| 00:26 | <Hixie_> | woah, the w3c archives archive across months now |
| 00:26 | <Hixie_> | sweet |
| 00:27 | <Hixie_> | othermaciej: can i link to http://esw.w3.org/topic/HTML/ChangeProposalTemplate from http://dev.w3.org/html5/status/issue-status.html ? |
| 00:27 | <othermaciej> | Hixie_: sure |
| 00:27 | <othermaciej> | Hixie_: that would be convenient and helpful in fact |
| 00:36 | <Hixie_> | done |
| 00:57 | <othermaciej> | Hixie_: have you had a chance to look at the I18N WG's proposal for multiple languages in meta, or Larry's for URLs? |
| 00:58 | <othermaciej> | Hixie_: no urgent deadline on those but your input could potentially save time if it allows dispensing with the formal bureaucracy |
| 01:01 | <MikeSmith> | Hixie_: I note that the spec says "The presence of a name attribute on an a element, if its value is not the empty string." .. so what should happen for the case of <a name=''> -- report it as an error? |
| 01:08 | <Hixie_> | othermaciej: do they have bugs filed? |
| 01:09 | <Hixie_> | MikeSmith: context? |
| 01:10 | <othermaciej> | Hixie_: Larry's thing is in the form of a Change Proposal |
| 01:11 | <othermaciej> | Hixie_: the I18N guys suggested a change in an informal way and said they talked to you about it |
| 01:11 | <othermaciej> | Hixie_: would you like that in the form of a bug? |
| 01:11 | <Hixie_> | i believe everything we spoke about in the i18n group has a bug already |
| 01:11 | <othermaciej> | Larry's thing has a bug as well as a Change Proposal I think (though the bug references a dated version) |
| 01:11 | <Hixie_> | right now my priority is getting the webapps draft to LC |
| 01:17 | <MikeSmith> | Hixie_: immediate context is http://bugzilla.validator.nu/show_bug.cgi?id=685 .. I'm trying to figure out what would be best to have validator.nu report for the case of <a name=''> . It currently just says, "Bad value for attribute name on element a: An ID must not be the empty string." My question is, what's the intent of the spec in distinguishing between the cases of <a name="foo"> and <a name=""> ? Is it useful for conforman |
| 01:17 | <MikeSmith> | ce checkers to report something different for those cases? |
| 01:18 | <Hixie_> | <a name="foo"> should warn that name="" is deprecated. <a name=""> should just say name="" is invalid, I think. |
| 01:18 | <Hixie_> | s/that name="" is deprecated/that id="" is considered a more reliable way of giving fragment identifier targets/ |
| 01:19 | <Hixie_> | <a name=""> should be an error because these obsolete features are only allowed to ease transition, and there's nothing to transition if the attribute's value is "". |
| 01:19 | <MikeSmith> | OK, that makes sense |
| 01:19 | <MikeSmith> | for the <a name=""> case, v.nu now reports the warning, "The name attribute is obsolete. Consider putting an id attribute on the nearest container instead." |
| 01:20 | <Hixie_> | (why the nearest container?) |
| 01:20 | <MikeSmith> | so for the <a name=""> , we can have it report the error and also report that same warning |
| 01:20 | <MikeSmith> | Hixie_: I dunno. That's verbatim from your spec text. |
| 01:20 | <Hixie_> | that works too |
| 01:21 | <Hixie_> | the term "nearest container" is not in html5 |
| 01:22 | <MikeSmith> | The logic of also reporting the warning is that somebody first gets the error, then changes their content to <a name=foo> and then gets a warning about, then that's just annoying -- they are going to say, you should have told me that the first time. |
| 01:22 | <MikeSmith> | Hixie_: hmm, I got it from somewhere. |
| 01:22 | MikeSmith | goes to look at current draft |
| 01:22 | <Hixie_> | agreed, the text for name="" should just say 'use id=""' |
| 01:22 | <Hixie_> | '...and give it a value!' |
| 01:23 | <MikeSmith> | erp, maybe I got that "nearest container" bit from Anne's HTML4-HTML5 differences doc |
| 02:01 | <TabAtkins__> | MikeSmith: The 'nearest container' bit still makes sense, as it was very common to just use an empty <a name> at the top of a container to target it. |
| 02:01 | <MikeSmith> | TabAtkins__: OK |
| 02:02 | <MikeSmith> | so I guess it's fine to keep it in there and the wording can be refined later if needed |
| 02:02 | <TabAtkins__> | Yeah. |
| 02:03 | <Hixie_> | maybe it's referring to the case of an empty <a name=""></a> ? |
| 02:03 | <MikeSmith> | yeah, I reckon so |
| 02:04 | <MikeSmith> | that seems like the most common way it's actually seen in existing content |
| 02:15 | <erlehmann> | Hixie_, is it possible that the figure construct eventually will be changed back again or are all arguments coming in on the list now re-hashes of ye olde conflict ? |
| 02:16 | <Hixie_> | othermaciej would be a better person to answer that question |
| 02:17 | othermaciej | assumes erlehmann was asking about the arguments on the whatwg list |
| 02:17 | <othermaciej> | since there are not any coming in on public-html afaik |
| 02:18 | <erlehmann> | othermaciej, did you inherit the honors of supreme HTML5 commander from Hixie somehow ? ;) |
| 02:18 | <erlehmann> | because i mean the whatwg list. |
| 02:19 | <erlehmann> | hmm, better would be "i'm meaning", would it ? |
| 02:19 | <Hixie_> | oh, i thought the discussion was on a w3c bug |
| 02:19 | <othermaciej> | there's different discussion on a w3c bug |
| 02:19 | <Hixie_> | on the whatwg list, rehashing old arguments without introducing new data will not have any effect |
| 02:19 | <erlehmann> | oh, sorry |
| 02:19 | <Hixie_> | i haven't checked to see if recent threads introduce new data |
| 02:19 | <othermaciej> | someone suggested restricting the content model for <figure> |
| 02:20 | <othermaciej> | based on what I think were originally incorrect assumptions (though I am not sure why the argument still continues) |
| 02:21 | <erlehmann> | i am not sure too. dt / dd may be ugly, but it gets the job done. |
| 02:21 | <Hixie_> | i need to get through this webapps stuff then i'll be able to do a better job of keeping on threads |
