IRC Log for #microformats on 2005-12-23

Timestamps are in UTC.

  1. [00:02:59] * TantekC (n=Tantek@dsl092-187-246.sfo1.dsl.speakeasy.net) Quit (Remote closed the connection)
  2. [00:10:34] * bear_lunch is now known as bear
  3. [00:31:17] <mfbot> [[hcard-examples]] M http://microformats.org/wiki?title=hcard-examples&diff=0&oldid=3526 * Tantek * (+993) Added Authors of Pages and Posts
  4. [00:33:20] * TantekC (n=Tantek@dsl092-187-246.sfo1.dsl.speakeasy.net) has joined #microformats
  5. [00:33:52] * hlb (i=hlb@CCCA.NCTU.edu.tw) Quit (Read error: 104 (Connection reset by peer))
  6. [00:33:57] * hlb (i=hlb@CCCA.NCTU.edu.tw) has joined #microformats
  7. [00:54:41] * kingryan (n=kingryan@dsl092-187-246.sfo1.dsl.speakeasy.net) Quit ()
  8. [00:56:54] * KevinMarks (n=Snak@pdpc/supporter/active/kevinmarks) Quit ("The computer fell asleep")
  9. [01:09:25] * tantek sets mode +o TantekC
  10. [01:11:42] * tantek (n=tantek@dsl092-187-246.sfo1.dsl.speakeasy.net) Quit ()
  11. [01:15:41] * TantekC (n=Tantek@dsl092-187-246.sfo1.dsl.speakeasy.net) Quit (Remote closed the connection)
  12. [01:16:33] * hober (n=ted@rrcs-67-52-157-90.west.biz.rr.com) Quit ("nil")
  13. [01:48:52] * bear (n=bear@68.81.93.59) Quit (Read error: 110 (Connection timed out))
  14. [01:56:54] * dustym_ (n=dustym@office.cavoksolutions.com) Quit ()
  15. [02:00:12] <_fil_> hello
  16. [02:00:18] <_fil_> no / are allowed in ids
  17. [02:00:36] <_fil_> what is recommended for the atomEntry id ?
  18. [02:02:07] <_fil_> id="post-60" is given as an example...
  19. [02:03:25] * kingryan (n=kingryan@dsl081-240-149.sfo1.dsl.speakeasy.net) has joined #microformats
  20. [02:03:40] * ChanServ sets mode +o kingryan
  21. [02:04:00] <_fil_> hi Ryan
  22. [02:04:03] <_fil_> 03:00 < _fil_> hello
  23. [02:04:03] <_fil_> 03:00 < _fil_> no / are allowed in ids
  24. [02:04:03] <_fil_> 03:00 < _fil_> what is recommended for the atomEntry id ?
  25. [02:04:03] <_fil_> 03:02 < _fil_> id="post-60" is given as an example...
  26. [02:05:57] <_fil_> I wanted to use the post URL as id...
  27. [02:17:44] * bear (n=bear@pcp01380088pcs.levtwn01.pa.comcast.net) has joined #microformats
  28. [02:20:11] * tantek (n=tantek@h-67-103-44-6.snfccasy.covad.net) has joined #microformats
  29. [02:20:12] * ChanServ sets mode +o tantek
  30. [02:40:28] * KevinMarks (n=Snak@h-68-164-88-162.snvacaid.dynamic.covad.net) has joined #microformats
  31. [02:46:27] * dustym (n=dustym@office.cavoksolutions.com) has joined #microformats
  32. [03:05:10] * tantek sets mode +o KevinMarks
  33. [03:05:13] <tantek> Hi Kevin
  34. [03:05:28] <tantek> here is a nice article on publish.com: http://www.publish.com/article2/0,1895,1904359,00.asp
  35. [03:05:42] <tantek> "For some developers, 2005 was the year of microformats."
  36. [03:05:45] <tantek> well done folks!
  37. [03:47:43] * fuzzyBSc (n=fuzzy@c210-49-82-193.rochd2.qld.optusnet.com.au) has joined #microformats
  38. [03:58:30] * tantek (n=tantek@h-67-103-44-6.snfccasy.covad.net) Quit ()
  39. [04:13:36] * tantek (n=tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net) has joined #microformats
  40. [04:13:36] * ChanServ sets mode +o tantek
  41. [04:31:27] * factoryjoe (n=cmessina@c-69-181-81-22.hsd1.ca.comcast.net) Quit ()
  42. [04:37:41] * TantekC (n=Tantek@adsl-216-102-88-235.dsl.snfc21.pacbell.net) has joined #microformats
  43. [04:55:13] * tantek (n=tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net) Quit ()
  44. [04:55:14] * TantekC (n=Tantek@adsl-216-102-88-235.dsl.snfc21.pacbell.net) Quit (Read error: 104 (Connection reset by peer))
  45. [04:56:58] * tantek (n=tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net) has joined #microformats
  46. [04:56:58] * ChanServ sets mode +o tantek
  47. [05:24:37] * TantekC (n=Tantek@adsl-216-102-88-235.dsl.snfc21.pacbell.net) has joined #microformats
  48. [05:24:46] * tantek sets mode +o TantekC
  49. [05:37:01] * dustym (n=dustym@office.cavoksolutions.com) Quit (Remote closed the connection)
  50. [05:44:16] * dustym (n=dustym@office.cavoksolutions.com) has joined #microformats
  51. [05:52:16] * TantekC (n=Tantek@adsl-216-102-88-235.dsl.snfc21.pacbell.net) Quit (Remote closed the connection)
  52. [06:06:43] * bear (n=bear@pcp01380088pcs.levtwn01.pa.comcast.net) Quit (Read error: 110 (Connection timed out))
  53. [06:12:38] * bear (n=bear@pcp01380088pcs.levtwn01.pa.comcast.net) has joined #microformats
  54. [06:14:12] * TantekC (n=Tantek@adsl-216-102-88-235.dsl.snfc21.pacbell.net) has joined #microformats
  55. [06:14:24] * tantek sets mode +o TantekC
  56. [06:28:04] * TantekC (n=Tantek@adsl-216-102-88-235.dsl.snfc21.pacbell.net) Quit (Remote closed the connection)
  57. [06:29:38] <mfbot> [[hcard-examples]] http://microformats.org/wiki?title=hcard-examples&diff=0&oldid=3527 * Tantek * (+3331) hCard and XFN
  58. [06:34:06] * bear (n=bear@pcp01380088pcs.levtwn01.pa.comcast.net) Quit (Read error: 110 (Connection timed out))
  59. [06:34:56] * bear (n=bear@pcp01380088pcs.levtwn01.pa.comcast.net) has joined #microformats
  60. [07:11:16] * TantekC (n=Tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net) has joined #microformats
  61. [07:11:25] * tantek sets mode +o TantekC
  62. [07:32:14] <Atamido> _fil_: You can use a /, but you have to escape it first.
  63. [07:32:41] <Atamido> It is a very evil way to do things, and I would recommend against it for anything that would touch the internet, but you can do it.
  64. [07:34:40] <Atamido> In fact, you can have darn near anything as an ID, including a single ID with a space in it, just by escaping characters. http://www.w3.org/TR/CSS21/syndata.html#q6
  65. [07:35:42] <Atamido> I think the only things you can't have are an ID that starts with a number or a hyphen followed by a number.
  66. [07:35:45] <Atamido> Oh, and no new lines.
  67. [07:36:06] <Atamido> "Backslash escapes are always considered to be part of an identifier or a string (i.e., "\7B" is not punctuation, even though "{" is, and "\32" is allowed at the start of a class name, even though "2" is not)."
  68. [07:36:29] <Atamido> Anyone that has a "{" as part of a class name deserves to be hurt.
  69. [07:37:06] * tantek smells a hack.
  70. [07:37:33] <Atamido> I would appreciate it if no one mentions this to anyone else. :P
  71. [07:47:00] <mfbot> [[Main Page]] http://microformats.org/wiki?title=Main_Page&diff=0&oldid=3528 * Manjari * (+593)
  72. [07:47:56] <tantek> Spammer!!!
