| 04:23 | <Hixie> | jgraham: always send mail :-) |
| 04:24 | <Hixie> | mpt: i don't think the problems are with html5 implementations, i think they're bugs with the implementations of other things (like xmlhttprequest, html4 forms, etc) |
| 04:49 | <mpt> | Hixie, perhaps, but it gives a bad impression nonetheless |
| 04:57 | <Hixie> | mpt: feel free to work out what workarounds i should use |
| 05:00 | <mpt> | No, sorry, I'm working |
| 05:10 | <Hixie> | mpt: ditto :-) |
| 06:32 | <Molly> | Hello |
| 07:21 | karlcow | wonders if Molly has a limit of 50 comments by post. I had replied and it has not been published. |
| 07:23 | <zcorpan_> | karlcow: she doesn't |
| 07:33 | <zcorpan_> | does anyone know already why http://www.whatwg.org/issues/ doesn't work in opera? :) |
| 07:36 | <zcorpan_> | status is undefined |
| 07:36 | <zcorpan_> | hmm |
| 07:36 | <othermaciej> | what aspect of it does not work? |
| 07:38 | <zcorpan_> | regestring doesn't work. because it fails to update the status |
| 07:38 | <zcorpan_> | which is because of a parsing bug, it seems. <p><form><p><span> |
| 07:38 | <zcorpan_> | vs <p></p><p><span> |
| 07:38 | <zcorpan_> | er |
| 07:38 | <zcorpan_> | <p></p><form><p><span> |
| 07:49 | <Whiskey_M> | 'lo |
| 07:50 | <zcorpan_> | hey Whiskey_M |
| 09:43 | <Hixie> | ok i worked around tehe parsing bug in opera |
| 09:43 | <Hixie> | (by putting in the missing </p>s) |
| 09:43 | <Hixie> | and made it detect the IEissues so it wouldn't fail later |
| 09:44 | <Hixie> | Lachy: btw dunno if you saw but someone else had a blog post ready |
| 09:44 | <Hixie> | (or so they tell me) |
| 09:44 | <Lachy> | yeah, I saw your message earlier. I just had to go get myself a new keyboard before I could do anything about it |
| 09:45 | <Hixie> | heh |
| 09:45 | <Hixie> | what did you do to your old one? |
| 09:45 | <Hixie> | also in case anyone missed my earlier comment, i'm experimenting with http://www.whatwg.org/issues/ |
| 09:45 | <Hixie> | i'll probably send a message about it to public-html tomorrow |
| 09:45 | <Hixie> | assuming it works well |
| 09:46 | <Lachy> | it was raining yesterday and it got wet when I pulled it out of my car, and even though I tried it, it appears that some of the wires inside had corroded and finally gave up |
| 09:46 | <Lachy> | s/tried/dried/ |
| 09:47 | <Hixie> | ah |
| 09:47 | <Hixie> | that sucks |
| 09:49 | <Lachy> | is that the list of issues created from your existing email archive folders? |
| 09:50 | <annevk> | hmm, 1035 new e-mails |
| 09:51 | <Lachy> | Hixie, who was it that emailed you about the blog entry? |
| 09:51 | <hsivonen> | annevk: welcome back |
| 09:51 | <hsivonen> | now it is my turn to go away: I'll be away from tomorrow until Sept 2 |
| 09:51 | <annevk> | hi all, thanks |
| 09:52 | <annevk> | oh, I will be away again starting tomorrow too... |
| 09:52 | <Lachy> | hi annevk |
| 09:52 | <Lachy> | where are you going? |
| 09:52 | annevk | will be back for "real" September 5 |
| 09:52 | <annevk> | some Dutch island (Schiermonnikoog) and after that hiking around the Mont Blanc (France, Switerzerland, Italy) |
| 09:53 | <hsivonen> | Lachy: Romania via Hungary |
| 09:53 | <Lachy> | hsivonen, for holiday or work? |
| 09:53 | <Lachy> | or other? |
| 09:53 | <hsivonen> | Lachy: holiday |
| 09:54 | <annevk> | have fun hsivonen! |
| 09:54 | <hsivonen> | annevk: thanks. you too |
| 09:56 | <Lachy> | I'm trying to figure out if this is a translation of a previous article or an original http://blog.whatwg.org/wp-admin/post.php?action=edit&post=48 |
| 09:56 | <Lachy> | the title of it is "Présentation : comment HTML 5 peut il être utilisé aujourd’hui ?" |
| 09:56 | <hsivonen> | Presentation: how can HTML 5 be used today? |
| 09:57 | <Lachy> | ah, right! it's a translation of http://blog.whatwg.org/html5-geekmeet |
| 09:57 | <annevk> | hmm, why did people oppose to dropping <input usemap>? |
| 09:58 | <Lachy> | because it has some purported accessibility benefits |
| 09:58 | <Lachy> | despite the fact that HTML4 didn't define how it worked and no implementation of it actually provded any accessibility benefits |
