/irc-logs / w3c / #testing / 2013-02-18 / end

Options:

  1. # Session Start: Mon Feb 18 00:00:00 2013
  2. # Session Ident: #testing
  3. # [00:55] * Quits: tobie (tobie@public.cloak) (Ping timeout: 60 seconds)
  4. # [01:54] * heycam is now known as heycam|away
  5. # [02:34] * heycam|away is now known as heycam
  6. # [05:24] * heycam is now known as heycam|away
  7. # [05:57] * heycam|away is now known as heycam
  8. # [06:08] * Joins: shepazutu (schepers@public.cloak)
  9. # [06:09] * Quits: shepazu (schepers@public.cloak) (Ping timeout: 60 seconds)
  10. # [06:09] * shepazutu is now known as shepazu
  11. # [07:59] * Quits: glenn (~gadams@public.cloak) (Client closed connection)
  12. # [08:20] * Joins: glenn (~gadams@public.cloak)
  13. # [08:52] * Quits: glenn (~gadams@public.cloak) (Client closed connection)
  14. # [09:25] * heycam is now known as heycam|away
  15. # [09:50] * Joins: darobin (rberjon@public.cloak)
  16. # [09:52] * Joins: Ms2ger (~Ms2ger@public.cloak)
  17. # [10:32] * Joins: tobie (tobie@public.cloak)
  18. # [11:42] * Quits: Ms2ger (~Ms2ger@public.cloak) ("bbl")
  19. # [12:09] * Joins: sstewart6 (~simons@public.cloak)
  20. # [12:09] <sstewart6> MikeSmith: now we're on monthly updates for the spec, is there anything we need to do to get a new WD published?
  21. # [12:09] * Joins: abarsto (~abarsto@public.cloak)
  22. # [12:09] * abarsto is now known as ArtB
  23. # [12:27] <sstewart6> :ci
  24. # [12:27] <sstewart6> Wrong channel. Sorry
  25. # [12:37] <MikeSmith> sstewart6: no, just need to pick a date
  26. # [12:38] <sstewart6> Well, my edits last time round are ok for me, and David is on paternity leave.
  27. # [12:38] <sstewart6> Any time is good :)
  28. # [12:50] <tobie> jgraham: http://www.w3.org/wiki/Testing/testharness.js_Improvements thoughts?
  29. # [12:52] * Joins: glenn (~gadams@public.cloak)
  30. # [12:52] <jgraham> tobie: The reason the arguments are in the order they are is because the title is an optional argument
  31. # [12:53] <jgraham> If you don't supply one it is taken from document.title
  32. # [12:53] <jgraham> And optional arguments are typically last
  33. # [12:53] <tobie> oh.
  34. # [12:53] <tobie> fair enough.
  35. # [12:54] <tobie> why is that offered as an option, btw?
  36. # [12:54] <tobie> Also, is it used?
  37. # [12:54] <jgraham> Why is what offered as an option?
  38. # [12:54] <tobie> using <title>.
  39. # [12:54] <jgraham> The pseudo-justification for underscores was that it makes the asserts look different to other js code and highly unlikely to conflict with anything, even though testharness stomps all over the global namespace
  40. # [12:54] <jgraham> Oh
  41. # [12:55] <jgraham> Because if you are writing a page with one one test it is more DRY
  42. # [12:55] <jgraham> I have certainly used it
  43. # [12:55] <jgraham> I think others have too
  44. # [12:55] <tobie> OK. well we can preserve this and still allow for title first for those who care.
  45. # [12:55] <jgraham> (sometimes having one test per page is necessary e.g. for testing navigation)
  46. # [12:56] <tobie> You have no idea how ridiculously annoying front-end devs can become about these things
  47. # [12:56] <tobie> So I think removing friction points is really useful to do.
  48. # [12:56] <jgraham> I certainly don't object to nicer rendering of results
  49. # [12:57] <tobie> :)
  50. # [12:57] * Quits: glenn (~gadams@public.cloak) (Ping timeout: 60 seconds)
  51. # [12:57] <jgraham> Or tests :)
  52. # [12:57] <tobie> heh
  53. # [12:57] <jgraham> I am slightly worried about adding lots of optional stuff to the API
  54. # [12:57] <jgraham> I mean saying "this argument can be here or here or missing, or..."
