| 00:02 | <gsnedders> | I got bored in a lecture about strings today. I ended up implementing a LL(1) parser instead. I'm sure this says something about my course… |
| 00:03 | <Philip`> | Strings as in string theory? |
| 00:04 | <gsnedders> | No, as in the string type in Python. |
| 00:04 | <jcranmer> | gsnedders: that's boring |
| 00:04 | <jcranmer> | go for an LALR(1) parser |
| 00:05 | <gsnedders> | jcranmer: WebIDL has a LL(1) grammar though |
| 00:08 | <jcranmer> | but LR(1) is more fun to code |
| 00:14 | <gsnedders> | jcranmer: Well, maybe. But less useful for me now. |
| 00:15 | <gsnedders> | (Well, I mean, I could use a LR parser, but it's needless complexity) |
| 00:17 | <jcranmer> | obviously, you live in Europe |
| 00:28 | <tabatkins> | So, anyone know a quick way I can spoof myself as being in the US so I can watch netflix? |
| 00:30 | <jamesr_> | VPN to work? |
| 00:41 | <Philip`> | tabatkins: Set up an EC2 virtual machine and install a proxy on it? |
| 00:41 | <Philip`> | (where "Set up" means "boot" and probably only takes ten minutes) |
| 02:24 | <JonathanNeal> | Can I write my own namespace and use it as a namespace uri? |
| 02:24 | <JonathanNeal> | Like document.createElementNS('http://localhost/jon.xml';, 'jon'); ? I've been looking through the docs on how to do this. |
| 02:28 | <Philip`> | As far as web browsers are concerned, namespaces are just meaningless strings |
| 02:28 | <Philip`> | It doesn't even need to be a syntactically valid URI |
| 02:37 | <JonathanNeal> | Philip`, when I make it null, I notice I lose the ability to work with the element in Firefox 2. |
| 02:39 | <JonathanNeal> | Basically --- I'm seeing if there's something someone has overlooked when it comes to getting html5 elements to work in firefox 2. |
| 02:42 | <JonathanNeal> | For instance, http://sandbox.thewikies.com/xmlns/ --- exact same code, difference being the xmlns. |
| 02:44 | <JonathanNeal> | In Chrome, Firefox 3, Firefox 2, the null namespaced element is totally bunk. |
| 02:51 | <tabatkins> | JonathanNeal: Why are you caring about FF2? |
| 02:51 | <JonathanNeal> | Well, it's a small, small niche population of browsers that got left out of html5. |
| 02:51 | <tabatkins> | Unrelated: I'm having fun playing around with the mandelbrot set in canvas. I'll have a new blog background in a few minutes. |
| 02:51 | <tabatkins> | JonathanNeal: ...and? |
| 02:51 | <JonathanNeal> | And I wanted to help 'em out. |
| 02:52 | <tabatkins> | Why? They're scarily insecure, not to mention out-of-date. |
| 02:53 | <nimbupani> | tabatkins: I knew there is a very minor population stuck of solaris machines with custom build of Firefox 2 |
| 02:53 | <tabatkins> | That doesn't change the above. Sucks that they're stuck with some custom version of a very old, insecure browser. They should change. |
| 02:54 | <JonathanNeal> | I could say the same for ie6, 7, and 8, but we took the time to hack them into presentational awareness of html5. |
| 02:54 | <JonathanNeal> | I'm just trying the same thing with Firefox 2. |
| 02:56 | <tabatkins> | You really can't say the same. They're certainly old and insecure, but they're not a tiny population. They demand catering purely due to their numbers. |
| 02:57 | <tabatkins> | You can do what you want, I'm just saying I think it's a waste of time. ^_^ |
| 02:58 | <JonathanNeal> | So I'm trying to learn more about how createElementNS works. |
| 02:58 | <JonathanNeal> | because document.createElementNS('http://www.w3.org/1999/xhtml';, 'section'); and document.createElementNS('html5', 'section'); produce very different kinds of elements. |
| 03:02 | <JonathanNeal> | I've been trying to get any xmlns elements to work in firefox 2, actually. |
| 03:16 | <tabatkins> | Reasonably happy with my new blog background. I'll at least keep it up for a little while: http://www.xanthir.com/blog/ |
| 03:34 | <aho> | tabatkins, http://i.imgur.com/RQLSj.png <- are you really satisfied with that? :> |
| 03:35 | <aho> | you probably shouldn't generate ~160kb of markup... especially not if it's that slow and uncompressed :f |
| 03:37 | <aho> | http://i.imgur.com/y5EVt.png <- looks a lot better while loading |
| 03:46 | <nimbupani> | tabatkins website is a new performance metric for canvas |
| 03:50 | <tabatkins> | aho: Interesting. The background isn't working in FF, but it's not throwing errors either. |
| 03:50 | <tabatkins> | Hrm. |
| 03:52 | <aho> | looks better now |
| 03:52 | <aho> | but still awful ;> |
| 03:52 | <tabatkins> | aho: And yeah, I need to actually implement an archiving functionality, rather than just sending out every post I've written. |
| 03:52 | <tabatkins> | Look at it in Chrome instead. ^_^ |
| 03:54 | <tabatkins> | aho: It'll look better in Firefox now, since FF isn't doing the canvas for some reason. |
| 03:55 | <aho> | that style block at the very bottom... put that at the top |
| 03:55 | <aho> | i'm using noscript |
| 03:55 | <tabatkins> | Nah, that's not it. FF isn't doing the canvas on my machine either. |
| 03:57 | <aho> | try window.onload |
| 03:57 | <tabatkins> | Ah, that does it. |
| 03:58 | <tabatkins> | It also reveals that Chrome's canvas handling is *way* faster than FF's. |
| 03:58 | <tabatkins> | So I need to put in some dynamic scaling. |
| 04:10 | <aho> | first you should put all your styles at the top |
| 04:10 | <aho> | it's better if the site looks right right off the bat |
| 04:11 | <aho> | and not 10, 20, or even 30 seconds later :> |
| 04:12 | <tabatkins> | My styles are at the top. The only thing at the bottom is the style for the <canvas>, also at the very bottom. |
| 04:13 | <aho> | all styles should be at the top |
| 04:13 | <aho> | google says so, yahoo says so, and i also say so :> |
| 04:14 | <tabatkins> | That, um, doesn't matter. The style at the bottom doesn't reflow anything. |
| 04:16 | <aho> | well, before you were setting the background to black there... that looked pretty bad without javascript, that is |
| 04:16 | <tabatkins> | Yup, and I fixed that. |
| 04:46 | <tabatkins> | Dynamic scaling works great now. FF properly throttles itself down to an extremely low level, while Chrome gets to render at closer to the initial settings on my comp. |
| 04:48 | <tabatkins> | That was a fun experiment. |
| 05:01 | <annevk> | while Skype Access is somewhat expensive, not having to deal with the silly login screens is awesome |
| 05:02 | <annevk> | also, yay for waking up ridiculously early |
| 22:45 | <gsnedders> | Hmm, interesting… |
| 22:46 | <gsnedders> | Accessing the HTML element in a same-origin frameset document is fine, but accessing the FRAMESET element in the same document throws SECURITY_ERR |
| 22:46 | <gsnedders> | Someone explain? |
| 22:57 | <gsnedders> | Oh, duh |
| 22:57 | <gsnedders> | It's not SECURITY_ERR |
| 22:57 | <gsnedders> | It's just the iframe hasn't loaded yet :) |
