03 | <virtuelv> | Do common desktop notification systems (read: Growl and the ones in Ubuntu/Gnome even support what that notification spec implies |
| 20:05 | <opdynamiks> | probably. The spec sounds like the think that Google Gears was doing with notifications (to alert the user when they were in a different App or Window) |
| 20:15 | <virtuelv> | I'm not sure |
| 20:16 | <virtuelv> | the HTML notifications seem like complete overkill |
| 20:17 | <virtuelv> | and I'm pretty certain neither Growl nor Ubuntu notifcations support it |
| 20:23 | <tantek> | recommend that those here who design/develop real world websites participate in this survey: http://surveys.webdirections.org/index.php?sid=69882 (has some good questions on HTML5) |
| 20:25 | <gratz|home> | heh |
| 20:48 | <opdynamiks> | virtuelv: I don't think it is overkill... and I'm not sure if it is technically considered HTML5 since it generally is a set of JavaScript functionality that ties in to the user's browser system (i.e. task tray on Windows and Gnome, notifications bar on Android) |
| 20:48 | <opdynamiks> | virtuelv: although such a thing may proove to be super annoying to users (think: advertising notifications popping up) |
| 21:13 | <FireFly> | opdynamiks, well, the browser could force you to accept or deny it on a per-host basis... I mean, I can see web apps like gmail sending notifications when you receive new mail being pretty useful |
| 21:15 | <opdynamiks> | FireFly: yeah thats true, that could easily be a browser setting (default: on or off for new sites). |
| 21:16 | <opdynamiks> | I could see it being cool for media player web apps (MP3 and streaming radio) displaying track information on song change |
| 21:16 | <opdynamiks> | does anyone know if the drag N drop thing in HTML5 would make drag N drop file uploads possible? |
| 21:31 | <opdynamiks> | nevermind, found that it does indeed work with file uploads. have to fool around with that tonight. can't wait until it becomes a reality |
| 21:56 | <jgraham> | I think notifications are generally pretty annoying to begin with. I certianly don't want full HTML notifications (meaning they could include audio, video, flash, whatever) |
| 22:02 | <opdynamiks> | No, I agree that they should be text only, maybe bold supported. |
| 22:20 | zcorpan | searches for "html5" in the flashplayer bts |
| 22:40 | <roc> | opdynamiks: drag-and-drop file upload is supported in Firefox 3.6 |
| 22:41 | <opdynamiks> | roc: really, no earlier huh. I'll have to try it out |
| 22:42 | <roc> | http://hacks.mozilla.org/2009/12/file-drag-and-drop-in-firefox-3-6/ |
| 23:18 | <Hixie> | Lachy: seamless can't work cross-site anyway, btw. But why do you need sandbox="" in that example? |
| 23:35 | <Sonja> | hi |
| 23:36 | <Sonja> | when will html5 become official? |
| 23:36 | <jonnybarnes> | what do you mean by official? |
| 23:38 | <Sonja> | ohhh "Ian Hickson, editor of the HTML5 specification, expects the specification to reach the W3C Candidate Recommendation stage during 2012, and W3C Recommendation in the year 2022 or later" |
| 23:38 | <jonnybarnes> | well, the requiremtns of it being a W3C recommendation is actually quite sctrict as far as i recall |
| 23:39 | <Sonja> | the new equivalent of Flash and silverlight will be canvas? |
| 23:39 | <TabAtkins> | Strict enough that HTML4 isn't a Rec under the current definition. ^_^ |
| 23:39 | <jonnybarnes> | html5 is quite well supported all ready |
| 23:39 | <jonnybarnes> | *already |
| 23:39 | <jonnybarnes> | particularly in modern browsers |
| 23:39 | <TabAtkins> | Sonja: <canvas> is meant to replace some of the things that Flash/Silverlight does. Some things that are currently done with Flash are solved more simply using other parts of HTML. |
| 23:39 | <jonnybarnes> | indeed, as TabAtkins says |
| 23:40 | <Sonja> | can you recommend a neat html5 test page to see it in action and test my browser? |
| 23:40 | <jonnybarnes> | for <canvas>, or html5 in general? |
| 23:40 | <Sonja> | i guess both ? |
| 23:40 | <Sonja> | or either |
| 23:41 | <jonnybarnes> | link for canvas game: well blog post about, it contains link for games |