| 09:59 | <Lachy> | eventually, the discussion essentially turned into a proposal for a new feature to provide those accessibility benefits, but no-one presented any use cases for using an image map for form submission |
| 09:59 | <annevk> | right... |
| 10:08 | <Lachy> | aargh! Who does Rob think he is, being able to tell people when they should or should not respond? http://lists.w3.org/Archives/Public/public-html/2007Aug/0716.html |
| 10:12 | <Hixie> | Lachy: it was someone posting about a web forms 2 impl |
| 10:13 | <Hixie> | Lachy: and i presume rob burns would be the most knowledgable about when to reply or not reply, since he replies to pretty much every thread |
| 10:14 | <Hixie> | and yes, the list mentioned above (whatwg.org/issues, for anne's benefit) is derived from my imap folders |
| 10:15 | <othermaciej> | Ironically, Rob seems unaware of his eponymous rules of order |
| 10:16 | <Lachy> | Hixie, I didn't see anything in the various WF3 lists about <input placeholder> |
| 10:16 | <Hixie> | Lachy: it's there somewhere |
| 10:16 | <Hixie> | might be in WF2 |
| 10:18 | <Lachy> | maybe none of the placeholder discussions have "placeholder" in the subject |
| 10:18 | <Lachy> | oh well, doesn't matter |
| 10:19 | <Hixie> | possibly |
| 10:36 | <zcorpan_> | <input disabled tabindex=0> -- focusable or not? |
| 10:38 | <Hixie> | zcorpan_: nothing disabled is focusable |
| 10:38 | <rburns> | othermaciej: hey. I'm not unaware of my pseudo namesake, I spelled it Robert's instead of Roberts' a subtle pun (I thought that would be clear to those in the know) |
| 10:39 | <rburns> | Anne: the question of the elimination of <input usemap> was concern that insufficient research had been done to justify eliminating it. |
| 10:41 | <othermaciej> | rburns: telling people they shouldn't have replied to you is not appropriate behavior on a mailing list; nor is it proper meeting procedure either |
| 10:41 | <othermaciej> | if you want to make a formal motion then the usual procedure would be to wait for the chair to recognize you |
| 10:41 | <rburns> | well I would say replying to a motion without a second is improper meeting procedure (especially when that norm is raised in the very email being replied to) |
| 10:41 | <othermaciej> | and it is the chair's job to make sure rules of order are followed, not any random person who says something |
| 10:42 | <othermaciej> | if you think the chair is not doing his job, you raise a point of order with the chair |
| 10:42 | <rburns> | the chair could rule my motion out of order and stop a seconding from happening. That would be the email analogous thing to happen |
| 10:43 | <rburns> | I think the chair is doing fine. |
| 10:43 | <othermaciej> | if you want to apply rules of procedure to a mailing list, even notwithstanding that this is a really bad idea and was not agreed to by any one, you can't pick and choose just the ones you like |
| 10:43 | <rburns> | othermaciej: You seem to be the one complaining that he didn't rule my motion out of order |
| 10:43 | <othermaciej> | no, I'm not complaining |
| 10:43 | <othermaciej> | I think you can phrase your email however you like |
| 10:44 | <othermaciej> | but the fact that you called it a motion doesn't give you the authority to tell people not to reply |
| 10:44 | <rburns> | othermaciej: I'mnot picking an choosing. I've read complaints about how the email list is unwieldy. I made a motion to help with that. That motion wasn't seconded. I don't see how that's out of order. |
| 10:44 | <othermaciej> | I'm not complaining about your motion email |
| 10:44 | <othermaciej> | I just think you were out of line in telling people they should not have replied to it |
| 10:44 | <rburns> | Well why would someone reply to it if they don't like the volume of email on the list and they have no intention of seconding it. Just let it die then. |
| 10:45 | <rburns> | othermaciej : Well I disagree |
| 10:45 | <Lachy> | rburns, I think the list needs much less bureaucracy |
| 10:45 | <othermaciej> | it's not the first time you've told people they shouldn't have replied to one of your emails, either |
| 10:45 | <rburns> | othermaciej: I think you're being over -sensitive here and missing the irony of the situtation, |
| 10:46 | <rburns> | othermaciej: It's not? |
| 10:46 | <othermaciej> | no, it's not |
