/irc-logs / w3c / #testing / 2015-01-19 / end

Options:

Previous day, Next day

  1. # Session Start: Mon Jan 19 00:00:00 2015
  2. # Session Ident: #testing
  3. # [01:49] * Quits: terri (~terri@public.cloak) (Ping timeout: 180 seconds)
  4. # [01:52] * Joins: terri_offline (~terri@public.cloak)
  5. # [01:52] * terri_offline is now known as terri
  6. # [01:58] * Quits: artb (~ArtB@public.cloak) ("Leaving.")
  7. # [02:19] * Joins: artb (~ArtB@public.cloak)
  8. # [02:23] * Joins: tripu (antonio@public.cloak)
  9. # [02:52] * heycam is now known as heycam|away
  10. # [03:03] * Quits: Lachy (~Lachy@public.cloak) ("My MacBook Pro has gone to sleep. ZZZzzz…")
  11. # [03:29] * Quits: ptressel (~chatzilla@public.cloak) ("bbl")
  12. # [03:40] * heycam|away is now known as heycam
  13. # [04:16] * Joins: ptressel (~chatzilla@public.cloak)
  14. # [05:20] * Quits: artb (~ArtB@public.cloak) ("Leaving.")
  15. # [05:42] * Joins: gitbot (~gitbot@public.cloak)
  16. # [05:42] -gitbot:#testing- [web-platform-tests] sideshowbarker deleted ElegantPig/dataset at 241b746: https://github.com/w3c/web-platform-tests/commit/241b746
  17. # [05:42] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  18. # [05:51] * Quits: terri (~terri@public.cloak) ("bye!")
  19. # [05:52] * Joins: terri_offline (~terri@public.cloak)
  20. # [05:52] * terri_offline is now known as terri
  21. # [06:23] * heycam is now known as heycam|away
  22. # [06:29] * Joins: artb (~ArtB@public.cloak)
  23. # [06:43] * heycam|away is now known as heycam
  24. # [06:47] * Quits: terri (~terri@public.cloak) (Ping timeout: 180 seconds)
  25. # [06:50] * Joins: terri_offline (~terri@public.cloak)
  26. # [06:50] * terri_offline is now known as terri
  27. # [06:53] * Quits: artb (~ArtB@public.cloak) ("Leaving.")
  28. # [07:34] * heycam is now known as heycam|away
  29. # [09:06] * Joins: dom (dom@public.cloak)
  30. # [09:10] * Quits: terri (~terri@public.cloak) (Ping timeout: 180 seconds)
  31. # [09:12] * Joins: terri_offline (~terri@public.cloak)
  32. # [09:12] * terri_offline is now known as terri
  33. # [09:20] * Joins: Lachy (~Lachy@public.cloak)
  34. # [09:27] * Joins: Ms2ger (~Ms2ger@public.cloak)
  35. # [09:28] * Quits: Lachy (~Lachy@public.cloak) ("My MacBook Pro has gone to sleep. ZZZzzz…")
  36. # [09:47] * Joins: gitbot (~gitbot@public.cloak)
  37. # [09:47] -gitbot:#testing- [web-platform-tests] zqzhang closed pull request #293: [ttwf shanghai 2013]FileAPI URL tests added. (master...submission/jcxia-fileapi-url) https://github.com/w3c/web-platform-tests/pull/293
  38. # [09:47] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  39. # [09:47] * Joins: gitbot (~gitbot@public.cloak)
  40. # [09:47] -gitbot:#testing- [web-platform-tests] zqzhang pushed 29 new commits to master: https://github.com/w3c/web-platform-tests/compare/630a3d3506e2...139aa09946cb
  41. # [09:47] -gitbot:#testing- web-platform-tests/master 46a472c Junchen Xia: FileAPI URL tests added.