  73. [07:48:05] <Atamido> Yep.
  74. [07:48:22] <mfbot> [[Special:Log/block]] http://microformats.org/wiki?title=Special:Log/block&diff=0&oldid=0 * Tantek * (+0) blocked "User:Manjari" with an expiry time of infinite: Spammer
  75. [07:48:34] <Atamido> Heh.
  76. [07:48:43] <mfbot> [[Main Page]] M http://microformats.org/wiki?title=Main_Page&diff=0&oldid=3529 * Tantek * (-593) Reverted edit of Manjari, changed back to last version by Tantek
  77. [07:48:58] <tantek> Seeya, wouldn't wanna be ya
  78. [08:07:20] * TantekC (n=Tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net) Quit (Remote closed the connection)
  79. [08:16:11] * Atamido (n=Atamido@user-0ccsqt9.cable.mindspring.com) Quit (Read error: 104 (Connection reset by peer))
  80. [08:30:11] * TantekC (n=Tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net) has joined #microformats
  81. [08:30:25] * tantek sets mode +o TantekC
  82. [08:32:52] <Frederic> hello
  83. [08:34:43] <tantek> hey Frederic
  84. [08:35:15] * Atamido (n=Atamido@user-0ccsqt9.cable.mindspring.com) has joined #microformats
  85. [08:35:59] <Frederic> Hi tantek
  86. [09:40:17] * kingryan (n=kingryan@dsl081-240-149.sfo1.dsl.speakeasy.net) Quit ()
  87. [09:50:42] * TantekC (n=Tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net) Quit (Remote closed the connection)
  88. [11:51:07] * bergie (n=bergie@cs78242134.pp.htv.fi) has joined #microformats
  89. [13:08:00] * amanuel_ (n=amanuel@24.150.141.166) has joined #microformats
  90. [13:41:00] * dglazkov (n=dglazkov@adsl-065-081-081-030.sip.bhm.bellsouth.net) has joined #microformats
  91. [13:58:00] * t1m (n=t1m@129.44.184.116) Quit (Read error: 101 (Network is unreachable))
  92. [14:07:24] * tantek (n=tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net) Quit ()
  93. [14:48:19] * t1m (n=t1m@pool-151-199-44-108.bos.east.verizon.net) has joined #microformats
  94. [14:51:02] <_fil_> Atamido: the blog-post hAtom entry id="xxxx" could be unique if the id were the full "stable" URL of the blog-post
  95. [14:51:34] <_fil_> but since slashes are not allowed, I'm a bit lost as to what is best
  96. [14:52:26] <_fil_> 'article-89' has no useful meaning outside the internal representation of the article, even if the article has a nice url like my.site.net/Articles-s-title.html
  97. [14:53:07] <fuzzyBSc> Hmm... aren't we using the rel=bookmark url as an entry's id? Using an anchor in the local document would usually mean multiple identities for a given article.
  98. [14:53:46] <_fil_> right, there is a rel-bookmark link for the "stable url"
  99. [14:54:16] <_fil_> but also an "id" is requested http://microformats.org/wiki/hatom
  100. [14:54:45] <_fil_> I would think they were the same thing, because an object *is* an URL
  101. [14:55:25] <fuzzyBSc> Are you looking at the "entry permalink" section?
  102. [14:55:37] <_fil_> in fact it's written:
  103. [14:55:40] <_fil_> "if practical, also define id="unique-identifier" to the Entry."
  104. [14:57:14] <_fil_> entry permalink is fine, I have no problem with it
  105. [14:57:47] <_fil_> it's just the "id" that bothers me... i started by equating it with the permalink, but it fails the W3C XHTML validator
  106. [14:58:07] <fuzzyBSc> I'm not sure that the id="unique-identifier" is meant to be globally unique, or that it have any place in an equivalent atom representation.
  107. [14:58:43] <_fil_> rereading the sentence above, I think it's in fact meant to convey internals such as the internal id in our database
  108. [14:58:49] <fuzzyBSc> I am guessing this is mean more so that javascript and the like on the page can refer to it.... but it seems a weak reason to include it in any standard.
  109. [14:59:00] <_fil_> indeed
  110. [14:59:38] <fuzzyBSc> It doesn't appear in the explainatory text, or I don't think it does.
  111. [14:59:55] <fuzzyBSc> Could this be an oversight carried over from earlier thinking on the topic?
  112. [15:00:07] <_fil_> I wasn't here then, can't say
  113. [15:01:05] <fuzzyBSc> The current hAtom2Atom.xsl does not use any "id" values. It only uses the rel=bookmark permalink. It is probably worth raising the question as to what to do with an id if it does exist, and decide whether or not it should be included in the standard.
  114. [15:02:10] <fuzzyBSc> I'm afraid I'm somewhat of a newcomer, also.
  115. [15:02:32] <_fil_> I don't know hAtom2Atom.xsl - in fact i don't know xsl at all :)
  116. [15:02:52] <fuzzyBSc> Heh :)
  117. [15:03:17] <fuzzyBSc> It is available as an implementation link off the main hAtom page. I'm using it to parse my blog as hAtom through liferea.
  118. [15:03:59] <fuzzyBSc> Version 0.0.6 is reasonably complete, but I have submitted an initial test case with further bugfixes. I think 0.0.7 should be out fairly shortly.
  119. [15:04:12] <_fil_> I guess i should start getting interested in xsl processing if I want to do more mf
  120. [15:04:43] <_fil_> cause doing it "by hand" with regexp and the like is long and painful
  121. [15:05:02] <fuzzyBSc> Heh. Yeah... regex isn't the best at handling xml.
  122. [15:05:44] <fuzzyBSc> Your document needs to be valid xml before it can be processed with an xslt. If it isn't, push the input through tidy -asxhtml first. It is pretty good at handling things after that.
  123. [15:06:30] <fuzzyBSc> The current method being used to determine whether a node has a matching css class is pretty untidy, but once the idiom is established the other occurances are straightforward to understand.
  124. [15:07:12] <fuzzyBSc> xslt is better at handling data that is in element and attribute names. Microformats aren't the best match for it... but it is still probably the best available tool.
  125. [15:08:14] <_fil_> I'm prgramming for the common webmaster, who doesn't necessarily have php/xsl installed
  126. [15:08:26] <_fil_> that's why I'm using regex
  127. [15:08:44] <_fil_> but sometimes it's pure a waste of time
  128. [15:08:59] <_fil_> they don't always have tidy either
  129. [15:10:12] <fuzzyBSc> Hrmmm... I would have thought that an xslt capability would be more common than a regex capability these days ;)
  130. [15:10:38] <fuzzyBSc> I believe that xsltproc and tidy are both standalone binaries, and quite small.
  131. [15:12:31] <_fil_> true for tidy
  132. [15:13:19] <_fil_> our "target" (for SPIP - www.spip.net) is the common webmaster, say my mother, who wants to make a website and doesn't tell the difference between html, php or else
  133. [15:13:51] <_fil_> so it's not so easy to explain that they have to include a binary on their free/shared web hosting :)
  134. [15:14:24] <_fil_> anyway... time to learn xsl
  135. [15:14:29] <fuzzyBSc> Well, it's only the reader of the hAtom that needs tidy and xslt.
  136. [15:14:51] <fuzzyBSc> The server could do it for a reader, I suppose.
  137. [15:15:19] <fuzzyBSc> Personally I publish in strict xhtml so no tidy is required, but I would be requiring my readers to parse the hAtom by whatever means appropriate.
  138. [15:15:44] <fuzzyBSc> I render static pages to my hosting service, so no runtime software support is required.
  139. [15:17:13] <_fil_> we're doing lots of rss syndication, that's the catch
  140. [15:17:35] <fuzzyBSc> So you're trying to convert the published works to rss (or atom)?