| 10:47 | <rburns> | Lachy: OK, that's fine. It seems a bit strange to complain about the list volume, but then not want to do anything about it. My proposal would have helped organize emails into meaningful categories. |
| 10:47 | <rburns> | othermaciej: It's not? |
| 10:47 | <hsivonen> | rburns: taxonomies don't lower the volume |
| 10:48 | <hsivonen> | rburns: not suggesting taxonomies doesn't add to the volume |
| 10:48 | <Hixie> | the best solution to the list volume issue (if one believes it to be an issue, which i don't) is to reduce the number of e-mails sent to hte list, e.g. by not posting more than 2 or 3 a day |
| 10:48 | <rburns> | hsivonen: no, taxonomies do not lower the volume but they help manage the volume |
| 10:49 | <othermaciej> | rburns: people substantively disagree with you that your specific proposal is a good way to address the issue of the list volume; that's not the same as not wanting to do anything about it |
| 10:49 | <Hixie> | sending an e-mail discussing the process for categorising e-mail isn't a solution to the e-mail volume)( |
| 10:50 | <rburns> | Othermaciej: that's fine to disagree. Though why continue a motion that hasn't even been seconded. It goes against the very concept of wanting to control list volume. |
| 10:50 | <othermaciej> | but in any case, telling them they shouldn't express their disagreement or alternate proposals because it would violate a rule that you proposed in that very email is out of line |
| 10:51 | <rburns> | Hixie: certainly it is. If the volume is high, and it's easy to separate announcements, from reviews, from motions, from general discussion (the bulk of the volume), then it's much easier to handle that volume. |
| 10:51 | <othermaciej> | it's like if I sent an email that says "I am now king of the HTML Working Group. I command that no one reply to this message, because I am king." |
| 10:51 | <othermaciej> | I hope if I did that, people would rightly see it as invalid and tell me so |
| 10:51 | <othermaciej> | and perhaps cart me off to the funny farm |
| 10:52 | <othermaciej> | but in any case I don't think the list has really been used for motions |
| 10:52 | <rburns> | Hixie: so that to me is a solution to the volume. It's like the volume of web content is a problem to deal with. But some folks thought of creating search engines to deal with that volume. The search engine is a solution to the volume of websites. No need to starting DO attacks on servers instead. |
| 10:52 | <rburns> | Othermaciej: but you do see the irony don't you? :-) |
| 10:53 | <Lachy> | rburns, conventions evolve over time, they're not dictated by authorities. Trying to force people to adopt a specific convention is not the right approach |
| 10:53 | <Hixie> | rburns: the problem people have with high volume is that they can't read all the e-mail. categorising it doesn't help with that. |
| 10:53 | <Hixie> | rb we already have threading mail clients that do a quite good job of categorising e-mails on a per-topic basis. |
| 10:53 | <Hixie> | [24~rb the volume of websites is di |
| 10:53 | <Hixie> | fferent pbecause you're not expected to read all websites |
| 10:53 | <rburns> | othermarcij: if you sent me that email I would just laugh. I thought my reply would cause the same reaction. |
| 10:53 | <othermaciej> | rburns: I do see the irony of you sending email to suggest ways to reduce email volume when you are one of the top posters |
| 10:53 | <othermaciej> | rburns: and I see the further irony of you replying to people who reply to your message, to tell them their replies needlessly add to list volume |
| 10:54 | <rburns> | othermaciej: I see that irony too. So what? |
| 10:54 | <rburns> | othermaciej: does that make it impossible to see the other irony I pointed out? |
| 10:54 | <Hixie> | wow, my client is chopping up my text real bad. silly lag issues. |
| 10:55 | <Hixie> | othermaciej: he's not "one of" the top posted, jhe has the honour of being the top poster, by a factor of 2 more volume than the next person. |
| 10:55 | <hsivonen> | rburns: well, there's a way to lower volume that is in your disposal and that doesn't even require the cooperation of the group |