  42. # [09:47] -gitbot:#testing- web-platform-tests/master a8301d8 Junchen Xia: title changed
  43. # [09:47] -gitbot:#testing- web-platform-tests/master 8e0e41e Junchen Xia: remove url.html
  44. # [09:47] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  45. # [09:50] <Ms2ger> \o/
  46. # [09:58] * Joins: Lachy (~Lachy@public.cloak)
  47. # [10:01] * Quits: Lachy (~Lachy@public.cloak) ("My MacBook Pro has gone to sleep. ZZZzzz…")
  48. # [10:03] * Joins: Lachy (~Lachy@public.cloak)
  49. # [10:03] * Quits: Lachy (~Lachy@public.cloak) (Client closed connection)
  50. # [10:07] * Joins: darobin (rberjon@public.cloak)
  51. # [10:12] * Joins: anssik (~uid10742@public.cloak)
  52. # [10:16] * Quits: tripu (antonio@public.cloak) (Ping timeout: 180 seconds)
  53. # [10:36] * Joins: gitbot (~gitbot@public.cloak)
  54. # [10:36] -gitbot:#testing- [web-platform-tests] Ms2ger created pr-294 (+1 new commit): https://github.com/w3c/web-platform-tests/commit/008092c4a1f5
  55. # [10:36] -gitbot:#testing- web-platform-tests/pr-294 008092c march1993: Progress events test for File API
  56. # [10:36] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  57. # [10:42] * Joins: zcorpan (~zcorpan@public.cloak)
  58. # [10:45] * Joins: gitbot (~gitbot@public.cloak)
  59. # [10:45] -gitbot:#testing- [web-platform-tests] Ms2ger pushed 2 new commits to pr-294: https://github.com/w3c/web-platform-tests/compare/008092c4a1f5...be0671c8658f
  60. # [10:45] -gitbot:#testing- web-platform-tests/pr-294 0a1f0c2 Ms2ger: Move test into place.
  61. # [10:45] -gitbot:#testing- web-platform-tests/pr-294 be0671c Ms2ger: Address comments.
  62. # [10:45] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  63. # [10:45] * Joins: gitbot (~gitbot@public.cloak)
  64. # [10:45] -gitbot:#testing- [web-platform-tests] Ms2ger opened pull request #1565: Progress events test for File API (master...pr-294) https://github.com/w3c/web-platform-tests/pull/1565
  65. # [10:45] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  66. # [10:46] * Joins: gitbot (~gitbot@public.cloak)
  67. # [10:46] -gitbot:#testing- [web-platform-tests] Ms2ger closed pull request #294: [ttfw shanghai 2013]Progress events test for File API (master...submissions/jinjiang) https://github.com/w3c/web-platform-tests/pull/294
  68. # [10:46] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  69. # [11:27] * Quits: zcorpan (~zcorpan@public.cloak) (Ping timeout: 180 seconds)
  70. # [11:33] * Joins: artb (~ArtB@public.cloak)
  71. # [11:33] * Joins: zcorpan (~zcorpan@public.cloak)
  72. # [11:39] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
  73. # [11:46] * Joins: tripu (antonio@public.cloak)
  74. # [11:49] * Joins: zcorpan (~zcorpan@public.cloak)
  75. # [12:12] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
  76. # [13:32] * Quits: terri (~terri@public.cloak) (Ping timeout: 180 seconds)
  77. # [14:07] * Joins: AutomatedTester (~AutomatedTester@public.cloak)
  78. # [14:16] * Joins: zcorpan (~zcorpan@public.cloak)
  79. # [14:18] * Joins: zcorpan_ (~zcorpan@public.cloak)
  80. # [14:23] * Quits: zcorpan (~zcorpan@public.cloak) (Ping timeout: 180 seconds)
  81. # [14:32] * Joins: terri_offline (~terri@public.cloak)
  82. # [14:32] * terri_offline is now known as terri
  83. # [14:39] * Quits: terri (~terri@public.cloak) (Ping timeout: 180 seconds)
  84. # [15:08] * Joins: Lachy (~Lachy@public.cloak)
  85. # [15:18] * Joins: terri_offline (~terri@public.cloak)
  86. # [15:18] * terri_offline is now known as terri
  87. # [15:19] <Ms2ger> jgraham, having critic guess the right new upstream commit is really nice
  88. # [15:20] <jgraham> Good :)
  89. # [15:22] * Joins: gitbot (~gitbot@public.cloak)
  90. # [15:22] -gitbot:#testing- [web-platform-tests] Ms2ger pushed 2 new commits to pr-288: https://github.com/w3c/web-platform-tests/compare/de2c68b1fb8c...50aeae071d99
  91. # [15:22] -gitbot:#testing- web-platform-tests/pr-288 82e8555 Ms2ger: Protect against stack overflow in Gecko.
  92. # [15:22] -gitbot:#testing- web-platform-tests/pr-288 50aeae0 Ms2ger: Address review comments.
  93. # [15:22] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  94. # [15:24] <jgraham> So I wonder how people feel about requiring that reftests have a rel="[mis]match"
  95. # [15:24] <jgraham> rather than also allowing files with a corresponding \1[-_]ref.\2 to be reftests
  96. # [15:25] <gsnedder1> minimal boilerplate, plz. esp. within the test file.
