| 01:28 | <annevk> | Lachy, "about:pugins" |
| 01:36 | <Lachy> | oops, I'll fix it |
| 02:38 | <Hixie> | Lachy: you should send a mail to the list about it then, because i know of at least two other people who have written or said they would write something |
| 02:40 | <Hixie> | annevk: i replied to the monolithic thing on ww-html |
| 02:41 | <Hixie> | what's with about: suddenly being something people want to write specs about |
| 02:43 | <karlcow> | and I know another person who has started a draft |
| 02:51 | <Hixie> | i guess all four of these people should get together :-) |
| 02:52 | <Lachy> | who else has drafted it? |
| 02:52 | <Hixie> | annevk talked about doing it, iirc, and i have one person who e-mailed me with a draft (i told them to post it to one of the lists, haven't checked if they have yet) |
| 02:53 | <Lachy> | well, I don't mind if someone takes over from me. I'll send mine to the list anyway and see what happens |
| 03:11 | <Lachy> | does public-html accept posts from non-subscribers? I just sent to it from the wrong email address, and want to know if I should resend from the correct address |
| 03:12 | <Lachy> | ah, no, the list serv wants my permission to publish it. I can just decline |
| 03:43 | <karlcow> | and someone sent me a direct message on twitter |
| 03:44 | <karlcow> | I'll reply to the person with the mail on the list |
| 07:44 | <zcorpan> | annevk: "The server-sent events feature (the event-source element)." - s/event-source/eventsource/ |
| 07:49 | <zcorpan> | annevk: "... but is not compatible with the more esoteric SGML features of HTML 4, such as <em/content/." - might read better as s#<em/content/#the NET syntax (i.e. <em/content/)# |
| 07:52 | <zcorpan> | annevk: "Using a meta element with a charset attribute that specifies the encoding as the first element child of the head element." - requirement changed to be just within 512 bytes |
| 07:52 | <zcorpan> | annevk: in the same bullet point it might be worth noting that the old syntax is also allowed |
| 07:55 | <zcorpan> | annevk: "DOCTYPEs from earlier versions of HTML were longer because the HTML language was SGML based and ..." - i think that should be s/SGML based/SGML-based/ |
| 07:58 | <zcorpan> | annevk: in 3.1, consider listing meter and progress next to each other |
| 08:05 | <zcorpan> | annevk: "The form and select elements (as well as the datalist element) have a data attribute ..." - wasn't this feature dropped? |
| 08:06 | <Hixie> | Lachy: as i understand it, the <!DOCTYPE> in XHTML5 is a purely XML-level concern |
| 08:06 | <Hixie> | it can point to a DTD, have an internal subset, whatever |
| 08:06 | <Hixie> | it's equivalent to the use or lack of use of <![CDATA[]]> or &#x....;, just part and parcel of XML |
| 08:07 | <zcorpan> | annevk: "You can now disable an entire fieldset by using the disabled attribute on it. This was not possible before." - this bullet point is phrased differently from the others |
| 08:22 | <zcorpan> | annevk: "size attribute on hr, input and select." - <input size> is back |
| 08:22 | <zcorpan> | and <select size> |
| 08:32 | <zcorpan> | annevk: "innerHTML in XML was slightly changed to improve round tripping." - i think that should be round-tripping |
| 08:41 | <zcorpan> | annevk: "The a element now allows nested flow content, e.g. a div element, but not itself. " - an element can't nest itself :) might want to reword to say interactive content or so |
| 08:45 | <zcorpan> | annevk: "The media elements now support just a single loop attribute." - might want to say which attributes were removed at the same time |
