| 00:17 | <Hixie> | daedb: if it's not too hard to get, sure |
| 00:21 | <daedb> | Hixie: I have one. It's not gonna be readable if it's scaled down to 400px width though (it's a 1680*1050 image). It can be scaled down somewhat and still be readable, but not down to 400px. |
| 00:22 | <Hixie> | that's ok, the shrunk down images are just to remind me of what they are |
| 00:25 | <daedb> | http://daed.site24.se/Screenshot20100420222809422.jpg is it. |
| 00:31 | <TabAtkins> | daedb: You could probably alter the image to work, by taking the full-size text and just transporting it into a scaled-down image. |
| 00:34 | <Hixie> | daedb: thanks |
| 00:34 | <daedb> | TabAtkins: http://daed.site24.se/dao_sub_cropped.png <-- like this? :) |
| 00:35 | <TabAtkins> | daedb: Yes. |
| 01:07 | <Hixie> | nessy: i've split it again, now there's three pages :-) |
| 01:10 | <Hixie> | nessy: is there a word to refer to a single segment of a timed track? whatever unit of text has a start and end time? |
| 01:10 | <nessy> | I've started calling them text cues |
| 01:11 | <Hixie> | cue, that works |
| 01:11 | <Hixie> | thanks |
| 01:16 | <nessy> | sorry, was on the phone |
| 01:17 | <nessy> | what's the third page now? |
| 01:17 | <Hixie> | one is use cases for timed tracks rendered over video by the UA, the other is use cases for API-level access to timed tracks. |
| 01:17 | <Hixie> | the main page has no images on it anymore |
| 01:18 | nessy | goes checking it out |
| 01:19 | <nessy> | ah I see |
| 01:19 | <nessy> | the links to the other pages are there, too, which got me confused yesterday |
| 01:19 | <Hixie> | links to the other pages? |
| 01:28 | <Hixie> | nessy: should a UA always grab all the tracks from the network, or should it only grab tracks when they're enabled? |
| 01:29 | <nessy> | external files: only when they are enabled |
| 01:29 | <Hixie> | interesting |
| 01:29 | <nessy> | save bandwidth where you can, I'd say |
| 01:29 | <nessy> | why? |
| 01:30 | <Hixie> | so we really need three modes when we have the API... on and displayed, on but only for hte API, and off |
| 01:30 | <nessy> | what do you mean by "on but only for the API"? |
| 01:30 | <Hixie> | suppose one of the timed tracks is a list of IDs of slides and times for when the slides turn on and off |
| 01:31 | <Hixie> | you wouldn't want the IDs actually shown on the video (or even listed in the UI) |
| 01:31 | <Hixie> | but you'd still want then "enabled" and triggering events for the script |
| 01:31 | <Hixie> | s/then/them/ |
| 01:32 | <nessy> | are you saying that external tracks should trigger events even if they are not active? |
| 01:32 | <Hixie> | what do you mean by active? |
| 01:32 | <Hixie> | (if they're triggering events, presumably they are active) |
| 01:32 | <nessy> | hmm… I guess there is a difference between being loaded and being displayed |
| 01:34 | <nessy> | maybe this has to be done with two different attributes just like on the media elements |
| 01:34 | <nessy> | something akin to @controls for being displayed and something akin to @preload for being "available" |
| 01:35 | <nessy> | though those names of course don't match |
| 01:35 | <Hixie> | i can't see any good way to do it other than something in attributes, yes |
| 01:35 | <Hixie> | but for now i'm just trying to work out requirements |
| 01:36 | <Hixie> | so that i can go through your proposals |
| 01:36 | <nessy> | sure - my brain jumped ahead |
| 01:36 | <nessy> | there is some need for turning on and off UA display |
| 01:37 | <nessy> | and there is some need for enabling / disabling a track and its events |
| 01:38 | <nessy> | it may even make sense to have that fourth state: displaying, but no events triggered |