  97. # [15:26] * jgraham weighs gsnedder1's opinion by the number of tests he has contributed in the last 12 months :p
  98. # [15:26] * gsnedder1 is now known as gsnedders
  99. # [15:26] <zcorpan_> jgraham: why?
  100. # [15:26] <gsnedders> jgraham: sssh!
  101. # [15:26] <jgraham> But seriously, the advantage is (I think) that CSS doesn't support that convention
  102. # [15:27] <jgraham> And I am trying to unify the way that reftests work so that CSS and wpt tests can both be run interchangably using the wpt infrastructure
  103. # [15:28] <Ms2ger> On the one hand, boilerplate
  104. # [15:28] <zcorpan_> jgraham: do you want to also adopt css's way of doing reftests where refs have refs?
  105. # [15:28] <gsnedders> the CSS tests don't follow that convention, or what?
  106. # [15:28] <Ms2ger> On the other hand, maybe more ref sharing?
  107. # [15:28] <gsnedders> do they only violate it when it comes to shared refs?
  108. # [15:29] <jgraham> zcorpan_: Yeah, so I think we have to
  109. # [15:29] <zcorpan_> encouraging shared refs seems like a good thing, and filename convension does the opposite
  110. # [15:30] <zcorpan_> jgraham: ok, seems fine to me
  111. # [15:30] <gsnedders> I think we want to encourage shared refs, but also want to allow boilerplateless tests using the naming convention?
  112. # [15:31] <gsnedders> (though as jgraham has correctly concluded, weighted on tests I've contributed in the past year, that's meaningless. :))
  113. # [15:32] <jgraham> I am hoping that the algorithm we can use to detect reftests is 1) for each file in the repo extract those with rel=[mis]match links 2) Construct a graph of these files 3) Identify the leaf nodes i.e. files that have rel="[mis]match" links, but are not pointed to by any such links 4) these are the tests
  114. # [15:32] <Ms2ger> gsnedders, you could finish some of the old Opera PRs and suddenly get veto power over everything :)
  115. # [15:32] <gsnedders> Ms2ger: soon! once html5lib 0.9999 or whatever we're at now has shipped!
  116. # [15:33] * zcorpan_ avoids arguing with gsnedders
  117. # [15:33] <Ms2ger> Sure
  118. # [15:34] <gsnedders> Ms2ger: like seriously just make sure I get it done then?
  119. # [15:34] <zcorpan_> jgraham: i thought css supported full circles
  120. # [15:34] <Ms2ger> gsnedders, lmk when you release? :)
  121. # [15:34] <gsnedders> Ms2ger: you'll probably know from my bitching at jgraham to review my code :)
  122. # [15:35] <jgraham> zcorpan_: Yeah, so
  123. # [15:35] <gsnedders> (can we not avoid testing some of the circle, because is some of it not implied?)
  124. # [15:36] <jgraham> I don't really know what to do there
  125. # [15:36] <gsnedders> run ALL THE TESTS?
  126. # [15:36] <jgraham> ISTM that making a *full* circle doesn't add anything useful over having A -> B, B-> C, C-> B
  127. # [15:36] <Ms2ger> zcorpan_, while you're not arguing with gsnedders: https://critic.hoppipolla.co.uk/r/3754
  128. # [15:37] <gsnedders> jgraham: agreed
  129. # [15:37] <zcorpan_> jgraham: maybe we can require one of them to have a <meta> or so saying that "i'm the test"
  130. # [15:37] <gsnedders> jgraham: though I'm slightly wondering if this is true with a mixture of == and != tests?
  131. # [15:37] <gsnedders> jgraham: certainly it doesn't hold with != chains
  132. # [15:38] * Joins: gitbot (~gitbot@public.cloak)
  133. # [15:38] -gitbot:#testing- [web-platform-tests] Ms2ger closed pull request #1549: Rewrite WorkerGlobalScope_nested_Worker.htm. (master...nested-worker) https://github.com/w3c/web-platform-tests/pull/1549
  134. # [15:38] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  135. # [15:38] * Joins: gitbot (~gitbot@public.cloak)
  136. # [15:38] -gitbot:#testing- [web-platform-tests] Ms2ger pushed 2 new commits to master: https://github.com/w3c/web-platform-tests/compare/139aa09946cb...f615e9d0ffbe
  137. # [15:38] -gitbot:#testing- web-platform-tests/master baf3865 Ms2ger: Rewrite WorkerGlobalScope_nested_Worker.htm.
  138. # [15:38] -gitbot:#testing- web-platform-tests/master f615e9d Ms2ger: Merge pull request #1549 from Ms2ger/nested-worker...