| 10:55 | <rburns> | Hixie: categorizing can help nonetheless. It can help decide how quickly to skim a message or even whether to skip it that isn't just based on the name it came from. |
| 10:55 | <Hixie> | i should point out that i have nothing against rob posting a lot |
| 10:56 | <Hixie> | i don't think we want working group members to be skipping any e-mail |
| 10:56 | <rburns> | hsivonen: except I'm not really interested in that way of lowering volume. |
| 10:56 | <Hixie> | that would not seem to be hte best way for working group members to stay up to speed on issues and f |
| 10:56 | <Hixie> | or us to get consensus on the spec |
| 10:57 | <rburns> | Hixie: I don't feel like we're getting consensus on the spec regardless of volume. |
| 10:57 | <Lachy> | I don't mind anyone posting a lot if their emails are actually productive |
| 10:57 | <rburns> | Lachy: that's why I suggested you shouldn't have replied. |
| 10:58 | <Lachy> | rburns, my email was a productive response to your suggestion pointing out existing conventions. Yours telling me not to respond was not |
| 10:58 | <Hixie> | rburns: oh i agree, we're years from the spec being near ready enough to have any sort of significant consensus |
| 10:58 | <Hixie> | html5 is a multi-year effort |
| 10:58 | <jgraham> | Seriously, the best way of lowering volume is for people to think more about what they are replying to before they post. Lots of discussion has been cyclical patterns of messages with tiny tiny increments of understanding per message. |
| 10:58 | <Hixie> | we've been doing it since 2004, i don't expect it to be done before 2014 (the spec, that is, before even considering test suites) |
| 10:59 | <rburns> | Lachy: if the message I posted said it needed a second or no further discussion was necessary,, how could a response to it that didn't second the message possibly be productive. What end could it serve,. for the WG that is. |
| 10:59 | <Hixie> | (which iwll probably take at least 5 more years beyond that) |
| 10:59 | <rburns> | jgraham: I agree. |
| 11:02 | <rburns> | jgraham: thtat's also why I felt tagging subject lines would help. Similar to distilling things to the wiki, general discussion could take place that might just be about building understanding among the participants in those discussions. Those would be the message everyone could safely ignore if they were not interested in the subject |
| 11:02 | <rburns> | jgraham: then as members reached some level of understanding of the topic, reviews, proposals or motions could be made. These would be the high-grade ore on the list. |
| 11:03 | <jgraham> | rburns: But saying "don't read stuff you don't think is interesting" isn't a good solution. People should be able to follow the discussions |
| 11:04 | <jgraham> | However that requires participants in the discussions to think hard about the other point of view and why it makes sense to the other person before posting |
| 11:04 | <jgraham> | and to prefer one well-researched post over many hastily written ones |
| 11:04 | <othermaciej> | One possible thing to think about is, "how far removed is the point I am about to make from the deliverables of this group" |
| 11:05 | <jgraham> | and to try, as far as possible, to express themselves concisely |
| 11:05 | jgraham | is especially bad the concise thing |
| 11:05 | <othermaciej> | I have at times been guilty of replying to emails only to make a meta-meta-meta-point that no longer has much concrete relevance to the spec, our other planned publications, the test suite, or any other deliverables |
| 11:06 | <othermaciej> | but I am certainly not the only one who needs to count the metas before hitting send |
| 11:06 | <rburns> | jgraham: I think conciseness has been a problem leading to more volume. I've seen half sentence replies that trigger dozens of misconstrued responses. |
| 11:08 | <jgraham> | Of course there is balance. But nothing puts me off a message like seeing it is tens of kb of text, especially if the discussion isn't producing much of value |
