Migrating to Positron, a next-generation data science IDE for Python and R
posit.co
posit.co
"Positron is licensed under the Elastic License 2.0, a source-available license. This license makes Positron available for free to everyone to use, build on, and extend for personal, academic, and commercial use. Its primary restriction is that you can’t host Positron as a service to third parties without Posit’s agreement. This restriction is necessary for us to build a sustainable business around Positron while also offering it free of charge to the community."
"You may not move, change, disable, or circumvent the license key functionality in the software, and you may not remove or obscure any functionality in the software that is protected by the license key."
This is not a recommendation on whether anyone should use Positron. I think it's fair that people know it's just another piece of proprietary software with a license key. Posit is a public benefit corporation, which sounds nice, but you're still subject to the same games any other for-profit private company plays with its customers.
(I use Posit Cloud in my teaching, so my interactions with their products are as a customer, but I use RStudio, which is open source.)
Feels really strange to lump this together with PowerBI...
Respectfully, I feel like we could all benefit from having better categories than you are offering. Good categories are tools to navigate an increasingly complex world, and this sort of reduction serves very little
> RStudio (now Posit) was founded in 2009 with the vision of creating high quality open-source software for data scientists. We’ve grown exponentially over time but our culture remains unchanged. We invest heavily in open-source development, education, and the community with the goal of serving knowledge creators 100 years from now.
> We want Posit to serve a meaningful public purpose and we run the company for the benefit of our customers, employees, and the community at large. That’s why we’re designated as a Public Benefit Corporation. As a Certified B Corp, we must meet the highest verified standards of social and environmental performance, transparency, and accountability. Our directors and officers have a fiduciary responsibility to address social, economic, and environmental needs while still overseeing our business goals.
So if maintaining RStudio is so much of a burden that it impedes the rest of their work, I don't think it's a bad idea to reduce the amount of work spent trying to compete with VSCode when that's an increasingly tough sell.
I'm not a fan of VSCode personally, but would probably be happy with a tmux setup with a console for R and some minimal output viewer, so people like me should be able to cobble something together that's a workable alternative to Posit.
There is Rkward[1], available from the repositories in many linux distributions.
Discovered it recently, because I'm currently learning R, and few linux distributions outside the Mandriva family offer Rstudio straight from the repositories, and I'm lazy to download and do a manual install every time.
> We anticipate many RStudio users will be curious about Positron.
Heavily implies it as a seperate thing that will continue to be maintained. They haven't said they're getting rid of support for Rstudio.
I think this is probably more that Posit have been trying to move more and more into the Python space, since that's where most data science is happenening. Rstudio has a great but is obviously very associated with R, so making a similarly intended project that is more explicit is supporting other languages isn't inherently a bad shout.
But the community can maintain it.
The Rstudio users can give a roadmap and ask for help.
> Look inside
> VS Code
Replacing Rstudio with something more reliable would be nice, because of some major design flaws Rstudio has. A lot of the UI stuff runs also in R, so when the R kernel dies, quite often I cannot save unsaved files. So I need to copy the file content to a different text editor when that happens. I also don't understand why the LLM-Chat window is running inside the R console, and then blocks running R code. That makes it completely unusable.
Coincidentally I was thinking of giving R another go, but honestly now... I'm good...
There are exceptions. E.g. Theia actually does enough stuff differently that I think it warrants being its own thing. At least it did. Looks like they have jumped on the AI bandwagon too.
Maybe this is the same; I haven't looked at it in detail. But "we have an IDE! (don't tell them it's vscode)" feels a lot like "we have an app! (don't tell them its a webview)".
RStudio is great when you are doing your own thing, but when it comes to more generic tools like Git, LLM, Autoformatting etc. it's a hard pass.
I think the future of data science will likely be something else, with the advent of WebGPU[1] (which isn't just a web technology) and the current quality/availability of GPUs in end user devices, and a lot of data computation clearly standing to benefit from this.