  139. # [15:38] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  140. # [15:38] <jgraham> zcorpan_: Well that works, obviously, but it is a special case that you have to add in the metadata
  141. # [15:38] * Quits: AutomatedTester (~AutomatedTester@public.cloak) (Client closed connection)
  142. # [15:38] <jgraham> I mean it's a special case that authors have to learn
  143. # [15:39] <gsnedders> jgraham: like A != B and B != C and A != C you do have to test all three, AFAICT
  144. # [15:40] <jgraham> gsnedders: True
  145. # [15:40] <jgraham> gsnedders: Given how rare that case is I wonder if it would be acceptable to say that you have to introduce D that's identical to A
  146. # [15:41] <zcorpan_> Ms2ger: r+
  147. # [15:41] <gsnedders> jgraham: I think we should just accept that sometimes we'll run redundant tests that are already implied by earlier results
  148. # [15:42] <jgraham> gsnedders: It's not "will we run redundant tests" that's worrying me
  149. # [15:42] <gsnedders> jgraham: it's getting into infinite loops?
  150. # [15:42] <jgraham> gsnedders: It's that I want to have a list of URLs that are "the tests"
  151. # [15:42] <Ms2ger> zcorpan_, takk
  152. # [15:42] <gsnedders> jgraham: how do you have a single URL for a reftest?
  153. # [15:42] * Joins: gitbot (~gitbot@public.cloak)
  154. # [15:42] -gitbot:#testing- [web-platform-tests] Ms2ger closed pull request #1551: Add test for FileReader multiple reads. (master...pr-288) https://github.com/w3c/web-platform-tests/pull/1551
  155. # [15:42] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  156. # [15:42] * Joins: gitbot (~gitbot@public.cloak)
  157. # [15:42] -gitbot:#testing- [web-platform-tests] Ms2ger pushed 1 new commit to master: https://github.com/w3c/web-platform-tests/commit/b10f9c3100a2e1178bd4a8737b7beeb03abe6451
  158. # [15:42] -gitbot:#testing- web-platform-tests/master b10f9c3 Ms2ger: Merge pull request #1551 from w3c/pr-288...
  159. # [15:42] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  160. # [15:43] <gsnedders> (am I just too out of the loop?)
  161. # [15:43] * Joins: gitbot (~gitbot@public.cloak)
  162. # [15:43] -gitbot:#testing- [web-platform-tests] Ms2ger deleted pr-288 at 50aeae0: https://github.com/w3c/web-platform-tests/commit/50aeae0
  163. # [15:43] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  164. # [15:43] <jgraham> gsnedders: If I have A == B, for example, then loading A tells me everything I subsequently need to load in order to run the test
  165. # [15:44] <jgraham> If I have A == B == A then loading A or B is equivalent so I don't have a *single* url I can point at and say "this is the test"
  166. # [15:45] <jgraham> Since being able to identify tests is important I want that to be possible, even if it's not strictly needed in order to *run* the tests
  167. # [15:45] <gsnedders> jgraham: we can cosntruct a graph of reftests, with URLs as vertices and the tests as edges (linking pairs of vertices), right? then we can pick one of the vertices for each edge, no? the question is how to pick that one vertex?
  168. # [15:45] <jgraham> gsnedders: Exactly
  169. # [15:46] <jgraham> gsnedders: My algorithm so far is "pick the vertex with no inbound edges"
  170. # [15:46] <gsnedders> which way are you directing them, and how are you directing them?
  171. # [15:46] <gsnedders> (I was describing an undirected graph, FWIW, as my initial model.)
  172. # [15:48] <jgraham> If A has rel="match" href="B" my graph is A -- == --> B. So I guess my imaginary graph has resources as vertexs and relations as edges
  173. # [15:48] <Ms2ger> I'd really appreciate it if someone took care of the remainder of https://critic.hoppipolla.co.uk/r/3777 btw
  174. # [15:48] <gsnedders> then for each edge you pick the vertex it's going from, no?
  175. # [15:49] <gsnedders> which deals with cases like {A -- == --> B, B -- == --> C, C -- == --> A}, does it not? you end up with {A, B, C} as tests for the cycle?
  176. # [15:49] <jgraham> gsnedders: By I don't want *each* pair to count as a test
  177. # [15:50] <zcorpan_> so the proposals so far are "<meta>" and ...? sort the URLs lexically and pick the first one?
  178. # [15:50] <gsnedders> jgraham: why not?