| 08:47 | <zcorpan> | annevk: "The load() method on media elements has been redefined as asynchronous. It also tries out files in turn now. " - instead of...? (which is, iirc, finding the most appropriate <source> by looking at type) |
| 08:50 | <zcorpan> | annevk: "User agents are required to support a headers attribute pointing to a td element, but authors are required to use th instead." - this sounds stupid on the face of it (it says UAs have to support one thing and authors have to do something else) |
| 08:50 | <zcorpan> | annevk: suggest to reword that UAs have to support both th and td but authors are only allowed to use th |
| 08:54 | <zcorpan> | annevk: "Author defined attributes have been added." - although this section might be frozen, i think it should be "Author-defined" |
| 08:59 | <zcorpan> | annevk: i'd like to have more focus and detail on the WF2 integration and changes (that and a pony) |
| 09:01 | <zcorpan> | annevk: the XML 1.0 reference says 4th ed |
| 09:05 | <zcorpan> | maybe i should email my comments to the list |
| 09:10 | <hsivonen> | wilhelm_: sorry about the bug report yesterday. It's invalid, because font-variant is not permitted in @font-face. |
| 09:17 | <roc> | phew! |
| 09:18 | <hsivonen> | it's a bit confusing to have things that look like CSS properties specced for two purposes in the same spec and then scroll to the wrong point when reading the spec piece-wise |
| 09:21 | <zcorpan> | Lachy: "3.1. about:blank Applications resolving this URI must return an empty resource, containing nothing." - an empty *text/html* resource (if the UA supports text/html) |
| 09:22 | <zcorpan> | Lachy: although i'm not sure if it should say anything about the quirkyness... |
| 09:25 | <hsivonen> | does about:blank have a well-defined charset? |
| 09:26 | <hsivonen> | (for the various DOM charset-related properties) |
| 09:27 | <Lachy> | zcorpan, I wasn't sure if about:blank needed to have an explicit MIME type, since it's interpretation would depend upon the nature of the application in which it's used. Although currently that may only include web brows.ers |
| 09:32 | <zcorpan> | hsivonen: dunno |
| 09:32 | <zcorpan> | Lachy: web pages typically enable designMode on about:blank iframes |
| 09:33 | <zcorpan> | Lachy: so in browsers for iframes it has to be text/html |
| 09:34 | <Lachy> | yeah, but I mean, shouldn't it be HTML5 that defines it to be interpreted as text/html? |
| 09:34 | <zcorpan> | likely :) |
| 09:35 | <zcorpan> | why have the emptyness requirement for about:blank? |
| 09:36 | <Lachy> | I'll add something explaining how to deal with the MIME type |
| 09:37 | <Lachy> | hsivonen, which DOM charset properties would be affected by the charset of about:blank, rather than the charset attributes on various elements? |
| 10:00 | <Philip`> | Lachy: It seems kind of nasty for HTML5 to determine the type of a resource not just from its HTTP Content-Type header, nor from the context from which it was requested, nor from sniffing its content bytes, but from comparing its URL against a hardcoded string |
| 10:22 | <zcorpan> | Philip`: html5 is nasty |
| 10:24 | <Philip`> | zcorpan: It can't shun necessary nastiness, but that doesn't mean we should encourage gratuitous nastiness |
| 10:41 | <annevk> | seems better to just define about:blank as an empty text/html resource in no quirks mode |
| 10:42 | <annevk> | thanks zcorpan, I was hoping you'd review it :) |
| 10:42 | <zcorpan> | annevk: welcome :) |
| 10:45 | <Lachy> | annevk, why does it need to be no quirks mode? |
| 10:46 | <Lachy> | Safari and Firefox use quirks mode. Opera uses no quirks mode. |
| 10:48 | <Lachy> | IE8 doesn't seem to want to let me execute javascrpit URIs in the address bar |
| 10:49 | <Lachy> | so javascript:alert(document.compatMode); doesn't tell me which mode it uses |
