| 02:13 | <Hixie> | acid3 updated a bit |
| 02:37 | <eseidel_> | Hixie: I'm confused by the presence of @media in Acid3 |
| 02:37 | <eseidel_> | didn't media queries get spec'd this summer? |
| 02:37 | <eseidel_> | Hixie: http://www.w3.org/TR/css3-mediaqueries/ or was there an earlier @media spec? |
| 02:42 | <othermaciej> | eseidel_: non-query media values are much older |
| 02:42 | <eseidel_> | sure, but he uses @media |
| 02:42 | <othermaciej> | eseidel_: also, http://www.w3.org/TR/2002/CR-css3-mediaqueries-20020708/ |
| 02:42 | <eseidel_> | @media all { foo: bar; } |
| 02:43 | <othermaciej> | (last CR was 2002) |
| 02:43 | <eseidel_> | ah |
| 02:43 | <eseidel_> | that's the one |
| 02:44 | <eseidel_> | othermaciej: however that can't be quite what Hixie was using since that doesn't support ',' syntax |
| 02:45 | <eseidel_> | also it rquires () |
| 02:45 | <othermaciej> | "Several Media Queries can be combined in a comma-separated list." |
| 02:45 | <othermaciej> | @media screen and (color), |
| 02:45 | <othermaciej> | projection and (color) { ... } |
| 02:45 | <othermaciej> | is one of the examples |
| 02:45 | <eseidel_> | interesting, the () remain required |
| 02:46 | <eseidel_> | so then this should be a CSS parse error, no? |
| 02:46 | <eseidel_> | style.appendChild(doc.createTextNode('@media all, all and color { #i { text-transform: uppercase; } }')); // matches |
| 02:46 | <eseidel_> | since it should be @media all, all and (color) { |
| 02:47 | <othermaciej> | I have no idea |
| 02:51 | <eseidel_> | othermaciej: I'm pretty sure they are. even though our error reporting is quite poor in that case. I updated http://bugs.webkit.org/show_bug.cgi?id=16796 and cc'd hixie |
| 02:51 | <eseidel_> | still 2 which we fail. one due to missing features, and one I'm not sure about |
| 03:52 | <Hixie> | eseidel_: media queries were in CR in 2002 |
| 03:52 | <eseidel_> | Hixie: yes, othermaciej pointed that out |
| 03:52 | <eseidel_> | they re-cr'd in 07? |
| 03:52 | <Hixie> | yeah, just saw that :-) |
| 03:52 | <Hixie> | fixing editorial problems |
| 03:52 | <Hixie> | it's common |
| 03:52 | <Hixie> | CSS 2.1 has been in CR at least 3 times |
| 03:53 | <Hixie> | (if you count the original CSS2 "REC") |
| 03:55 | <eseidel_> | Hixie: btw, assert(test( would be better if test() just did the assert, IMO. Then test could provide what exception was actually thrown |
| 03:56 | <Hixie> | can you send me mail with that bug # and the assert feedback? i'm about to run off |
| 03:56 | <eseidel_> | Hixie: sure... |
| 03:56 | <eseidel_> | you should see mail about the @media but anyway, you're cc'd |
| 03:56 | eseidel_ | writes hixie a mail |
| 03:56 | <Hixie> | thanks |
| 07:15 | <MikeSmith> | does IE5 support getElementById ? |
| 07:17 | <MikeSmith> | er, make that IE4 . |
| 07:20 | <MikeSmith> | ah, nair mind |
| 07:33 | <hubick> | Using Firefox on my Fedora 8 x86_64 box spontaneously reboots the whole machine instantly on http://planet.mozilla.org/ when viewing a post about Acid3 test. Can anyone reproduce? |
| 09:39 | <krijnh> | I'm on vacation till the 21st, so sorry for the missing logs if I disconnect and don't reconnect again :) |
| 09:42 | <Philip`> | Vacation away from IRC? How dreadful :-( |
| 11:46 | <Hixie> | crap, i don't have the build toold installed on my mac yet |
| 12:49 | <gsnedders> | They tend to be useful :P |
| 18:30 | <gsnedders> | Hixie: any reason you used XML-RPC not SOAP for Pingback? |
| 18:55 | <Dashiva> | gsnedders: Do you need a reason not to use SOAP? ;) |
| 18:56 | <gsnedders> | Dashiva: any reason to use/not use either would be nice, especially in the context of things like Pingback |
| 21:03 | <Hixie> | gsnedders: SOAP is more insane than XML-RPC. But using XML-RPC was pretty dumb too. |
| 21:04 | <gsnedders> | Hixie: From my experience with RSS I tend to be a bit concerned about any standard written by Dave Winer |
| 21:04 | <Dashiva> | He was outdone this time :) |
| 21:41 | <gsnedders> | Hixie: Content-Type sniffing is currently type-sensitive. It shouldn't be. |
| 21:41 | <gsnedders> | Hixie: for the MIME types, that is. |
| 21:42 | <gsnedders> | Hixie: I assume it is safe to change step one to be totally case-insensitive? |
