These are chat archives for opal/opal

29th
May 2015
ylluminate
@ylluminate
May 29 2015 07:13
@elia @catmando @fkchang i was thinking of something along these lines: http://f.cl.ly/items/0n1p3c1k0a1S2G322s3i/Opal%20Logo%20-%20R3.1.png
obviously not the 3d rendered version i'd mentioned, but this gives an idea of what i was talking about.
Elia Schito
@elia
May 29 2015 07:28
I like the fill color, yet I prefer a more unique shape
Mitch VanDuyn
@catmando
May 29 2015 10:14
I like it
Clear message
ylluminate
@ylluminate
May 29 2015 14:10
I was going to change shape too @elia, but then I kept getting stuck on the importance about how we want folks to recognize that this IS "ruby" (I think a la @catmando's "clear message" conclusion). So that's why I thought I'd send this for some chatter before going any further. I do perhaps think, at the very least, I need to somehow raise the contrast of this or perhaps "opalize" the facets separately. Hmmm.
Elia Schito
@elia
May 29 2015 14:18
gradients and textures are not very scalable (svg-wise)
Vais Salikhov
@vais
May 29 2015 14:38
@ylluminate in my past life I did a short stint as a web dev at a jewelry business. Opal's never cut to have facets, it's normally more of a smooth, oblong affair. So, yes, making it look like ruby is problematic :smile:
ylluminate
@ylluminate
May 29 2015 14:42
LOL, that crossed my mind @vais, but that was thrown into the wind when the name "opal" was chosen for a ruby implementation. ;)
of course these shapes did cross my mind: http://cl.ly/image/2S1r0q1u2l44
Vais Salikhov
@vais
May 29 2015 14:45
Here's a stream of consciousness.... From wikipedia "The internal structure of precious opal makes it diffract light; depending on the conditions in which it formed, it can take on many colors. "
:wink2:
ylluminate
@ylluminate
May 29 2015 14:47
lol, cute
Vais Salikhov
@vais
May 29 2015 14:47
(with a ruby shape instead of a triangle of course)
Oops, that's actually not an example of diffraction :sweat_smile:
ylluminate
@ylluminate
May 29 2015 14:50
one problem with logos is that when we get too many mental concepts going on it wrecks havoc on the user's ability to obtain meaning. i once did a logo design for an organization about 8 years ago that caused me to not do any designs since for anyone buy my own projects because a committee had to decide on it. i went through about 200 iterations until i finally gave up, used a logo design app and used a nike-like swoosh default and they freaking loved it. -.- i was not a happy camper, but c'est la vie
Vais Salikhov
@vais
May 29 2015 14:51
Hehe...
Nike-like swoosh... that's not a bad idea!
ylluminate
@ylluminate
May 29 2015 14:53
:P
Vais Salikhov
@vais
May 29 2015 14:53
LOL
ylluminate
@ylluminate
May 29 2015 14:53
let's just use a white dot
(no outline, just a white dot... :shipit: )
Vais Salikhov
@vais
May 29 2015 14:58
That's already taken - I'm using it for Diamond, my Opal implementation on top of Windows Scripting Host.
Anyway, long day, it's time for me to shutdown :zzz: Have fun with it!
Elia Schito
@elia
May 29 2015 14:59
good night!
ylluminate
@ylluminate
May 29 2015 14:59
lol, i thought that was "Glass"
sounds like it's pretty fragile... ;) night
Forrest Chang
@fkchang
May 29 2015 15:34
@ylluminate I like logo, w/the intentional ruby shape
Christian Käser
@dfyx
May 29 2015 15:53
@elia Did you see that the specs for opal-rails fail on travis?
Elia Schito
@elia
May 29 2015 15:56
@dfyx yes, I think it's a weird issue with travis' file system
Christian Käser
@dfyx
May 29 2015 15:56
Alright
Just saw it because my boss finally lets me migrate some of our web stuff to rails and what better way is there to start a new rails app than with opal on the front end?
Elia Schito
@elia
May 29 2015 15:57
:tada:
Christian Käser
@dfyx
May 29 2015 15:58
(Would have loved to use volt but I don't think it fits that specific application)
Ryan Stout
@ryanstout
May 29 2015 16:54
@adambeynon hey, how are you? Have you had any more thoughts about making opal-rspec wait on returned promises?
is there a way to do async/run_async in opal rspec for before/after right now?
Mitch VanDuyn
@catmando
May 29 2015 17:33
@ryanstout can you explain your opal-rspec question again? I am having what I think is a similar problem, so its something I have been thinking about as well.
Ryan Stout
@ryanstout
May 29 2015 17:35
@catmando basically, I have some stuff that returns a promise, and I want the it spec to wait until it resolves before continuing. (Since I setup .then blocks that run more expect's)
Mitch VanDuyn
@catmando
May 29 2015 17:35
@ylluminate how about putting a small standard red ruby symbol inside the current logo (instead of the triangle at the back)
@ryanstout - exactly the same problem I having. For now I just have one giant spec with a bunch of expects inside that concluding in the final async { expect ...
Mitch VanDuyn
@catmando
May 29 2015 17:41
@ryanstout this might be a quick easy fix: have each async block do a busy wait until the previous async block completes. Busy-wait normally bad, but acceptable in a test environment.
Ryan Stout
@ryanstout
May 29 2015 17:41
@catmando you can't busy wait in JS
Mitch VanDuyn
@catmando
May 29 2015 17:41
I am suggesting this function be added to aync
Ryan Stout
@ryanstout
May 29 2015 17:41
well, you can, but the other thing won't get done :-)
for (;;) {}
Mitch VanDuyn
@catmando
May 29 2015 17:42
perhaps I should have said "poll"
Ryan Stout
@ryanstout
May 29 2015 17:42
you can't do that either in this situation, because a setTimeout would move on to the next spec
afaik
Mitch VanDuyn
@catmando
May 29 2015 17:43
you would add the polling to the async block so yes it would "move on" but what you would actually create is a chain of async blocks that will run in order
Mitch VanDuyn
@catmando
May 29 2015 17:50