| 02:21 | <Hixie_> | if anyone sees anyone being even remotely rude on the whatwg list, please let me know asap |
| 02:22 | <othermaciej> | the recent figure bug was orthogonal to dt/dd I think |
| 02:22 | <erlehmann> | oh, i was remotely rude today. |
| 02:22 | <erlehmann> | misunderstood TabAtkins, but we reconciled pretty fast. |
| 02:23 | <erlehmann> | othermaciej, what was it ? browsers aren't doing strange things with the dom again, are they ? |
| 02:24 | <othermaciej> | erlehmann: I'm actually not entirely sure what the motivation for the bug was |
| 02:25 | <erlehmann> | bet aesthetical reasons, hehe |
| 02:25 | Hixie_ | is still unsure how to handle some of the htmlwg feedback (especially shelley's "remove XYZ" bugs) |
| 02:55 | <TabAtkins__> | Restricting <figure>'s content model won't work. A common use of figures in, frex, academic journals is to hold tables. |
| 02:56 | <webben> | That bug would benefit from documentation of examples. |
| 02:59 | <TabAtkins__> | It's possible I'm misunderstanding the purpose of <figure>, of course. My impression of it, though, was for content that is part of the main content, but whos exact location within the content is not important. |
| 02:59 | <TabAtkins__> | That is, after all, what figures are. They're referred to, and are part of articles, but where in the article they're actually placed doesn't matter. |
| 03:00 | <TabAtkins__> | This is in opposition to <aside>, which is content that is only tangentially related to the main content, and could be removed without changing the meaning of the main content. |
| 03:00 | <webben> | that's part of what figures are. |
| 03:00 | <webben> | well, actually, I wouldn't define them like that |
| 03:01 | <webben> | often the main content requires the figure - the only thing that allows them to be removed is the reference that points to them. |
| 03:01 | <TabAtkins__> | Alternate conversation: Can anyone think of *any* sane reason why I'm being given a forbidden error when trying to run this .pl file, but if I change it to a .txt or .php file it works fine? |
| 03:01 | <TabAtkins__> | webben: True, but it's also very common practice to do so - to remove it from the main flow and insert a reference. |
| 03:02 | <TabAtkins__> | The point is that they represent additional data that is not part of the flow of the article. |
| 03:02 | <TabAtkins__> | But is still *part* of the article and is necessary for full understanding of it. |
| 03:02 | <webben> | I'm still not sure that's right. |
| 03:03 | <webben> | depends what "flow" refers to - they can be entirely essential to the flow of ideas. |
| 03:03 | <TabAtkins__> | Well, my definition happens to correspond pretty exactly to the current spec text, luckily enough. ^_^ |
| 03:04 | <TabAtkins__> | (I didn't actually realize it was that close.) |
| 03:04 | <erlehmann> | TabAtkins__, messed up .htaccess |
| 03:04 | <erlehmann> | ?? |
| 03:05 | <TabAtkins__> | erlehmann: Not that I can *tell*. In fact, I have just *added* an htaccess, as there wasn't one there before. All it contains is an "Options ExecCGI" line. |
| 03:05 | <erlehmann> | well, then … |
| 03:05 | <erlehmann> | no idea ^_^ |
| 03:05 | <TabAtkins__> | (That would be a sane reason, though, although it's wrong.) |
| 03:07 | <TabAtkins__> | webben: Sometimes captioned images are indeed part of the normal flow of an article, I agree, and these are well-served by <figure> as well. |
| 03:13 | <webben> | but anyway I more meant - Shelley's claim is allowing tables and code listings and other text in "figure" is confusing because that's not what people call figures. If there's evidence to the contrary (as seems likely to me) would be good to cite specific examples. (I see only vague references in the ticket.) |
| 03:14 | <TabAtkins__> | I can go grab textbooks off my shelf that have code samples as figures. |
| 03:15 | <webben> | here's a specific example: http://tinyurl.com/ylrv4qe |
| 03:15 | <TabAtkins__> | Heh, that works. |
| 03:16 | <TabAtkins__> | And here's a table used as a figure: http://books.google.com/books?id=Fsz234KPVz8C&pg=PT120&dq=%22Figure+5%22&ei=I4wUS8raFJjQNPHJ2ZQL#v=onepage&q=%22Figure%205%22&f=false |
| 03:16 | <TabAtkins__> | I'll add these to the bugs. |
| 03:17 | <webben> | here's another data table labelled as a figure: http://tinyurl.com/yfksktb |
| 03:19 | <TabAtkins__> | Darn, I should have waited. I just submitted it. |
| 03:19 | <TabAtkins__> | Still, it takes 30 seconds to find many examples. |
| 03:19 | <TabAtkins__> | So saying that using tables in figures will cause confusion is silly. |
| 03:20 | <webben> | or at least, saying that people don't put tables in "figures" is empirically false. |
| 03:21 | <webben> | (it might be a real practice, but still be confusing I guess) |
| 03:21 | <TabAtkins__> | True, but it's common enough that I highly doubt there is any confusion. |
| 03:29 | <TabAtkins__> | Okay, what are the reasons the cgi-bin folder on a linux box would be forbidden? It has 755 permissions, and has both owner and group set to the values specified in apache2.conf |
| 04:17 | <Sirisian> | When you submit a review comment in the HTML 5 draft do people actually read those? Is that the most effective way to give feedback? |
| 04:25 | <MikeSmith> | Sirisian: you mean submitting it using the built-in form in the whatwg version of the spec? |
| 04:25 | <Sirisian> | yeah |
| 04:28 | <MikeSmith> | Sirisian: yeah, people do read all those. When you submit a comment that way, it actually causes a new issue to be opened in the W3C bugzilla |
| 04:29 | <Sirisian> | yeah I saw that. I added a comment about the canvas element. Just wanted to make sure that wasn't like a "this is for spelling mistakes" kind of thing |
| 04:30 | <Sirisian> | MikeSmith, so is it just one person in their spare time "Ian" that goes through and makes the changes? |
| 04:30 | <MikeSmith> | Sirisian: so you can also submit a comment directly using http://www.w3.org/Bugs/Public/enter_bug.cgi?product=HTML%20WG |
| 04:31 | <Sirisian> | yeah I saw that |
