Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Datalore by JetBrains (jetbrains.com)
3 points by gamesbrainiac on Oct 20, 2020 | hide | past | favorite | 1 comment


I continue to be underwhelmed (and slightly concerned) by Datalore. As a paying customer who spends most of my time using PyCharm and a regular user of Jupyter notebooks... I would much rather they be focusing these efforts on better support for notebook functionality inside PyCharm itself and deliver a first class experience in the IDE that was interoperable with ideally existing cloud vendors but at the least just any Jupyter Lab/Notebook server that you have permission to connect to.

If they wanted to enhance something with “data” in the name there’s a lot they could do to improve DataGrip. How about broad support for NoSQL data stores like DynamoDB, CosmosDB Redis, and ElasticSearch. Not keen on adding a new there’s plenty of sharp edges left in the MongoDB functionality that need more work. Not flashy enough? Ok how about some sort of built in magic Presto/DIY AWS Athena that lets me use DataGrip to query practically any data I like?

I understand why they need to be building online tools like TeamCity which fit into their overall product line, but between their Source collaboration tools and DataLore I feel like they are leaving their core competencies and building things that have much lower real world value. Collaboration are usually pushed into a company by upper management and unless they want to fully integrate with G-Suite or Office365 and become a competitor to the terrible Microsoft Teams app then I don’t see them gaining more than a pittance when it comes to collaboration tools market.

The story around notebooks just doesn’t mesh the same way into a software development lifecycle and I just don’t see the value of DataLore... it’s not hosted in the same cloud as my data so I have to deal with all the security and management overheads of external data for no real gain when all three major cloud platforms have hosted Notebook and/or “workspace” equivalent services or you can just host your own Jupyter server secure along with your data for easy access.

I know they are smart people and I have no doubt they do plenty of cost benefit analysis and aren’t wasting money on these things, but these things just feel worryingly like distractions.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: