| 00:31 | <AryehGregor> | "with" sometimes appears in golfing? |
| 00:33 | <AryehGregor> | I mean, I imagine it appears in many contexts, being a common English word. |
| 00:33 | <Dashiva> | Code golf |
| 01:52 | <hsivonen> | I find it most odd that HTTP gurus oppose the use of POST for ping on safety grounds |
| 01:53 | <othermaciej> | what are the safety grounds? |
| 01:53 | <Hixie> | if anyone ever understand the objection, please let me know, so i can fix it |
| 01:53 | <othermaciej> | can ping produce POSTs that can't be produced by posting a form? |
| 01:53 | <hsivonen> | othermaciej: see Roy's Change Proposal |
| 01:53 | <Hixie> | i've tried many times to understand the objection, but not been able to |
| 01:54 | hsivonen | avoids summarizing in order to avoid missummarizing |
| 01:55 | <othermaciej> | hsivonen: ah, when you say "safety grounds" you just mean the official HTTP classification of what methods are labeled "safe" |
| 01:57 | <hsivonen> | othermaciej: yes |
| 01:57 | <othermaciej> | I wonder if it is a relevant consideration that a submit button can be styled to look just like a link |
| 01:57 | <othermaciej> | (let alone the fact that a POST can be issued programmatically from an onclick handler) |
| 01:58 | <Hixie> | http://damowmow.com/playground/demos/http/002/ is relevant |
| 01:58 | <othermaciej> | what Content-Type does the ping post use? |
| 01:58 | <Hixie> | note that the test.html file on that page has no script |
| 01:59 | <Hixie> | yet it can trick people (even highly experienced http experts) into causing POSTs to be submitted |
| 01:59 | <othermaciej> | clever, that one doesn't even require form controls to be stylable |
| 02:00 | <Dashiva> | othermaciej: It doesn't seem that anything is a relevant consideration except the notion that POST is unsafe |
| 02:00 | <Hixie> | i guess if POST is unsafe and POST is part of HTTP that means HTTP is unsafe! |
| 02:02 | <Dashiva> | Only in browsers, because browsers never do anything right |
| 10:19 | <mitnavn> | Is there a wrapping element for HTML 5? I need a container of all my elements to set a side width, should I just use a div? |
| 10:19 | <Hixie> | yes |
| 10:19 | <Hixie> | <div> is that element |
| 10:20 | <mitnavn> | Thanks |
| 11:09 | <gsnedders> | jgraham: yt? |
| 11:39 | <jgraham> | gsnedders: Yes |
| 15:04 | <Dashiva> | How many times does someone have to repeat obvious falsehoods and be corrected before you assume bad faith? |
| 15:06 | <Steve^> | 6 or 7 |
| 15:13 | <virtuelv> | http://qdb.us/295884 |
| 19:08 | <Dashiva> | http://www.android.com/ has a <link rel="stylesheet" href="//google.com/css/inlay.css">; |
| 19:08 | <Dashiva> | I wonder if that href type is going to catch on |
| 19:10 | <hsivonen> | Dashiva: I expect redirect from http to https catch on first |
| 19:12 | <Dashiva> | How does that look? |
| 19:14 | <hsivonen> | Dashiva: just a 301 respons with an https URL in the Location header |
| 19:24 | <webben> | hsivonen: How does the server know to offer a redirect to an HTTPS alternative? Referer header? |
| 19:26 | <Philip`> | Dashiva: Slashdot has used that style for ages |
| 19:28 | <hsivonen> | webben: by seeing which port was used for the request |
| 19:30 | hsivonen | wishes the chairs would point out to Shelley that the HTML WG charter covers application in addition to documents |
| 19:30 | <webben> | hsivonen: But if you serve up HTML response over SSL, but link to a resource with http:// won't it use port 80 when requesting that resource? |
| 19:31 | <webben> | s/it/the UA/ |
| 19:31 | <hsivonen> | webben: I meant not serving up HTML as http but redirecting to https at that point |
| 22:06 | <boblet> | hey all, re: abbr/acronym, is there a way now for authors to indicate whether the term should be read as a word (like FUD) or as initials (like CIA)? Or is that up to AT? |
| 22:10 | <zcorpan> | boblet: it's up to the AT, as it has always been |
| 22:11 | <boblet> | zcorpan: ok, thanks. I was under the impression that historically that was the distinction between abbr/acronym. good to know |
| 22:13 | <Philip`> | As far as I'm aware it should be the same as for non-marked-up abbreviations, and the same as for all non-abbreviated collections of characters too |
| 22:14 | <Philip`> | i.e. it's the AT's job to work out how to read text |
| 22:25 | <boblet> | another random question; it’s fine to have an id and a class with the same name, right? Spec doesn’t forbid it (id value unique amongst all ids in subtree), but wondering if eg browser bug or other gotcha |
| 22:26 | <zcorpan> | it's fine |
| 22:26 | <boblet> | thanks zcorpan |
| 23:12 | <zcorpan> | so anyone's got any comments on the xml-stylesheet draft? |
| 23:59 | <AryehGregor> | Ooh, Chrome has neat UI for untrusted SSL certs: it writes "https" in red, but has a little slash through it too. |
| 23:59 | <AryehGregor> | Very noticeable. |