| 04:31 | <MikeSmith> | Sirisian: Ian is Hixie_ and yeah he is the one who actually makes the changes to the spec |
| 04:31 | <MikeSmith> | to that spec |
| 04:31 | <MikeSmith> | but he does it full time, not spare time |
| 04:32 | <Sirisian> | ah cool. I like that idea. |
| 04:32 | <MikeSmith> | there are quite a number of other people who follow all the bugzilla issues |
| 04:32 | <Sirisian> | yeah I noticed the issue rating system and such |
| 04:34 | <Sirisian> | Also this might seem silly, but is there something similar to the HTML 5 draft but for javascript and how browsers implement and interact with it? I've noticed over the years that browsers seems to use it slightly differently for input related things especially the mouse. |
| 04:35 | <MikeSmith> | Sirisian: not for Javascript itself, not that I know of |
| 04:37 | <MikeSmith> | but for existing APIs that get exposed to Javascript, there is some documentation in HTML5 that addresses improving interoperability of some of those |
| 04:37 | <MikeSmith> | and there's also the beginning of a "Web DOM Core" draft that attempts to document more |
| 04:38 | <MikeSmith> | Sirisian: http://simon.html5.org/specs/web-dom-core |
| 04:40 | <MikeSmith> | but if you mean differences in the way that browsers handle Javascript other than how they expose standard APIs, then that's more an issue of either bugs in particular browsers, or bugs/ambiguities in the Javascript spec that need to be fixed |
| 04:41 | <Sirisian> | nah not so much dom, but I was thinking of the mouse input and the event model associated with it. That's a good idea. |
| 04:41 | Sirisian | searches for some kind of spec |
| 04:41 | <erlehmann> | if anyone wants to test my wordpress plugin generating microdata licensing info I coded in the last few hours (thanks, foolip) => http://github.com/erlehmann/cc-figure |
| 05:00 | <MikeSmith> | Sirisian: as far at the event model, you might want to look at the DOM3 Events draft |
| 05:01 | <MikeSmith> | http://dev.w3.org/2006/webapi/DOM-Level-3-Events/html/DOM3-Events.html |
| 05:04 | <Sirisian> | ah thank you |
| 05:04 | <Sirisian> | I wasn't sure if it would be in the Dom stuff |
| 05:04 | <MikeSmith> | that document in part attempts to document how event handling actually works in current browsers |
| 05:04 | <MikeSmith> | so-called DOM Level Zero |
| 05:06 | <Sirisian> | Why? |
| 05:07 | <Sirisian> | Shouldn't it just list what the browsers should do? |
| 05:07 | <Sirisian> | Or is that what you meant? |
| 05:11 | <MikeSmith> | Sirisian: there would not be much point in it listing what browsers should do if browser instead actually did something different |
| 05:12 | <MikeSmith> | the DOM Level 0 stuff is basically what browsers have always done but that was never actually specified |
| 05:12 | <MikeSmith> | but there are some differences among browsers with regard to some of it |
| 05:14 | <MikeSmith> | so the value is speccing it at this point is that anybody developing a new browser would not need to reverse engineer it |
| 05:16 | <Sirisian> | Why is the spec so vague about the mouse stuff? Even the default action stuff is extremely vague. No wonder IE handles the 3 mouse buttons as 7 buttons and why opera has no prevent default in middle mouse or right mouse button (speaking of the context menu). It's not even defined. http://dev.w3.org/2006/webapi/DOM-Level-3-Events/html/DOM3-Events.html#events-Events-MouseEvent-button talk about vague |
| 05:18 | <MikeSmith> | Sirisian: the spec is still in development. If you find things it doesn't address but should, the best place to comment is on the public-webapps⊙wo mailing list |
| 05:18 | <Sirisian> | I don't do mailing lists. I just lose track of messages and get confused. |
| 05:26 | <MikeSmith> | Sirisian: OK, you can comment through bugzilla |
| 05:26 | <MikeSmith> | http://www.w3.org/Bugs/Public/enter_bug.cgi?product=WebAppsWG |
| 05:26 | <MikeSmith> | and select component "DOM3 Events" |
| 05:27 | <Sirisian> | Thank you for being pacient with :) |
| 05:31 | <MikeSmith> | Sirisian: "enlightened self interest" .. the easier it is for people to review and comment, the greater the chance of getting more good feedback and possibility of improvements |
| 05:32 | <Sirisian> | then why did they use bugzilla? |
| 05:33 | <MikeSmith> | Sirisian: point taken |
| 05:34 | <MikeSmith> | (if your point is that bugzilla is not an easy thing for most people to use) |
| 05:34 | <MikeSmith> | but some people would rather submit through bugzilla than join a mailing list |
| 05:34 | <Sirisian> | yeah that was my point. It took me a while to figure out how to view all of the issue. Someone told me about the product:html thing though and it made things easier. |
| 05:34 | <MikeSmith> | perhaps more specs should have the built-in comment form like the HTML5 spec does |
| 05:35 | <Sirisian> | yeah that's a clever setup they have there. |
| 05:36 | <MikeSmith> | they is Hixie, in this case |
| 05:37 | <Hixie_> | yeah i like it |
| 05:37 | <Hixie_> | we should do that more |
| 05:46 | <othermaciej> | I only wish the built-in comment form had a way for people to identify themselves if they choose |
| 05:51 | <MikeSmith> | othermaciej: yeah, me too |
| 05:51 | <MikeSmith> | though some people have just taken to including their name or initials in the text of the comment |
| 06:03 | <Hixie> | why does the originator matter? |
| 06:03 | <Hixie> | i ignore it anyway when dealing with feedback |
| 06:14 | <MikeSmith> | Hixie: it matters on that some people would like to actually identify themselves when submitting a comment |
| 06:15 | <MikeSmith> | yeah, I know they could just use bugzilla directly to do that |
| 06:15 | <MikeSmith> | but doing it from the built-in comment form is more convenient |
| 06:16 | <MikeSmith> | having it be an optional second input field in the form would seem to be the best of both worlds |
| 06:18 | <Sirisian> | What's that feature in browsers when you middle click? autoscroll? |
| 06:19 | <Hixie> | MikeSmith: i see that they do, but why do they do? |
