Migrate to 16.04

[Kaixhin]

All images should move from Ubuntu 14.04 LTS to 16.04 LTS, except for CUDA images where versions <= 7.5 should remain with 14.04 and versions >= 8.0 should migrate (see https://github.com/NVIDIA/nvidia-docker/issues/110).

• [ ] brainstorm
• [ ] caffe
• [ ] cuda-brainstorm
• [ ] cuda-caffe
• [ ] cuda-digits
• [ ] cuda-keras
• [ ] cuda-lasagne
• [ ] cuda-mxnet
• [ ] cuda-neon
• [ ] cuda-pylearn2
• [ ] cuda-ssh
• [ ] cuda-theano
• [ ] cuda-torch
• [ ] cuda-vnc
• [ ] digits
• [ ] fglab
• [ ] fgmachine
• [ ] keras
• [ ] lasagne
• [ ] localtunnel
• [ ] mxnet
• [ ] neon
• [ ] neuron (LXTerminal not working)
• [ ] pylearn2
• [x] ros
• [x] samba
• [x] spearmint
• [x] ssh
• [x] sshx
• [ ] theano
• [ ] torch
• [ ] vnc (LXTerminal not working)
• [ ] vnc-ros (LXTerminal not working)

[Kaixhin]

Closing for now as Ubuntu 16.04 support for lots of software is still questionable.

[jorpilo]

Hey I really love this docker but I really think it's time to ship to 16.04 or even 18.04

14.04 is getting obsolete and tones of packages aren't even available for that release.

Also updating with latest cuda would be nice.

I can try to help you if you want.

[Kaixhin]

I am no longer actively maintaining this repo but am able to take contributions. If you are able to add updated Dockerfiles that build and run successfully then I will accept them and also build them locally to push to the Docker hub. I think that having some 16.04 images (with CUDA 9.2 where applicable) would be the right level to target for now.

[jorpilo]

Sure I will do that. Do you have any tests I can run to check if the image runs properly rather than just see a compilation sucess?

If you want I can upload the images myself if you allow me to do it.

[Kaixhin]

Unfortunately I don't have any automated tests, but the general way I do it is to start up a container and then run a basic relevant task, such as run a piece of example code provided with a framework - that's usually enough to catch if something broke. If you could send separate PRs per framework (e.g. caffe and cuda-caffe) then it would make it easier to sort out any potential issues.

That's a very kind offer 😄 Unfortunately, letting you upload images would be irresponsible on my part. In any case, I think it would be best for me to independently build and test your PRs, just as an additional check for errors, particularly as I don't have automated tests.

[jorpilo]

Hahahaha I wasn't going to hack everyones PC but I totally get your point :satisfied: really good tweet!

I will try to find an example code for it.

No problem I will create different PR for each thing

Let's see when I get some free time to do it :smile: