[13:48] <elopio> la_juyis: ready?
[13:48] <elopio> I have 45 minutes before my next meeting.
[13:49] <elopio> alesage: you wanted to pair-program today too, right?
[14:10] <la_juyis> elopio, oops, sorry ,just saw this
[14:10] <la_juyis> let's do it
[14:12] <elopio> la_juyis: ok, I'll call you.
[14:42] <alesage> elopio, oops a little early for me :/
[15:08] <elopio> ping ubuntu-qa: can somebody review this one, please?
[15:08] <elopio> https://code.launchpad.net/~elopio/ubuntu-ui-toolkit/slider-autopilot_helper/+merge/222311
[15:08] <robotfuel> elopio: I will
[15:08] <elopio> thanks.
[15:27] <elopio> la_juyis: on test doubles: http://blog.8thlight.com/uncle-bob/2014/05/14/TheLittleMocker.html
[15:37] <la_juyis> elopio, thanks :)
[15:39] <robotfuel> elopio: review is finished with a small needs fixing,
[15:49] <elopio> thanks robotfuel
[15:49] <elopio> but I don't get your comment.
[15:50] <elopio> you would like a single test that selects the 4 values ?
[15:50] <robotfuel> elopio: you are already testing a negative number with -10 so the -5 is redundant
[15:52] <elopio> robotfuel: no, -10 and 10 are the extreme values.
[15:53] <elopio> -5 and 5 are intermediate values. They need to be tested too because sometimes we could be able to select the extremes and not the others. Or viceversa.
[15:53] <robotfuel> you have a value in the middle with the float. so -10 covers the negative number test
[15:54] <elopio> the float covers a different case, that's being able to select a float that falls into an integer pixel slot.
[15:55] <elopio> in order to check that my assumption that the slider is a linear equation of the values with respect to the pixels, I think that having the two extremes and two intermediate values is the safe bet.
[15:59] <robotfuel> elopio: can you make the negative number a negative float and the postitve number the float?
[16:00] <robotfuel> elopio: that makes it more interesting with 1 less test
[16:01] <elopio> robotfuel: I can, but I still feel more confident having a test for 5 and -5.
[16:01] <elopio> actually, after your last message I would like to add one extra for a negative float
[16:02] <robotfuel> if you feel like we would be able to select a float and not an integer then keep the -5 5 tests
[16:04] <elopio> robotfuel: I think that for some combinations of slider size and screen resolution, and integer value might not fall into a pixel slot. Thus, we could select some floats but not some ints.
[16:04] <robotfuel> elopio: ok add the - float and I will +1 :D
[16:05] <elopio> I haven't found an example of that, but for that and to check the formula for quadruplicate, I don't mind having to wait an extra 3 seconds.
[16:10] <robotfuel> elopio: if we test a random integers between -9 -1 and 1 9 for the tests, it could give us wider coverage.
[16:10] <robotfuel> og oy
[16:10] <robotfuel> oops
[16:12] <elopio> robotfuel: we shouldn't put random values on the tests because then they are no longer reproducible.
[16:12] <elopio> if we are not confident with the partitions that we have selected to test, we should run tests for all the possible values.
[16:13] <elopio> but in this case, I'm confident with the partition. And if an unexpected thing happens with a value that we are not testing, it's easy to add a regression test for it.
[16:26] <robotfuel> balloons: lp:~vthompson/ubuntu-filemanager-app/fix-null-error-1316677 are you the person to talk to about making sure this lands?
[16:28] <balloons> robotfuel, sure I can make sure that happens
[16:28] <balloons> it's not proposed
[16:28] <balloons> why does it need to land?
[16:29] <balloons> fixes the null error.. so the qml error.. good I guess