  141. [15:17:48] <_fil_> the other way around
  142. [15:17:49] <fuzzyBSc> And you have to do it on the web hosting server because you aren't doing static rendering...
  143. [15:18:02] <_fil_> say you publish (by whatever means) a hAtom feed
  144. [15:18:09] * fuzzyBSc nods
  145. [15:18:24] <_fil_> I would like my rss parser to read it and import it as if it were an atom feed
  146. [15:18:43] <_fil_> the web hosting server is all we have, yes
  147. [15:19:19] <fuzzyBSc> In that case it is probably worth tracking the progress of hAtom2Atom.xsl if you have the necessary tools at your disposal. It'd be a shame to fork efforts at this time.
  148. [15:19:20] <_fil_> so I could try and parse the hAtom via regex (I'll spend a week on this project)
  149. [15:19:46] <fuzzyBSc> I know Luke is keen on seeing more test cases written to cover different aspects of hAtom parsing.
  150. [15:19:49] <_fil_> I could ask the user to use a hatom2atom proxy (but that would defeat the initial purpose)
  151. [15:20:41] <_fil_> or I could include the hAtom parsing only for those who have php5 (or libxslt) installed
  152. [15:21:02] <fuzzyBSc> The main hold-up with hAtom2Atom.xsl is that hAtom is not a completed specification. The output is not valid hAtom because elements like id an updated are missing at the feed level. hAtom doesn't currently specify what these should look like. Presumably one with have some influence over the other.
  153. [15:21:07] <_fil_> which would solve the problem for, say, 2007, which is fine with me
  154. [15:21:39] <fuzzyBSc> On the other hand, it is presently "valid enough" to get my feed reader to make use of it.
  155. [15:22:11] <_fil_> what i mean is that the feed reader itself should parse hAtom
  156. [15:22:18] * bergie (n=bergie@cs78242134.pp.htv.fi) Quit ()
  157. [15:22:49] <_fil_> so let's say I try Luke's xsl file
  158. [15:23:31] <_fil_> basically I have to read this file as xsl stylesheet, and the hatom page as source, process it via the xsl command, and treat the result as an Atom feed ?
  159. [15:23:52] <_fil_> is that correct? that's going to be one hour's work, i can try it
  160. [15:25:12] <fuzzyBSc> I don't know about other feed readers, but liferea allows any feed to passed through an arbitary command before parsing. You can either give it the command "xsltproc hAtom2Atom.xsl -", or "tidy -asxhtml | xsltproc hAtom2Atom.xsl -" for non-xhtml input.
  161. [15:25:40] <_fil_> I have to do it in php :)
  162. [15:25:44] <_fil_> if i can !
  163. [15:25:57] <fuzzyBSc> Assuming you have xslt processing at your disposal the hAtom2Atom.xsl should give you the best available atom file to then go and parse as per usual.
  164. [15:26:58] <fuzzyBSc> With the spec in the state it is in I wouldn't go trying to write a native parser of hAtom at this stage. Once it is fully defined and prototyped with the xsl transform I would think about it.
  165. [15:27:26] <_fil_> let's try
  166. [15:27:35] <_fil_> is there a nice hatom page somewhere?
  167. [15:28:06] <fuzzyBSc> Heh. Try mine: http://members.optusnet.com.au/benjamincarlyle/benjamin/blog/
  168. [15:33:03] <fuzzyBSc> So where are you from _fil_?
  169. [15:33:18] <fuzzyBSc> I don't usually see #microformats active in my time zone.
  170. [15:34:35] <_fil_> paris
  171. [15:34:37] <_fil_> france
  172. [15:35:02] <fuzzyBSc> :)
  173. [15:35:25] <fuzzyBSc> I flew over continental europe a few months back, but unfortunately did not get to stop until I hit the UK.
  174. [15:36:02] <_fil_> well no luck
  175. [15:36:09] <_fil_> I do the simplest thing
  176. [15:36:09] <_fil_> $xsl = new DOMDocument;
  177. [15:36:09] <_fil_> $xsl->load('hAtom2Atom.xsl');
  178. [15:36:09] <_fil_> var_dump($xsl);
  179. [15:36:24] <_fil_> and I get an empty object :(
  180. [15:36:52] <fuzzyBSc> Hrmm... hAtom2Atom.xsl should be valid xml, well and truely.
  181. [15:37:11] <_fil_> (maybe var_dump is not good with objects)
  182. [15:37:21] <fuzzyBSc> Is your path to hAtom2Atom.xsl correct? Would php throw an exception on failure?
  183. [15:38:00] <_fil_> ok it works
  184. [15:38:12] <fuzzyBSc> :)
  185. [15:38:14] <fuzzyBSc> That's good.
  186. [15:39:18] <_fil_> it 'finds' 40 elements on your page
  187. [15:39:50] <fuzzyBSc> Hmm... that sounds like it is in the right ballpark.
  188. [15:39:58] <fuzzyBSc> That would be atom entries?
  189. [15:40:25] <_fil_> yep
  190. [15:41:12] <_fil_> Intelligent Design
  191. [15:41:12] <_fil_> http://members.optusnet.com.au/benjamincarlyle/benjamin/blog/2005/12/22#intelligentDesign
  192. [15:41:15] <_fil_> and so on
  193. [15:41:17] <_fil_> RESTful Blogging
  194. [15:41:20] <_fil_> http://members.optusnet.com.au/benjamincarlyle/benjamin/blog/2005/12/21#restfulBlogging
  195. [15:41:23] <_fil_> ain't that great? :)
  196. [15:41:46] <_fil_> ok so that was 20 minutes in all
  197. [15:41:49] <_fil_> not so bad
  198. [15:41:55] <_fil_> thanks php5
  199. [15:42:58] <_fil_> whew... on http://blog.davidjanes.com/ I get so many errors
  200. [15:43:25] <fuzzyBSc> Heh :)
  201. [15:44:43] <fuzzyBSc> I haven't had too much luck with the other implementation examples. In both cases I have concluded it is due to a lack of appropriate information in the source pages. If you're getting actual errors, though, it is probably not a valid xhtml page. It may be plain old xml, in which case you'll need to run tidy (in some form) over it first.
  202. [15:45:03] <fuzzyBSc> s/plain old xml/plain old html/
  203. [15:45:26] <_fil_> yes
  204. [15:45:37] <_fil_> that's not very nice methinks
  205. [15:46:14] <fuzzyBSc> This is one of the advantages and disadvantages of microformats. They are mean to work in either xhtml or in html. That tends to mean you have to support parsing of both. Actual xhtml is rare enough still that supporting only it would limit your parser's applicability.
  206. [15:46:14] <_fil_> normally you should be strict in what you publish but lax in what you read
  207. [15:46:23] <_fil_> all those XML functions are strict in reading
  208. [15:47:42] <fuzzyBSc> This is true, but the SGML->XML tipping point came about naturally. It just took too much code to correctly parse all of the bad html out there. Mobile devices couldn't really cope. XML introduces a "reasonable" amount of strictness in interpretation... but is still often mispublished.
  209. [15:48:55] <_fil_> tidy -ashxtml is not enough: I get Entity 'agrave' not defined in /Users/fil/Sites/spip/CACHE/atom-tidy.xml, line: 24 in /Users/fil/Sites/spip/hatom.php on line 17
  210. [15:49:11] <_fil_> and all other entities (nbsp and more)
  211. [15:49:44] <_fil_> "still often mispublished" means you need tidy n the reading end
  212. [15:50:40] <fuzzyBSc> My "hAtom2Atom" script always runs tidy over the content before running the xslt. I think that's reasonable on desktop hardware :)
  213. [15:51:04] <_fil_> yes and no
  214. [15:51:14] <_fil_> yes it's reasonable in terms of processing power
  215. [15:51:16] <fuzzyBSc> It's a shame entities aren't being processed correctly. You would assume they would be skipped over should they fail to be understood.