  55. # [12:58] <jgraham> I know DWIM is popular with people raised on jQuery
  56. # [12:58] <jgraham> But it does add to the internal complexity, and testharness.js is already too complex
  57. # [12:58] * Joins: Ms2ger (~Ms2ger@public.cloak)
  58. # [12:59] <tobie> Agreed, though we actually have three possible args to tests:
  59. # [12:59] <tobie> a function, a string and an options object.
  60. # [12:59] <tobie> implementation complexity here isn't that high
  61. # [13:00] <tobie> Anyway, this is food for thought.
  62. # [13:00] <jgraham> It also makes tests harder to read, since you have to work out which set of cenventions this author decided to use
  63. # [13:00] <tobie> sure.
  64. # [13:01] <tobie> jgraham: as most things software, it's a cost/benefit analysis.
  65. # [13:03] <tobie> Having spent hours arguing about testharness with influential and smart front-end devs, I'd rather deal with the added complexity than have these conversations yet another time.
  66. # [13:03] <tobie> :D
  67. # [13:04] <tobie> I'm tired of it to the point I'm actually considering writing a script to built ASTs of the whole test repo and rewrite the args in a less anger-inducing order.
  68. # [13:04] <jgraham> Being angry about argument order is just sad
  69. # [13:05] <jgraham> But of course it would have been nice if I had been able to talk to the front-end devs directly
  70. # [13:05] <tobie> being forced to discuss it to length is sadder.
  71. # [13:05] <jgraham> Well sure
  72. # [13:05] <jgraham> Or
  73. # [13:05] <jgraham> I think that the one that causes the problem is sadder
  74. # [13:06] <jgraham> But I agree that it's unfortunate that it has taken up time
  75. # [13:06] <tobie> where "the one" needs to be defined
  76. # [13:06] <tobie> could be the software, the person, the culture, the community, the dogmas, etc.
  77. # [13:06] <tobie> jgraham: you have no idea.
  78. # [13:06] <jgraham> People getting angry because the argument order doesn't match some different library they once used
  79. # [13:07] <jgraham> Oh right
  80. # [13:07] <tobie> unfortunate. but true.
  81. # [13:07] <tobie> I can probably put a price tag on it too.
  82. # [13:08] <tobie> Trust me, fixing the lib the test repo is already worth the money.
  83. # [13:08] <jgraham> Well, only if these people actually become contributers
  84. # [13:09] <tobie> Nope.
  85. # [13:09] <tobie> Bad press is equally an issue.
  86. # [13:10] <jgraham> OK, imagine I said "only if this stops people becoming contributers"
  87. # [13:10] <jgraham> even if it isn't the same people
  88. # [13:10] <tobie> Then yes, absolutely. And that is a concern.
  89. # [13:11] <jgraham> Well, like I say, it would be nice to be involved in these conversations if possible
  90. # [13:13] <jgraham> But I am convincable that this could change, but it isn't just a case of a random bad decision. It was one of the things that was actually decided to fit with convention.
  91. # [13:17] * Ms2ger wonders what this is about
  92. # [13:20] <jgraham> http://www.w3.org/wiki/Testing/testharness.js_Improvements
  93. # [13:20] <tobie> jgraham: going back to read those tweets is slightly sickening.
  94. # [13:21] <tobie> most of my 2012 april folder is full of those.
  95. # [13:22] <jgraham> Pointer?
  96. # [13:23] <tobie> I'd have put my tweet archives online
  97. # [13:23] <jgraham> Oh
  98. # [13:23] <tobie> which I haven't done yet.
