Rob Patro
@robp.bsky.social
Associate Professor of CS @ University of Maryland. Proud Rust advocate! I ♥ science & compiled, statically-typed programming languages! Views are my own. Tech stack: https://github.com/rob-p/tech-stack.
It is an old 27" iMac Pro with an 8 core intel CPU and 64GB of RAM. So both the Mac Studio M4 Max or the AMD threadripper would be massive upgrades.
January 16, 2026 at 8:27 PM
It is an old 27" iMac Pro with an 8 core intel CPU and 64GB of RAM. So both the Mac Studio M4 Max or the AMD threadripper would be massive upgrades.
The question is, will I make use of that power in a desktop if I have a beefy server for large-scale work? 3/3
January 16, 2026 at 8:16 PM
The question is, will I make use of that power in a desktop if I have a beefy server for large-scale work? 3/3
The biggest arguments in favor of the Mac are that I use keynote for all of my presentations, and I make a fair number (since I teach every semester) and also my laptop is a Mac. The biggest argument in favor of linux is I like the dev env better & the machine will be more powerful. 2/3
January 16, 2026 at 8:16 PM
The biggest arguments in favor of the Mac are that I use keynote for all of my presentations, and I make a fair number (since I teach every semester) and also my laptop is a Mac. The biggest argument in favor of linux is I like the dev env better & the machine will be more powerful. 2/3
Yea, so we are getting a new server too for heavy workloads, and that will be a 2 socket machine with tons of cores. This is just my desktop. At the same time, I want to be able to do development, interactive work, and some analysis on it locally. 1/3
January 16, 2026 at 8:16 PM
Yea, so we are getting a new server too for heavy workloads, and that will be a 2 socket machine with tons of cores. This is just my desktop. At the same time, I want to be able to do development, interactive work, and some analysis on it locally. 1/3
Any thoughts on the MacOS + M4 Max vs Linux + Threadripper?
January 16, 2026 at 7:18 PM
Any thoughts on the MacOS + M4 Max vs Linux + Threadripper?
I hope they figure out how to block LLM access...
January 15, 2026 at 2:27 PM
I hope they figure out how to block LLM access...
More than I thought, honestly.
January 15, 2026 at 4:54 AM
More than I thought, honestly.
Back up now for me. It was down for me for about 2.5 hours this morning.
January 14, 2026 at 5:50 PM
Back up now for me. It was down for me for about 2.5 hours this morning.
I think it went pretty well!
January 14, 2026 at 3:43 AM
I think it went pretty well!
Basic support for "post monomorphization" macros would allow the user to stably build their own solutions, while we wait for reasonably capable const-generic to hit stable. 2/2
January 13, 2026 at 2:01 AM
Basic support for "post monomorphization" macros would allow the user to stably build their own solutions, while we wait for reasonably capable const-generic to hit stable. 2/2
Most immediately, I'm interested in solving things like this (users.rust-lang.org/t/could-macr...) without requiring the unstable const generic expressions. The problem with nightly, is that they regularly change the const-generic expression behavior, so things stop working. 1/2
Could macros be used to work around the lack of support for const generics arithmetic?
I don't understand why const generics arithmetic is so hard to implement as a language feature, and I don't think I ever will. I couldn't begin to understand the intricacies of the Rust compiler (what...
users.rust-lang.org
January 13, 2026 at 2:01 AM
Most immediately, I'm interested in solving things like this (users.rust-lang.org/t/could-macr...) without requiring the unstable const generic expressions. The problem with nightly, is that they regularly change the const-generic expression behavior, so things stop working. 1/2
Thanks for the suggestion, Yann! I've decided to go with DP as well, rather than NP-completeness. I made lots of nice pictures of edit graphs and will attempt to explain global string alignment :). I'll let you know how it goes!
January 12, 2026 at 10:00 PM
Thanks for the suggestion, Yann! I've decided to go with DP as well, rather than NP-completeness. I made lots of nice pictures of edit graphs and will attempt to explain global string alignment :). I'll let you know how it goes!
Yea, valid point. I think maybe one would need to restrict the output of these "post monomorphization" macros somehow. However, I've already solved one of the most important problems for a 3rd party crate; the name. It should be called "postmon" (and be a play on postman).
January 12, 2026 at 3:19 PM
Yea, valid point. I think maybe one would need to restrict the output of these "post monomorphization" macros somehow. However, I've already solved one of the most important problems for a 3rd party crate; the name. It should be called "postmon" (and be a play on postman).
Yes, I'm asking why not, to allow working around the extreme restrictions to const generic expressions, to have an optional annotation to allow some macros to run after monomorphization.
January 12, 2026 at 3:10 PM
Yes, I'm asking why not, to allow working around the extreme restrictions to const generic expressions, to have an optional annotation to allow some macros to run after monomorphization.
What do you mean by "faculty transition" here?
January 10, 2026 at 9:33 PM
What do you mean by "faculty transition" here?