-
Notifications
You must be signed in to change notification settings - Fork 20
[WIP] Pull streams #64
Conversation
@@ -2,7 +2,7 @@ sudo: false | |||
language: node_js | |||
node_js: | |||
- 4 | |||
- 5 | |||
- stable |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Add Node.js 5 tests too
c422ea8
to
3229064
Compare
Got an error, basically due to assumed ordering
Strange that it is not ordered for this example. I would say that ordering is not important, but this might actually be a bug. //cc @dignifiedquire |
@diasdavid what do we want to do about the ordering, just make the tests not care about it? |
Also phantom fails on the last test of Importer with
|
Why would this be a bug? it's just traversing/emitting in a different order |
In PhantomJS
|
Also first time seeing this:
|
I have no idea what phantomjs has for a problem, except that it might be too slow to do anything right :( I can't debug Phantom on my machine currently as every time I start it, it will die with
|
@diasdavid fixed the first issue |
found the reason of the second issue: https://www.npmjs.com/package/buffer#one-minor-difference and pull-block relies on this :( not sure how this worked before as the chunker in |
PhantomJS exists in the pipeline to verify that the ES5 transpiled version works as expected, in this specific case it is not a transpilation issue, but an 'old browser' problem. We haven't declared what are the official minimum browser version to run js-ipfs, but it will be certainly a browser that at least requires WebRTC (which brings it to the class of modern browsers, where we won't hit this issue). I've been wanting to be able to disable PhantomJS for a while (ipfs/aegir#48), seems that now it is the time :) |
Migrate
fixed-size-chunker
exporter
importer