| 06:19 | <MikeSmith> | Sirisian: my pointing device doesn't have a middle click |
| 06:20 | <MikeSmith> | Hixie: I guess in part to save the trouble of posting a comment to the bug later and needing to say, "I'm the one who submitted this comment." |
| 06:20 | <Sirisian> | I almost did that. |
| 06:20 | <MikeSmith> | and/or so that others know clearly who to contact if they want to get clarification outside of bugzilla -- e.g., on IRC |
| 06:21 | <MikeSmith> | so if Sirisian files a bug and I see it, I can ping him here to say, "Hey Sirisian, about that bug you just posted, I have a quick question.." |
| 06:22 | <Hixie> | ah |
| 06:22 | <othermaciej> | Hixie: when I see an IP address comment that's insightful, sometimes I wonder who it is |
| 06:22 | <Hixie> | well if anyone wants to submit a patch to do something different, i'm happy to let them have the code :-) |
| 06:23 | <othermaciej> | Hixie: I don't make IP address comments myself, because I want myself and others to be able to find them later |
| 06:23 | <Hixie> | (personally though i don't really care who wrote the comment :-) ) |
| 06:25 | <Sirisian> | Hixie, so you're really the person that reads all the issues and makes the changes? |
| 06:25 | <Hixie> | yup |
| 06:25 | <Sirisian> | How did you get picked? |
| 06:37 | <Hixie> | Sirisian: i volunteered |
| 06:38 | <Sirisian> | ah cool |
| 06:40 | <Hixie> | should i call the web database spec Web SQL Database, Web SQL Database API, Web SQLite Database, or something else? |
| 06:47 | <nessy> | not SQLite - the others sound ok |
| 06:49 | <Mannerisky> | for what it's worth, I think WSQLDAPI is a long-enough acronym |
| 06:51 | <Hixie> | i went for Web SQL Database |
| 06:52 | <MikeSmith> | good |
| 06:53 | <MikeSmith> | I think the API part isn't necessary in spec titles if the Web qualifier it in the title |
| 06:53 | <MikeSmith> | e.g., Geolocation API could well be "Web Geolocation" |
| 06:53 | <MikeSmith> | maybe |
| 06:54 | <Hixie> | i just stick "Web" on all my specs so that they start with a W instead of an X |
| 06:54 | <MikeSmith> | heh |
| 06:54 | <MikeSmith> | just a long as you don't start starting them with "WS-", please |
| 07:30 | <MikeSmith> | one feature that's in git and mercurial that I will I really had with subversion is the way you can interactively select specific parts of changes to commit |
| 07:30 | <MikeSmith> | "git add --patch" and "hg record" |
| 07:31 | <MikeSmith> | hsivonen: another good reason for moving to mercurial or git ↑ |
| 07:50 | <MikeSmith> | Hixie: did the spec previously recommend doing "a img, img[usemap] { border: 0; }" in place of using the border attribute? |
| 07:51 | <Hixie> | i don't think html5 ever recommended specific CSS rules |
| 07:52 | <MikeSmith> | hmm, this is another case of I wonder where it came from then -- validator.nu warning about it advises to do "a img, img[usemap] { border: 0; }" instead |
| 07:53 | <MikeSmith> | similar to the advice about putting an id on nearest container instead of a/@name |
| 07:53 | <Sirisian> | heh. https://connect.microsoft.com/IE/SearchResults.aspx?FeedbackType=2&Status=1&Scope=0&SortOrder=15&TabView=1 |
| 07:53 | <Sirisian> | number 1 bug is that the canvas tag doesn't work. |
| 07:54 | <MikeSmith> | I guess hsivonen included that advice in the initial version of the warnings-reporter when he added it |
| 07:57 | jgraham | grumbles about getting 36 bugzilla emails in 9 hours, all from the same bug |
| 07:57 | <foolip> | w3c spec bug, perchance? |
| 07:58 | <Hixie> | this is a terrible article http://en.wikipedia.org/wiki/Push_technology |
| 07:58 | <jgraham> | foolip: Indeed |
| 08:00 | <foolip> | not that I'm a masochist, but is there a way to follow those bugs with manually CC:ing yourself on all those which interest you? |
| 08:00 | <foolip> | er, manually adding yourself to the CC list rather |
| 08:00 | <jgraham> | Yeah, ou can do that from the bugzilla page |
| 08:01 | <foolip> | jgraham: care to link to that particular flame war for my breakfeast entertainment? |
| 08:01 | <jgraham> | http://www.w3.org/Bugs/Public/show_bug.cgi?id=8404 |
| 08:02 | <jgraham> | Not sure if it is a flame war, I haven't actually read any of the mail yet |
| 08:02 | <foolip> | go <figure> |
| 08:02 | <Hixie> | can anyone find a reference to the term "long polling" before the HTTP technique? |
| 08:03 | <Hixie> | did someone just make that up for HTTP for no good reason? or does it actually mean something? |
| 08:03 | <foolip> | I thought it quite specifially referred to keeping HTTP connections open |
| 08:25 | <Sirisian> | I've never heard of it used except in the context of the HTTP thing. It's a common practice. I've used it before since it allows you to communicate back and forth with a client using ajax and pretend like it's a socket. Then again you can only keep the connection open for 30 seconds. |
| 08:26 | <Hixie> | if the term was invented for HTTP specifically, then "long polling" is a pretty dumb name for it |
| 08:27 | <Hixie> | then again, the web is full of dumb names |
| 08:27 | <Hixie> | DHTML, REST, AJAX, Comet... |
| 08:55 | <Philip`> | About identifying bugs: I do it so that I can later search for the bugs I filed and see if they were addressed satisfactorily or not |
| 08:57 | <Hixie> | yeah, but that just needs a unique string, it doesn't need an identifier |
| 09:01 | <Philip`> | A unique string is an identifier :-) |
| 09:03 | <Hixie> | i mean, a person identifier |
| 09:03 | <Hixie> | that is recognisable as such |
| 09:18 | <gsnedders> | We need some spec for normalizing web addresses |
| 09:19 | <gsnedders> | As there are limits to how much percent encoding normalization you can do |
| 09:19 | <gsnedders> | You can't, for example, normalize pct-encoded & and = without breaking sites |
| 09:20 | <Hixie> | when would this be done? |
| 09:20 | <gsnedders> | Comparing IRIs for caching purposes, for example |
| 09:20 | <Hixie> | seems like something the IRI spec should define; send mail to larry |