| 10:51 | <Lachy> | but using <iframe src=about:blank> and a script, IE8 uses quirks mode too |
| 10:57 | <annevk> | just seems nicer, especially since about:blank is always used with designMode |
| 10:57 | <annevk> | does it by any chance depend on other factors as well then? seems likely that Opera would break stuff if it's that different |
| 10:58 | <Lachy> | does HTML5 currently say anything about using standards mode for about:blank? I couldn't find anything like that |
| 11:53 | <annevk> | http://www.openmedianow.org/ |
| 12:19 | <Lachy> | Hixie, yt? |
| 12:20 | <annevk> | zcorpan, sorry about the pony, but the rest is done |
| 12:24 | <Lachy> | Hixie, I have a functioning polyglot UTF-8/UTF-16LE JavaScript file, as you requested http://krijnhoetmer.nl/irc-logs/whatwg/20090123#l-239 |
| 12:29 | <annevk> | that sounds cool, where? |
| 12:31 | <Lachy> | it's not published yet. I'm just working out how to demo it effectively |
| 12:31 | <Lachy> | hmm. Do scripts served without a charset inherit the encoding of the document? |
| 12:32 | <annevk> | <script charset=utf-8 src=...></script> <script charset=utf-16 src=...></script> |
| 12:32 | <annevk> | yeah, that's the idea |
| 12:32 | <Lachy> | ok, I'll do that |
| 12:32 | <annevk> | but using the charset attribute is easier |
| 12:32 | <Lachy> | is the charset attribute supported? |
| 12:32 | <annevk> | it should be, yes |
| 12:32 | <annevk> | though this might have been in part to try to find out to what extent it is |
| 12:33 | <zcorpan> | annevk: cool, will check |
| 12:36 | <Lachy> | ah, ok. In that case, I could serve the document as ISO-8859-1 and ensure that the script will only execute if the script is parsed as UTF-8 or UTF-16LE |
| 12:37 | <Lachy> | currently, it uses an ASCII function name for UTF-8 which will work in 8859-1. I'll make it non-ASCII |
| 12:43 | <hsivonen> | Lachy: I weant the charset, characterSet and defaultCharset properties of the HTMLDocument interface |
| 12:44 | <Lachy> | oh, dear. serving the doc as 8859-1 won't work cause the UTF-16LE interpretion of the script uses a function name that can't be represented in 8859-1. |
| 12:45 | <hsivonen> | Moving to an apartment building built in 2001 is an interesting experience after living in a building built in 1946. The new one has *manuals*. |
| 12:45 | hsivonen | has never had manuals for an apartment before |
| 12:47 | <annevk> | Lachy, load another script that calls the function? |
| 12:48 | <Lachy> | annevk, yeah, that's what I thought too |
| 12:48 | <annevk> | Lachy, I suppose you can't use entities inside <script>? does JavaScript allow escapes everywhere like CSS? |
| 12:48 | annevk | thinks the answer to both is "no" |
| 12:48 | <Lachy> | I don't remember the rules about JS excapes |
| 12:56 | <Lachy> | last thing I need to do is make apache override my normal charset settings so it doesn't send charset=UTF-8 for the JS |
| 12:58 | <annevk> | AddDefaultCharset None |
| 12:58 | <annevk> | iirc |
| 13:00 | <Lachy> | yeah, that and RemoveCharset |
| 13:00 | <Lachy> | except it's Off instead of None |
| 13:03 | <Lachy> | hsivonen, about:blank and .characterSet returns UTF-8 and WebKit and Firefox, utf-16 and Opera and undefined in IE. .defaultCharset returns undefined in Firefox and Opera, ISO-8859-1 in Safari and windows-1252 in IE |
| 13:03 | <Lachy> | s/UTF-8 and/UTF-8 in/ |
| 13:05 | <hsivonen> | seems like the closest thing to consensus is to say you parse a zero-length stream whose media type is text/html; charset=utf-8 |
| 13:05 | <Lachy> | yeah, that was my assessment too |
| 13:06 | <Lachy> | annevk, http://lachy.id.au/dev/2009/polyglot/ |
| 13:07 | <zcorpan> | hsivonen: but that gives quirks mode which sucks :) |
