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.
Why is it that, despite being behind a perpetual cloud flare fence, @biorxivpreprint.bsky.social feels like it’s down an awful lot?
January 14, 2026 at 3:01 PM
Why is it that, despite being behind a perpetual cloud flare fence, @biorxivpreprint.bsky.social feels like it’s down an awful lot?
Reposted by Rob Patro
Anyone here know any relevant RFCs, Pre-RFCs, internals.rlo threads, or anything else?
Here's a rust question for folks out there in the community. Why not have an annotation for macros that signifies that they should be evaluated after constants/generics are placed rather than before? This would let macros fill a big gap that const generics are currently missing.
January 12, 2026 at 3:14 PM
Anyone here know any relevant RFCs, Pre-RFCs, internals.rlo threads, or anything else?
Here's a rust question for folks out there in the community. Why not have an annotation for macros that signifies that they should be evaluated after constants/generics are placed rather than before? This would let macros fill a big gap that const generics are currently missing.
January 12, 2026 at 2:40 PM
Here's a rust question for folks out there in the community. Why not have an annotation for macros that signifies that they should be evaluated after constants/generics are placed rather than before? This would let macros fill a big gap that const generics are currently missing.
Reposted by Rob Patro
Things Claude has been
Excellent with recently: trait bounds, GAT constraints, modern C++ metaprogramming with variadic generics and SFINAE patterns.
Poor at recently: generating correct and efficient concurrent code, and removing the artifacts of old and incorrect designs as code evolves.
Excellent with recently: trait bounds, GAT constraints, modern C++ metaprogramming with variadic generics and SFINAE patterns.
Poor at recently: generating correct and efficient concurrent code, and removing the artifacts of old and incorrect designs as code evolves.
January 10, 2026 at 5:38 PM
Things Claude has been
Excellent with recently: trait bounds, GAT constraints, modern C++ metaprogramming with variadic generics and SFINAE patterns.
Poor at recently: generating correct and efficient concurrent code, and removing the artifacts of old and incorrect designs as code evolves.
Excellent with recently: trait bounds, GAT constraints, modern C++ metaprogramming with variadic generics and SFINAE patterns.
Poor at recently: generating correct and efficient concurrent code, and removing the artifacts of old and incorrect designs as code evolves.
Ok, I am giving a talk (~45 min) to a high school audience next week. I haven't really done this before. I'd like to teach them 1 deep concept (NP-completrness, I think). What strategy/example would you use for this audience? Any suggestions for making it stick/be impactful?
January 10, 2026 at 4:35 PM
Ok, I am giving a talk (~45 min) to a high school audience next week. I haven't really done this before. I'd like to teach them 1 deep concept (NP-completrness, I think). What strategy/example would you use for this audience? Any suggestions for making it stick/be impactful?
Hey @biorxivpreprint.bsky.social; what's up with the dupes today?
January 8, 2026 at 4:09 PM
Hey @biorxivpreprint.bsky.social; what's up with the dupes today?
De Bruijn graphs for what now? www.preprints.org/manuscript/2...
A De Bruijn Graph Formulation of Quantum Entanglement
Quantum entanglement is commonly characterized through global state descriptions on tensor product spaces, correlation measures or algebraic constructions, while local consistency constraints play no explicit structural role. We formulate entanglement as a combinatorial structure of overlapping local descriptions, drawing on De Bruijn graphs, where nodes represent overlapping contexts and paths encode globally coherent assemblies. We construct a graph whose nodes represent reduced quantum states on subsystems of fixed size and whose edges encode admissible extensions consistent with quantum mechanical compatibility conditions. Global many body states correspond to paths on this graph, while entanglement is reinterpreted as a property of graph connectivity and path multiplicity, rather than as a standalone numerical quantity. This formalism allows a separation between constraints imposed purely by local quantum consistency and additional structure introduced by dynamics, symmetries or boundary conditions, also clarifying how large-scale structural features may arise from local compatibility alone. Our graph-based formulation provides several advantages over conventional approaches. Supporting a unified treatment of static entanglement structure and dynamical evolution, it incorporates finite order locality and memory effects. Entanglement growth can be interpreted as path proliferation, while decoherence and noise correspond to the removal of admissible transitions. Our approach leads to testable hypotheses concerning the scaling of admissible state extensions, the robustness of entangled structures under local perturbations and the emergence of effective geometry from overlap constraints. Potential future directions include applications to many body reconstruction problems and comparative analysis of different classes of quantum states within a single combinatorial language.
www.preprints.org
January 8, 2026 at 5:38 AM
De Bruijn graphs for what now? www.preprints.org/manuscript/2...