| 09:22 | <Lachy_> | gsnedders, how would it break sites if it's being done for comparison purposes? |
| 09:23 | <gsnedders> | Lachy_: You use the cache for the wrong page due to over-normalizing |
| 09:23 | <gsnedders> | So you display the wrong page to the user |
| 09:23 | <gsnedders> | Which seems _really_ bad. |
| 09:24 | <Lachy_> | ok, can you give an example of over normalising? |
| 09:24 | <gsnedders> | http://www.publicbroadcasting.net/mpr/.eventsfeed?cat=Rock%2C%20Roots%2C%20%26%20Other&start=10122009&end=12312010 --- decode the escaped ampersands in there, and you get a different resource |
| 09:25 | <Lachy_> | ah, ok |
| 09:25 | <gsnedders> | If you normalize before cache lookup/save, that can break badly. |
| 09:26 | <Lachy_> | normalising should never decode ampersands, so I don't get why you're calling it over-normalising. But I get what you mean anyway. |
| 09:26 | <gsnedders> | Lachy_: Why shouldn't it? |
| 09:26 | <gsnedders> | Lachy_: ampersands are just another character in the query string |
| 09:27 | <gsnedders> | Lachy_: Point me to the bit in the spec that says I shouldn't normalize pct-encoded ampersands. |
| 09:27 | <Lachy_> | well, on the server side, I guess. |
| 09:27 | <Lachy_> | on the client side, you'd never want %26 to be wrongly interpreted as & in a query sting |
| 09:28 | <Lachy_> | outside of the query string, decoding it would be fine though |
| 09:29 | <Lachy_> | similar issue with %3F for ? |
| 10:00 | Philip` | is disappointed to find that <!DOCTYPE html PUBLIC “-//WHATWG//HTML 20100401"> is almost but not quite quirky, by the rules in http://philip.html5.org/docs/quirks.txt |
| 10:01 | <Philip`> | s/“/"/ |
| 10:02 | <Philip`> | (Disappointed because it was almost a practical demonstration of why making up random strings is a bad idea) |
| 14:45 | <Philip`> | Someone less lazy than me should suggest proposing <meta name="designed-for-specs" content="http://www.w3.org/TR/html5/">; or content="http://www.w3.org/TR/2009/WD-html5-20090825/ http://www.w3.org/TR/2009/WD-rdfa-in-html-20091015/"; etc |
| 14:45 | <Philip`> | instead of trying to squeeze the information needlessly into the most complex syntactic feature of HTML5 with the most compatibility problems |
| 14:46 | <TabAtkins> | Okay. |
| 14:47 | <TabAtkins> | I can't play this "You're wrong", "That's your opinion", "You're wrong that it's an opinion" game with Shelley for too long. |
| 14:47 | <gsnedders> | TabAtkins: Can you stop filling up that mailbox |
| 14:48 | <Philip`> | It seems unlikely to be productive to say "You're wrong" to people |
| 14:48 | <TabAtkins> | Philip`: And when they're actually wrong? And keep saying things that are wrong, and using those wrong things as justification for their arguments? |
| 14:48 | <Philip`> | Better to just state your arguments and reasoning, and trust that reasonable people will agree with you if you're right |
| 14:48 | <TabAtkins> | It's really hard to see what one can do except point out, "Uh, yeah, here are examples proving you wrong." |
| 14:48 | <Philip`> | (and trust that reasonable people will determine what goes in the spec) |
| 14:49 | <jgraham> | Philip`: It's not clear that the processed is biased in that way |
| 14:49 | <jgraham> | *process |
| 14:49 | <TabAtkins> | gsnedders: Sorry, but to be fair, I'm not the largest source of emails on that thread. |
| 14:50 | Philip` | looks at the thread again |
| 14:50 | <Philip`> | Oh, Shelley already basically said what I just said |
| 14:50 | <TabAtkins> | Philip`: It's just impossible to actually *argue* with someone who is saying things that are demonstrably wrong. If you don't correct them, then *at best* you're shouting over them and hoping to reach the choir. |
| 14:51 | <annevk> | You could just ignore it |
| 14:51 | <Philip`> | So don't waste so much effort trying to argue with them :-) |
| 14:51 | jgraham | has now got 56 messages in 18 hours or so |
| 14:52 | <Philip`> | s/argue with/respond to/ |
| 14:52 | <jgraham> | s/18/20 |
| 14:52 | <jgraham> | From that bug |
| 14:54 | <TabAtkins> | annevk: Yeah, I could. But then I'd have to address the arguments later on when they start wielding the process like a hacksaw. And in the meantime it's possible that people are swayed over to the wrong side due to them not realizing that there are false arguments being employed. |
| 14:56 | <jgraham> | TabAtkins: You may as well ignore all bugs. Eventually the controversal ones will all get turned into change proposals at which point they have to be unignored |
| 14:57 | <jgraham> | Nothing you say upfront can possibly prevent that happening |
| 14:57 | <TabAtkins> | Can quirks mode trigger parsing changes? That is, can a page produce a different DOM based on which mode it's in? |
| 14:57 | <annevk> | yes |
| 14:57 | <jgraham> | TabAtkins: Yes, there is one difference |
| 14:57 | <TabAtkins> | jgraham: Sigh, I guess you're right. But they're Wrong On The Internet. |
| 14:57 | <TabAtkins> | annevk, jgraham: k thanks. |
| 14:58 | <TabAtkins> | Philip`: Suggestion made. |
| 15:02 | <AryehGregor> | TabAtkins, the process doesn't require anyone to talk to each other. If the group doesn't come to consensus, you just have to write the proposals to convince the chairs, as far as I can tell. |
| 15:03 | <TabAtkins> | AryehGregor: Sure, it doesn't require it. But discussion is the whole point of having a list. |
| 15:03 | <Philip`> | TabAtkins: Thanks |
| 15:04 | <AryehGregor> | TabAtkins, well, no. Getting a better spec is the point of having the list. Discussion is only a means to an end. |
| 15:04 | <TabAtkins> | Sorry, I'll clarify. Discussion is typically a vital part of producing a better spec. |
| 15:14 | <AryehGregor> | Some types of discussion are. Some discussion is just noise. |
| 15:17 | <daedb> | I can't believe there's actually a suggestion to allow iframe in figure, but not directly allow tables/code/etc in there... |
| 15:19 | <annevk> | currently those are allowed |
| 15:20 | <daedb> | I know, I'm reading the bug that wants to disallow them |