| 01:38 | <Hixie> | oh? |
| 01:39 | <nessy> | no? |
| 01:39 | <nessy> | if the Web developer knows they don't need the events why should the UA have to make the effort? |
| 01:40 | <othermaciej> | nessy: no events of what kind? |
| 01:40 | <othermaciej> | nessy: like DOM events for JavaScript? |
| 01:40 | <Hixie> | nessy: an event every few seconds isn't a big deal |
| 01:40 | <othermaciej> | nessy: browsers can optimize to avoid dispatching the event at all if there are no listeners |
| 01:40 | <nessy> | we're just talking about the cues from external text resources for media elements |
| 01:40 | <othermaciej> | if it is a performance issue to dispatch it |
| 01:41 | <nessy> | ok - I was just thinking through the logical possibilities and whether three states will do or whether there are two independent dimensions |
| 01:42 | <othermaciej> | what are the three states being considered? |
| 01:45 | <nessy> | on and displayed, on but only for hte API, and off |
| 01:47 | <Hixie> | nessy: anything look obviously wrong on http://wiki.whatwg.org/wiki/Timed_tracks#Architecture ? |
| 01:49 | <nessy> | checking … did you consider tracks inside the video resource? |
| 01:50 | <Hixie> | oh, forgot about those, good point |
| 01:51 | <nessy> | they are essentially the same, but linked through the media resource |
| 01:51 | <Hixie> | (reload) |
| 01:54 | <Hixie> | bbiab - if you have any other comments feel free to mention them here, i'll look at backlog |
| 01:54 | <Hixie> | next stage is to go through your proposals |
| 01:54 | <Hixie> | to learn what parts of this they cover |
| 01:54 | <nessy> | looks ok |
| 01:55 | <nessy> | our proposals are rather bare on the javascript and rendering side of things - but solve the top, basically |
| 01:55 | <nessy> | I wonder about one thing: what if the timed text track has styling information? |
| 01:56 | <nessy> | CSS and Web Fonts cannot be the only part to trigger styling - it might come from the text resources |
| 01:59 | <nessy> | looks like you are solving the cue ranges problem as well with this ;) |
| 02:08 | <Hixie> | yeah hopefully we can hit cue ranges as well, that would be nice |
| 02:09 | <Hixie> | dunno what to do with the styling thing... are we sure we don't just want it to be CSS? |
| 02:09 | <Hixie> | i mean most systems don't let the author style the captions at all, from what i've seen |
| 02:09 | <Hixie> | e.g. TVs these days put that under teh control of the user |
| 02:10 | <Hixie> | and HTML itself has moved towards putting all styling in CSS |
| 02:10 | <Hixie> | it would be weird to make subtitles an exception |
| 02:10 | <Hixie> | but i don't know, depends what the use cases are |
| 02:58 | <nessy> | some of the formats have styling |
| 03:01 | <nessy> | like the QTText example that I emailed: |
| 03:01 | <nessy> | {QTtext} {size:16} {font:Lucida Grande} {width:320} {height:42} |
| 03:01 | <nessy> | {language:0} {textColor:65535,65535,65535} {backColor:0,0,0} |
| 03:01 | <nessy> | {doNotAutoScale:off} {timeScale:100} {timeStamps:absolute} |
| 03:01 | <nessy> | {justify:center} |
| 03:01 | <nessy> | but often this is ignored |
| 03:02 | <nessy> | where styling is on the elements rather than on the full text cue, we cannot simply ignore it I think |
| 03:02 | <nessy> | formats is a whole problem by itself and I am tending towards developing this new one that we have been discussing on WHATWG and then mapping everything else to that |
| 03:17 | <nessy> | or look at SSA (substation alpha) format, see http://en.wikipedia.org/wiki/SubStation_Alpha |
| 03:17 | <nessy> | probably the most comprehensive subtitling format around |
