because method chaining in Polars is much more composable and ergonomic than SQL once the pipeline gets complex which makes it superior in an exploratory "data wrangling" environment.
"revolutionary"? It just copied and pasted the decades-old R (previous "S") dataframe into Python, including all the paradigms (with worse ergonomics since it's not baked into the language).
No other modern language will compete with R on ergonomics because of how it allows functions to read the context they’re called in, and S expressions are incredibly flexibly. The R manual is great.
To say pandas just copied it but worse is overly dismissive. The core of pandas has always been indexing/reindexing, split-apply-combine, and slicing views.
It’s a different approach than R’s data tables or frames.
> allows functions to read the context they’re called in
Can you show an example? Seems interesting considering that code knowing about external context is not generally a good pattern when it comes to maintainability (security, readability).
I’ve lived through some horrific 10M line coldfusion codebases that embraced this paradigm to death - they were a whole other extreme where you could _write_ variables in the scope of where you were called from!
I can write code like:
penguin_sizes <- select(penguins, weight, height)
Here, weight and height are columns inside the dataframe. But I can refer to them as if they were objects in the environment (I., e without quotes) because the select function looks for them inside the penguins dataframe (it's first argument)
This is a very simple example but it's used extensively in some R paradigms
Dataframes first appeared in S-PLUS in 1991-1992. Then R copied S, and from 1995-1996-1997 onwards R started to grow in popularity in statistics. As free and open source software, R started to take over the market among statisticians and other people who were using other statistical software, mainly SAS, SPSS and Stata.
Given that S and R existed, why were they mostly not picked up by data analysts and programmers in 1995-2008, and only Python and Pandas made dataframes popular from 2008 onwards?
Exactly. I was programming in R in 2004 and Pandas didnt exist. I remember trying Pandas once and it felt unergonomic for fata analysis and it lacked the vast library of statistical analysis library.
cool, but doesn't sound that great when you close your eyes and just listen. Other synths beat this hands down especially at > $1000, and can easily bring in the physical world already, including live workflows. The issue is when we get into the physical analogue world, craftsmanship, materials, shape, often age, and of course the varied kinetic interactions with the sound solicitor, bring depth and richness which no little electrically-excited xylophone will ever get anywhere close to.
Ultimately what's happening here is AI is undermining trust in remote contributions, and in new code. If you don't know somebody personally, and know how they work, the trust barrier is getting higher. I personally am already ultra vigilant for any github repo that is not already well established, and am even concerned about existing projects' code quality into the future. Not against AI per se (which I use), but it's just going to get harder to fight the slop.
I don't understand why Apple doesn't offer a headless MacOS or at least a path to a minimal install. Those mac minis make a great little server box but losing 8GB to hundreds of processes, before you've done anything, just feels wasteful and inelegant.
Worse, there's sales in NOT doing it. When I buy a Mac, I get extra memory "just in case." I would've been fine with 24 gigs on my MacBook Pro, but I got 48.
It wasn't an absence of a market. Those of us that had to manage OSX Server soon found out the software was marked by several high-profile bugs, technical debt, and a perceived decline in reliability. I migrated a large number of Macs to Ubuntu Server software. The hardware was great.
I fear the quality of macOS is deteriorating today in the same manner than befell OSX Server.
Right, but I could see an alternate timeline where OS X Server took off, and within a decade took a path similar to Windows Server (pared down services, headless flavor, etc)
I am not sure iOS popularity would justify macOS as a server. What would be the use case? It's not app development; that is done just fine on the standard desktop macOS. It's not backend; that is done just fine on Linux servers, even in Swift if that's your thing.
Those Mac minis are a pain in the ass of a server box that auto-enable FileVault after annual releases, and getting LaunchDaemons just right compared to a Linux OS feels like perpetual iterations. trying to figure out why my apache didn't start after the last reboot. Oh, must have been the Mac log rotator messing with the file permissions again
It's a shame, because I love how efficiently MacOS runs and the form factor/design language of a Mac mini is not something I feel the need to hide in a dark corner
You'll have to leave virtual desktop enabled, and will definitely be using it semi-regularly aside SSH
> Those mac minis make a great little server box but losing 8GB to hundreds of processes
It doesn't matter because all the extra stuff just goes to swap. And you can't disable virtual memory anyways. So in the end you're not really losing anything. Those hundreds of processes are ultimately basically mostly just using up a little bit of your SSD, not your RAM, so it's not a concern.
Yeah but they mostly barely do, and only the memory they actually access gets used, as opposed to everything they've allocated. You can observe the actual aggregate usage in Activity Monitor. This is why it's no problem at all to run something actively using 10-12 GB of memory on a 16 GB Mac.
Can speak for someone who managed a racked [1] Mac Mini racked as a server in 2015~2018:
* Profile Manager / MDM
* Caching server for OS updates
* Thunderbolt connection to a PCIe RAID HBA for a JBOD for NAS file sharing
All these things/use cases now are depreciated, MDM via JAMF in the cloud, internet is fast enough that caching OS updates isn't necessary, and NAS's are much better suited.
>> "The adversary has a limited ability to persuade users to trust the adversary’s agents - thus the number of social connections between the adversary’s agents and the rest of the network is limited." [1]
This seems like a fairly reasonable assumption for the Briar case. An adversary would have to get a user accept a Briar link in error. Contrast with things like Signal[1] that base their trust on phone numbers.
Note that Briar groups are controlled by one moderator. Other participants of the group can not add members. Note also that Briar has the concept of introductions. So it is easy to avoid making the sort of identity errors commonly made by other schemes.