| 15:20 | <annevk> | oh that |
| 15:22 | <zcorpan_> | is "-//H//T M//L" a legal FPI? |
| 15:22 | <daedb> | I just stared in disbelief when I saw "another option would be to allow iframe"... |
| 15:23 | <annevk> | zcorpan_, does it matter? |
| 15:24 | <zcorpan_> | in other news the secret xml-stylesheet draft will hopefully soon be published as a FPWD |
| 15:24 | <Philip`> | daedb: Where was <iframe> suggested? (I only remember <object>) |
| 15:25 | <gsnedders> | zcorpan_: I think not, but I haven't a copy of the spec here |
| 15:25 | <Philip`> | Oh, in the bug |
| 15:25 | <daedb> | http://www.w3.org/Bugs/Public/show_bug.cgi?id=8404#c28 "Another option is to add one more element: iframe" |
| 15:25 | <zcorpan_> | gsnedders: why? because "L" is not a legal language tag? |
| 15:26 | <zcorpan_> | or is "NONSGML" an allowed string and "T" is not? |
| 15:27 | AryehGregor | wants to see what happens when some of these Change Proposals actually get past the bureaucratic machinery and get resolved |
| 15:27 | <Philip`> | I usually use "figure" to express the semantics "this would cross over a column/page break if I inserted it inline" |
| 15:27 | <gsnedders> | zcorpan_: http://xml.coverpages.org/tauber-fpi.html |
| 15:30 | <zcorpan_> | public_text_class = |
| 15:30 | <zcorpan_> | CAPACITY | CHARSET | DOCUMENT | DTD | ELEMENTS | ENTITIES | LPD | NONSGML | NOTATION | SHORTREF | SUBDOC | SYNTAX | TEXT |
| 15:30 | <Philip`> | Use "-////TEXT HT//ML" |
| 15:32 | <zcorpan_> | Maltese? |
| 15:32 | <Philip`> | Malayalam |
| 15:32 | <Philip`> | obviously |
| 15:32 | <zcorpan_> | ah |
| 15:32 | <Philip`> | Oh, how confusing |
| 15:33 | <Philip`> | http://www.w3.org/WAI/ER/IG/ert/iso639.htm puts ml next to Maltese in the 3-letter section |
| 15:33 | <Philip`> | but it's MT in the 2-letter section |
| 15:34 | <Philip`> | Guess that's a bug in the W3C page |
| 15:37 | <Philip`> | Anyway, I guess it doesn't matter since Maltese and Malayalam are equally accurate descriptions of the markup language |
| 15:45 | <foolip> | AryehGregor: had any luck writing a change proposal of your own? |
| 15:45 | <AryehGregor> | foolip, no, nothing much I want changed that the editor doesn't. |
| 15:46 | <foolip> | AryehGregor: I mean the counter-proposal, the "empty change proposal" as it were |
| 15:46 | <AryehGregor> | I haven't tried writing one yet. |
| 15:46 | <foolip> | k |
| 15:46 | <AryehGregor> | I've thought of it, but I don't have the time to do it justice, and am not really the most qualified person on any given issue that I've seen. |
| 15:47 | <foolip> | sorry, I'm being a retard, I mistook you for TabAtkins just ehn |
| 15:47 | <AryehGregor> | I'm mostly interested in what will happen when people figure out that the HTMLWG is more or less just a dictatorship of the chairs instead of the editor, in the end. Just with a lot more politics and bureaucracy. |
| 15:47 | <Philip`> | foolip: Easy mistake to make, since they're both CamelCase names |
| 15:48 | <foolip> | Philip`: exactly! |
| 15:48 | <TabAtkins> | foolip: Sorry, I was at work until 10pm last night raging against this server. I"m doing final troubleshooting with our network admin before we wipe the fucker and start over. |
| 15:48 | <TabAtkins> | It's planned for lunch today instead. |
| 15:48 | <foolip> | haha, we can always request a new deadline in a week if things are too busy |
| 15:48 | <TabAtkins> | If necessary, I'll do that. But I think I can get it out today. |
| 18:00 | <TabAtkins> | I'm still not sure what it is that Lief is suggesting with this "caption" idea. |
| 18:00 | <TabAtkins> | In his examples he seems to just be renaming <figure> to <cap> and then throwing in <object> for some reason. |
| 18:30 | <foolip> | <hascaption><img><caption></hascaption> :) |
| 18:36 | <gsnedders> | IE creates a single tag for <span<span> right? |
| 18:36 | <gsnedders> | s/tag/element/ |
| 18:37 | <gsnedders> | Whereas Firefox and WebKit create two (and Opera agrees with IE) |
| 19:19 | <hober> | TabAtkins: I'm usually unsure of what Lief is proposing in his emails, on any subject... |
| 19:19 | <TabAtkins> | I suppose that's true. |
| 19:29 | <othermaciej> | Lief often says that I proposed something which to me seems counter to the statements I actually made |
| 19:30 | <Philip`> | s/Lief/Leif/ |
| 19:31 | <gsnedders> | You need the g flag. |
| 19:33 | <Philip`> | No I don't, these are magic regexps |
| 20:09 | <jgraham> | gsnedders: You still there? |
| 20:32 | <TabAtkins> | Dammit, perl is hard to learn when you're doing so by debugging a mildly complex working program. |
| 20:36 | <Philip`> | Easier than debugging a program that doesn't work |
| 20:37 | <Dashiva> | I didn't know it was possible to write perl that's only mildly complex |
| 20:40 | <TabAtkins> | This is a pretty straightforward cgi processing a form and sending an email. It's only mildly complex. |
| 20:41 | <TabAtkins> | I also find it strange that when a cgi script fails you get a "file not found" message from apache. |
| 20:49 | <charles_> | Is there any way to get the data of a media element in javascript (html5)? |
| 20:49 | <charles_> | Something like getDataURL from canvas?? |
| 20:51 | <Philip`> | charles_: You could draw a <video> onto a <canvas> and then use toDataURL |
| 20:51 | <Sonja> | hullo |
| 20:51 | <Philip`> | though I'm not sure whether it'd return the current frame or not |
| 20:51 | <charles_> | Philip, thanks but had audio in mind |
| 20:51 | <Philip`> | Oh, okay |
| 20:51 | <Philip`> | No way to access that |
| 20:52 | <charles_> | Seems to be lacking from standard and what if you wished to make an audio editor?? |
| 20:52 | <Philip`> | There have been discussions about creating some kind of 'audio canvas' for editing |
| 20:52 | <charles_> | Philip, there are other uses as well |
| 20:52 | <Philip`> | but nothing concrete yet |
| 20:52 | <Dashiva> | The usual way is to spec simple features first, and return and add complex ones later |
| 20:53 | <charles_> | THANKS, I'd love to know about it and will keep an eye open |