  216. [15:52:01] <_fil_> one micro-error and your connection is lost
  217. [15:52:24] <fuzzyBSc> _fil_: It depends on whether your mantra is "be lax in what you read", or "It's about time everyone learned to produce output that is valid at some level or another".
  218. [15:52:47] <_fil_> You can't put pressure on "everyone"
  219. [15:53:03] <_fil_> if I need the information, I take what's out there
  220. [15:53:31] <_fil_> it's not a competition of standards engineers, it's really still about having people publish and exchange things
  221. [15:54:04] <_fil_> I try tidy -numeric -utf8 -asxhtml
  222. [15:54:14] <_fil_> but it still doesn't process the entities
  223. [15:59:16] * fuzzyBSc tries the link
  224. [16:00:27] <fuzzyBSc> Using the command-line tidy and xsltproc I get plenty of warnings, but in the end a reasonble atom feed version of the input.
  225. [16:02:19] <fuzzyBSc> Hrmm... I don't actually get warnings on entities, either... just html -> xhtml warnings.
  226. [16:02:27] <_fil_> can you process this one? http://fil.rezo.net/hatom.html
  227. [16:02:31] * RobertBachmann (n=RobertBa@M2430P020.adsl.highway.telekom.at) has joined #microformats
  228. [16:02:37] <_fil_> it's my test site (on localhost)
  229. [16:02:46] <_fil_> I just copied the page over to online
  230. [16:04:52] <fuzzyBSc> Currently, the output is:
  231. [16:04:54] <fuzzyBSc> <feed xmlns="http://www.w3.org/2005/Atom" xml:lang="fr">
  232. [16:04:54] <fuzzyBSc> <title>Un cyber conflit judiciaire d’URL à 500.000 euros</title>
  233. [16:04:54] <fuzzyBSc> </feed>
  234. [16:05:08] * fuzzyBSc has a look at the input
  235. [16:05:28] <_fil_> that's what I get too
  236. [16:05:45] <fuzzyBSc> You're using atomentry for your entries...
  237. [16:05:54] <fuzzyBSc> The xsl is expecting "entry".
  238. [16:05:57] * fuzzyBSc checks the spec.
  239. [16:06:06] <_fil_> oh ho
  240. [16:06:38] <_fil_> right!!
  241. [16:06:47] <RobertBachmann> greetings
  242. [16:07:09] <fuzzyBSc> Yeah, I think all of the atomfeed/atomentry was dropped before publication.
  243. [16:07:14] <fuzzyBSc> Morning, RobertBachmann :)
  244. [16:07:22] <_fil_> evening, RobertBachmann :)
  245. [16:07:37] <_fil_> the urls are not there either
  246. [16:08:44] <fuzzyBSc> _fil_: My copy contains a rel=bookmark, which should come out as something :)
  247. [16:08:57] <fuzzyBSc> I'm not sure whether the relative URL will be handled correctly by the current transform.
  248. [16:09:15] <Atamido> _fil_: As I was saying, slashes are allowed, they simply have to be escaped.
  249. [16:10:26] <fuzzyBSc> _fil_: Also check whether it would be wise to include an "updated" entry. Published is not required by updated is.
  250. [16:10:57] <fuzzyBSc> The use of address is not enough to be parsed as meaning "author". You'll need to put class="author" onto it to make that work.
  251. [16:11:08] <_fil_> ok
  252. [16:11:30] <fuzzyBSc> The rest looks good on a cursary inspection :)
  253. [16:11:50] <_fil_> the URL is only part of the <content><h2><a href="URL">title<////>
  254. [16:12:07] <_fil_> in the resulting atom page
  255. [16:12:20] <_fil_> I don't get the <link> thing
  256. [16:13:27] <fuzzyBSc> I think I still have the old version. You should get both an <id> and <link> element in the output if your rel=bookmark is working.
  257. [16:13:56] <_fil_> do you get them? I don't
  258. [16:14:08] <fuzzyBSc> Oh...
  259. [16:14:20] <fuzzyBSc> Why does your h2 have a class=content against it?
  260. [16:14:29] <_fil_> isn't it content ?
  261. [16:14:52] <fuzzyBSc> The title isn't content...
  262. [16:15:16] <fuzzyBSc> Well... it could be part of the content... but that's not how things are currently set up.
  263. [16:15:22] <fuzzyBSc> ... in the xsl, that is.
  264. [16:15:39] <_fil_> ok
  265. [16:15:47] <_fil_> well, it's getting better
  266. [16:15:55] <_fil_> let me send a new version
  267. [16:16:06] <fuzzyBSc> The title (either class=title, or h?) is a separate node in the atom, and hAtom sort of assumes it is separate also. The xsl won't parse within it if it thinks it is content :)
  268. [16:16:15] * bergie (n=bergie@cs78242134.pp.htv.fi) has joined #microformats
  269. [16:16:39] <_fil_> try http://fil.rezo.net/hatom.html
  270. [16:17:06] <_fil_> this content-title thing might be a bug in the XSL file
  271. [16:17:14] <_fil_> now to author
  272. [16:17:29] <fuzzyBSc> I'm getting title, id, link, content... yep :)
  273. [16:17:56] <fuzzyBSc> Putting class=author onto your address element should do the trick, methinks.
  274. [16:18:02] * tantek (n=tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net) has joined #microformats
  275. [16:18:03] * ChanServ sets mode +o tantek
  276. [16:18:18] <_fil_> you have it now
  277. [16:18:59] <fuzzyBSc> You won't get an email address out of that, of course... but I now have author.name and author.url
  278. [16:19:12] <fuzzyBSc> Hrrm.. I think that should be uri :) Another bug.
  279. [16:20:15] <fuzzyBSc> fil: It looks to me like everything you're putting in is now coming back out of the xsl :)
  280. [16:20:32] <_fil_> well, in one hour we get it working, proof of concept and implementation
  281. [16:20:46] <_fil_> I think I deserve a tea and a break
  282. [16:20:49] <_fil_> and you too :)
  283. [16:21:36] <_fil_> fuzzyBSc: for atomentry vs entry, I trusted http://microformats.org/wiki/hatom-issues#Why_atomentry.3F
  284. [16:22:21] * _fil_ boils tea
  285. [16:22:44] <fuzzyBSc> :)
  286. [16:23:29] <fuzzyBSc> Yes, I think some of the supporting documentation is a little inconsistent with the current spec. Even the spec itself has a few items that seem contradictary still.
  287. [16:24:42] <_fil_> we have to ask tantek
  288. [16:24:48] <fuzzyBSc> :)
  289. [16:27:44] <RobertBachmann> regarding Atom's id element and hAtom: I think we should document our concerns on the http://microformats.org/wiki/hatom-issues page
  290. [16:28:03] <_fil_> my current (functioning) code is at http://fil.rezo.net/hatom.phps with a few comments
  291. [16:28:51] * fuzzyBSc starts a writeup (unless someone else is doing it?)
  292. [16:29:36] <_fil_> I'm not, unless you want it in french
  293. [16:29:43] <_fil_> don't link to my files, it's unstable
  294. [16:29:57] <_fil_> I still have to get a blog up and running :)
  295. [16:30:00] <tantek> good morning
  296. [16:30:13] <RobertBachmann> hi tantek
  297. [16:30:19] * tantek sets mode +o RobertBachmann
  298. [16:31:03] <tantek> _fil_, please go ahead and add issues to the hatom-issues document
  299. [16:32:09] <_fil_> good evening tantek
  300. [16:32:35] <tantek> David Janes has done some great work, but it does still need review and feedback!