| 13:07 | <Lachy> | oops, that's buggy still |
| 13:08 | <zcorpan> | iirc, i've found one opera bug that would be solved by changing about:blank to quirks mode, but it could also be solved by changing from <p> to <div> as default block in contenteditable |
| 13:08 | Philip` | has a script which can distinguish utf-8 and utf-16 and iso-8859-1 |
| 13:08 | <zcorpan> | (the bug was that on enter you got a "line break" before the first line) |
| 13:11 | <Philip`> | http://philip.html5.org/demos/charset/polyglot-script.html |
| 13:17 | <Lachy> | I don't understand why no browser works with my version. They're all interpreting the polyglot script file as UTF-8 |
| 13:18 | <Philip`> | Firefox 2 says "UTF-16LE" on http://lachy.id.au/dev/2009/polyglot/polyglot-utf-16le.html |
| 13:18 | <Philip`> | FF3 too |
| 13:18 | <zcorpan> | seems webkit ignores charset |
| 13:18 | <Lachy> | says UTF-8 for me |
| 13:19 | <map> | Is there an easy way to hook in a URI resolver/normalizer into html5lib? |
| 13:20 | <map> | For URI attribute values like href and src etc. |
| 13:20 | <Philip`> | Lachy: That's a bit odd |
| 13:20 | <Lachy> | woah, if I use a clean profile in Firefox, it works. But it doesn't work with my normal profile |
| 13:21 | <Lachy> | I'll try clearing the cache to see if that fixes it |
| 13:21 | <Philip`> | Lachy: Strange config settings? |
| 13:22 | <Philip`> | map: As far as I'm aware, html5lib doesn't know anything about which attributes are URIs, so you'd most likely have to parse the document then write code to walk the document tree and update all the relevant attributes |
| 13:23 | <Philip`> | map: (which gets a bit complex if you want to resolve URIs correctly when there's a <base href>) |
| 13:24 | <map> | Thanks. I was thinking of something like parse.py --resolve-uris --normalize-uris "--document-location=" |
| 13:29 | <Philip`> | map: That sounds like it could be a useful feature to add as some kind of filter in html5lib, but I don't think anyone's tried implementing it yet |
| 13:31 | <map> | Yeh, I personally want to use it to test normalization ideas (like when to leave %HH alone or when to normalize wide characters to UTF-8 %HH etc.). |
| 13:31 | <zcorpan> | annevk: [DOCTYPE] should say "Activating Browser Modes with Doctype" |
| 13:32 | <Philip`> | map: Why do you want to do it as part of html5lib, rather than as a feature in whatever application is using the parsed document? |
| 13:35 | <zcorpan> | annevk: i wonder if there's something that has been added that should be included in the Impact on Web Architecture list |
| 13:36 | <map> | Phillip: I basically want to experiment with normalizing and resolving similar to how browsers do. I can do that without a parser and just use test strings, but I wanted to compare with browsers and how they do it. Using html5lib could help with that. |
| 13:38 | <map> | Philip`: Part of the reason for this is I want to come up with a solution to solve *particular* normalizing/resolving bugs/differences between browsers. |
| 13:38 | <Lachy> | Philip`, clearing my cache fixed it |
| 13:39 | <Lachy> | It works in Safari, but only if you clear the cache before looking at the other one |
| 13:39 | <map> | Philip`: Hooking into htmlserializer.py seems like it might be the easiest way. |
| 13:40 | <map> | But I don't know the code at all. |
| 13:40 | <Philip`> | Lachy: Ah, right |
| 13:41 | <Philip`> | Lachy: In IE6 in Wine I get "Invalid character" on the utf-16 page, and "Object expected" on the utf-8 page |
| 13:42 | <Philip`> | map: Hmm, I'm only familiar with the parser code so I don't know what would be a good way to work with other parts of html5lib |