Harbaugh out; FINALLY!!
January 6, 2026 at 11:20 PM
Harbaugh out; FINALLY!!
TIL that the final level pattern table commonly taught with Jacobson’s rank (incl by me!) does not, actually appear in Jacobson’s thesis (that has only 2 levels). The first actual description I can find is Munro et al. 2001 www.sciencedirect.com/science/arti.... Is anyone aware of an earlier ref?
Space Efficient Suffix Trees
We give first the representation of a suffix tree that uses nlgn+O(n) bits of space and supports searching for a pattern string in the given text (fro…
www.sciencedirect.com
January 6, 2026 at 3:03 PM
TIL that the final level pattern table commonly taught with Jacobson’s rank (incl by me!) does not, actually appear in Jacobson’s thesis (that has only 2 levels). The first actual description I can find is Munro et al. 2001 www.sciencedirect.com/science/arti.... Is anyone aware of an earlier ref?
Amazing, but Jesus Christ, please don't let Donny or RFK Jr. hear about these promising developments, lest they destroy them in the name of reckless disregard of human life and accomplishment.
January 4, 2026 at 4:27 AM
Amazing, but Jesus Christ, please don't let Donny or RFK Jr. hear about these promising developments, lest they destroy them in the name of reckless disregard of human life and accomplishment.
Dear Senators @vanhollen.senate.gov & @alsobrooks.senate.gov — given the dangerous, illegal and inflammatory action by the President in instituting regime change in Venezuela, I ask you, in the strongest possible terms, to work with your colleagues in the house to impeach & remove him ASAP.
January 3, 2026 at 6:40 PM
Dear Senators @vanhollen.senate.gov & @alsobrooks.senate.gov — given the dangerous, illegal and inflammatory action by the President in instituting regime change in Venezuela, I ask you, in the strongest possible terms, to work with your colleagues in the house to impeach & remove him ASAP.
Reposted by Rob Patro
Fast, accurate construction of multiple sequence alignments from protein language embeddings https://www.biorxiv.org/content/10.64898/2026.01.02.697423v1
January 3, 2026 at 7:46 AM
Fast, accurate construction of multiple sequence alignments from protein language embeddings https://www.biorxiv.org/content/10.64898/2026.01.02.697423v1
While we've been somewhat successful in spurring adoption of alevin-fry/simpleaf for scRNAseq processing, an impediment Dongze brought to our attention (now working directly w/ many experimentalists) was the need for a nice QC report for it's output; hence QCatch www.biorxiv.org/content/10.1... 1/x
QCatch: A framework for quality control assessment and analysis of single-cell sequencing data
Motivation: Single-cell sequencing data analysis requires robust quality control (QC) to mitigate technical artifacts and ensure reliable downstream results. While tools like alevin-fry and simpleaf (...
www.biorxiv.org
January 3, 2026 at 6:18 PM
While we've been somewhat successful in spurring adoption of alevin-fry/simpleaf for scRNAseq processing, an impediment Dongze brought to our attention (now working directly w/ many experimentalists) was the need for a nice QC report for it's output; hence QCatch www.biorxiv.org/content/10.1... 1/x
Let us never forget that the Trump administration axed all funding for pet immortality research... and we were so, so close!
January 3, 2026 at 12:50 AM
Let us never forget that the Trump administration axed all funding for pet immortality research... and we were so, so close!