| 03:23 | <nessy> | found this positioning-interesting subtitle http://img3.imageshack.us/img3/6743/vlcsnap090208083027134xq5.png |
| 03:25 | <annevk> | vertical text strikes back |
| 03:25 | <annevk> | aaah |
| 03:27 | <Hixie> | nessy: i don't think we need to map everything to a single format; if the UA supports two formats, it can just support them both, that seems fine |
| 03:27 | <nessy> | ok, though I think there may be more than 2 in the future seeing all the different formats that exist |
| 03:27 | <nessy> | just found this, btw http://img254.imageshack.us/ifs/7581/img410/3/fabulousab2.png |
| 03:27 | <nessy> | has coloured text done in SSA and Matroska - want it on the page, too? |
| 03:28 | <Hixie> | sure |
| 03:28 | <Hixie> | karaoke i assume? |
| 03:28 | <Hixie> | wait, that's a demo of SSA and Matroska |
| 03:28 | <Hixie> | we don't want demos |
| 03:28 | <Hixie> | only things that people actually use |
| 03:28 | <Hixie> | the idea is to find out what the use cases really are |
| 03:29 | <Hixie> | i added some more thoughts here: http://wiki.whatwg.org/wiki/Timed_tracks#Declaring_timed_tracks |
| 03:29 | <Hixie> | based on your proposals |
| 03:29 | <Hixie> | a few questions arising: |
| 03:29 | <Hixie> | - what's the use case for media="" on timed tracks? |
| 03:29 | <Hixie> | - how do we handle sign language video overlays? |
| 03:30 | <Hixie> | - how do we handle alternative audio tracks? |
| 03:30 | <nessy> | not a demo - SSA and Matroska are not demos, they are fully in use |
| 03:30 | <Hixie> | i gotta go for dinner, but if you have any thoughts on the above, don't hesitate to mention them |
| 03:30 | <Hixie> | SSA and Matroska are not demos, but that file is a demo of SSA and Matroska |
| 03:30 | <Hixie> | if the filename is any indication |
| 03:30 | <Hixie> | anyway, gotta go, bbiab |
| 03:31 | <nessy> | enojy your dinner |
| 03:31 | <nessy> | @media is supposed to be what it is for video: help filter a file for a particular device |
| 03:32 | <nessy> | for example if some caption file is written for a specific width, but wouldn't work on mobile, then there might be a max-width on it |
| 03:33 | <nessy> | as for sign language videos, they are the same problem as actual audio descriptions (the ones that are recorded voice) |
| 03:33 | <nessy> | they both are not time-aligned text, but time-aligned other media resources |
| 03:34 | <nessy> | the browser vendors have all told me that they don't want to implement such media synchronisation at this point in time, in particular if they come from different servers |
| 03:34 | <nessy> | (btw: did you see my test at http://blog.gingertech.net/2010/02/12/audio-track-accessibility-for-html5/ ?) |
| 03:35 | <nessy> | however, the idea is that the <track> element will allow for such media files in the future, too |
| 03:35 | <nessy> | so, the proposal already covers it |
| 07:45 | <Hixie> | nessy: do sign language files and recorded audio descriptions have built-in timings? |
| 08:15 | <zcorpan> | Hixie: hmm. should the binary frame length parsing algorithm bail out if the first length byte is 0x80? |
| 08:17 | <annevk> | meh, CORS is complex |
| 08:17 | <annevk> | I should have said no way more often to the Mozilla guys... |
| 08:19 | hsivonen | doubts about all the anime fansubbing effects being hard requirements |
| 08:20 | <annevk> | HTML and CSS should address most of it |
| 08:20 | <Hixie> | yes let's put <video> in the subtitles! nothing could go wrong there |
| 08:21 | Hixie | points annevk to <annevk> I should have said no way more often to the Mozilla guys... |
| 08:21 | <Hixie> | zcorpan: the server guys are most likely to just ignore us on that front, and i don't see why the client guys would care about a 0x80 prefix |