| 21:42 | <othermaciej> | type-sensitive? |
| 21:43 | <gsnedders> | s/type/case/ |
| 21:44 | <gsnedders> | Hixie: I just noticed step 2 makes it lowercase |
| 21:44 | gsnedders | headdesks |
| 21:52 | <Hixie> | i have no idea what you're asking :-) |
| 22:04 | <gsnedders> | Hixie: ignore me. I can't read the spec :) |
| 22:05 | <gsnedders> | Hixie: (assuming you read this :)) well, just the above bit |
| 22:20 | <Lachy> | hmm, I got request to write some articles/tutorials for developertutorials.com - has anyone heard of that site before? |
| 22:22 | <Hixie> | not off-hand |
| 22:26 | <jwalden> | Hixie: see <http://spreadsheets.google.com/pub?key=pNgBCwWdyRTT2JeiZn4B2Yw>;, the notes next to 19 and 85 |
| 22:27 | <Hixie> | fixed 85 |
| 22:28 | <Hixie> | and for 19, i was using DOM2, which seems to justify having it on the prototype |
| 22:28 | <Hixie> | i think we want the constants on all objects, no? |
| 22:29 | <Lachy> | which browser is that spreadsheet listing results for? |
| 22:29 | <jwalden> | Firefox trunk |
| 22:34 | <jwalden> | Hixie: I don't mind either way, but Moz had thought constants on non-ctors wasn't per spec |
| 22:34 | <jwalden> | Hixie: just thinking about this because of the conversation in <http://groups.google.com/group/mozilla.dev.tech.dom/browse_frm/thread/8f7feeb90be81c99/02a92f4574ca032a?lnk=st&q=class+properties+prototype+properties#02a92f4574ca032a>; |
| 22:36 | <Hixie> | well i don't mind changing it to test the constructor |
| 22:36 | <jwalden> | Hixie: see bz in #9 there: "They might also break in future Mozilla versions (e.g. if we stop using XPConnect for DOM classes). " |
| 22:36 | <Hixie> | that's just gonna show even more bugs in IE... |
| 22:37 | <gsnedders> | But IE6 supports 100% of it! |
| 22:37 | jwalden | tries to decide whether this makes him laugh or just be sad |
| 22:50 | <gsnedders> | Hixie: if you were to create Pingback from the ground up today with no regard for backwards compat., what would you use in place of XML-RPC? |
| 22:52 | <Hixie> | the Referer header and a GET request |
| 22:53 | <gsnedders> | Okay, that doesn't map to the slightly different use-case of tracking comments on other sites. |
| 22:53 | <gsnedders> | (where you need to send at least the commenter's name + comment) |
| 22:54 | <gsnedders> | Ideas for transport media are welcome :) |
| 22:55 | <Lachy> | Hixie, wouldn't Referer spam, which has been happening since long before pingback, have made that a bad idea? |
| 22:55 | <gsnedders> | (over HTTP, obviously) |
| 22:55 | <gsnedders> | Lachy: it's done using Pingback, what difference does it make? |
| 22:58 | <Lachy> | dunno, it just seems like it would be worse |
| 22:58 | gsnedders | thinks we'll never know |
| 23:02 | <Hixie> | Lachy: the techniques used to avoid it for pingback can be used for referer spam. |
| 23:03 | <Hixie> | Lachy: spammers do trackback spam, pingback spam, and referer spam, they don't care what the exact protocol is. |
| 23:03 | <Hixie> | gsnedders: for tracking comments, i recommend NNTP |
| 23:04 | gsnedders | ponders how that'd work |
| 23:04 | <gsnedders> | Hixie: <http://groups.google.com/group/habari-dev/browse_thread/thread/5a5fabe48110df2b>; is the current proposal |
| 23:05 | <gsnedders> | NNTP would probably be overkill anyway. |
| 23:22 | <gsnedders> | Hixie: if you have any good thoughts about such a thing, could I possibly con you into replying to that within the next month or so? or is that too ambitious of a timescale? |
| 23:22 | <Hixie> | i have no thoughts |
| 23:23 | <Hixie> | i think http is a dumb protocol to do threaded communication over |
| 23:23 | <gsnedders> | Think, Ian, think. :) |
| 23:24 | <gsnedders> | I think requiring blog software to implement a whole protocol like NNTP is unrealistic. |
| 23:25 | <gsnedders> | It almost certainly has to be done over a protocol they already support: HTTP & XML-RPC is already widely supported, however unidealistic it is. |
| 23:27 | <Hixie> | i wouldn't implement a whole protocol like NNTP |
| 23:27 | <Hixie> | i would just use NNTP |
| 23:27 | <Hixie> | or SMTP |
| 23:27 | <Hixie> | what's wrong with e-mail? |
| 23:29 | gsnedders | doesn't bother going into use-cases |
| 23:29 | gsnedders | needs to sleep |