I’ve been playing around with Datalore to understand how it compares to the Jupyter Notebook environment. Can you tell me if all of my statements below are correct?
- If you upload a data file, it is only available to that workbook, and is not available to other workbooks.
- Some actions are only available by using keyboard shortcuts. For example, if I want to delete a block, I have to use a keyboard shortcut, since “delete block” is not an option in the main menu or in the right click menu.
- There’s no way to move a block up or down. The only efficient way to do this is by using cut and paste.
- IPython magic functions are not supported.
You got it right, all your statements are correct. We have some ideas on how to improve UX for the first three points – I’ll ask the devs to check if some of them are cheap to implement and can be done now.
Great! Points 1 through 3 all make it more difficult for me to use, as compared to the Jupyter Notebook, so any UX improvements in those areas would be great.
Speaking of the keyboard shortcuts (point 2), I find it hard to navigate the Datalore shortcuts because I don’t understand how they are organized. They’re not in alphabetical order, but they are not grouped into sections either. I find Jupyter’s help window to be easy to follow, and it’s nice that Jupyter has a separate, searchable command palette. Being able to browse the Datalore shortcuts in logical sections is crucial for someone coming from Jupyter, because the Datalore shortcuts are 100% different from Jupyter shortcuts.
Hi Kevin. I’ve just wanted to mention that you can select all content in the cell (Cmd+A twice) and then delete. That said, there’s a keyboard shortcut for deleting cell (sort of).
UPD: sorry got your question wrong - you’ve meant you know about shortcuts
@igro I have a few more questions about Datalore for you:
- The 10 GB of disk space that is available on the Community plan is 10 GB per user account, not 10 GB per workbook - is that correct?
- If I leave a workbook open but it’s inactive (all computations are complete), after how much time will Datalore kill my session?
- Is there a maximum session length? In other words, if my workbook is active (edits and/or computations), is there a maximum amount of time that my Datalore session can run?
That’s right, you can check disk usage at the Billing Report tab in the Account Settings dialog:
(and we definitely need to improve this report to show which workbooks are consuming space – this task is in our backlog)
Datalore will ask you to turn on the “long computation” mode if there is no activity for 50 minutes and after another 10 minutes will close the connection and terminate the agent.
Currently, we do not have a seamless redeployment, so now almost every day the application is unavailable for several minutes during the update – this also forces users to reload already opened pages, if any. But I believe you can work all day long without any session-expiration related issues.