  301. [16:32:43] <_fil_> two issues to report: one is atomentry vs entry, which is not clear
  302. [16:32:45] <tantek> I myself still have to try it out on my blog
  303. [16:33:02] <tantek> will almost certainly do so starting with next months posts
  304. [16:33:20] <_fil_> the other is that class="author" was lost on me (was it written somewhere? dunno)
  305. [16:33:48] <_fil_> I remember I searched for it and didn't find it
  306. [16:35:02] <_fil_> oh, david jane's blog is now read by my code :)
  307. [16:36:27] <mfbot> [[hatom-issues]] http://microformats.org/wiki?title=hatom-issues&diff=0&oldid=3530 * FuzzyBSc * (+602) Add Identification section
  308. [16:36:33] <_fil_> fuzzyBSc: I don't think "lang" is processed correctly, although it is present in http://fil.rezo.net/hatom.html
  309. [16:39:48] <fuzzyBSc> _fil_: Under "Recent Changes" is the text, "Entry Author most explicitly be marked class="author"". Also, see http://microformats.org/wiki/hatom#Entry_Author
  310. [16:40:14] <fuzzyBSc> Hrrm... I haven't touched the lang handling and haven't really looked at how it works.
  311. [16:40:20] * fuzzyBSc reads up on it.
  312. [16:40:47] <fuzzyBSc> I know it is currently marked as not necessarily being consistent with the spec, or the spec being unclear at the time of writing.
  313. [16:41:31] <RobertBachmann> The hatom draft says nothing about lang. This is just an extension for hatom2atom.xsl
  314. [16:42:09] <fuzzyBSc> Ok... it is looking for an xml:lang attribute. It doesn't seem to look for plain "lang".
  315. [16:42:10] <tantek> BTW, I encourage you to also add your name to the "Contributors" section when adding issues to hatom-issues
  316. [16:42:26] <fuzzyBSc> tantek: Ok.
  317. [16:44:08] <mfbot> [[hatom-issues]] http://microformats.org/wiki?title=hatom-issues&diff=0&oldid=3531 * FuzzyBSc * (+83) Add my name, as per tantek's suggestion
  318. [16:44:34] <RobertBachmann> yes, it only looks for xml:lang. The reason is that tidy will automaticaly add "xml:lang" to all elements which have a lang attribute. But I think I should change it to look for "lang" when there's no "xml:lang"
  319. [16:45:01] <_fil_> is it the same concept?
  320. [16:45:20] <mfbot> [[press]] M http://microformats.org/wiki?title=press&diff=0&oldid=3532 * Tantek * (+170) December -
  321. [16:45:37] <tantek> Robert, yes
  322. [16:45:44] <fuzzyBSc> fil: Yes. I believe that newer xhtml drafts drop the lang, but earlier ones include it because xml:lang (a separate standard) had not been completed.
  323. [16:45:57] <tantek> "lang" is more cross-mime-type compatible
  324. [16:46:27] <_fil_> I also included "dir" as I'm working on many very-multilingual sites
  325. [16:46:30] <tantek> "lang" works in HTML 4 docs, and XHTML 1.0 docs, whether served as text/html or application/xhtml+xml
  326. [16:46:31] <fuzzyBSc> The recommendations I have seen for good xhtml say "use both" when publishing :) That way everyone sees what they should.
  327. [16:46:39] <_fil_> with arabic and farsi
  328. [16:46:54] <tantek> whereas xml:lang only works in XHTML when served as application/xhtml+xml
  329. [16:47:30] <tantek> fuzzyBSc, XHTML 1.1 drops the "in-language" 'lang' attribute
  330. [16:47:41] <tantek> but XHTML 1.1 is not practical to serve today for various reasons
  331. [16:48:33] * fuzzyBSc nods
  332. [16:49:18] <tantek> _fil_ even if your files are unstable, it's still worth linking to them, and explicitly pointing out that caveat
  333. [16:49:34] <_fil_> I'd rather first set things up a bit
  334. [16:49:46] <_fil_> I've added xml:lang now
  335. [16:49:57] <_fil_> ok, my daughters are back
  336. [16:50:02] <_fil_> time to log off
  337. [16:50:10] <hlb> hello?
  338. [16:50:21] <tantek> hello
  339. [16:50:27] <hlb> anyone plays with structured blogging?
  340. [16:50:29] <fuzzyBSc> Morning, hlb :)
  341. [16:50:36] <tantek> _fil_, if you're still around, feel free to add your implementation link to http://microformats.org/wiki/hatom#Implementations
  342. [16:50:39] <hlb> tantek: hello, tantek :)
  343. [16:50:57] <tantek> hello hlb
  344. [16:51:00] <hlb> fuzzyBSc: midnight here :)
  345. [16:51:22] <tantek> New Zealand?
  346. [16:51:28] <fuzzyBSc> It's just coming on three am in my locality. I'm on holiday hours.
  347. [16:51:35] <hlb> tantek: no, Taiwan
  348. [16:51:39] <tantek> ah
  349. [16:52:50] <hlb> I just install the new structured blogging plugin for wordpress.
  350. [16:53:12] <hlb> and find it so.... complicated.
  351. [16:54:47] <hlb> maybe the microformats should be transparent to (blog) users?
  352. [16:55:08] <tantek> hlb, ideally yes
  353. [16:55:25] <tantek> the user should be able to use the easiest user interface possible
  354. [16:55:34] <tantek> while the code should write out the richest microformats it can
  355. [16:58:03] <hlb> but not now? :)
  356. [16:59:04] <tantek> it's good to start somewhere
  357. [16:59:10] <tantek> and rapidly iterate and improve
  358. [16:59:31] <mfbot> [[press]] M http://microformats.org/wiki?title=press&diff=0&oldid=3533 * Tantek * (-25) December -
  359. [16:59:57] * bergie (n=bergie@cs78242134.pp.htv.fi) Quit ()
  360. [17:01:22] <mfbot> [[press]] M http://microformats.org/wiki?title=press&diff=0&oldid=3534 * Tantek * (+109) December -
  361. [17:01:32] <fuzzyBSc> hlb: Organisationally structured blogging is a separate initiative to microformats. I have used microformats, but not SB. Perhaps if you have specific complaints they would be best made nearer to the author of the software... ;)
  362. [17:01:56] <fuzzyBSc> Alternatively, my understanding is they are open source and open to involvement by interested users such as yourself.. :)
  363. [17:02:52] <RobertBachmann> regarding press: Do we only list articles published in English?
  364. [17:03:18] <hlb> haha, ya.
  365. [17:03:39] * fuzzyBSc grins
  366. [17:04:17] * hlb but still try hard to make my blog m17n :p
  367. [17:05:08] <mfbot> [[press]] M http://microformats.org/wiki?title=press&diff=0&oldid=3535 * Tantek * (+57)
  368. [17:05:19] <tantek> Robert, no we include others also
  369. [17:05:34] <tantek> see http://microformats.org/wiki/press#July for example
  370. [17:05:55] <tantek> m17n?
  371. [17:06:01] <tantek> ?def m17n
  372. [17:06:03] <fuzzyBSc> Hrrm.... yes. Multilingual is hard. The protocols support it, but the tools often don't.
  373. [17:06:13] <tantek> hmm... where did jibot go?
  374. [17:06:22] <hlb> what's different between rel-home & <link rel="start"> ?
  375. [17:06:24] <tantek> Kevin, how do wwe reinvite jibot?
  376. [17:06:37] <tantek> hlb, they can be the same
  377. [17:06:52] <fuzzyBSc> hlb: rel-home is the home page of a site. rel-start is the first page in a sequence of pages.
  378. [17:06:56] <tantek> if for example your web page only has one series of pages
  379. [17:07:13] <tantek> however if you have several "collections" of pages, each of which has their own "start" or "first" page
  380. [17:07:30] <tantek> then you can have multiple pages on your site referenced by rel="start"
  381. [17:07:43] <tantek> yet you should probably have only one rel="home" per "site"
  382. [17:07:44] <hlb> I am trying to adapt the polyglot plugin (wp) in my own website.