| 13:43 | <Philip`> | Lachy: (I get "Invalid character" on the utf-16 bit of http://philip.html5.org/demos/charset/polyglot-script.html too, but the other two say "utf-8" and "iso-8859-1" respectively) |
| 13:44 | <Lachy> | Philip`, ok. I'll look at it later. i've got other stuff I need to do ow |
| 13:44 | <Lachy> | *now |
| 13:51 | <map> | Philip`: Thanks. I'll investigate. And, I'll try to come up with some examples of what I'm trying to do to make things clearer. |
| 13:52 | <annevk> | zcorpan, URLs |
| 13:53 | <zcorpan> | annevk: ok |
| 13:57 | <Philip`> | Amazon emails me to say "Amazon.co.uk's January Sale Must End Sunday 1st February", presumably just in time for the start of their February sale |
| 14:05 | <hsivonen> | I like <!DOCTYPE html SYSTEM "about:blank"> |
| 14:05 | <hsivonen> | I'm also amused by the implications of being strict about the content type when derefercing it. |
| 14:06 | <Lachy> | yeah, about:blank could work |
| 14:06 | <Lachy> | it's even more memorable than sgml-compat and already widely used |
| 14:07 | <Philip`> | I want <!DOCTYPE html SYSTEM "about:mozilla"> |
| 14:07 | <Lachy> | or about:robots :-) |
| 14:08 | <hsivonen> | about:marca no longer works |
| 14:08 | <Lachy> | what was about:marca? |
| 14:08 | <hsivonen> | Lachy: IIRC, it redirected to Marc Andreessen's home page on people.netscape.com |
| 14:09 | <Lachy> | ok |
| 14:09 | <Lachy> | in NN4? |
| 14:09 | <annevk> | about:jwz |
| 14:09 | <hsivonen> | Lachy: and earlier |
| 14:09 | <Philip`> | http://www.jwz.org/gruntle/blowme.html |
| 14:09 | <Lachy> | annevk, which browser does that one work in? |
| 14:13 | Philip` | appreciates how Opera understands if you type "/." into the address bar |
| 14:49 | <karlcow> | hmmm Opera (at least 10.00 alpha) rewrites about: into opera: |
| 14:52 | <Lachy> | karlcow, yes, Opera has always done that |
| 14:52 | <Lachy> | I don't know why though |
| 15:18 | <zcorpan> | Lachy: "If the selectors parameter is set to either null, or undefined in the ECMAScript language binding, the implementation must behave as if an empty string had been passed instead. " - doesn't the idl already require that? |
| 15:23 | <Lachy> | zcorpan, sure, but restating it in prose to make it clearer doesn't hurt |
| 15:30 | zcorpan | remembers studying technical drawings at university where it was forbidden to have over-defined dimensions |
| 15:31 | <zcorpan> | Lachy: for some reason i see it as an error to have the same thing required twice in a spec |
| 15:32 | <zcorpan> | your requirement might be subtly different from that in web idl |
| 15:35 | <Philip`> | When writing canvas tests, I didn't like it if a requirement existed in more than one place, since I liked having a one-to-one mapping between spec phrases and testable criteria |
| 15:35 | <annevk> | as co-editor, I agree with zcorpan :) |
| 15:36 | <Philip`> | because that avoids the chance of conflicting requirements and confusion |
| 15:36 | <Philip`> | But it's nice to have a non-normative note pointing out relevant behaviour that is normatively defined elsewhere |
| 15:36 | <annevk> | ah, [NoInterfaceObject] is aleady there, nice |
| 15:37 | <annevk> | Lachy, it should probably use ImplementedOn (or whatever it was called) instead of prose as well |
| 15:38 | <hsivonen> | http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2009Jan/0106.html |
| 15:42 | <Lachy> | annevk, please send mail |
| 15:42 | <Lachy> | I'm working through my backlog of feedback and I will get to it by the end of today |
| 15:53 | <Lachy> | oh, I hate when I find out that I never received an important mail, presumably cause it got caught by one of my spam filters :-( |
| 15:55 | <krijnh> | Lachy: a job offer at Google? :) |
