[Debian-science-sagemath] ipywidgets: java material

[Ximin Luo]

Julien Puydt:
> Hi,
> 
> On 03/10/2016 14:47, Jeroen Demeyer wrote:
>> On 2016-10-03 14:43, Ximin Luo wrote:
>>> IMO it's easier to motivate extra volunteers if they can see a nice
>>> mostly-working sagemath Debian package
>>
>> And how is a non-functioning ipywidgets better than not having
>> ipywidgets at all? I would say that not having ipywidgets is better,
>> since you will at least get an explicit and clear error when you try to
>> use ipywidgets.
> 
> I have to agree here : having something partially functional is ok, having somethings where the packaging isn't quite right is ok -- in experimental, of course! But having something just broken doesn't really push things further.
> 
> In fact, those missing JavaScript pieces is the reason why I don't have ITPed ipywidgets myself...
> 

It pushes things further, in the sense that there are less remaining tasks to do.

Jerome has already done the work, and it does help to unblock some work on the sagemath package. So I think the current discussion is a bit pointless. We are all in agreement that it would not be suitable for a final release to Debian.

In terms of the javascript stuff, we can stretch policy for npm-related crap. It is probably not necessary to package *every single npm package* as its own Debian package. For example if developer J releases packages npm-left-pad-string-with-1-zero and npm-left-pad-string-with-2-zeros, I would just put them both in the Debian package npm-string-utils or something like that; use your own judgement of course. See node-ws for an example, although I probably over-engineered that one slightly.

X

-- 
GPG: ed25519/56034877E1F87C35
GPG: rsa4096/1318EFAC5FBBDBCE
https://github.com/infinity0/pubkeys.git