  179. # [15:50] <jgraham> Because A -> B -> C should be interpreted as "A must match B and C"
  180. # [15:50] <gsnedders> jgraham: isn't that A == B, A == C?
  181. # [15:50] <jgraham> (assuming those are all == edges)
  182. # [15:51] <gsnedders> rather than A == B, B == C?
  183. # [15:51] <jgraham> gsnedders: Those seem equivalent
  184. # [15:51] <gsnedders> I'd say {A, B} are tests in the latter case, and {A} in the former?
  185. # [15:52] <jgraham> Now I don't know what you're talking about
  186. # [15:52] <gsnedders> They're equivalent in what they test, yes.
  187. # [15:53] <gsnedders> I'm not sure I agree that they're equivalent in what the "test URLs" are, though.
  188. # [15:56] <jgraham> gsnedders: I am assuming that we want A == B == C to have a test url of A (one other complication is that you can have A == B == C OR A == D == E as two possible conditions. I quite strongly think that we don't want to run D == E if the first part of the OR holds because that's making a statement about which implementation choice has been made in a case where a CSS handwove various options)
  189. # [15:57] <Ms2ger> handwove?
  190. # [15:57] <jgraham> zcorpan_: Yeah, but the "sort and pick one" proposal seems unsatisfying because it creates even weirder action at a distance than this whole system already has. If I have B == C == D == B then B is my test. Renaming a file so that D becomes A makes A (nee D) my test.
  191. # [15:57] <gsnedders> past tense of hand-wave, I presume
  192. # [15:58] <gsnedders> jgraham: then we need optional boilerplate to help break ties, otherwise do so arbitrarily, IMO
  193. # [15:58] <zcorpan_> jgraham: i agree. that leaves <meta> :-)
  194. # [15:58] <gsnedders> jgraham: or dont' do it arbitrarily and have lint.py bitch at the submitter
  195. # [15:58] <gsnedders> jgraham: but only need to require it when we have chains?
  196. # [15:58] <zcorpan_> (<meta> + lint check for circles without it)
  197. # [15:58] <jgraham> Well yeah, I was going to say, option B is disallow complete cycles
  198. # [15:59] <jgraham> (there is presumably a technical term for what I mean by "complete" here)
  199. # [15:59] <Ms2ger> That seems better
  200. # [15:59] <zcorpan_> is that acceptable for css people?
  201. # [15:59] <Ms2ger> Hamiltonian?
  202. # [15:59] <gsnedders> what's an incomplete cycle?
  203. # [16:00] <zcorpan_> your mom
  204. # [16:00] <jgraham> gsnedders: A -> B -> C -> B
  205. # [16:01] <gsnedders> jgraham: a closed walk I think is what we're actually meaning?
  206. # [16:01] <jgraham> By "complete" I mean with every edge having an inbound and an outbound vertex
  207. # [16:01] <jgraham> No replace edge and vertex by each other in that sentence
  208. # [16:01] * Ms2ger always messes those up
  209. # [16:03] * gsnedders thinks we're meaning a closed walk, given we need to include cases where you have to visit vertices/edges multiple times to fully explore the cycle
  210. # [16:08] <jgraham> I guess I can try and figure out what the tests in CSS currently do
  211. # [16:27] * Quits: terri (~terri@public.cloak) ("bye!")
  212. # [16:28] * Joins: terri_offline (~terri@public.cloak)
  213. # [16:28] * terri_offline is now known as terri
  214. # [16:48] * Quits: tripu (antonio@public.cloak) ("Leaving")
  215. # [16:56] * Joins: gitbot (~gitbot@public.cloak)
  216. # [16:56] -gitbot:#testing- [web-platform-tests] sideshowbarker pushed 1 new commit to sideshowbarker/cc-tools-dir: https://github.com/w3c/web-platform-tests/commit/258eadbc52601f375dfad849c2488507ec029cf4
  217. # [16:56] -gitbot:#testing- web-platform-tests/sideshowbarker/cc-tools-dir 258eadb Michael[tm] Smith: Backported commit 2169425 to fix merge conflict.
  218. # [16:56] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  219. # [16:57] * Joins: gitbot (~gitbot@public.cloak)
  220. # [16:57] -gitbot:#testing- [web-platform-tests] zcorpan closed pull request #1558: Address review comments for #567. (master...disabled) https://github.com/w3c/web-platform-tests/pull/1558
  221. # [16:57] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  222. # [17:02] * Quits: zcorpan_ (~zcorpan@public.cloak) (Client closed connection)
  223. # [17:25] * Joins: gitbot (~gitbot@public.cloak)
  224. # [17:25] -gitbot:#testing- [web-platform-tests] sideshowbarker pushed 4 new commits to master: https://github.com/w3c/web-platform-tests/compare/73af92f5cd74...2c8eb227271b
  225. # [17:25] -gitbot:#testing- web-platform-tests/master d7bc574 Michael[tm] Smith: Deal with merge conflict.