  99. # [13:23] <tobie> eg: https://twitter.com/wycats/status/189493328872607744
  100. # [13:24] <jgraham> I thought maybe twitter provieded that kind of service
  101. # [13:24] <jgraham> But what do I know
  102. # [13:25] <tobie> https://twitter.com/tobie/statuses/187685921926295552
  103. # [13:25] <tobie> https://twitter.com/tobie/statuses/187689638306131969
  104. # [13:26] <tobie> https://twitter.com/tobie/statuses/187660065438838785
  105. # [13:26] <tobie> etc.
  106. # [13:26] <tobie> ad nauseam.
  107. # [13:27] <jgraham> I see. Thanks
  108. # [13:28] <jgraham> Seems like there is some serious NIH there. Or at least "Not invented by one of the cool kids"
  109. # [13:28] <tobie> I don't want to debate this more.
  110. # [13:28] <jgraham> Sorry
  111. # [13:28] <tobie> :)
  112. # [13:28] <tobie> No seriously.
  113. # [13:28] <tobie> This is solvable through a bunch of small changes.
  114. # [13:29] <tobie> Where everyone let's go of a couple of things he cares about
  115. # [13:29] <tobie> and we end with a solution that works and doesn't create shit-storms.
  116. # [13:30] <tobie> So we can focus on what matters.
  117. # [13:30] <jgraham> I am totally open to changes that actually help people
  118. # [13:30] * Joins: darobin_ (rberjon@public.cloak)
  119. # [13:30] * Quits: darobin (rberjon@public.cloak) (Client closed connection)
  120. # [13:31] <jgraham> But I would much prefer that we do that with the people who are complaining actually engaged
  121. # [13:31] <tobie> jgraham: oh, sure.
  122. # [13:33] <tobie> This is by no means urgent either.
  123. # [13:41] * darobin_ is now known as darobin
  124. # [13:42] * Ms2ger is perfectly happy with what we have now, fwiw
  125. # [13:47] <darobin> well, one thing that's clear to me is that we'd get the exact same complaints if we'd picked something out there
  126. # [13:47] <darobin> people would go "Oh but you used mocha+expect.js, those suck, you should really use QUnit+should!"
  127. # [13:48] <darobin> I know that for a fact because it's actual feedback I've got on some of my projects
  128. # [13:49] <darobin> the way people bitch about test frameworks you'd think you're forcing them to use emacs to edit Java on Win 3.1
  129. # [13:49] <darobin> and commit the whole thing using hg
  130. # [13:50] <jgraham> s/emacs/ed/ and s/hg/VSS/ I think :)
  131. # [13:50] <darobin> no no, I picked those very deliberately :)
  132. # [13:51] * jgraham tries M-x cuss-darobin
  133. # [13:51] <darobin> tee hee
  134. # [13:59] * Quits: Lachy (~Lachy@public.cloak) ("Bye")
  135. # [14:03] <darobin> so when is it that IE switches to WebKit?
  136. # [14:03] * darobin tired of watching VMs load...
  137. # [14:06] * Quits: sstewart6 (~simons@public.cloak) (sstewart6)
  138. # [14:07] * Joins: sstewart6 (~simons@public.cloak)
  139. # [14:10] <tobie> http://www.paulgraham.com/identity.html
  140. # [14:22] <jgraham> darobin: As a counter example, the Apache Cordova (sp?) people who were at TTWF were very open to trying new things; they had been using Jasmine, they tried testharness, found that some things that they expected to work didn't work, asked a few questions, discovered that there was a different way of doing it, and proceeded to use that
  141. # [14:26] <jgraham> Of course it is quite possible that they went away from there and never wrote another testharness test again
  142. # [14:27] <tobie> jgraham: so these are actually smart folks working for Adobe who has a vested interest in seeing this happem
  143. # [14:27] <tobie> not your traditional dev.
  144. # [14:28] <tobie> the reason they stopped contribution thereafter was due to the awful reviewing process...
  145. # [14:28] <tobie> which means their tests haven't been accepted anywhere yet.
  146. # [14:30] <tobie> :( or they haven't figured out where to get them reviewed. I'm unsure.
  147. # [14:30] <darobin> jgraham: yeah those people were paid to make that experiment happen
  148. # [14:31] <tobie> Anyway, I think we should close this subject for now. jgraham that was a heads up. I'll have a more concrete proposal, with code in a while.