The real next generation of data science tools will likely involve tools that are GPU first and try to keep as much work in the GPU as possible. I definitely think we'll see some new languages eventually emerge to abstract much of the overhead of batching work but also forces people to explicitly consider when they write code that simply won't run on the GPU, like sequential operations that are nonlinear, nonassociative/noncommutative (like highly sequential operations like processing an ordered block of text).
I think WebGPU is going to make this a lot easier.
That said I'd imagine for larger compute workloads people are going to continue to stick with large CUDA clusters as they have more functionality and handle a larger variety of workloads. But on end user devices there's an opportunity to create tools that allow data scientists to more trivially do this kind of work when they compute their models, process their datasets.
[1] Other compute APIs existed in the past, but WebGPU might be one of the most successful attempt to provide a portable (and more accessible) way to write general GPU compute code. I've seen people say WebGPU is hard, but having given it ago (without libraries) I don't think this is all that true, compared to OpenGL there are no longer specialised APIs to load data into uniforms everything is just a buffer. I wonder if this has more to do with non JS bindings for use outside the browser/node or the fact you're forced to consider memory layout of anything your loading into the GPU from the start (something that can be abstracted and generalised), just in my experience after my first attempt at writing a compute shader it's fairly simple IMO. Like stuff that always complicated in rendering like text is still complicated, but at least its not a state based API like web/opengl.
As long as silicon is scarce it would make sense to hoard it and rent it out (pricing as a means of managing scarcity); if we end up in a scenario where silicon is plentiful, then everyone would have powerful local GPUs, using local AI models, etc.
There are aspects of data science which is iterative and you're repeatedly running similar computations with different inputs, I think there's some value in shaving off time between iterations.
In my case I have a temporal geospatial dataset with 20+ million properties for each month over several years each with various attributes, it's in a nonprofessional setting and the main motivator for most of my decisions is "because I can and I think it would be fun and I have a decent enough GPU". While I could probably chuck it on a cluster, I'd like to avoid if I can help it and an optimisation done on my local machine would still pay off if I did end up setting up a cluster. There's quite a bit of ETL preprocessing work before I load it into the database, I think are portions that might be doable on the GPU. But it's more so the computations I'd like to do on the dataset before generating visualisations in which I think I could reduce the iteration wait time for processing for plots, ideally to the point I can make iterations more interactive. There's enough linear operations you could get some wins with a GPU implementation.
I am keen to see how far I'll get, but worst case scenario I learn a lot, and I'm sure those learnings will be transferrable to other GPU experiments.
Since AWS/etc are quite complicated, there are now a swarm of startups trying to make it easier to take baby steps into the cloud (eg. Modal, Runpod, etc) and make it very easy for the user to get a small slice of that GPU pie. These services have drastically simpler server-side APIs and usage patterns, including "serverless" GPUs from Modal, where you can just "submit jobs" from a Python API without really having to manage containers. On the client side, you have LLM coding agents that are the next evolution in UI frontends -- and they're beginning to make it much much easier to write bespoke code to interact with these backends. To make it abundantly clear what target audience I'm referring to: I imagine they are still mostly using sklearn (or equivalents in other languages) and gradient boosting with Jupyter notebooks, still somewhat mystified by modern deep learning and stuff. Or maybe those who are more mathematically sophisticated but not software engg sophisticated (eg: statisticians / econometricians)
To inspire you with a really concrete example: since Modal has a well documented API, it should be quite straight-forward ("soon", if not already) for any data scientist to use one of the CLI coding agents and
1. Implement a decent GPU-friendly version of whatever model they want to try (as long as it's not far from the training distribution i.e. not some esoteric idea which is nothing like prior art)
2. Whip up a quick system to interface with some service provider, wrap that model up and submit a job, and fetch (and even interpret) results.
----
In case you haven't tried one of these new-fangled coding agents, I strongly encourage you to try one out (even if it's just something on the free tier eg. gemini-cli). In case you have and they aren't quite good enough to solve your problem, tough luck for now... I anticipate their usability will improve substantially every few months.
Or at least a positive version of it...