| 15:56 | <Lachy> | krijnh, no selectors api feedback http://lists.w3.org/Archives/Public/public-webapps/2008OctDec/0245.html |
| 15:56 | <Philip`> | You should whitelist relevant mailing lists |
| 15:56 | <Lachy> | the only reason I found out about it was because I have a reply to a message I don't have |
| 15:57 | <Lachy> | most of the time, it hasn't caused any problems at all. I'm not even sure which filter of mine caught it |
| 15:57 | <Lachy> | but I will see if I can add whitelists to my server side filter |
| 15:57 | <Lachy> | though I just wish people would stop sending email that looks like spam |
| 16:30 | <Lachy> | oh crap. A whole bunch of emails just disappeard from my IMAP server. WTF?! |
| 16:31 | <jcranmer> | it hates you |
| 16:32 | <Lachy> | ah, actually, something did screw up, but I didn't lose as many as I thought. There's only 3 missing that I know of |
| 16:39 | <Lachy> | oh, it's ok. Crisis is over :-). It was just my email client's cache that got messed up. Clearing it helped :-) |
| 16:39 | <Philip`> | http://www.agrifirm.com/ - hooray, neurotic namespaces in text/html |
| 16:39 | <Philip`> | (Look for xmlns:wedoit) |
| 16:39 | <hsivonen> | is there another well-known name for jis_x0212-1990? |
| 16:43 | <Philip`> | http://philip.html5.org/data/xmlns-bindings.txt |
| 16:43 | <Philip`> | http://philip.html5.org/data/xmlns-bindings-per-tag.txt |
| 16:44 | <Philip`> | http://philip.html5.org/data/unbound-prefix-elements.txt |
| 16:44 | <Philip`> | http://philip.html5.org/data/unbound-prefix-attributes.txt |
| 16:44 | <Philip`> | I'm not at all sure what any of that is useful for |
| 16:44 | <Philip`> | but I wanted to try it anyway |
| 16:46 | <Philip`> | I suppose one observable point is that while hundreds of pages contain the string xmlns:rdf, only 18 have it in a place where an HTML parser sees it |
| 16:46 | <Philip`> | (and presumably all the rest are in comments) |
| 16:47 | <hsivonen> | Philip`: what proportion looks Trackback-motivated? |
| 16:48 | <Philip`> | hsivonen: No idea |
| 16:48 | Philip` | notices 14 xmlns:v="urn:scheman-microsoft-com:vml" (sic), 1 xmlns:v="urn:schemas-microDsoft-com:vml" |
| 16:50 | <Philip`> | Those 14 look like unrelated sites |
| 16:50 | <Philip`> | so I guess they copied-and-pasted from one of the four pages listed on http://www.google.com/search?q=%22scheman-microsoft-com%22 |
| 16:50 | <Lachy> | annevk, if you send a response to this mail regarding removing that conformance statement, I'll look into tonight http://lists.w3.org/Archives/Public/public-webapps/2008OctDec/0374.html |
| 16:51 | <Philip`> | Oh |
| 16:51 | <Philip`> | Those 14 are not unrelated |
| 16:51 | <Philip`> | They're all Ford dealerships, who presumably got sites built from the same template |
| 16:52 | <Philip`> | Oh, actually they're all Ford dealerships whose domains redirect to subsites of dealerconnection.com |
| 17:01 | <annevk> | "10.2. html5lib" cool |
| 17:01 | <annevk> | from http://diveintopython3.org/ |
| 17:03 | <Philip`> | Someone's got to port html5lib to Python 3 first :-) |
| 17:03 | <Lachy> | oh, cool. maybe I should learn python 3 as he publishes it |
| 17:03 | <Lachy> | yeah, the backwards compat issues with python are annoying |
| 17:04 | <Philip`> | He'll have to write chapter 19 before starting on 10.2 |
| 17:08 | <gsnedders> | backwards incompatibility is a feature, damnot! |
| 17:08 | <annevk> | "This site is optimized for Lynx just because fuck you." except it starts by talking about PapayaWhip :p |
| 17:08 | <gsnedders> | Bus routes are crazy |
| 17:17 | <Philip`> | gsnedders: They're just reflecting the craziness of humans, who always want to be somewhere other than where they are |