| 11:08 | <rburns> | jgraham: however, by categorizing the messages, I'm not saying others should be allowed to join in and read the message or reply to them. I'm saying those without tags could be more safely ignored than the others. And you can't tell me messages aren't being ignored anyway, because I see the responses to threads I'm not even necessarily involved in and I can tell the responder hasn't even read the thread. Categorizing would make the p |
| 11:08 | <hsivonen> | rburns: "Categorizing would make the p |
| 11:08 | <hsivonen> | " |
| 11:08 | <hsivonen> | rburns: you got cut off again. (ironically :-) |
| 11:08 | <rburns> | othermaciej: I really don't see a lot of problems with out-of-scope emails. The scope HTML5 has set for itself is pretty far-reaching. |
| 11:10 | <rburns> | " Categorizing would make the process of filtering easier." |
| 11:11 | <rburns> | othermaciej: the most seirous volume problems I've seen is members going round and round not understanding what other members are saying. and each time incrementing the number of emails. |
| 11:12 | <othermaciej> | I agree that arguing in circles is a problem |
| 11:12 | <rburns> | othermaciej: I've intervened from time to time and illuminated the situation and the thread ends, because the participants see what the disagreement was about. |
| 11:12 | <othermaciej> | but I don't think labelling such discussions with a one-letter code will in any way help |
| 11:12 | <rburns> | othermaciej: that's fine. I can see the drawbacks to that too. |
| 11:13 | <othermaciej> | two things I'm going to try to help with that are: (a) read the whole thread before replying; and (b) if I want to reply to many messages, try to consolidate my reply |
| 11:13 | <othermaciej> | those are in general good practices to avoid adding too much noise to a discussion |
| 11:13 | <rburns> | othermaciej: I'm not bothered if no one agrees with the proposal. I was simply taking feedback and remarks I'd seen and trying to formulate a solution |
| 11:14 | <rburns> | othermaciej: that sounds like good advice. I try to do the same (particularly the combining part). |
| 11:27 | <rburns> | annevk: all but 5 character references have to be handled by the next level above XML. XML is unaware of any other entity references. |
| 11:27 | <Hixie> | http://whatwg.org/issues/ now has a companion http://whatwg.org/issues/top |
| 11:27 | <Hixie> | i'll mail the list about them tomorrow |
| 11:28 | <rburns> | s/character references/entity references |
| 11:28 | <Hixie> | now i must sleep |
| 11:28 | <hsivonen> | rburns: not correct. entity references are expanded on the XML except when a non-validating XML processor has skipped an external entity |
| 11:29 | <hsivonen> | s/on the XML/on the XML layer/ |
| 11:29 | <rburns> | hsivonen: Wouldn't these be non-validating XML processors? |
| 11:30 | <hsivonen> | rburns: what's the HTML 5 spec to say what mode to run the XML processor in? |
| 11:31 | <hsivonen> | rburns: though in practice, in sane browser implementations, the XML processor is non-validating |
| 11:31 | <rburns> | hsivonen: I guess I'm saying that IF the XHTML5 processor is acting as a non-validating XML processor then... |
| 11:31 | <hsivonen> | rburns: there's also the other condition: the non-validating processor has to skip an external entity for the loop hole to kick in |
| 11:32 | <hsivonen> | rburns: generally, XML processors cannot be configured to skip external entities conditionally |
| 11:32 | <rburns> | hsivonen: it's not a big deal, but I feel like when I do encounter draconian error-handling its often because of an over-zealoous exclusion of entity references |
| 11:32 | <hsivonen> | (implementation limitation) |
| 11:33 | <rburns> | hsivonen: Also I wasn't suggesting no error, I was just suggestion a non-fatal-error |
| 11:33 | <hsivonen> | rburns: like I explained before, the loophole does not kick in in expat in Gecko, because as far as expat is concerned, it has not skipped anything |
| 11:33 | <rburns> | hsivonen: like a big black cockroach or something where the entity should be :-) |
| 11:34 | <hsivonen> | rburns: what you are asking is rigging both browsers and content is a very specific way in order to exploit a loophole in the XML spec |
| 11:34 | <rburns> | hsivonen: expat? |
| 11:34 | <hsivonen> | rburns: the XML parser |
| 11:34 | <rburns> | hsivonen: oh |
| 11:35 | <rburns> | hsivonen: OK. thanks for the info (I didn't know they named their xml parser) :-). |
