Conversation
- stopped in the middle - probably CPU buffer should be added to child DataBuffer rather than to Buffer, especially because DataBuffer.__init__ does not call Buffer.__init__
|
did not pass the tests again... and i have difficulty reading the test reports to determine where is the problem |
|
Hm this may be something with conda/conda-forge on the python 3.7 environments. On the python 2.7 environments it looks like you need to fix the docstring to mention the new kwargs. |
|
Blame me, will issue a hot fix instantly... |
|
Hang on until this is ready: #1608 |
|
@thomasdeneux could you try merging with master, pushing to this same branch, and seeing if the tests pass. You'll also need to update the docstring for the buffer to include your new kwarg. |
|
1/2 !!! |
|
@thomasdeneux Looking at the logs there is some flake8 errors. |
|
@thomasdeneux You can actually click on the "Details" from this pull request and click on the failing environments and see what is failing: |
|
got it, i fix it |
…ta pending_data property
|
@thomasdeneux There is a newline still missing at the eof. |
|
@djhoese Could we have a test run just with flake8 which breaks instantly the whole suite? |
|
We could run it earlier in the relevant environments, but I'm not sure we can wait for one flake8 environment before running all other environments. Especially since travis gives us parallel processing for free (and building the flake environment would take minutes). |
|
No, not wait, but break as soon as the flake env fails. If we take a Travis python env with just flake8 it would not take much time to fire up, or do I'm missing something. |
|
Ah good point. I'm not sure how to do that in travis. |
|
Seems that is most wanted, but still not implemented: travis-ci/travis-ci#2062 |
|
@thomasdeneux Green lights. Thanks for adding this! |
djhoese
left a comment
There was a problem hiding this comment.
Would it be possible to add some tests for this?
|
Ok, i will add some tests, probably next week |
|
@thomasdeneux Any chance you have time to merge with current master and we can see how the tests perform? |
|
oh i am busy these days... will try next week however |
|
@thomasdeneux Any chance you can make the change? |
|
@thomasdeneux How busy are you a year after the last time I asked? 😜 |
|
@djhoese sorry i didn't know what to answer: i am really busy all the time and this CPU buffer improvement is a feature that is complete but which, at the end, i did not use; so it is not fresh in my mind, i expect it would take me 1~2 hours to go back into it and have the test passed; i would like to take this time, yet it cannot be this week. I was wondering if the best was not to forget about it, i.e. close the branch. Let's say if i find the time in the coming weeks i will still try to do it, otherwise it will remained closed... |
|
Let's keep this open for now. I'll close it after a more thorough review and if I can't see an easy way forward. Thanks. |
# Conflicts: # vispy/gloo/buffer.py
Follow-up from previous PR #1538.