  149. # [14:31] <jgraham> Right, but it turns out that people who are apid can still be difficult to work with :)
  150. # [14:32] <jgraham> tobie: Sure, that sounds great
  151. # [14:33] <jgraham> I think we all have the same goal at least; getting interoperability via tests
  152. # [14:34] <jgraham> and if we can make some changes that help that I'm totally in favour. I just want to make changes that actually help :)
  153. # [14:36] <tobie> ++ on the end goals.
  154. # [14:37] <tobie> I also want to pick my fights carefully. And defending testharness isn't a fight worth fighting imho, when a bunch of changes to the API can avoid the fight altogether.
  155. # [14:38] <jgraham> Sure, if your premise (that trivial API differences are the underlying problem) are the real issue.
  156. # [14:39] <jgraham> s/are the real issue/is correct//
  157. # [14:40] <tobie> Right. Total loss of time otherwise.
  158. # [14:41] <tobie> I'm thinking of selling testharness as a simplified version of Qunit.
  159. # [14:41] <tobie> with all of the crazy scoped stuff ripped right out.
  160. # [14:42] <jgraham> Yes. I think it's important to understnad that if you're testing the browser itself "simplified" is an important goal
  161. # [14:42] <tobie> ++
  162. # [14:42] <jgraham> You don't want to be running into interop issues in the harness itself
  163. # [14:42] <jgraham> (which is also why #whatwg uses the live dom viewer rather than jsfiddle)
  164. # [14:43] <tobie> not sure what the live dom viewer
  165. # [14:43] <tobie> … is.
  166. # [14:43] <Ms2ger> bit.ly/livedom
  167. # [14:43] <jgraham> http://software.hixie.ch/utilities/js/live-dom-viewer/
  168. # [15:01] <tobie> ty
  169. # [16:01] * Joins: darobin_ (rberjon@public.cloak)
  170. # [16:01] * Quits: darobin (rberjon@public.cloak) (Client closed connection)
  171. # [16:11] * darobin_ is now known as darobin
  172. # [17:02] * Joins: shepazutu (schepers@public.cloak)
  173. # [17:03] * Quits: shepazu (schepers@public.cloak) (Ping timeout: 60 seconds)
  174. # [17:04] * shepazutu is now known as shepazu
  175. # [17:05] * Joins: plh (plehegar@public.cloak)
  176. # [17:10] * Joins: glenn (~gadams@public.cloak)
  177. # [17:19] * Quits: Ms2ger (~Ms2ger@public.cloak) ("bbl")
  178. # [18:08] * Parts: sstewart6 (~simons@public.cloak) (sstewart6)
  179. # [18:08] * Joins: darobin_ (rberjon@public.cloak)
  180. # [18:08] * Quits: darobin (rberjon@public.cloak) (Client closed connection)
  181. # [18:39] * Joins: jhammel (~jhammel@public.cloak)
  182. # [18:40] * Quits: jhammel (~jhammel@public.cloak) ("leaving")
  183. # [18:49] * Quits: darobin_ (rberjon@public.cloak) (Client closed connection)
  184. # [20:16] * Joins: Ms2ger (~Ms2ger@public.cloak)
  185. # [20:50] * Joins: darobin (rberjon@public.cloak)
  186. # [21:18] * Quits: Ms2ger (~Ms2ger@public.cloak) ("nn")
  187. # [21:30] * Quits: darobin (rberjon@public.cloak) (Client closed connection)
  188. # [21:32] * Joins: Lachy (~Lachy@public.cloak)
  189. # [21:51] * Quits: Lachy (~Lachy@public.cloak) ("Bye")
  190. # [21:51] * Joins: Lachy (~Lachy@public.cloak)
  191. # [22:57] * heycam|away is now known as heycam
  192. # [23:09] * Quits: plh (plehegar@public.cloak) ("Leaving")
  193. # Session Close: Tue Feb 19 00:00:00 2013

The end :)