| 11:35 | <hsivonen> | rburns: now, the question is, why do you want to use XML, if you don't like the way XML is to the point of wanting to standardize a particular loop hole exploitation? |
| 11:36 | <hsivonen> | and the way XML is is that the speccers of XML gave us 5 reliable entities and UTF-8 |
| 11:36 | <rburns> | hsivonen: I just think the draconian part is taken too far there with the entity references. Most of the other fatal-errors make sense from the XML perspective. They have wider implications for the whole document. These unknown character references don't seem to so much. |
| 11:37 | <takkaria> | but you should be using Unicode anyway |
| 11:37 | <rburns> | hsivonen: perhaps most of the problem is being fixed anyway. It wasn't long ago when the major browsers would choke even on known HTML entity references when processing XHTML |
| 11:37 | <hsivonen> | rburns: I think that the situation sucks too, but the thing is that the value of XML is in XML parsers (however configured!) being ubiquitous. for this reason, XML has to be what it is and we don't guess to second guess |
| 11:38 | <rburns> | hsivonen: Now they seem to be repairing those bugs for the most part. My feeling was just why choke on any well-formed though invalid entity reference. |
| 11:39 | <rburns> | hsivonen: well we do second-guess a lot of other standards tat I think have wider problems than well-formed though invalid entity references in XML (authoritative content-type headers for example). |
| 11:39 | <hsivonen> | rburns: they weren't bugs to begin with and the issue has not been "fixed". there's just an ugly workaround for a handful of common cases |
| 11:40 | <hsivonen> | rburns: yes we do, but second-guessing XML has been implicitly out-of-bounds |
| 11:40 | <rburns> | hsivonen: I'm not sure what you mean. The XHTML 1.0 recommendation includes a complete list of HTML entity references. |
| 11:40 | <hsivonen> | XML 1.0 is kinda sacred |
| 11:40 | <rburns> | hsivonen: Yeah I understand. It's just HOW sacred. |
| 11:41 | <hsivonen> | rburns: well, second-guessing it defeats the point of using XML |
| 11:41 | <hsivonen> | rburns: since the rationale for using XML is to leverage the network effects of the installed base of XML tooling |
| 11:42 | <rburns> | hsivonen: I don't think so. Not in these few isolated areas. As long as their are not wider-reaching implication (which it is sometimes hard to determine). |
| 11:43 | <hsivonen> | rburns: what was "I don't think so" in reference to? to the network effects? |
| 11:44 | <hsivonen> | rburns: don't you consider micromanaging the parser configuration of every XHTML5 consuming XML parser far reaching? |
| 11:44 | <rburns> | hsivonen: no that was in reference to defeating the purpose (which I guess than you added to it about the xml tooling) |
| 11:44 | <rburns> | hsivonen: but to implement any namespace in XML without retrieving schema, ,an application has to do something else with the XML just to be able to handle entity references. So it can't just be using the purse XML stuff. |
| 11:44 | <hsivonen> | rburns: what's your purpose for using XML? |
| 11:45 | <hsivonen> | purse? |
| 11:45 | <othermaciej> | I assume he meant "pure" |
| 11:45 | <hsivonen> | oh |
| 11:46 | <hsivonen> | rburns: the XML processor is a sacred black box. you get to write app code outside it but you don't get to hack its internals |
| 11:46 | <rburns> | hsivonen: To me, extensibility. common parsing of elements. separating well-formedness from validity. The fatal error handling has some benefits, I'm just not quite sure how they relate to entity references and these more isolated parts of the parsing behavior. |
| 11:46 | <rburns> | othermaciej: yes "pure" |
| 11:47 | <hsivonen> | rburns: they do relate and we don't get to pick and choose which parts of XML we like |
| 11:47 | <rburns> | hsivonen: Like I said before. I think it should be an error. It could even be an obvious ugly error on an XHTML page. I'm just not quite sure what making it a fatal error does for anyone. (in the entity reference case) |
| 11:47 | <othermaciej> | as I understand it, the main reason to have an XML serialization is to interoperate with all of the deployed XML stuff that's out there |