| 17:17 | <gsnedders> | Philip`: Yes, but on a bus service going pretty much due west it seems stupid to spend the entire time going north and south |
| 17:17 | <gsnedders> | (And there are bigger better roads heading due west too) |
| 17:18 | <Lachy> | gsnedders, does it have many stops along the way? |
| 17:18 | <gsnedders> | Lachy: Yeah |
| 17:18 | <gsnedders> | Lachy: That's the excuse, obviously |
| 17:18 | <gsnedders> | Philip`: No, he doesn't have to write chapter 19 first, he just needs to know how to do it himself |
| 17:19 | <Philip`> | gsnedders: Only if he's going to port html5lib himself |
| 17:19 | <gsnedders> | Philip`: Who else will? |
| 17:19 | <Philip`> | gsnedders: Magical elves will |
| 19:05 | <webben> | Hixie: http://www.whatwg.org/specs/web-apps/current-work/#audience : "defined in this specificaton" (sic for specification) |
| 19:25 | gsnedders | has almost no idea about what he wants to do at uni |
| 19:25 | gsnedders | can't even decide whether to do an arts degree or a science degree :P |
| 19:29 | <Philip`> | Arts degrees are a waste of time, therefore you should do a science degree |
| 19:30 | <Philip`> | It's an easy decision when you look at it the right way |
| 19:30 | <gsnedders> | Philip`: A science degree in what, then? |
| 19:33 | <Philip`> | gsnedders: A degree in, uh, scientificness |
| 19:33 | <Philip`> | They're all the same really |
| 19:33 | <svl> | gsnedders: doesn't really matter. Something which'll keep your interest for the duration. The important thing it'll teach you is a way of thinking and approaching problems. |
| 19:34 | <gsnedders> | svl: The problem is I have no idea what'll keep my interest :) |
| 19:34 | <gsnedders> | svl: I applied to do physics starting this year, but my interest (despite having been strong for many years) has now mainly gone |
| 19:35 | <svl> | gsnedders: what're the options? (You have a preferred university, right?) Something broad or something applied probably has the highest chance of containing a path through that'll see you complete it. |
| 19:36 | <webben> | gsnedders: Do you know how come you lost interest? |
| 19:37 | <gsnedders> | svl: Well, I'm not really sure I want to go to any of the places I have offers from, thus the entire question about applying again (and what for) |
| 19:37 | <gsnedders> | svl: Also, esp. in England/Wales (far more so than Scotland) degrees really aren't very broad |
| 19:37 | <gsnedders> | webben: Too much electronics and atomic-level physics this year |
| 19:38 | <webben> | gsnedders: Hmm. So what parts of physics do you prefer? |
| 19:38 | <gsnedders> | s/electronics/electromagnetism/ |
| 19:38 | <gsnedders> | webben: Stuff on a bit bigger scale:) |
| 19:38 | <webben> | :) |
| 19:38 | <gsnedders> | Y'know, stuff we can see? |
| 19:39 | <gsnedders> | If I drop a ball, I care about gravity. I don't care about a charge from the air transferring to the ball. |
| 19:39 | <gsnedders> | (Though a magnetic field may influence it, to be fair to electromagnetism) |
| 19:40 | <gsnedders> | The only place I'm particularly tempted to go from which I've received an offer for is Edi. for computational physics |
| 19:41 | webben | wonders if that's a failing of the course for not connecting electromagnetism/atomic-level physics to real-world applications. |
| 19:41 | <gsnedders> | But that is a stupidly easy course to get into and I could easily get an offer for it again, if I re-applied for elsewhere too |
| 19:42 | <gsnedders> | (i.e., normally the number of offers made is identical to the number of applicants) |
| 19:42 | <gsnedders> | (and I well exceed the entrance requirements) |
| 19:42 | <svl> | gsnedders: _any_ studies will start with a lot of background material and a broad overview that'll include a lot of things you'll not be interested in and won't ever need. |