currently

async "test returns_a_promise" do
   returns_a_promise.then do 
      async { expect ... }
   end
end

async "test returns_another_promise" do
  returns_another_promise.then do
    async { expect ... }
  end
end

Today it may call returns_another_promise before returns_a_romise completes
(just confirming the problem statement is correct)

If each async test block just waits (heck probably doesn't need to poll) for the previous block finish its async test, then I think we are all set

in fact we can add a optional parameter called :test_chain so that you can set up independent chains of tests
This message was deleted
Mitch VanDuyn
@catmando
May 29 2015 17:56
async "test 1", test_chain: :chain_1 do
# I will start right away
end
async "test 2", test_chain: :chain_2 do
# I will also start right away
end
async "test 3" do
# and I will also start right away
end
async "test 4" do
# but I wont' start until test3 finishes
end
async "test 5", test_chain: :chain_1 do
# and I won't start until test 1 finishes
end
How about that?
I think it can all be done today without updating opal-rspec just as a test:
Mitch VanDuyn
@catmando
May 29 2015 18:05
before(:all) do
  @test_chains = Hash.new(Promise.new.tap { |p| p.resolve})
end
async "test 1" do
   @test_chains[:default].then do
   @test_chains[:default] = Promise.new
    # I will start right away
    async {@test_chains[:default].resolve; expect ... }
  end
end
async "test 2", test_chain: :chain_2 do
   etc
   # I will also start right away
end
async "test 3" do
# and I will also start right away
end
async "test 4" do
# but I wont' start until test3 finishes
end
async "test 5", test_chain: :chain_1 do
# and I won't start until test 1 finishes
end
Adam Beynon
@adambeynon
May 29 2015 18:25
@ryanstout no, havne’t had a chance Im sorry. In honesty, im not sure if its something Im going to look to fix anytime soon. not really having time to work on web/opal stuff atm
Ryan Stout
@ryanstout
May 29 2015 19:31
@adambeynon no worries. I know how that goes. Would you take a PR for it?
Mitch VanDuyn
@catmando
May 29 2015 19:40
@ryanstout - fyi I also started on a change for this... let me know if you want to wait for me to finish it up, else I can use what you have in mind....I too want to get this fixed :-)
Ryan Stout
@ryanstout
May 29 2015 19:42
@catmando so the thing I'm trying to get working is making it so specs run the same in MRI and opal, so we may not have the same use cases. On the server, I can have the promises block when created.
Mitch VanDuyn
@catmando
May 29 2015 19:43
@ryanstout but you use async?
Ryan Stout
@ryanstout
May 29 2015 19:45
yea, but they are still in promises.
long story
anyway, let me know if you start on it, maybe I can help
Mitch VanDuyn
@catmando
May 29 2015 19:54
@catmando so roughly right should the outline I gave above work for you? I.e. async case is blocked by a promise until the previous case finishes?