| 20:53 | <Sonja> | will html5 make it possible for me to write a little roguelike browser game with Ruby without needing to use Flash or othetr RIA? |
| 20:53 | <charles_> | For me the toDataURL would be all I need |
| 20:54 | <charles_> | Sonja, as a game developer, I'd say YES |
| 20:54 | <Sonja> | cool |
| 20:54 | <Sonja> | the person in control of html5 is Ian Hickson, right? |
| 20:54 | <charles_> | Sonja, canvas for the well canvas and audio for sound effects |
| 20:54 | <Dashiva> | Sonja: That should be possible now |
| 20:54 | <Dashiva> | It's just animated ascii art |
| 20:55 | <Sonja> | Dashiva how would i animate the ascii ? using ajax ? |
| 20:55 | <charles_> | Sonja, you could even do it with custom fonts if you wish |
| 20:55 | <Sonja> | ah, like tile fonts |
| 20:55 | <Dashiva> | Sonja: Yes |
| 20:55 | <charles_> | Sonja, ajax is communication, you would use a timer event |
| 20:56 | <TabAtkins> | Heh, doing a roguelike with actual monospace text would be fun. Nuts to <canvas>, ascii is where it's at. |
| 20:56 | <charles_> | Sonja, if it is a standalone game, you could load the entire thing and then play no need for ajax unless you really need it |
| 20:57 | <charles_> | Sonja, with custom font loaded that is most of the battle |
| 20:57 | <TabAtkins> | Just load it as the textContent of a <pre> and you're done. |
| 20:57 | <charles_> | TabAtkins, need anim with timer for so fun stuff |
| 20:57 | <TabAtkins> | Though I suppose you may want to color the characters. In that case you just wrap them in spans and such. |
| 20:58 | <TabAtkins> | charles_: Yeah, of course, but on each frame you're just replacing the textContent. |
| 20:58 | <charles_> | or divs with fixed position |
| 20:58 | <charles_> | or spans of course |
| 20:58 | <TabAtkins> | Sure, but that's making things too difficult. |
| 20:58 | <Sonja> | each time the player hits an arrow key or a letter, it would load a new html page? |
| 20:59 | <charles_> | Sonja, nope, capture input events via javascript and away you go |
| 20:59 | <Sonja> | sounds like i need to learn javascript then. cool! |
| 20:59 | <Sonja> | and i can make javascript talk to ruby where the actual game is written in |
| 21:00 | <charles_> | Sonja, that would be the core of the game, using javascipt you could change the fonts all over the place and anim them too |
| 21:00 | <TabAtkins> | Sonja: Yeah, definitely learn javascript. Anything else would be silly in terms of latency. Learn jQuery to make it easy. |
| 21:00 | <TabAtkins> | And then yes, you can use ajax to talk back to the ruby program on the server and receive instructions. |
| 21:00 | <Sonja> | thanks. you guys are all very smrt and friendly!! |
| 21:00 | <charles_> | Sonja, maybe later ruby or lua or etc |
| 21:01 | <charles_> | Do it for a living and so have no choice ;) |
| 21:02 | <charles_> | Sonja, it is very important to do games in html5 (browser) because if you can do games you can do apps and that is a goal of html5 |
| 21:03 | <charles_> | Thanks all will look for the audio canvas and see what I can find. |
| 21:03 | <Philip`> | It's important to do games because games are fun |
| 21:04 | <Dashiva> | You could make a javascript roguelike proxy |
| 21:04 | <Philip`> | You don't need to justify it with any other reasoning :-p |
| 21:04 | <Dashiva> | That just sends raw key events and receives a string array containing the screen |
| 21:06 | <Sonja> | neat. this makes javascript look like ruby. http://github.com/jedediah/prettyscript |
| 21:06 | <Philip`> | Just use a JS Telnet emulator |
| 21:07 | <jcranmer> | you say that as if it's a good thing |
| 21:08 | <Sonja> | haha aw |
| 21:08 | <Dashiva> | >Blocks are translated to anonymous functions (yay!) |
| 21:09 | <Dashiva> | Sounds efficient |
| 21:13 | <TabAtkins> | Not their fault javascript doesn't allow direct syntax hacking. |
| 21:13 | <TabAtkins> | All your higher-order functions have to be done with closures. |
| 21:13 | jgraham | always thought games were just designed as stress tests for implementations |
| 21:13 | <Sonja> | what are some good books to learn javascript/jquery from scratch? |
| 21:14 | <jgraham> | TabAtkins: You are veering dangerously close to lisp weenie territory :) |
| 21:14 | <TabAtkins> | jgraham: I am so far into that territory it isn't even funny. |
| 21:14 | <jgraham> | TabAtkins: No need o embaras yourself in public though ;) |
| 21:15 | <TabAtkins> | There's nothing wrong with being a lisp weenie. We have to keep the rest of you imperative-lovers honest. |
| 21:16 | jgraham | always had the impression that lisp allowed for imperative-style programming |
| 21:16 | <jgraham> | at least compared to, say, haskell |
| 21:16 | <jgraham> | Sonja: Supposedly "javascipt: the good parts'" is OK |
| 21:17 | <Sonja> | thanks |
| 21:17 | <jgraham> | At least I seem to recall people sying that |
| 21:17 | jgraham | has never really read a book about javascript |
| 21:18 | <jgraham> | Sonja: I think there is a John Resig book about jQuery if you care about that in particular |
| 21:18 | <TabAtkins> | jgraham: Yeah, it certainly allows it. It just really hurts to do so. |
| 21:19 | <TabAtkins> | jgraham: And then there's things that people *call* imperative, like LOOP, which aren't. LOOP's just a delicious DSL. |
| 21:19 | <Sonja> | well TabAtkins suggested jquery "to make it easy" |
| 21:19 | <jgraham> | And he is releasing "Secrets of the Javacript Ninja" sometime. Which could be good. Or terrible |
| 21:20 | <jgraham> | Sonja: jQuery is mainly useful as a low-pain alterntaive to direct DOM manipulation |
| 21:20 | <TabAtkins> | Sonja: Yeah, the thing is that while javascript is a great language, the DOM (the functions you use to communicate with/manipulate the page) is *not*. jQuery papers all that over with a beautiful, consistent syntax. |
| 21:20 | <TabAtkins> | I've never learned DOM scripting, and don't intend to. jQuery is just too beautiful. |
| 21:21 | <jgraham> | The DOM is indeed eye-bleedingly ugly but it is not exactly hard |