| 08:22 | <annevk> | the comparison with CORS would be more to allow HTML/CSS but require a bunch of extra syntax to enable all their functionality |
| 08:22 | <annevk> | rather than not allowing HTML/CSS |
| 08:24 | <Hixie> | the needs i've seen so far are to allow ruby and italics. that's it. making subtitles use all of html and css is so far beyond overkill i don't even know where to begin. |
| 08:26 | <othermaciej> | Hixie: <video> in <video> subtitles sounds like awesome demo fodder |
| 08:26 | <othermaciej> | Hixie: I cannot wait to tweet someone's demo based on creatively abusing such a capability |
| 08:27 | <annevk> | the karaoke seemed to require some more |
| 08:27 | <Hixie> | which one? |
| 08:27 | <annevk> | the one with the rainbow text or just text gradually changing color |
| 08:27 | <othermaciej> | what about color? |
| 08:28 | <Hixie> | the rainbow colour one is a demo of what you can do with a subtitle format |
| 08:28 | <Hixie> | it's not a use case |
| 08:28 | <annevk> | i've seen it in the wild |
| 08:29 | <Hixie> | please put a screenshot or link up so i can study it |
| 08:29 | <annevk> | i mean in a bar |
| 08:30 | <Hixie> | go there and take a photo :-) |
| 08:30 | <annevk> | i'd love to if someone covers my ticket to nicaragua :p |
| 08:30 | <annevk> | though here in tokyo it should prolly be relatively easy to find |
| 08:31 | <annevk> | actually yesterday we mike and I much weirder stuff but no camera at hand |
| 08:31 | <Hixie> | (i don't think you can ever do what that example does with css currently... however, if we use css pseudo-elements to style subtitles in general, then that would handle that case once we have text paint servers) |
| 08:32 | <annevk> | yeah, that seems reasonable |
| 08:33 | <Hixie> | ok bed time |
| 08:33 | <Hixie> | nn |
| 08:45 | <nessy> | ah bummer missed Hixie |
| 08:45 | <nessy> | does Hixie read back? |
| 08:46 | <othermaciej> | I think he does |
| 08:46 | <othermaciej> | at least if you mention his name |
| 08:46 | <othermaciej> | or you can email him |
| 08:55 | <nessy> | just emailed him :) |
| 09:29 | <gsnedders> | So I went out and ordered alcoholic drinks last night. I didn't even get ID'd. |
| 09:30 | <othermaciej> | do you look old enough? |
| 09:34 | <annevk> | not really :p |
| 09:34 | <othermaciej> | lol |
| 09:42 | <jgraham> | He doesn't really look old enough to be out without his mother |
| 09:42 | <hsivonen> | well, at a dinner at TPAC in Boston, MikeSmith was ID'd (and denied alcoholic beverage) and I was not ID'd |
| 09:44 | <zcorpan> | yay, now i can remove img { border:0 } from my style sheets |
| 09:44 | <annevk> | but Mike has a mustache |
| 09:44 | <hsivonen> | annevk: I had a collar on my shirt :-) |
| 09:46 | <hsivonen> | plinss had a collar, too, and was not ID'd. IIRC, everyone wearing t-shirts was ID'd regardless of facial hair |
| 09:47 | <annevk> | interesting |
| 09:47 | <annevk> | though Mike was telling me they also do |
| 09:47 | <hsivonen> | and then the waiter refused to accept Mike's Japanese alien card as ID |
| 09:48 | <annevk> | euh oops; anyway, that was supposed to end in US airport trouble |
| 09:48 | <othermaciej> | MikeSmith looks way too old to ID |
| 09:48 | <othermaciej> | even I look too old to ID these days |
| 09:48 | <annevk> | ah yeah, I was on that table I think |
| 09:49 | <othermaciej> | I was getting ID'd regularly as recently as 2 years ago and I was 32 then |
| 09:49 | <annevk> | but you seem to have a much better memory of it :) |
| 09:49 | <othermaciej> | or 31? |
| 09:49 | <othermaciej> | I can't remember how old I am any more |
| 09:49 | gsnedders | has wine at TPAC in Mandelieu |