| 11:47 | <hsivonen> | rburns: cause if you hack the internals of the XML processor, it ain't an XML processor anymore |
| 11:48 | <othermaciej> | having an almost-XML serialization would not really have the same benefits |
| 11:48 | <hsivonen> | othermaciej: exactly |
| 11:48 | <hsivonen> | othermaciej: hence "network effects" above |
| 11:48 | <rburns> | hsivonen: it would to me. Even more benefits I think. |
| 11:49 | <othermaciej> | at which point, you may as well have only the HTML serialization, since it already works in clients and with well-defined parsing rules is feasible to process with tree-based tools (instead of just with text processing tools) |
| 11:50 | <othermaciej> | XML does also have the benefit (some would say) of namespaces |
| 11:50 | <rburns> | othermaciej: I don't mean an almost-xml serialization. I mean an all xml serialization that is processed without fatal errors for entity references. In other words the HTML XML application would define an anyname entity reference mapped to the Unicode replacement character. |
| 11:50 | <othermaciej> | but grafting namespaces onto HTML is a better way to get namespaces + looser parsing than loosening XML parsing rules |
| 11:51 | <rburns> | tothermaciej: but were already going to have XML and namespaces, whatever we do with namespaces for the text/html serialization. |
| 11:51 | <rburns> | othermaciej: adding namespaces to the text/html serialization sounds scary to me. |
| 11:51 | <rburns> | othermaciej: though I think it would be great if we could pull it off. |
| 11:53 | <rburns> | I know other schemas have used an anyAttributeName in their schemas. This doesn't seem that different. It's just an NCName for any entity reference rather than any attribute name. |
| 11:54 | <rburns> | s/NCName/NCNames (as in the infinity of them) |
| 13:36 | <Lachy> | anyone who speaks French in here? What would be an appropriate post slug for the article "Présentation: comment HTML 5 peut-il être utilisé aujourd'hui ?" |
| 13:37 | <Lachy> | (the post slug is the path in the URL on the blog, e.g. in http://blog.whatwg.org/html-contre-xhtml the post slug is "html-contre-xhtml") |
| 13:37 | <Lachy> | hsivonen? |
| 13:38 | <hsivonen> | html5-aujourd-hui perhaps |
| 13:38 | <Lachy> | thanks, that'll do |
| 13:38 | <Lachy> | http://blog.whatwg.org/html5-aujourd-hui |
| 15:40 | Lachy | hopes his latest post to public-html raises the bar for quality on public-html, and that others follow the example for their own proposals |
| 15:40 | <gsnedders> | Lachy: quality? who needs no stinkin' quality!? |
| 15:41 | <Lachy> | gsnedders, well, we could continue with the current state of unproductive discussion if you like? |
| 15:41 | <gsnedders> | Lachy: I doubt whether quality posts will help |
| 15:41 | <Lachy> | I have my doubts too, but it never hurts to have a little hope |
| 15:41 | <Whiskey_M> | ahhh, that is what is meant by placeholder ;-) |
| 19:49 | <aroben> | Does anyone know of a way to determine whether a document was interpreted as HTML or XHTML from within the document itself? |
| 19:51 | <aroben> | Rather, a way to do so in an HTML5-compliant user agent |
| 19:52 | <annevk> | isHtml = document.documentElement.tagName == "HTML" ? true : false |
| 19:53 | <aroben> | annevk: thanks! |
| 19:53 | <annevk> | yw |
| 19:55 | <Philip`> | <html xmlns="http://www.w3.org/1999/xhtml"><body onload="alert('This is '+(document.getElementById('this-is-html')?'':'X')+'HTML')"><script>document.write('\x3cp id=this-is-html>\x3c/p>')</script></body></html> |
| 20:06 | <Hixie> | Lachy: i specifically didn't add a vote button to the "top" page so that we wouldn't have issues get skewed votes |
| 20:06 | <Lachy> | ok |
| 20:08 | <Hixie> | and yeah, i don't do any encoding converstion |
| 20:09 | <Hixie> | if you know of a perl library that would do that for me let me know, 'll plug it in |
| 20:09 | <Hixie> | http://www.cssglobe.com/post.asp?id=863: "A cross-browser implementation that has been adopted by the W3C HTML Working Group of the WHATWG Web Forms 2.0 specification." |
| 20:12 | <Lachy> | aroben, the more accurate test would be |
| 20:12 | <Lachy> | var isHTML = (function() { |