  226. # [17:25] -gitbot:#testing- web-platform-tests/master 626bd1f Michael[tm] Smith: Merge branch 'master' into sideshowbarker/cc-tools-dir
  227. # [17:25] -gitbot:#testing- web-platform-tests/master 80a8ca1 Michael[tm] Smith: Fixed merge conflict.
  228. # [17:25] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  229. # [17:25] * Joins: gitbot (~gitbot@public.cloak)
  230. # [17:25] -gitbot:#testing- [web-platform-tests] sideshowbarker closed pull request #1498: cc Moved some scripts up into safe tools dir. (master...sideshowbarker/cc-tools-dir) https://github.com/w3c/web-platform-tests/pull/1498
  231. # [17:25] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  232. # [17:28] * Quits: darobin (rberjon@public.cloak) (Client closed connection)
  233. # [17:29] * Joins: gitbot (~gitbot@public.cloak)
  234. # [17:29] -gitbot:#testing- [web-platform-tests] sideshowbarker deleted sideshowbarker/cc-tools-dir at 258eadb: https://github.com/w3c/web-platform-tests/commit/258eadb
  235. # [17:29] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  236. # [17:32] * Quits: Lachy (~Lachy@public.cloak) ("My MacBook Pro has gone to sleep. ZZZzzz…")
  237. # [17:45] * Quits: terri (~terri@public.cloak) (Ping timeout: 180 seconds)
  238. # [17:48] * Joins: terri_offline (~terri@public.cloak)
  239. # [17:48] * terri_offline is now known as terri
  240. # [17:48] <Ms2ger> Down to 89!
  241. # [18:02] * Joins: zcorpan (~zcorpan@public.cloak)
  242. # [18:02] <Ms2ger> MikeSmith, r?
  243. # [18:02] * Joins: gitbot (~gitbot@public.cloak)
  244. # [18:02] -gitbot:#testing- [web-platform-tests] Ms2ger opened pull request #1566: Remove tabs from url/. (master...url-tabs) https://github.com/w3c/web-platform-tests/pull/1566
  245. # [18:02] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  246. # [18:04] <MikeSmith> Ms2ger: looking now
  247. # [18:06] <MikeSmith> hmm I guess I was the one who put those tabs in that file
  248. # [18:06] <MikeSmith> bad me
  249. # [18:07] <Ms2ger> Bad Aryeh :)
  250. # [18:07] <MikeSmith> oh it wasn't me after all
  251. # [18:07] <MikeSmith> good me
  252. # [18:08] <MikeSmith> the lint.py script takes a long time to run
  253. # [18:08] * MikeSmith tries to remember just how long
  254. # [18:08] <jgraham> Probably because it took a short time to write
  255. # [18:09] <MikeSmith> heh
  256. # [18:09] <MikeSmith> 1 min 26 sec
  257. # [18:09] * Joins: gitbot (~gitbot@public.cloak)
  258. # [18:09] -gitbot:#testing- [web-platform-tests] sideshowbarker closed pull request #1566: Remove tabs from url/. (master...url-tabs) https://github.com/w3c/web-platform-tests/pull/1566
  259. # [18:09] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  260. # [18:10] <MikeSmith> I'll fix the tabs and WS problems in the conformance-checkers dir too, so that we can un-whitelist that
  261. # [18:10] <MikeSmith> that whitelist is too long
  262. # [18:21] * hober2 is now known as hober
  263. # [18:21] <Ms2ger> jgraham, mostly because it has to run a bunch of regexps on a lot of code, I'd guess
  264. # [18:22] <Ms2ger> I guess the parsing will be slowest
  265. # [18:22] <jgraham> Yeah
  266. # [18:23] <jgraham> Would have been faster to run / slower to write in Java or Rust or some other language with a faster parser
  267. # [18:25] <Ms2ger> jgraham, and now for something completely different: any thoughts on tests that run in window/worker scopes both?
  268. # [18:34] <gsnedders> jgraham: JavaScript!
  269. # [18:34] <Ms2ger> gsnedders, on second thought, let's not go there, it's a silly place
  270. # [18:36] <gsnedders> Ms2ger: But JS VMs have really quick regexp impls!