  383. [17:07:49] <tantek> that's my guess at the difference ;)
  384. [17:08:25] <tantek> hlb, how is the Traditional Chinese translation of the Knowledge at Wharton article? http://knowledge2.wharton.com.cn//index.cfm?fa=article&articleid=1203&specialid=58&l=4&languageid=5
  385. [17:08:33] <hlb> http://fredfred.net/skriker/index.php/polyglot # not best, but good enough...
  386. [17:08:50] <hlb> tantek: checking
  387. [17:08:59] <mfbot> [[press]] M http://microformats.org/wiki?title=press&diff=0&oldid=3536 * RobertBachmann * (+163) June -
  388. [17:09:15] <fuzzyBSc> hlb: Nice.
  389. [17:09:35] <hlb> tantek: not bad :)
  390. [17:09:47] <hlb> tantek: i wrote one too: http://hlb.yichi.org/blog/2005/08/19/8
  391. [17:11:07] <hlb> fuzzyBSc: still waiting wordpress 2.0 release... hate to merge...
  392. [17:11:40] <tantek> Robert that is very cool - I hadn't seen that
  393. [17:12:23] <tantek> hlb, nice post! i wish i could read chinese though.
  394. [17:13:18] <hlb> tantek: ya, it's my reason for m17n :-/
  395. [17:14:08] <hlb> tantek: people say 'good' but can't read :p
  396. [17:14:44] <tantek> what does m17n stand for?
  397. [17:14:52] <hlb> tantek: there is a interesting translation about ur name, in that post.
  398. [17:15:06] <hlb> tantek: multilingual
  399. [17:15:13] <fuzzyBSc> tantek: multilingualization
  400. [17:15:36] <tantek> that's a mouthful!
  401. [17:15:48] <hlb> truly.
  402. [17:15:51] <fuzzyBSc> :)
  403. [17:16:46] <RobertBachmann> see also http://en.wikipedia.org/wiki/I18N#Alternative_names
  404. [17:20:03] <fuzzyBSc> RobertBachmann: It is something a little more than internationalisation, though. Internationalisation typically puts data into some language neutral form for localisation for a single people group. Tools like gettext tend to be written from the perspective that internationalised data will be transformed into a single local version.
  405. [17:20:26] <hlb> it's always hard to put more than one language in a website.
  406. [17:20:41] <fuzzyBSc> The multilingual approach is to support different languages simultaneously. This is pretty important on the server side when there is no "neutral" format for content... only a collection of localised versions.
  407. [17:21:30] * fuzzyBSc has felt this pain recently in his professional workload :)
  408. [17:21:37] <hlb> !
  409. [17:21:59] <fuzzyBSc> I write SCADA systems :)
  410. [17:22:52] <fuzzyBSc> There is no actual web involved most of the time, but the princples and tools are the same. Sometimes you sell to portugal, and need to support five languages out of the same set of servers ;)
  411. [17:24:09] <hlb> I thought the interface translation is not too hard, since we have the great gettext.
  412. [17:24:39] <hlb> but the content is the other thing :(
  413. [17:25:48] <hlb> for example, if I blog (or write) something in chinese, and I want to write in another language again.
  414. [17:26:32] <hlb> there is always difference between the two version.
  415. [17:27:21] <fuzzyBSc> Well, say you are generating an event. You get your english version "The current on that circuit breaker just jumped to high", but you also have four other versions for different languages. You need to store them on disk in their localised formats. Either that, or keep your gettext catalogue around forever.... events have to be stored for long periods for legal reasons.
  416. [17:28:09] <fuzzyBSc> If you are using gettext you have to switch your program's locale five times for each event, and that kills all of gettext's caching. It's a bit of a mess :)
  417. [17:28:21] <hlb> haha.
  418. [17:29:28] <fuzzyBSc> We go for something similar to your polyglot approach and keep a bundle of localised messages together in the one data structure.
  419. [17:30:14] <hlb> I think put all localised messages into one entity is a reasonable choice :)
  420. [17:31:01] <hlb> since they are all variants for one entity.
  421. [17:31:44] <hlb> the polyglot plugin uses <html_en>hi</html_en> <html_zh>...</html_zh>...
  422. [17:31:55] <hlb> I think it is good enough for me.
  423. [17:32:21] <hlb> # of course I won't write my blog in 10 languages :p
  424. [17:33:10] <fuzzyBSc> I think that ideally you would have xml:lang tags on each to allow different zh variants, for example. Not knowing wp though, I don't know what the tradeoffs would be.
  425. [17:33:35] <hlb> and I think microformats is cool, because it use html & we don't need to worry about the language problem again & again.
  426. [17:34:06] <fuzzyBSc> Yeah, if you have a solution for html you have one for microformats.
  427. [17:34:21] <hlb> fuzzyBSc: u mean put all languages into one web page?
  428. [17:34:49] <hlb> fuzzyBSc: and use <p lang="xx"> to implement m17n?
  429. [17:37:35] <fuzzyBSc> hlb: You couldn't do that with stright html. The clients would just interpet subsequent paragraphs as being different languages that you still want to see. In the wp input, though, ideally you would be able to get specific with variants such as TW for your zh and even GB for my en.
  430. [17:38:59] <fuzzyBSc> I'm sure you can answer this more definitively than I, but my understanding is that someone who writes simplified chinese often cannot read traditional and vice versa... so this information is more important than in english where we basically understand each other regardless of strange spelling and grammar :)
  431. [17:39:58] <hlb> fuzzyBSc: no, people write simplified chinese can read traditional too (mostly)
  432. [17:40:19] <hlb> fuzzyBSc: and many people in Taiwan can read simplified chinese too.
  433. [17:41:34] <fuzzyBSc> Ok. Well, I guess simplified only came about fairly recently in the greater scheme of things.
  434. [17:42:24] <hlb> simplified chinese characters come from politics reason...
  435. [17:43:25] <hlb> just because the china government wants to make people write faster & learn easier.
  436. [17:43:56] <hlb> but we have computers and don't need to 'write faster'.
  437. [17:44:20] <hlb> and the simplified chinese characters somewhat make learning harder, in my point of view.
  438. [17:44:23] <hlb> :p
  439. [17:44:25] * fuzzyBSc nods
  440. [17:44:31] * fuzzyBSc grins
  441. [17:44:42] <fuzzyBSc> I don't want to get too deep into that argument ;)
  442. [17:44:49] <hlb> ya :)
  443. [17:45:31] <fuzzyBSc> I can certainly see how the radicals get lost in simplified. The history of each character is written into it in traditional which should make it easier to remember.
  444. [17:46:47] <hlb> btw, if I use xml:lang for the plugin, then I can't use xml:lang in the post...?
  445. [17:47:09] <fuzzyBSc> Oh, and sorry about Alexander Downer. We Australians are embarrassed every time he opens his mouth.
  446. [17:47:50] <hlb> O_o
  447. [17:48:46] <fuzzyBSc> hlb: I'm not familiar enough with wp to really make a statement. I would assume that an xml:lang on a html element would be the only one you need in general... but it can be overridden by more deeply-nested elements if you need it to be.
  448. [17:48:56] <hlb> tantek: I think the article in Knowledge@Wharton is readable, but it reads like 'english article'.
  449. [17:49:18] <tantek> like it was translated?
  450. [17:49:30] <tantek> and it is an interview in English, so that kind of makes sense
  451. [17:50:00] * RobertBachmann (n=RobertBa@M2430P020.adsl.highway.telekom.at) Quit (Read error: 104 (Connection reset by peer))
  452. [17:50:57] <hlb> tantek: ya, and some words are not in well-known chinese.
  453. [17:51:23] <hlb> tantek: but.. well, good enough :p
  454. [17:52:35] <hlb> fuzzyBSc: since wp is a blogging system, we just put the entry content in it, not entirely html.
  455. [17:52:53] <hlb> fuzzyBSc: so there is no html element in it.
  456. [17:54:12] <hlb> fuzzyBSc: therefore I have to use some delimiter (like html_xx) to split different translation.