| 21:22 | <TabAtkins> | jgraham: I'll note, though, that I'm only a few steps away from being a Haskell weenie too. I keep dipping my toes in the pool, and then backing out and just hacking the functionality I want into my lisp utils file instead (like lazy lists, woo!). |
| 21:23 | jgraham | is about three chpters in to "Real World Haskell" |
| 21:23 | <jgraham> | Which is really fun but not something I spend enough time on |
| 21:23 | <TabAtkins> | It feels like http://learnyouahaskell.com/ is an awesome resource. |
| 21:23 | <Sonja> | cool |
| 21:23 | <TabAtkins> | In addition to being ridiculous in general. |
| 21:23 | <Sonja> | so i'd learn javascript and jquery instead of javascript and dom |
| 21:24 | <TabAtkins> | Sonja: Yeah. |
| 21:24 | <Sonja> | you guys are such a n00b-friendly channel. thanks! <3 |
| 21:24 | <jgraham> | TabAtkins: That is clearly _why inspired. Anything that is _why inspired is inherently good. Even thouh I know no ruby... |
| 21:25 | <jgraham> | "This work is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 Unported License because I couldn't find a license with an even longer name. " |
| 21:26 | <Sonja> | _why quit the internets! :-O |
| 21:26 | <Sonja> | jgraham haha |
| 21:27 | <jgraham> | Sonja: indeed :( |
| 21:27 | <Sirisian> | Does w3c have their IRC server/channel? |
| 21:28 | <TabAtkins> | Sirian: Yeah, irc://irc.w3.org:80 |
| 21:28 | <TabAtkins> | And then #htmlwg, though that's mostly dead but for the telecons. |
| 21:28 | <TabAtkins> | ^^^ I meant Sirisian. |
| 21:29 | <Sirisian> | oh 80. I was trying 6665 like it said on a random site |
| 21:29 | <Philip`> | You mean #html-wg |
| 21:29 | <TabAtkins> | jgraham: Yeah, definitely _why-inspired, and definitely great for that. It pulls off the ridiculous properly as well. You can't just be silly and expect to be good. |
| 21:29 | <TabAtkins> | Philip`: Sure. |
| 21:29 | <Sirisian> | TabAtkins, is it down? |
| 21:30 | <TabAtkins> | No? I'm there right now, in the #css channel. |
| 21:30 | <TabAtkins> | Oh, *but* a bunch of people just got kicked a few minutes ago. |
| 21:31 | <TabAtkins> | So you may not have a route to it. |
| 21:31 | <Sirisian> | TabAtkins, ah I got it. My IRC client didn't understand when I was using a semi-colon. Definitely wanted a space. Thank you |
| 21:31 | <Sirisian> | er colon* |
| 21:32 | <TabAtkins> | Sirisian: Ah, k. You can just drop the irc:// in the front, and just do a /server irc.w3.org:80 |
| 21:32 | <TabAtkins> | Or wait, you mean the port colon? |
| 21:32 | <TabAtkins> | Weird. |
| 21:32 | <Sirisian> | x-chat is odd. /server irc.w3.org 80 |
| 21:32 | <TabAtkins> | Strange. |
| 22:01 | <TabAtkins> | othermaciej: Where should I ask for an extension on writing a change proposal for issue 76? I'm writing it tonight, but I want to run it past foolip first, and don't want to run the risk of overrunning the tomorrow deadline. |
| 22:02 | <othermaciej> | TabAtkins: in an email reply to the call for counter-proposals |
| 22:02 | <othermaciej> | TabAtkins: request for extension will probably be viewed more favorably if you link a rough draft and say you need some extra time to finalize it (with a specific deadline for finishing) |
| 22:03 | <TabAtkins> | Cool, I'll do so tonight. |
| 22:03 | <othermaciej> | TabAtkins: thank you sir |
| 22:17 | <TabAtkins> | Argh she has no idea what a cowpath is why god why. |
| 22:17 | <Sirisian> | wow who doesn't know what that is? >_> |
| 22:18 | <TabAtkins> | It's over on the htmlwg list. |
| 22:18 | <Philip`> | TabAtkins: A cowpath is when you lay out a long strip of concrete for the cows to walk down, even if the cows almost never bother doing so |
| 22:18 | <Dashiva> | Well, if you don't know the term it's not that outlandish to assume cowpath = any path that has been walked, as opposed to paved roads |
| 22:19 | <Philip`> | and even if the cows that do walk down it get lost and end up falling into a ditch |
| 22:20 | <TabAtkins> | Laughing at that mental image probably makes me a bad person. |
| 22:47 | <shelleyp> | TabAtkins, you don't like me. Fine. But you have to question if you're responding to me because you disagree, or just because its me |
| 22:47 | <TabAtkins> | shelleyp: I respond because I disagree. My like or dislike of you has nothing to do with it. |
| 22:48 | <shelleyp> | I find that unlikely, when I see the petty sniping that goes on in this group |
| 22:48 | <shelleyp> | You have a problem with what I write on the email list, take it up with me, face to face |
| 22:49 | <TabAtkins> | We've been over this before. I do challenge you on something when it's important. When it's not and I'm merely annoyed, there's no need to. |
| 22:50 | <shelleyp> | Then you can take it up with me via email. But I'm tired of having people point out to me the latest nasty little quip about me on this children's hour called an IRC channel |
| 22:50 | <TabAtkins> | Then... don't listen about the IRC channel? It's not like anyone's insulting you. There aren't low blows. |
| 22:51 | othermaciej | notes irony of shelleyp's last statement |
| 22:51 | <TabAtkins> | What, that the reason I don't take it up in email is precisely because it's not important enough to take up in email? |
| 22:52 | <TabAtkins> | Or the nasty little quip followed by "children's hour called an IRC channel"? |
| 22:53 | <othermaciej> | the fact that she used a nasty little quip to complain about nasty little quips |
| 22:53 | <TabAtkins> | Perhaps I should point it out to her. Nah, she says she's tired of that. |
| 22:56 | <TabAtkins> | shelleyp: You'll read this later, so I can address it to you here. I've discussed with you before (over email!) the differences between high-latency, high-cost communications like email and low-latency, low-cost communications like IRC. Some things are appropriate in one that are not appropriate in the other. I route all my important communications through email where I can collect my... |
| 22:56 | <TabAtkins> | ...thoughts best, so feel free to ignore me in IRC. |