Juno is an iPad client for Jupyter Notebook — it lets you connect to a remote Jupyter notebook server and do everything you do with Jupyter on desktop, anywhere.
No, you need a notebook server that will do the actual computing. There are many options of running a notebook server: it can be your own laptop in your local network, AWS node, or even a dedicated powerful GPU-enabled machine configured as a public server or accessible via tunneling. Alternatively, you could use one of cloud computing services that are based on Jupyter, such as CoCalc or Azure Notebooks.
Yes, absolutely! Your PC will need to be switched on and accessible to your iPad at all times though. There are a couple of ways of connecting to it: if your iPad is on the same local network, you can just use your PC’s IP address. If you would like to connect over the internet, you would need to configure your PC as a public server (see Running a public notebook server in Jupyter documentation).
You can access your server in Juno if you have SSH tunnel established on your iPad; however, you may need a 3rd party app to configure the port forwarding part. There are apps on the AppStore that will let you establish a tunnel with port forwarding (for example, this one), and Juno will then work by pointing it at
localhost and mapped port.
We plan to include SSH tunneling capability in Juno in one of our future releases.
Usually this means that your server uses a self-signed SSL certificate for HTTPS which doesn’t work on iOS due to security restrictions — neither in Safari nor in Juno. Check out our step-by-step guide on how to configure certificates and enable SSL, so that you are able to connect to your server from iPad.
Sadly, not yet. We would love to support Google Colaboratory as a cloud computing service, however, we are currently pending some additional info from the Google team. Meanwhile, I’d say it wouldn’t hurt to ping them through customer support, so that they know there is interest in Juno integration. 😉
Juno can do a lot on top of what Safari has to offer! For instance, it has responsive and less intrusive interface and benefits from full hardware keyboard support.
Juno has additional binding to Control key, which duplicates Esc key and lets you enter command mode, for instance.
Unfortunately, there is a known issue with assisted input using latin characters (something similar to IME in Windows, where you use latin characters to input in Japanese, Korean or Chinese). Apparently, this issue is caused by one of the components Jupyter is using, and this issue has been reported, although not resolved yet — see discussion here. We are looking into a workaround for Juno, and hope to fix or at least improve international input in our future releases. Sorry for inconvenience meanwhile.
Editing of text files is switched off for now, but we are looking into ways of either allowing it the way Jupyter normally does, or at least letting you open copies of those files in other apps. Sorry for inconvenience meanwhile.
Juno doesn’t support any JupyterLab specific features yet; however, all JupyterLab servers also include “classic” Jupyter Notebook as a fallback option, which Juno works perfectly with. You can access it by appending
/tree to your JupyterLab server URL.
As of version 1.2, Juno should be able to detect JupyterLab servers and automatically redirect you to
If your workflow involves Jupyter notebooks and:
Please, send an inquiry to [email protected], and we will be happy to help.
Please, report all defects and send feature requests to [email protected].
Make sure to run a recent version of Jupyter Notebook, preferably 4.1.0 or higher.