  457. [18:01:39] * RobertBachmann (n=RobertBa@M2462P005.adsl.highway.telekom.at) has joined #microformats
  458. [18:08:38] * fuzzyBSc nods
  459. [18:08:43] <fuzzyBSc> Well, I'd better get some sleep.
  460. [18:08:53] <RobertBachmann> I've fixed the problem with lang in hatom2atom.xsl. however I spotted some more bugs.
  461. [18:09:08] <hlb> fuzzyBSc: thank you :)
  462. [18:09:10] <fuzzyBSc> RobertBachmann: Oh?
  463. [18:09:18] <fuzzyBSc> :)
  464. [18:10:43] <fuzzyBSc> My list currently includes,
  465. [18:10:43] <fuzzyBSc> * Missing published in output
  466. [18:10:43] <fuzzyBSc> * link element gets rel="bookmark" instead of rel="alternate"
  467. [18:11:03] <fuzzyBSc> * vcard processing doesn't normalize-space on fn
  468. [18:11:21] <fuzzyBSc> * contributor template has select instead of match
  469. [18:11:40] <fuzzyBSc> * vcard url should be uri.
  470. [18:12:05] <fuzzyBSc> Yourself?
  471. [18:12:41] * bergie (n=bergie@cs78242134.pp.htv.fi) has joined #microformats
  472. [18:13:49] <RobertBachmann> had this too: contributor template has select instead of match
  473. [18:13:49] <RobertBachmann> and the link generation is realy broken.
  474. [18:13:49] <RobertBachmann> <link rel="alternate">
  475. [18:13:49] <RobertBachmann> <xsl:for-each select="@*">
  476. [18:13:49] <RobertBachmann> <xsl:copy/>
  477. [18:14:08] <_fil_> sorry I missed the m17n discussion -- I've spent 1 yr doing it for SPIP, and there were so many nice things to consider
  478. [18:14:16] <fuzzyBSc> I have put together one test case for Luke, but more are required. Also definition of what is going at the feed level needs to be finished in the standard.
  479. [18:14:35] <_fil_> for multilingual content we finally opened two possibilities:
  480. [18:14:45] <_fil_> either different articles with translation links
  481. [18:15:06] <_fil_> or "<multi> blocks", which are written:
  482. [18:15:21] <_fil_> <multi>[fr]Chose [en]Thing</multi>
  483. [18:15:43] <hlb> SPIP?
  484. [18:15:54] <_fil_> the idea is that if you need to print this block when you are in a german context, you'll use the first of the two strings
  485. [18:16:01] <fuzzyBSc> RobertBachmann: How does that link generation break things?
  486. [18:16:04] <_fil_> if it's [en]glish, you'll use "Thing"
  487. [18:16:11] <_fil_> hlb: www.spip.net
  488. [18:16:36] <hlb> wow.
  489. [18:16:54] <hlb> checking
  490. [18:16:58] <_fil_> it's a bit late for the Chinese translation
  491. [18:16:59] <_fil_> see http://www.spip.net/rubrique4.html
  492. [18:17:21] <_fil_> I especially love the RTL/LTR thing
  493. [18:17:35] <fuzzyBSc> _fil_: So this is an xml rather than a html thing? The parser knows which one to pick?
  494. [18:17:47] <_fil_> like in http://www.spip.net/fa_article2077.html
  495. [18:18:12] <_fil_> it's just a block that you embed ni your content
  496. [18:18:35] <_fil_> and depending on the current language context you select the right language/translation
  497. [18:19:00] <_fil_> the context itself obeys different rules
  498. [18:19:25] <_fil_> usually, for example, an article has a fixed language, so the context is always its language
  499. [18:19:41] <_fil_> so <multi> blocks are not really relevant there
  500. [18:19:53] <_fil_> however an author's language context is not fixed
  501. [18:20:37] <_fil_> so if <multi>[fr]George [ar]&#456;</multi> "signs" an article in French, it will be signed as "George", and if the article is in arabic it will be written as &#456;
  502. [18:21:04] * bergie (n=bergie@cs78242134.pp.htv.fi) Quit ()
  503. [18:21:07] <_fil_> (except if we wish otherwise, I'm just explaining the default behaviour here)
  504. [18:21:20] <fuzzyBSc> And that is translated by your templating engine into html that matches the content of the current page?
  505. [18:21:31] <hlb> the icon is soooo cute.
  506. [18:21:34] <_fil_> yes
  507. [18:22:18] <_fil_> you have fun in the back-office when you need to view arabic content in an esperanto context :)
  508. [18:22:21] <fuzzyBSc> I wonder if something cute could be done with css to filter out content that doesn't apply to the current user language.
  509. [18:22:27] <_fil_> and it works
  510. [18:22:42] <_fil_> "the current user language" is typically a bad idea imho
  511. [18:23:00] <_fil_> for example my web browser is set to "french, english, other"
  512. [18:23:34] <_fil_> but I need sometimes to check pages in japanese, and I want to see the "original" page
  513. [18:23:53] <_fil_> best thing is to have a nice navigation, but fixed pages
  514. [18:24:01] <_fil_> however ymmv
  515. [18:24:29] <RobertBachmann> fuzzyBSc: <a class="foo bar" rel="bookmark" href="..."> becomes <link class="foo bar" rel="alternate" href="...">
  516. [18:24:43] <RobertBachmann> which isn't proper Atom
  517. [18:25:34] <hlb> _fil_: ya, I agree with u.
  518. [18:26:04] <_fil_> for content-based websites, at least; I'm not interested in e-commerce :)
  519. [18:26:11] <fuzzyBSc> RobertBachmann: I've submitted a patch that filters out the rel, but I haven't filtered out the class. Are there any other fields that are valid in a html <a> element but not in an atom <link>?
  520. [18:26:28] <fuzzyBSc> Hrmm.. feel free to submit your own patch :) I might just go to sleep as promised.
  521. [18:26:46] <_fil_> is there any common place where all this is developed? Subversion?
  522. [18:27:02] <hlb> _fil_: it's good to auto identify the user's language prefs, but never restrict people to view one language only :)
  523. [18:27:11] <_fil_> scripts, xsl sheets, and so on?
  524. [18:27:20] <fuzzyBSc> _fil_: I've just been submitting directly back to Luke. It is only one file at present, so the CM doesn't have to be complicated.
  525. [18:27:27] <_fil_> yes hlb
  526. [18:28:10] <_fil_> fuzzyBSc: in my view it does not have to be complicated to need a revision/history/collaborative repository
  527. [18:28:38] <_fil_> (I'm willing to contribute a SVN+trac repo if it can be of any use)
  528. [18:28:41] <fuzzyBSc> hlb, _fil_: Although the RESTful approach might be to say http://example.com/page is a suitable representation based on your browser's language preferences, while http://example.com/page/jp is always the japanese version.
  529. [18:29:38] <fuzzyBSc> _fil_: Just think of it as a drcs based on diff -u patches of the file from different locations, with Luke as the build manager ;)
  530. [18:29:38] * factoryjoe (n=cmessina@c-69-181-81-22.hsd1.ca.comcast.net) has joined #microformats
  531. [18:29:48] <_fil_> I don't really like it, as it makes impossible to exchange an email saying: "Hi Jane why do you write "<div>" and not "<div/>" on this URL?
  532. [18:30:10] <_fil_> if "this URL" contents depends on Jane's browser and mine
  533. [18:30:39] <_fil_> fuzzyBSc: hence the "fuzzy" part of your pseudo:)
  534. [18:30:55] <hlb> fuzzyBSc: i'm working on that :)
  535. [18:31:10] <fuzzyBSc> _fil_: That's the debugger's point of view, but forwarding a link on and having your japanese friend see it in their own language might also be of some benefit :)
  536. [18:31:36] <RobertBachmann> fuzzyBSc: yes for example lang, id, ... These and most other attributes from html don't make sense in that context of Atom
  537. [18:31:39] <hlb> _fil_: mondediplo.com is interesting example... :)
  538. [18:31:42] <hlb> # http://mondediplo.com/2005/12/01torture
  539. [18:32:06] <_fil_> disclosure: I'm the webmaster for mondediplo :)
  540. [18:32:12] <hlb> wow.