| 09:49 | <gsnedders> | *had |
| 09:49 | <gsnedders> | Apparently the age in France has been raised to 18 |
| 09:49 | <annevk> | I went into a bar once in Palo Alto where I wasn't ID'd but chaals was |
| 09:49 | <gsnedders> | hah |
| 09:49 | <annevk> | and I was under 21 at the time :p |
| 09:50 | <jgraham> | You're assuming they want to catch you |
| 09:50 | <annevk> | (of course, the fully story is that the bouncer was in the restroom while I entered) |
| 09:50 | <annevk> | s/fully/full/ |
| 09:51 | <nessy> | it's a compliment if you're getting ID-ed at 40 ;) |
| 09:52 | <annevk> | I'm not sure chaals would find it a compliment as he isn't 40 yet (afaik) :) |
| 09:55 | <annevk> | 8 days until a new Ubuntu comes out |
| 09:55 | <annevk> | somewhat excited about this release |
| 09:56 | <jgraham> | I am considering installing it early just so I can change the theme as fast as possible |
| 09:56 | <jgraham> | s/fast/soon/ |
| 09:57 | <jgraham> | (the buttons-on-the-left thing I can probably cope with, but srsly, what's with the dark grey) |
| 09:58 | <gsnedders> | Even the guy who jgraham calls emo dislikes the darkness of the theme. |
| 10:10 | <hsivonen> | I thought the default theme was supposed to be Light |
| 10:12 | <hsivonen> | do Qt apps still look totally different on Lucid? |
| 10:14 | <jacobrask> | should I escape & to & in links in HTML5? |
| 10:15 | <hsivonen> | jacobrask: yes |
| 10:16 | <hsivonen> | you don't have to in most cases, though |
| 10:16 | <jacobrask> | hm? :P |
| 10:17 | <jacobrask> | why and when should I |
| 10:17 | <hsivonen> | always, so you don't accidentally fail to escape when it's required |
| 10:17 | <jacobrask> | ok |
| 10:18 | <jacobrask> | when is it required then |
| 10:18 | <othermaciej> | per the latest spec, it's only required if the & is followed by something that matches an entity name |
| 10:19 | <jacobrask> | ok, and I won't remember all of those, so yes then I'll escape it all the time |
| 11:34 | <Dashiva> | It is mindboggling that bug 8833 is still going |
| 11:38 | <Dashiva> | Or maybe that's only if you assume rational intent... |
| 11:46 | <Philip`> | "There was no way that the group could come up with a title for the spec when the group was not told what the spec included." |
| 11:46 | <Philip`> | Couldn't the group have, like, read the spec to see what it included? |
| 11:46 | <zcorpan> | that would involve reading the spec |
| 11:46 | <othermaciej> | see my comments for copious evidence that the group was told what the spec included |
| 11:47 | <othermaciej> | without even having to read it even |
| 11:52 | <Dashiva> | It's a good thing we have a solid process in place to neutralize poisonous people! |
| 12:54 | <zcorpan> | hmm it's snowing |
| 12:55 | <jgraham> | zcorpan: Yes |
| 13:17 | <Lachy> | it was snowing briefly here in Oslo too. Seems to have stopped now though |
| 13:19 | <jacobrask> | in Gothenburg as well, for 10 minutes or so |
| 13:19 | <jacobrask> | weird |
| 13:35 | <Lachy> | I wonder if it has anything to do with all the volcanic ash messing with the weather. |
| 15:38 | <jarib> | does anyone maintain a complete IDL document from HTML5? |
| 15:38 | <jarib> | i'm giving up on automatically extracting it from the markup |
| 15:38 | <asmodai> | Geez, I wonder if I should give up on this Lorentz build. |
| 15:39 | <asmodai> | It might protect Firefox from plugins, but so far it just crashes randomly without even touching it. |
| 15:39 | <jarib> | about to do extract the IDL parts manually, but now looking forward to it :P |
| 15:39 | <asmodai> | Now or not? :) |
| 15:39 | <jarib> | not! :) |
| 15:40 | <asmodai> | Nothing in the logs for the channel/ |
