Spyder: New colour scheme results in unreadable tracebacks

[dhirschfeld]

On spyder master a recent change to the color scheme has resulted in unreadable tracebacks:

This seems similar to what was being used in jupyterlab:

jupyterlab/jupyterlab#807
jupyterlab/jupyterlab#2384

In particular, the discussion in 2384 seems relevant.

[ccordoba12]

I can't reproduce this, at least on Linux, sorry.

Please take a look at it on Windows.

[ccordoba12]

Do you have set some options about this in your IPython config files?

[dhirschfeld]

This was on my work computer on windows x64, Anaconda Python 3.6.

I don't set anything to do with colours for ipython - just use whatever the defaults are. Will check on my home computer also Win x64 / py36

[dhirschfeld]

On my home PC (latest master) I'm not seeing it. Will double check at work and post more details...

[ccordoba12]

Ok, please do. I think the problem could arise because we're not detecting the background color of the current color scheme correctly. But we're just using a Pygments function for that...

[dhirschfeld]

So on my work PC I checked to see if the problem was still there and it was. I then updated to the latest master and the problem appears to be gone.

C:\dev\packages\spyder [master ≡]
λ git pull
remote: Counting objects: 402, done.
remote: Compressing objects: 100% (10/10), done.
Receiving objects:  93% (374/402), 84.00 KiB | 66.00 KiB/s
Receiving objects: 100% (402/402), 146.24 KiB | 66.00 KiB/s, done.
Resolving deltas: 100% (308/308), completed with 70 local objects.
From https://github.com/spyder-ide/spyder
   243b390d..b7d651bb  master     -> upstream/master
   3507b148..38518184  3.x        -> upstream/3.x
Updating 243b390d..b7d651bb
<snip>

So it looks like somewhere between 243b390d and b7d651bb the problem has been fixed.

Since I can't repro on master I'll close this as fixed.