Python 3 - please test!!!

[Andrey Novoseltsev]

Hello!

The future is almost here - SageMathCell is working with current SageMath beta using Python 3. If you care - please test it here

http://cosmos.mat.uam.es:8888/

and see how bad your code is broken. If you don't - please test anyway and help others to discover problems!

Already known problems:

1) Permalinks don't work - will take care of this, surely some Unicode issues.

2) Threejs does not work for me in Firefox, but works in Chromium. If someone has insights into this - please share.

I would like to pull the trigger in a couple weeks, so that those who rely on it for classes in January have the winter break to adapt. If Sage 9 is not ready by then - I will use a beta version until the official release.

Thank you for your help!

Andrey

[Samuel Lelièvre]

Sun 2019-12-09 01:45:37, Andrey Novoseltsev:

>
> Already known problems:
> 1) Permalinks don't work - will take care of this,

> surely some Unicode issues.
> 2) Threejs does not work for me in Firefox, but works in

> Chromium. If someone has insights into this - please share.

A similar issue (but concerning Windows -- might have nothing

to do with the issue you see) was recently solved by Erik Bray:

https://trac.sagemath.org/ticket/28801

[Andrey Novoseltsev]

Not sure it is the same, I am getting "THREE.WebGLRenderer: Error creating WebGL context." so it seems that scripts are loaded.

[Samuel Lelièvre]

Mon 2019-12-09 03:24:35 UTC, Andrey Novoseltsev:

> Not sure it is the same, I am getting

> "THREE.WebGLRenderer: Error creating WebGL context."

> so it seems that scripts are loaded.

Could the answer to this Stack Overflow question help?

  https://stackoverflow.com/q/34547208

Note: no problem with Three.js on Firefox 71.0

on macOS 10.14.6 here.

[Andrey Novoseltsev]

Ok, so I am the only one with Threejs issues in Firerox? Just entering "cube()" does not work for me. It is version 68.2 ESR. The graphics card is from 2008, which may be the reason for "Refused to create native OpenGL context because of blacklist entry: FEATURE_FAILURE_OLD_FGLRX" Well, I guess nobody is going to address that issue.

Is everything else working? Would be nice ;-) And I have updated the test version to 9.0.beta9.

[Samuel Lelièvre]

Le lun. 9 déc. 2019 à 15:53, Andrey Novoseltsev:

>
> Ok, so I am the only one with Threejs issues in Firerox?
> Just entering "cube()" does not work for me. It is
> version 68.2 ESR. The graphics card is from 2008,
> which may be the reason for "Refused to create
> native OpenGL context because of blacklist entry:
> FEATURE_FAILURE_OLD_FGLRX" Well, I guess
> nobody is going to address that issue.

There is a similar report on the Mozilla support forum,
see the 15 Nov 2019 post which is extremely similar:

https://support.mozilla.org/en-US/questions/1266637

> Is everything else working? Would be nice ;-)

I'm hoping to do more testing this week.

> And I have updated the test version to 9.0.beta9.

Excellent, thank you!

[kcrisman]

cube() works fine for me on FF on Mac.

[David Austin]

cube() works on firefox on linux (ubuntu 18.04).

On Mon, Dec 9, 2019 at 9:21 PM kcrisman <kcri...@gmail.com> wrote:

cube() works fine for me on FF on Mac.

--
You received this message because you are subscribed to the Google Groups "sage-cell" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sage-cell+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sage-cell/cff208e0-3d77-4792-be29-eae9dd6e69d4%40googlegroups.com.

[Andrey Novoseltsev]

I was able to make threejs work for me as well after tweaking settings. And it is actually not working for me without tweaking settings in the old version as well, so it is an issue unrelated to Python 3.

Please do test your favourite interacts - this is where I expect to uncover some problems if there are any left!

[Andrey Novoseltsev]

Hello,

Please do test the test server! I have resolved the problem with permalink and plan to switch on Christmas Eve (using the latest Sage-9.0beta).

For the record - I had a lot of issues with my interacts that had to be addressed. No idea how typical my problems were, but to encourage you to look at yours, here is the list of problems I've ran into so far:

1) Works!

2) Several caught exceptions must be in parentheses.

3) Works!

4) map does not return a list anymore.

5) print is not a statement,

map does not return a list anymore.

6) dict.keys() is not a list,

map does not return a list anymore.

7) map does not return a list anymore,

zip does not return a list anymore

8) dict.keys() is not a list,

map does not return a list anymore.

9) Works!

10) Warnings due to \( as math delimiters and invalid escape sequence \m in \mathrm

11) function('y', x) does not work anymore (was already broken by 8.9)

12) Warnings for math delimiters and invalid escape sequences

map does not return a list anymore.

The code

v0, v1 = 1, 2

[eval("v%d" % i) for i in range(2)]

does not work anymore as inside list comprehension locals() are different, had to tweak it to

v0, v1 = 1, 2

current_locals = locals()

[eval("v%d" % i, None, current_locals) for i in range(2)]

[Kwankyu Lee]

On Monday, December 23, 2019 at 6:34:15 AM UTC+9, Andrey Novoseltsev wrote:

Please do test the test server! I have resolved the problem with permalink and plan to switch on Christmas Eve (using the latest Sage-9.0beta).

I could test if the server is served over https...