| 23:41 | <jonnybarnes> | http://blog.nihilogic.dk/2008/04/super-mario-in-14kb-javascript.html |
| 23:41 | <Sonja> | "other parts of HTML". do you mean like javascript? |
| 23:42 | <jonnybarnes> | html5 video demo, needs safari or google chrome though, http://jilion.com/sublime/video |
| 23:42 | <TabAtkins> | In part, yes. But also, say, internet video is done with <video> now. |
| 23:42 | <jonnybarnes> | thats less an html5 issue, more a what video codec will the browser support |
| 23:43 | <Sonja> | so do you thnk youtube and such sites will migrate to a new format? |
| 23:43 | <Sonja> | using <video> and such? |
| 23:43 | <TabAtkins> | They already have, in fact! |
| 23:44 | <Sonja> | neat |
| 23:44 | <jonnybarnes> | yeah, youtube.com/html5 is a demo for it i think |
| 23:44 | <Sonja> | i'm guessing ff will support <video> soon enough |
| 23:45 | <Sonja> | or is mozilla trying to kill html5 video? |
| 23:45 | <jonnybarnes> | its an issue with what codec to use |
| 23:46 | <jonnybarnes> | when you use the <video> tag the browser decodes the video natively |
| 23:46 | <jcranmer> | ff supports <video> since 3.5, if not 3.0 itself |
| 23:46 | <jonnybarnes> | sites like youtube are using a codec caleed h264 |
| 23:46 | <TabAtkins> | Mozilla *definitely* wants the <video> tag. But they want it with codecs that don't require royalties, so their users are safe. |
| 23:46 | <jonnybarnes> | very good, but licensed(sp?) |
| 23:46 | <jcranmer> | just youtube doesn't support FF's video |
| 23:47 | <Sonja> | h264 is not royalty-free? |
| 23:47 | <jonnybarnes> | mozilla won't5 pay the money for financial and politicaln reasons |
| 23:47 | <jcranmer> | bwahahaha no |
| 23:47 | <jonnybarnes> | nope |
| 23:47 | <jonnybarnes> | MPEG-LA own the license |
| 23:47 | <jcranmer> | it requires like $1M+ in fees, and I don't know if that's per-annum |
| 23:47 | <Sonja> | i'm reading up now.... http://weblogs.mozillazine.org/roc/archives/2010/01/video_freedom_a.html |
| 23:48 | <TabAtkins> | Yeah, ROC's post is a good one. |
| 23:48 | <jonnybarnes> | yeah, thast a good article about it |
| 23:48 | <jcranmer> | I'm willing to bet that if MS implements <video> in IE 9, it won't support H.264 for a large number of computers |
| 23:48 | <roc> | TabAtkins: not just users, but content providers and everyone else |
| 23:48 | <jonnybarnes> | why not? what would MS use instead? |
| 23:48 | <roc> | jcranmer: more like $5M for us |
| 23:48 | <TabAtkins> | roc: Sorry, was meaning the term broadly like that. ^_^ |
| 23:48 | <roc> | per year |
| 23:49 | <jcranmer> | jonnybarnes: I believe MS only includes the H.264 codec with Win 7 |
| 23:49 | <jcranmer> | roc mentioned the version at some point |
| 23:49 | <jonnybarnes> | jcramnmer: fair enough |
| 23:49 | <Sonja> | maybe this will make people switch from ff to chrome |
| 23:49 | <roc> | jcranmer: that is correct |
| 23:49 | <jcranmer> | that's the oddest misspelling of my name that I've seen |
| 23:50 | <roc> | maybe it will |
| 23:50 | <jcranmer> | I doubt it |
| 23:50 | <jonnybarnes> | lol, shows how tired i am |
| 23:50 | <jonnybarnes> | as long as flash exists as a faalback |
| 23:50 | <jonnybarnes> | most users wont noptice a difference |
| 23:50 | <jcranmer> | most users don't care about HTML 5 video support |
| 23:50 | <Sonja> | it defeats the purpose of having html5 to go back to flash tho |
| 23:51 | <jcranmer> | and many of those that do are likely to be evangelist for Theora |
| 23:51 | <roc> | currently, that's true, but that could change |
| 23:51 | <Sonja> | one of the appeals of html5 for me is that i won't need to use 3rd party thingy from adobe |
| 23:51 | <Sonja> | and i can use some sort of open format |
| 23:51 | <Sonja> | or maybe i'm misunderstanding what html5 will do |
| 23:52 | <roc> | no, that's totally true |
| 23:53 | <Sonja> | <acronym> is gone? i guess <abbr> is used instead ? |
| 23:54 | <Sonja> | i was using it with title="spell it out in full" |
![http://example.com/test.png"></object>](http://example.com/test.png"></object>){kind=link}