| 20:12 | <Lachy> | var htmlElement = document.getElementsByTagName("html").item(0); |
| 20:12 | <Lachy> | if (htmlElement) { |
| 20:12 | <Lachy> | return (htmlElement.tagName == "HTML"); |
| 20:12 | <Lachy> | } else { |
| 20:12 | <Lachy> | return false; |
| 20:12 | <Lachy> | } |
| 20:12 | <Lachy> | })(); |
| 20:12 | <Lachy> | That handles obscure XML cases like <HTML xmlns="http://example.org/NotHTML">; |
| 20:14 | <Lachy> | Hixie, there's probably email clients that are written in perl that would use libraries like that, though I don't know of any. |
| 20:14 | <annevk> | or simply |
| 20:14 | <annevk> | isHtml = document.documentElement.tagName == "HTML" && document.documentElement.namespaceURI = "http://www.w3.org/1999/xhtml"; ? 1 : 0 |
| 20:15 | <Lachy> | that will work for HTML5, but not current implementations, especially IE |
| 20:15 | <Hixie> | isHTML = document.createElement('html').tagName == 'HTML'; |
| 20:16 | <Lachy> | Hixie wins! :-) |
| 20:17 | <annevk> | "Rather, a way to do so in an HTML5-compliant user agent" |
| 20:17 | <annevk> | fwiw |
| 20:17 | <Philip`> | isHTML = true; and then you'll be right 99.99% of the time |
| 20:19 | <annevk> | you can now DOS-attack the (X)HTML5 tracker easier if you know the source code |
| 20:21 | <annevk> | this is considered to be a feature, until it's abused so heavily it will be removed :) |
| 20:21 | <Lachy> | annevk, can I have a copy of the source code? |
| 20:23 | <annevk> | sure: http://html5.googlecode.com/svn/trunk/web-apps-tracker/web-apps-tracker |
| 20:24 | <annevk> | the bit with form["limit"] is relevant |
| 20:24 | <Lachy> | annevk, there's a bug in Firefox. because you have <input ... value= name=from>, it displays name=from in the text field |
| 20:25 | <annevk> | hmm |
| 20:25 | <annevk> | that's silly |
| 20:25 | <Lachy> | you need to add quotes to value="" when there's no value, or omit the attribute |
| 20:25 | <annevk> | that seems rather obvious, yes |
| 20:25 | <Philip`> | Why parse the SVN log text rather than using "svn log --xml"? |
| 20:26 | <annevk> | not sure if dreamhost supports --xml, it might though |
| 20:27 | <annevk> | Lachy, that parsing behavior seems consistent with IE7 |
| 20:27 | <annevk> | Lachy, it seems that after the equal sign whitespace is ignored |
| 20:27 | <Lachy> | yes. I believe it's also what HTML5 has specced |
| 20:28 | <annevk> | and what every browser implements |
| 20:28 | <annevk> | fixed... |
| 20:29 | <annevk> | Philip`, you're welcome to provide enhancements |
| 21:46 | <gsnedders> | kingryan: the htmlOrFeed.json isn't being parsed by PHP. Can't work out why, though :\ |
| 21:46 | <kingryan> | which json parser are you using? |
| 21:46 | <gsnedders> | PHP 5.2's own |
| 21:46 | <kingryan> | hmm |
| 21:50 | <gsnedders> | kingryan: line 22 |
| 21:50 | <kingryan> | what about it? |
| 21:51 | <gsnedders> | kingryan: it seems to be one of several lines causing it to fail |
| 21:51 | <gsnedders> | kingryan: it's the \digits that is causing it |
| 21:52 | <kingryan> | oh, whoops |
| 21:52 | <kingryan> | yeah, that's not valid JSON |
| 21:52 | <gsnedders> | PHP's parsing is very draconian |
| 21:53 | <gsnedders> | kingryan: you want me to fix that in SVN? |
| 21:53 | <kingryan> | I'll fix it now |
| 21:54 | <kingryan> | I see only 2 occurences of that |
| 21:54 | <kingryan> | are you getting more than 2 errors? |
| 21:54 | <gsnedders> | kingryan: lines 33 and lines 22 |
| 21:54 | <kingryan> | got those |
| 21:55 | <gsnedders> | just those two lines |
| 21:56 | <kingryan> | alright, fix checked in |
| 21:57 | <gsnedders> | right. my paranoia about now properly using my untested PHP implementation was right |
| 22:04 | <gsnedders> | http://geoffers.no-ip.com/hg/foleeo?f=89f835292a7b;file=sniffed.php;style=gitweb is my implementation, BTW. still fails various tests, though |
| 22:07 | <gsnedders> | well, that revision's child passes all of the tests, now |
| 22:09 | <kingryan> | gsnedders: cool |
| 22:09 | <kingryan> | you have tests for binary? |
| 22:39 | <gsnedders> | kingryan: no, I mean those tests |
| 22:39 | <kingryan> | ah, gotcha |
| 22:39 | <gsnedders> | kingryan: if I had tests, they'd be in SVN |
| 22:40 | gsnedders | sighs. school tomorrow. |