  541. [18:32:39] <_fil_> fuzzyBSc: I don't think so, if the navigation links are okay
  542. [18:32:52] <hlb> a question, how do i use URL to present the translation for interface & content?
  543. [18:33:13] <hlb> for example, somebody may want to see chinese content with english interface?
  544. [18:33:26] <_fil_> yeah, this is a crasy thing
  545. [18:33:30] <hlb> i'm wondering that for 3~4 days.
  546. [18:33:31] <fuzzyBSc> RobertBachmann: It might be a good idea to go back to the link code from 0.0.5 then, before my largeish patch was applied. It seems a defined subset of attributes should be passed through. My problem with the earlier version was that it defined things like title as empty when they didn't occur in the source document. Perhaps you could supply a patch that solves both problems? :)
  547. [18:33:35] <_fil_> in SPIP you have two modes
  548. [18:33:51] <_fil_> for language interface; one is user-driven, the other is content-driven
  549. [18:33:52] <hlb> i even think about http://blah.blah/en/blah/blah/zh/
  550. [18:33:52] <hlb> XD
  551. [18:34:11] <_fil_> I prefer the content-driven approach, but users were demanding the other approach too
  552. [18:34:16] <_fil_> so you can have both
  553. [18:34:26] <_fil_> usually you'll add a "language menu"
  554. [18:34:55] <_fil_> if selected, in user-driven approach, the interface will follow the user's language
  555. [18:35:05] <hlb> i know drupal is actually working on it.
  556. [18:35:17] <RobertBachmann> fuzzyBSc: I'll try to fix these bugs.
  557. [18:35:17] <RobertBachmann> BTW Luke has a Subversion repositry ... I don't have access yet nor do I know the server
  558. [18:35:30] <hlb> http://drupal.org/project/i18n
  559. [18:35:32] <_fil_> good for them :) we've had it for 2 yrs now
  560. [18:36:01] <hlb> ya, finally.
  561. [18:36:08] <_fil_> it's weird that SPIP is not well known outside France
  562. [18:36:21] <_fil_> probably because we're mostly French programmers...
  563. [18:36:25] <_fil_> anyway, this is off topic
  564. [18:37:15] <_fil_> hlb: I also did http://cn.mondediplo.com
  565. [18:37:29] <_fil_> check the simplified/traditional switch
  566. [18:37:57] <_fil_> it has a bug, now I see it ;-(
  567. [18:38:12] <hlb> by iconv?
  568. [18:38:38] <hlb> _fil_: ha, kind of 'never live demo'?
  569. [18:38:40] <fuzzyBSc> hlb: To get to your chinese-in-english web page problem... well you only have to disambiguate between multiple existing versions of the page. If the navigation is in english I would call that the english version of the page. It happens to be viewing chinese content, but is the english version of the page that views chinese content ;) Which content you are viewing depends on the URL.
  570. [18:38:49] <fuzzyBSc> Anyway, I'm really off to sleep.
  571. [18:38:51] <fuzzyBSc> Goodnight, all.
  572. [18:39:08] <_fil_> not iconv, it didn't work, i had to find a perl script
  573. [18:39:14] * fuzzyBSc (n=fuzzy@c210-49-82-193.rochd2.qld.optusnet.com.au) has left #microformats
  574. [18:39:18] <_fil_> but i'm doing it blind alas
  575. [18:39:55] <_fil_> fuzzyBSc makes an interesting point here
  576. [18:40:27] <_fil_> I would add that if you positively select the /en version of a page, you should then stay in that /en mode
  577. [18:40:38] <hlb> piconv(perl) is much useful for big5/gb translation.
  578. [18:40:43] <_fil_> so you have to add a cookie
  579. [18:41:07] <_fil_> in fact it's not big5 or gb, it's all utf-8
  580. [18:41:33] <_fil_> the button is a bit misleading, but it's the clearest way we found (with our Chinese translator) to express it
  581. [18:41:35] <hlb> hey, good.
  582. [18:42:50] <hlb> i am just wondering.. maybe somebody want to view content in one language, but use interface in another?
  583. [18:43:40] <_fil_> fwiw i use /usr/bin/b2g.pl
  584. [18:43:40] <hlb> i know it's strange... :p
  585. [18:43:52] <_fil_> yes hlb it happens
  586. [18:44:14] <_fil_> example: I want to add a keyword on an article in chinese
  587. [18:44:15] <hlb> oh, autrijus wrote it.
  588. [18:44:18] <_fil_> i don't read chinese
  589. [18:44:32] <_fil_> so i'm using the french interface, to add my keyword
  590. [18:44:41] <_fil_> to chinese content
  591. [18:44:56] <_fil_> yes, that's the autrijus script
  592. [18:46:02] * hober (n=ted@rrcs-67-52-157-90.west.biz.rr.com) has joined #microformats
  593. [18:46:24] <hlb> i think it would be a choice & maybe i should put it in.
  594. [18:50:41] * bergie (n=bergie@cs78242134.pp.htv.fi) has joined #microformats
  595. [18:51:16] * tantek sets mode +o RobertBachmann
  596. [18:52:20] * TantekC (n=Tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net) has joined #microformats
  597. [18:54:04] * tantek sets mode +o TantekC
  598. [18:54:43] <TantekC> Frederic, still there?
  599. [19:05:18] * bergie (n=bergie@cs78242134.pp.htv.fi) Quit ()
  600. [19:27:28] <_fil_> hAtom: I think it would make sense to consider that the title is content
  601. [19:27:47] <_fil_> i.e. <h1 class="content title">Title</h1>
  602. [19:29:58] <_fil_> well, never mind, I can still get to it via the css ".entry .title {}"
  603. [19:31:04] * darkavengr (n=darkaven@sboote.demon.co.uk) has joined #microformats
  604. [19:33:29] * darkavengr (n=darkaven@sboote.demon.co.uk) has left #microformats
  605. [19:33:53] * RobertBachmann (n=RobertBa@M2462P005.adsl.highway.telekom.at) Quit ("bye")
  606. [19:42:44] * bergie (n=bergie@cs78242134.pp.htv.fi) has joined #microformats
  607. [19:42:55] * dglazkov (n=dglazkov@adsl-065-081-081-030.sip.bhm.bellsouth.net) Quit ()
  608. [20:26:13] * dustym (n=dustym@office.cavoksolutions.com) Quit ()
  609. [20:33:40] * edsu (n=esummers@66.187.134.52) has joined #microformats
  610. [21:24:34] <amanuel_> ?herald
  611. [21:25:51] <amanuel_> ?def amanuel_
  612. [21:37:06] * bergie (n=bergie@cs78242134.pp.htv.fi) Quit ()
  613. [22:27:33] * factoryjoe (n=cmessina@c-69-181-81-22.hsd1.ca.comcast.net) Quit ()
  614. [22:36:59] * bergie (n=bergie@cs78242134.pp.htv.fi) has joined #microformats
  615. [22:39:26] * bergie (n=bergie@cs78242134.pp.htv.fi) Quit (Client Quit)
  616. [22:54:53] * factoryjoe (n=cmessina@c-69-181-81-22.hsd1.ca.comcast.net) has joined #microformats
  617. [23:33:42] * hober (n=ted@rrcs-67-52-157-90.west.biz.rr.com) Quit ("nil")
  618. [23:47:57] * hober (n=ted@rrcs-67-52-157-90.west.biz.rr.com) has joined #microformats
  619. [23:55:00] * TantekC (n=Tantek@adsl-63-195-114-133.dsl.snfc21.pacbell.net) Quit ("via the nearest exit")

These logs were automatically created by mflogbot on chat.freenode.net using a modified version of the Java IRC LogBot.

See http://microformats.org/wiki/mflogbot for more information.