| 15:40 | <asmodai> | ? |
| 15:41 | <jarib> | hmm, i'll look |
| 15:56 | <hendry> | jarib: syntax.whattf.org/relaxng/ is probably the closest |
| 15:59 | <gsnedders> | That's not IDL |
| 16:22 | <Philip`> | jarib: Someone had set up a thing to automatically validate the spec's IDL |
| 16:22 | <Philip`> | which I assume includes the extraction of it |
| 16:26 | <Philip`> | jarib: http://www.w3.org/2009/07/webidl-check / http://dev.w3.org/2009/webidl-checker/webidl-check does the extraction, I think |
| 17:03 | <jarib> | Philip`: thanks, unfortunately just extracting the <pre class=idl> will still be incomplete |
| 17:03 | <jarib> | since some elements just has "DOM Interface: Uses HTMLElement" |
| 17:05 | <annevk> | why would you need to extract that for a IDL overview? |
| 17:06 | <annevk> | there's also some index table with IDL info |
| 17:06 | <jarib> | annevk: i'm generating code based on the IDL for a testing tool |
| 17:07 | <annevk> | oh sweet |
| 17:07 | <jarib> | basically i just need a map of tag_name => typed_attribute_list |
| 17:07 | <annevk> | browser testing? |
| 17:07 | <jarib> | yeah |
| 17:08 | <annevk> | maybe grab all IDL blocks, grab the element index -> IDL interface name table and mash it up? |
| 17:09 | <jarib> | yeah, that sounds like the best approach |
| 17:09 | <jarib> | i've tweaked it by hand for now, but that table will be useful in thefuture |
| 18:36 | <AryehGregor> | "Feel free to REOPEN if you think luring people to Berlin and telling them stories about volcanoes to make them stay is not an acceptable way of setting up a branch office." |
| 18:36 | <gsnedders> | I find it disturbing that I can find mistakes in a friend's English work, and know enough of a quote to be able to google it and find the exact quote |
| 18:37 | <gsnedders> | (in some other secondary work) |
| 18:59 | <daedb> | Wow, bug 8833 is really delicious :) |
| 19:12 | <TabAtkins> | daedb: I... I'm so confused. What *possible* benefit could Apple, Google, and Maciej and Ian personally, gain from the name of the spec? |
| 19:13 | <TabAtkins> | The answer, of course, is "mu", because the dude is clearly an internet crazy. |
| 19:14 | <Dashiva> | TabAtkins: They force their bloated and useless and dangerous misfeatures onto the innocent real world filled with people who only ever write static documents |
| 19:15 | <Dashiva> | The personal gain is of course entertainment |
| 19:15 | <TabAtkins> | Ah, I see. They get to laugh at the monkeys dancing on their fiendish strings. |
| 19:15 | <TabAtkins> | Makes sense. |
| 19:19 | AryehGregor | adds himself to the CC list |
| 19:20 | <daedb> | TabAtkins: It's the first step to total world domination, *obviously* :p |
| 19:21 | <daedb> | First you control the spec, then you control the language, and then somehow you win :) |
| 19:21 | <TabAtkins> | Then you get de wimmin. |
| 19:21 | <daedb> | That too. |
| 19:26 | <Hixie> | TabAtkins: i get paid by a secret group organised by swiss bankers for each day that the spec remains called "HTML5" -- the amount I make goes down for each character away from HTML5 |
| 19:28 | <Hixie> | TabAtkins: that's why Jeremy was so concerned when I renamed the spec, he was worried I would lose out on that money |
| 19:29 | <Hixie> | jesus, what is joe doing |
| 19:31 | <miketaylr> | I KNEW IT |
| 20:15 | <zcorpan> | http://www.h-online.com/open/news/item/High-performance-Theora-codec-for-Firefox-on-OMAP3-previewed-980720.html |
| 20:35 | <TabAtkins> | AryehGregor: Where is the "luring people to Berlin" thing from? |
| 20:35 | <jgraham> | So, good people, is there any way to convince windows that a path can be > 255 characters long other than making it an absolute path starting \\?\ |
