from Hacker News

Show HN: Nextjournal – seamless data science for teams

by kvlr on 5/6/19, 8:36 AM with 33 comments

  • by kvlr on 5/6/19, 8:36 AM

    I’m one of the founders of Nextjournal and I’m really excited that after almost three years in Private Beta we’re finally opening signups to everyone today!

    Nextjournal is a computational notebook platform and our goal is to make computation more accessible and automatically reproducible, so it becomes easier to collaborate and build on top of each others work.

    If you'd like to know more, check out our launch blog post at https://nextjournal.com/mk/public-beta or sign up and give it a try!

  • by kfk on 5/7/19, 1:40 PM

    So I do BI/analytics at a big company with a team of 6 people, here is my take. We need something like this aimed at business analysts with little to no coding experience and we need it to be priced in the $100-300 per year, not more. Such tool would compete with the MS Office package and would be great. Most of the stuff is available in various open source packages, it would be about putting all together in 1 easy desktop install, adding a nice gui interface on top of various functions (like ipywidgets but more high level). For instance, we could totally add a basic gui on top of altair to do some basic charting, that basic charting is 80% of business needs when it comes to explorative analysis.

  • by MartinMond on 5/7/19, 8:47 AM

    I've been researching what data science tools to use at my company.

    How is Nextjournal different from Jupyter or Google Colaboratory?

  • by r3tex on 5/6/19, 9:26 AM

    Nextjournal is really how notebooks were meant to be used - for sharing one's code, its output, and all the reasoning in-between with great looking presentation. I'm very happy that my articles turned out so good looking on the platform.

  • by sandGorgon on 5/7/19, 12:08 PM

    How is this comparable to Google Colab or Azure ML notebooks for python only ? (i know that nextjournal supports many more languages)

    especially pricing per resources (its not clear from the website)

  • by reacharavindh on 5/7/19, 7:54 AM

    I wish I could leverage such polished interfaces for my research group. But, we have lot of contracts that bind us to keep our research data in house. We cannot simply "run something in the cloud".

    So, Jupyterhub and manual tinkering to get such polish for now.

  • by refset on 5/7/19, 11:17 AM

    This is really neat - great work! It took me less than 10m to figure out how to copy a Crux tutorial into Nextjournal using the Clojure template: https://nextjournal.com/crux/a-bitemporal-tale

    The only issue I encountered was that adding comments after the final close parens in the code sections creates EOF errors.

  • by ZeroCool2u on 5/7/19, 1:39 PM

    Having just gone through an evaluation for platforms just like Nextjournal, there are a lot of companies that make similar claims, but very few that deliver in reality.

    In the end, the only one we found that delivered on the promises of reproducibility and managing the entire data science life cycle end to end, facilitating collaboration, and getting stuff done was Domino Datalab[1].

    Can you compare and contrast Nextjournal to DD? Better yet, do you feel you're competing in the same areas or are you really more focused just on reproducibility? Even if you're not now, it feels like eventually, all these types of products seem to converge to this state eventually just by nature of the sales process and promising more and more features to customers.

    Regardless, it looks really solid, so best of luck!

    [1]: https://www.dominodatalab.com/

  • by cw on 5/7/19, 4:43 PM

    really excited to see you guys launch!