<rdopieralski> tsufiev: you are right about the xstatic package for jasmine not containing ajax.js

<rdopieralski> tsufiev: I also agree with your conclusion that if we have to add a package to have that, we can as well choose the better one

<rdopieralski> tsufiev: but I don't think your conclusion about rewriting qunit into jasmine is properly motivated or explained ;)

<tsufiev> rdopieralski, personally, I don't like QUnit very much

<rdopieralski> it's not perfect, but it's better than the natural-language-pretending jasmine

<tsufiev> rdopieralski, well, if we care about lessening the number of supported dependencies, why not try drop something :)?

<rdopieralski> which tries to look like it's stateless, when it obviously isn't

<rdopieralski> tsufiev: it would be awesome to have one well documented way of writing js tests, with a rich set of examples for different situations -- no matter what library that uses

<rdopieralski> tsufiev: by examples I mean existing tests

<rdopieralski> alas, we only have a handful of tests for the stuff that was easiest to test

<tsufiev> rdopieralski, that means such document should stick to the syntax of some existing framework

<rdopieralski> and no real guidelines or examples

<tsufiev> rdopieralski, or do you suggest writing it in some testing pseudo-code?

<rdopieralski> no, no, what I mean is that I could live with Jasmine if it meant cleaner tests :)

<rdopieralski> I have no idea how hard it would be to rewrite the existing tests, and/or write new tests for non-angular stuff

<tsufiev> rdopieralski, okay. The possible reason I don't like QUnit tests much is the fixtures it uses - they're really ugly. Maybe there is some way to get rid of them?...

<rdopieralski> tsufiev: I'm sure there are specialized libs for fixtures similar to sinon.js

<tsufiev> rdopieralski, well, that means adding another dependency :). Are you fine with that?

<rdopieralski> I'm a great fan of small libraries that do one thing well

<tsufiev> +1

<rdopieralski> to be honest, I don't know what are the future plans for Horizon

<rdopieralski> should we ultimately rewrite evertyhing to angular?

<rdopieralski> if so, then just using jasmine would make sense

<rdopieralski> that is definitely a topic that we should discuss on the summit

<tsufiev> rdopieralski, agree

<tsufiev> rdopieralski, I have been thinking for a while about unifying/simplifying JS testing in Horizon, perhaps writing a doc/tutorial summarizing existing use cases would be a good way to deal with this issue

<rdopieralski> tsufiev: I think that we should start with cleaning up the existing tests and writing some more, for the difficult cases

<rdopieralski> tsufiev: so that we can point people to examples

<rdopieralski> and good practices

<rdopieralski> then, when we actually know what the good practices are, write the docs

<tsufiev> rdopieralski, and what if in order to write test for some difficult case additional library will be needed, just as in the case with Sinon.JS?

<rdopieralski> tsufiev: obviously, we add it

<rdopieralski> tsufiev: if it's needed, it's needed

<tsufiev> rdopieralski, okay :)

<rdopieralski> tsufiev: I'm not against adding dependencies or anything like that, I just want them to be properly reviewed and versioned

<rdopieralski> tsufiev: and to have someone maintaining them