| 20:35 | <TabAtkins> | Install linux. |
| 20:36 | <jgraham> | Sadly Windows is a critical feature in this case |
| 20:36 | <TabAtkins> | Then I give up. |
| 20:36 | <Philip`> | jgraham: As far as I'm aware, Windows doesn't have that requirement |
| 20:36 | <Philip`> | though some Windows tools do |
| 20:36 | <Philip`> | like xcopy |
| 20:36 | <jgraham> | Philip`: Subversion seems to |
| 20:37 | <Philip`> | You can create more deeply nested directories in Explorer |
| 20:37 | <Philip`> | so it's probably a bug in the tool if it doesn't like them |
| 20:37 | <Dashiva> | TabAtkins: It's a mediawiki bug |
| 20:37 | <jgraham> | Specifically the Win32 API seems to have a function with that limitation even though NTFS allows much longer paths |
| 20:37 | <Dashiva> | They had a convention in Berlin or somewhere in Germany, and got stranded because of the volcano |
| 20:38 | <jgraham> | Philip`: Sadly subversion is also a critical component in this setup |
| 20:38 | <Dashiva> | TabAtkins: https://bugzilla.wikimedia.org/show_bug.cgi?id=23223 |
| 20:39 | <Philip`> | jgraham: Can you make the path shorter, by removing some parent directories? |
| 20:39 | <TabAtkins> | jgraham: I had some tool that had similar problems. We solved it by starting our path in c:/x/ |
| 20:40 | <Philip`> | Or, can you install Cygwin? (since maybe that avoids the API limitation) |
| 20:40 | <Dashiva> | It doesn't help that the default documents folder is nested quite deep |
| 20:40 | <jgraham> | Philip`: Possibly |
| 20:41 | <jgraham> | Philip`: I fear that the same problem would just happen again (I don't really control the filenames so I can only trim the prefixes) |
| 20:42 | <jgraham> | TabAtkins: That is kind of my plan but it is not quite trivial in this case because the checkout is initiated by a different machine that doesn't know about the absolute paths on the remote end |
| 20:42 | <Dashiva> | Have you tried a different svn implementation? |
| 20:43 | <jgraham> | Dashiva: No. I don't really understand the subversion-on-windows situation |
| 20:44 | <jgraham> | Are there material differences between the various binaries? |
| 20:44 | Philip` | sees http://svn.haxx.se/users/archive-2005-02/1088.shtml says "if you feed Subversion an absolute path, [...] you *should* be okay" |
| 20:44 | <Dashiva> | There are several different svn libraries at least |
| 20:45 | <Philip`> | There's a big difference if you use TortoiseSVN |
| 20:46 | <Philip`> | though they still all use the same backend library as far as I'm aware |
| 20:46 | <Philip`> | because only crazy people would reimplement it from scratch |
| 20:48 | <Dashiva> | Lots of crazy people out there |
| 20:48 | <Dashiva> | There are two implementations just in java, even |
| 20:49 | <Philip`> | Java people are particularly crazy |
| 20:50 | <Philip`> | and I guess they like to not rely on native C libraries |
| 20:50 | <Philip`> | Are there any in other languages? |
| 20:50 | <Dashiva> | One is a JNI interface to the C lib, another is a full reimplementation |
| 22:03 | <Dashiva> | UMP keeps trucking along... |
| 22:04 | <Dashiva> | It seems it will be blocked waiting for two implementations until CORS finishes, so I don't see the hurry |
| 22:05 | <JonathanNeal> | I was wondering if I could get some feedback from you guys on http://sandbox.thewikies.com/learn-html/ a project I'm writing to teach kids HTML. |
| 22:06 | <JonathanNeal> | Or maybe 9th graders (not little ones) |
| 22:08 | <TabAtkins> | Interesting. |
| 22:09 | <Dashiva> | I think those kids are too young to get "all your base" :) |