  271. # [18:37] * Quits: dom (dom@public.cloak) ("")
  272. # [18:51] <jgraham> gsnedders: And really — HTML parsers?
  273. # [18:51] <jgraham> Ms2ger: What kind of thoughts?
  274. # [18:53] <Ms2ger> jgraham, should we make it easier to write tests that way?
  275. # [18:53] <jgraham> Do we expect there to be lots of tests that work in both places seamlessly?
  276. # [18:54] <Ms2ger> I'm thinking most of the FileAPI/XHR/websockets tests
  277. # [18:54] <jgraham> I guess my answer is "I'm not going to stop you doing that":)
  278. # [18:54] <Ms2ger> Probably the WebCrypto too
  279. # [18:54] <jgraham> Well unless it adds lots of complexity
  280. # [18:54] <Ms2ger> Right
  281. # [18:55] <Ms2ger> I guess we could do something along the lines of .worker.js
  282. # [18:55] <jgraham> I mean if you can create a system where you can write a foo.window-worker.js and it creates both, that seems fine
  283. # [18:55] <jgraham> Yeah
  284. # [18:57] <jgraham> It would be a gentle introduction to having > 1 test per file, which is supposed to be supported
  285. # [19:03] <gsnedders> jgraham: oh, yeah, right…
  286. # [19:07] * Quits: ptressel (~chatzilla@public.cloak) ("restart FF")
  287. # [19:13] * Joins: gitbot (~gitbot@public.cloak)
  288. # [19:13] -gitbot:#testing- [web-platform-tests] Ms2ger opened pull request #1567: Remove a tab from vibration/. (master...vibration-tab) https://github.com/w3c/web-platform-tests/pull/1567
  289. # [19:13] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  290. # [19:18] * Joins: gitbot (~gitbot@public.cloak)
  291. # [19:18] -gitbot:#testing- [web-platform-tests] Ms2ger opened pull request #1568: Whitelist fonts from linting. (master...whitelist-fonts) https://github.com/w3c/web-platform-tests/pull/1568
  292. # [19:18] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  293. # [19:59] <jgraham> Hmm, so unless I did this wrong — which is more than possible — CSS has 59 reftest cycles with no edge from a vertex outside the cycle of which 13 are length 1 and 46 are length 2
  294. # [19:59] <Ms2ger> Length 1? Is that a == a?
  295. # [19:59] <jgraham> Yes
  296. # [19:59] <Ms2ger> And a == a-ref; a-ref == a
  297. # [20:00] <jgraham> Yes
  298. # [20:00] <Ms2ger> So you should've checked first before the discussion earlier :)
  299. # [20:01] <jgraham> If this isn't wrong — and I stress again the implausibility of that statement — I conclude that this feature is either being used in ways that have no value or are actively buggy
  300. # [20:02] <jgraham> This is based on manifest.py-like parsing
  301. # [20:02] <Ms2ger> I'm not sure I follow
  302. # [20:02] <Ms2ger> Hm
  303. # [20:02] <jgraham> Well A == B == A isn't more useful than A == B
  304. # [20:02] <Ms2ger> No
  305. # [20:03] <jgraham> And A == A is just buggy
  306. # [20:03] <Ms2ger> But you're saying you didn't find any A == B == C == A
  307. # [20:03] <jgraham> Right
  308. # [20:03] <Ms2ger> That doesn't actually seem that weird
  309. # [20:03] <Ms2ger> Except that it doesn't sound that weird, and this is the CSSWG, which makes it weird again
  310. # [20:04] <jgraham> There might be A == B == C == D == B, I "carefully" eliminated those
  311. # [20:04] <Ms2ger> Right
  312. # [20:04] <Ms2ger> Those seem more sensible too, no?
  313. # [20:04] <jgraham> Yep, that's really the point
  314. # [20:05] <Ms2ger> So the conclusion is "the really weird stuff isn't used in practice"
  315. # [20:05] <jgraham> Yes
  316. # [20:05] <jgraham> Or that my script is wrong :)
  317. # [20:05] <Ms2ger> Works for me
  318. # [20:06] <jgraham> For my next trick I will try to figure out the CSSWG's manifest format and see if I am getting a very different graph when I parse the files compared to them.
  319. # [20:07] <Ms2ger> I mean, the script could be wrong, but the conclusion doesn't seem to be particularly unlikely
  320. # [20:08] <jgraham> Right, that's the most worrying kind of conclusion though. If it was very surprising you would check it carefully :)
  321. # [20:09] <Ms2ger> Fair :)
  322. # [20:09] <Ms2ger> I guess what I'm saying is that this wouldn't get you into a peer-reviewed journal :)
  323. # [20:18] <jgraham> Yeah, well. I have spent enough energy worrying about whether things were correct enough to go in a peer reviewed journal for one lifetime.