I overlapped with Brenna for several years at Stony Brook before we both moved. She does brilliant work and this is devastating to read and makes me even further ashamed of our scientific establishment.
www.nytimes.com/2026/01/02/s...
www.nytimes.com/2026/01/02/s...
She Wanted to Improve Genetic Medicine
www.nytimes.com
January 2, 2026 at 3:50 PM
I overlapped with Brenna for several years at Stony Brook before we both moved. She does brilliant work and this is devastating to read and makes me even further ashamed of our scientific establishment.
www.nytimes.com/2026/01/02/s...
www.nytimes.com/2026/01/02/s...
Reposted by Rob Patro
I was about to post my annual "Games I finished this year" thread, but I only finished 5 games; busy year! 😅 So I'll add audiobooks I finished as well!
1) Final Fantasy XVI
2) Baten Kaitos: Eternal Wings and the Lost Ocean
3) Clair Obscur: Expedition 33
4) Baten Kaitos Origins
1) Final Fantasy XVI
2) Baten Kaitos: Eternal Wings and the Lost Ocean
3) Clair Obscur: Expedition 33
4) Baten Kaitos Origins
December 31, 2025 at 5:43 PM
I was about to post my annual "Games I finished this year" thread, but I only finished 5 games; busy year! 😅 So I'll add audiobooks I finished as well!
1) Final Fantasy XVI
2) Baten Kaitos: Eternal Wings and the Lost Ocean
3) Clair Obscur: Expedition 33
4) Baten Kaitos Origins
1) Final Fantasy XVI
2) Baten Kaitos: Eternal Wings and the Lost Ocean
3) Clair Obscur: Expedition 33
4) Baten Kaitos Origins
Ok bioinformatics peeps. Anyone have any idea how to make ISMB paper time more fun by making an OUP template for @typst.app ?
December 31, 2025 at 5:51 PM
Ok bioinformatics peeps. Anyone have any idea how to make ISMB paper time more fun by making an OUP template for @typst.app ?
Have a really cool project cooking that should be ready to go in Jan. It started a couple of years ago with an exceptional REU student (Noah Cape, now at Princeton CS for his PhD), and we brought on an amazing UMD MS student (Elan Fisher) to do serious perf. engineering. Excited for this to land!
December 31, 2025 at 5:38 PM
Have a really cool project cooking that should be ready to go in Jan. It started a couple of years ago with an exceptional REU student (Noah Cape, now at Princeton CS for his PhD), and we brought on an amazing UMD MS student (Elan Fisher) to do serious perf. engineering. Excited for this to land!
Holy crap; Cleveland Browns; holy crap! Thank you!
December 28, 2025 at 9:06 PM
Holy crap; Cleveland Browns; holy crap! Thank you!
Ok, so why aren't the Ravens starting Tyler Huntley even when Lamar Jackson isn't ruled out? That game was incredible. And Henry is simply a generational talent... what a beast.
December 28, 2025 at 4:47 AM
Ok, so why aren't the Ravens starting Tyler Huntley even when Lamar Jackson isn't ruled out? That game was incredible. And Henry is simply a generational talent... what a beast.
If you value the craft of software engineering & want programmers to learn to respect/treat the craft as such, I find it hard to believe that designing a language like Go is aligned with those goals. Its simplicity to me seems not the simplicity of elegance, but the absence of well-learned lessons.
December 28, 2025 at 12:31 AM
If you value the craft of software engineering & want programmers to learn to respect/treat the craft as such, I find it hard to believe that designing a language like Go is aligned with those goals. Its simplicity to me seems not the simplicity of elegance, but the absence of well-learned lessons.
I love Rust, but I think my biggest growing gripe is its current lack of reasonable compile-time capabilities. TIL that you can’t use an associated const of a trait as a const generic parameter. It counts as a const generic “expression” even though it’s just a value. WTH?!
December 27, 2025 at 4:04 AM
I love Rust, but I think my biggest growing gripe is its current lack of reasonable compile-time capabilities. TIL that you can’t use an associated const of a trait as a const generic parameter. It counts as a const generic “expression” even though it’s just a value. WTH?!