| 22:11 | <Dashiva> | It's interesting, but I'm not sure if <em> and <strong> are good example elements. They're fairly abstract, and very similar (semantically, not visually) |
| 22:14 | <JonathanNeal> | Dashiva, sure, what other elements would be useful suggestions? I picked <em> because it was short. |
| 22:14 | <JonathanNeal> | I would have picked <b> or <i> but I really didn't want to encourage that. |
| 22:18 | <Dashiva> | You could use one of em/strong, and another elements with a different type. E.g. <sup> |
| 22:18 | <TabAtkins> | Or <a>, though that's probably too complex for a first lesson. |
| 22:19 | <Dashiva> | <hx> and <p> are also common starting elements, in my experience |
| 22:21 | <TabAtkins> | JonathanNeal: Honestly, teaching kids <b> and <i>, and having them be misused to mark up emphasis and importance, is better than teaching them <strong> and <em>, and having them be misused to mark up citations, foreign words, random bold text, etc. |
| 22:21 | <TabAtkins> | Having not enough semantics is better than having wrong semantics. |
| 22:22 | <JonathanNeal> | Dashiva, headings and paragraphs would be the next lesson, I think. |
| 22:23 | <Dashiva> | I would tend to agree with what TabAtkins said there |
| 22:23 | <JonathanNeal> | I wanted to go for inline elements first, then either block elements or attributes (img, a, etc) |
| 22:24 | <JonathanNeal> | I was thinking of including "Solemn Oaths" after each lesson, where they agree not to use <strong> to bold random text, or <em> to italicize. |
| 22:24 | <Dashiva> | Then they'll ask "But how do we do it then?" |
| 22:25 | <Philip`> | You should start with <video> |
| 22:25 | <Dashiva> | And then you'll say "You'll have to wait a few weeks until we learn CSS" |
| 22:25 | <JonathanNeal> | Include an example in the oath! |
| 22:25 | <Philip`> | otherwise they'll get bored too soon |
| 22:25 | <Dashiva> | And then they'll say "That's boring" and use it anyway |
| 22:26 | <JonathanNeal> | Well, the reality is that I'm teaching them HTML as a content tool and not as a design tool. |
| 22:26 | <Dashiva> | Then <em> and <strong> aren't really that important |
| 22:27 | <JonathanNeal> | Well, name other useful inline elements, like <sup> |
| 22:28 | <Philip`> | When we had an introduction to HTML in IT lessons at school, it took about ten minutes for knowledge of <marquee> to spread throughout the whole room and then it was used on pretty much everybody's first pages |
| 22:28 | <Dashiva> | I had one person in my class who wrote each line as a <h1> or <h2> |
| 22:28 | <Dashiva> | And also forgot the / in closing tags |
| 22:29 | <Philip`> | JonathanNeal: Your page looks cooler if I tell it my name is <marquee> |
| 22:29 | <Philip`> | (Maybe you want to escape it before inserting it in the output) |
| 22:29 | <JonathanNeal> | Oh Philip` you dog! |
| 22:29 | <Dashiva> | What if you pretend to be little bobby tables? |
| 22:30 | <JonathanNeal> | Philip`, I've fixed that. |
| 22:33 | <TabAtkins> | Aw, I missed out on the fun. Now you just call me 1337. ;_; |
| 22:40 | <jgraham> | Practically speaking there is no difference between <em> and <i> or <b> and <strong> |
| 22:40 | <TabAtkins> | Argh, I have an almost irrestible urge to contact the makers of this Facebook game and offer to rewrite all their HTML. |
| 22:40 | <TabAtkins> | The different font metrics on Linux screw tons of stuff up because they're doing it wrong. |
| 22:40 | <jgraham> | I strongly doubt any general purpose tools make a distinction |
| 22:44 | <JonathanNeal> | jgraham, well in that case, I guess I'd just pick elements that make a distinction in their look and feel. |
| 22:44 | <JonathanNeal> | look of element name and look of default rendered appearance. |