  324. # [20:47] * Joins: AutomatedTester (~AutomatedTester@public.cloak)
  325. # [20:47] * Quits: AutomatedTester (~AutomatedTester@public.cloak) ("Leaving...")
  326. # [20:48] * Quits: anssik (~uid10742@public.cloak) ("Connection closed for inactivity")
  327. # [20:52] * Joins: AutomatedTester (~AutomatedTester@public.cloak)
  328. # [21:02] * Joins: zcorpan_ (~zcorpan@public.cloak)
  329. # [21:02] * Quits: zcorpan (~zcorpan@public.cloak) (Client closed connection)
  330. # [21:03] * Joins: darobin (rberjon@public.cloak)
  331. # [21:03] * Joins: plh (plehegar@public.cloak)
  332. # [21:09] * Quits: zcorpan_ (~zcorpan@public.cloak) (Ping timeout: 180 seconds)
  333. # [21:19] * Quits: terri (~terri@public.cloak) (Ping timeout: 180 seconds)
  334. # [21:21] * Joins: terri (~terri@public.cloak)
  335. # [21:29] * terri is now known as terri_offline
  336. # [21:36] * Quits: terri_offline (~terri@public.cloak) (Ping timeout: 180 seconds)
  337. # [21:45] * heycam|away is now known as heycam
  338. # [21:46] * Joins: gitbot (~gitbot@public.cloak)
  339. # [21:46] -gitbot:#testing- [web-platform-tests] Ms2ger pushed 2 new commits to master: https://github.com/w3c/web-platform-tests/compare/95ea00a62390...26c05ca05f5a
  340. # [21:46] -gitbot:#testing- web-platform-tests/master 65afbd9 Ms2ger: Whitelist fonts from linting.
  341. # [21:46] -gitbot:#testing- web-platform-tests/master 26c05ca Ms2ger: Merge pull request #1568 from Ms2ger/whitelist-fonts...
  342. # [21:46] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  343. # [21:46] <Ms2ger> r? https://critic.hoppipolla.co.uk/r/3784
  344. # [21:52] <gsnedders> Ms2ger: r+. just to prove I can review stuff still. :)
  345. # [21:53] <Ms2ger> I'm very grateful :)
  346. # [21:53] * Joins: gitbot (~gitbot@public.cloak)
  347. # [21:53] -gitbot:#testing- [web-platform-tests] Ms2ger closed pull request #1567: Remove a tab from vibration/. (master...vibration-tab) https://github.com/w3c/web-platform-tests/pull/1567
  348. # [21:53] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  349. # [21:53] * Joins: gitbot (~gitbot@public.cloak)
  350. # [21:53] -gitbot:#testing- [web-platform-tests] Ms2ger pushed 2 new commits to master: https://github.com/w3c/web-platform-tests/compare/26c05ca05f5a...d7106a1dec7a
  351. # [21:53] -gitbot:#testing- web-platform-tests/master 681352a Ms2ger: Remove a tab from vibration/.
  352. # [21:53] -gitbot:#testing- web-platform-tests/master d7106a1 Ms2ger: Merge pull request #1567 from Ms2ger/vibration-tab...
  353. # [21:53] * Parts: gitbot (~gitbot@public.cloak) (gitbot)
  354. # [22:24] * Quits: darobin (rberjon@public.cloak) (Client closed connection)
  355. # [22:36] * Joins: terri_offline (~terri@public.cloak)
  356. # [22:36] * terri_offline is now known as terri
  357. # [22:45] * Quits: plh (plehegar@public.cloak) ("Leaving")
  358. # [23:14] * heycam is now known as heycam|away
  359. # [23:16] * heycam|away is now known as heycam
  360. # [23:27] * Joins: darobin (rberjon@public.cloak)
  361. # [23:31] * heycam is now known as heycam|away
  362. # [23:31] * Quits: Ms2ger (~Ms2ger@public.cloak) ("nn")
  363. # [23:43] * heycam|away is now known as heycam
  364. # [23:48] * Quits: darobin (rberjon@public.cloak) (Client closed connection)
  365. # [23:58] * Joins: darobin (rberjon@public.cloak)
  366. # Session Close: Tue Jan 20 00:00:00 2015

Previous day, Next day

Think these logs are useful? Then please donate to show your gratitude (and keep them up, of course). Thanks! — Krijn