| 19:43 | <svl> | You'll just have to accept that you'll need to go through that to get to the final year or two in which you're doing solely and purely the things you absolutely love. |
| 19:44 | <svl> | So with determining what to go for, focus on the things you do like, not on what you don't like. |
| 19:44 | gsnedders | is also an indecisive bastard :) |
| 19:45 | karlcow | wonders which countries Philip` and svl are from. Specifically when I read things like "[14:35] <Philip`> Arts degrees are a waste of time, therefore you should do a science degree" |
| 19:46 | svl | is from the Netherlands. Based on Philip`'s comment, I suspect he's from the USA. :) |
| 19:46 | <karlcow> | which somehow shows that the way of thinking was not that successful "The important thing it'll teach you is a way of thinking and approaching problems." |
| 19:47 | <gsnedders> | svl: Philip` is UK |
| 19:47 | <svl> | shows what I know. :) |
| 19:47 | gsnedders | does have the advantage of having me him :) |
| 19:50 | <Philip`> | gsnedders: You have you me? |
| 19:51 | <gsnedders> | s/me/met/ |
| 20:37 | <Hixie> | thanks to lachy and philip, http://www.hixie.ch/tests/adhoc/html/script/encodings/001-demo.html now supports utf-16 too |
| 20:58 | <Hixie> | and the rdfa community continues to discuss syntax issues without at all explaining to me what rdfa is for (and thus how to evaluate its usefulness in html5) |
| 20:58 | <Hixie> | i don't know how much clearer i can make my request |
| 22:32 | <takkaria> | hm, IE8 now in feature freeze |
| 22:32 | <takkaria> | I wonder if they've prefixed vendor-specific extensions yet |
| 22:58 | <Hixie> | i love when i'm trying to work out what i should spec in some edge case and the browser i use to find out what The Platform does today... crashes |
| 23:03 | <gsnedders> | Hixie: So, I take it you decide to not copy that behaviour, then? |
| 23:33 | <annevk> | http://www.0xdeadbeef.com/weblog/?p=977#comment-176228 |
| 23:39 | <Lachy> | MikeSmith, yt? |
| 23:40 | <MikeSmith> | yeah |
| 23:40 | <Lachy> | it looks like all you did to write the audience section was to move the vague, overly broad statement from the scope section to a new audience section. That doeesn't make it any less vague or overly broad |
| 23:41 | <Hixie> | it's no broader than html5's, the problem now is just that the spec itself is far too narrow to match the given audience statement |
| 23:42 | <Lachy> | MikeSmith, is there a public CVS repo for your spec |
| 23:50 | <MikeSmith> | Lachy: No. That's an issue with hosting anything on www.w3.org. Which is why I think dev.w3.org is better |
| 23:51 | <MikeSmith> | I guess I need to just move it over to dev.w3.org |
| 23:52 | <MikeSmith> | and I need to push for getting some more horsepower behind dev.w3.org to address that problems that systeam has pointed out |
| 23:52 | heycam | wonders what specs dev.w3.org has |
| 23:53 | <Lachy> | heycam, lots from WebApps, HTML, and CSS WGs |
| 23:53 | <heycam> | as in speed/memory, not which web specifications :) |
| 23:53 | <Lachy> | oh |
| 23:53 | <Hixie> | you gotta admit, that was an ambiguous query :-P |
| 23:54 | <heycam> | yeah =) |
| 23:54 | <MikeSmith> | I don't know what kind of machine it is, but I think it's extremely low-spec |
| 23:55 | <MikeSmith> | I think it started out as just being a place to publicly manage the source for a couple of tools |
| 23:56 | <MikeSmith> | But it then evolved to a lot more from there |
| 23:56 | <MikeSmith> | As these kinds of things tend to do. |
| 23:58 | <Lachy> | MikeSmith, can't they just move it to the same set of servers that they run w3.org from? |
| 23:58 | <annevk> | nn |
| 23:59 | <Hixie> | nn |