mgreenbe

8 ms·

1.
by mgreenbe 6y ago
I'd say it's conventional to reuse an existing parser to generate facts. The AST point is a subtle one. Classic Datalog (the thing that characterizes PTIME computation) doesn't have "constructors" like the ADTs (alg
2.
by mgreenbe 6y ago
I think the solvers _are_ tuned for PP. But we're comparing CSA and PP on the queries that Formulog issues... which don't really match well with the DFS discipline that the PP stack aligns with. I think CSA beats PP in our experim
3.
by mgreenbe 6y ago
Datafun is quite cool! Formulog and Datafun seem similar---both combine logic programming and pure functional programming---but they take wildly different approaches. Datafun is a foundational re-imagining of what a Datalog could look like:
4.
by mgreenbe 6y ago
We do use incremental solving. check-sat-assuming is generally better than push/pop, though, because Datalog's bottom-up search isn't DFS. If you're interested, check out our ICLP 2020 extended abstract: https:/&#x
5.
by mgreenbe 6y ago
Thanks! :) We should be very clear that the bulk of the work is Aaron Bembenek's. I think Formulog would work great for analyzing the shell---as would any other Datalog, though SMT-based string reasoning will certainly come in handy. I
6.
by mgreenbe 6y ago
See also: http://siek.blogspot.com/2012/07/crash-course-on-notation-in...
7.
by mgreenbe 17y ago
Fair enough. I would play down the significance where you played it up. Considering the equanimitous tone of the whole article, it seems unfair---or, at least, ungenerous---to take him to task for a colorful, one-off analogy.
8.
by mgreenbe 17y ago
Dynamically typed languages are automobiles and statically typed are bicyles?? Not exactly a neutral comparison and hypocracy in light of his complaints against the strong vs. weak labels. Good thing he didn't make it. Ple
9.
by mgreenbe 17y ago
First, Coq isn't an automatic theorem prover, it's a proof assistant. Adequacy is an issue, but (particularly) for computer-science applications it's sometimes possible to use the Coq artifact as the implementation directly. Adam Chilpala
10.
by mgreenbe 17y ago
A neat trick, but---as one of the commenters points out---manual dictionary passing is a bit much. I would love Haskell, but for laziness...
11.
by mgreenbe 17y ago
Good advice. Transaction costs and, in particular, startup costs are some of the most difficult things I face in my (graduate student) day-to-day. Simply getting up to the whiteboard can be difficult. I'm definitely going to start using a
12.
by mgreenbe 17y ago
You forgot to add the argument that it's (strictly speaking) more accurate, since Yehuda is just one of the tribes. But s/Hebrew/Colored/. Same arguments apply, but---here we are.
13.
by mgreenbe 17y ago
There's one particular thing about Christianity though. Well OK, a couple of things, one it's kind of a modern less complex flavor of Judaism, no offense meant to any Hebrews reading this. First, the similarity between the
14.
by mgreenbe 17y ago
There are no French programming languages, not even Pascal or Eiffel uses French words for the syntax, and all the documentation is in English. OCaml is implemented by le French , and I've seen code that uses French variable
15.
by mgreenbe 17y ago
Original article: http://www.nature.com/neuro/journal/vaop/ncurrent/abs/nn.238... I was inclined to be skeptical---breaking news, shocking the brain during sleep impairs performance---but they used three different groups of rats: ...w
16.
by mgreenbe 17y ago
AOP was the next big thing five years ago. I've never heard of a serious, non-dog-food project written in any (pointcut-rich) implementation of aspects. The fundamental problem with aspect-oriented programming is that it makes it impossibl
17.
by mgreenbe 17y ago
To be fair, Astarte and Ashtoreth are probably the same---just a vocalization error. :)
18.
by mgreenbe 17y ago
Python is just JS with extra syntax? The languages have very different object and meta-object/overloading models, as well as other divergent features: JavaScript can mutate closures, for example. I agree, though, that a stdlib should be pa
19.
by mgreenbe 17y ago
Did you not read the talk? You can provide the car/cdr interface on top of tree-based implementations. Yes, you pay a cost to use that interface...but you pay a different (Guy Steele would argue more expensive) cost if you write your algo
20.
by mgreenbe 17y ago
Your fingers are going to be crossed for a long time. I think the Steering Committee has too many divergent interests on it for there ever to be agreement on a serious standard library. Schemers are very DIY; for example, see http://phil
21.
by mgreenbe 17y ago
Neat. I'm not familiar enough with typeflow analysis to know how practical it is. Can it handle polymorphic functions, e.g., map ? What about dependent functions (just kidding)? Is there a way to annotate functions so Pyntch will give
22.
by mgreenbe 17y ago
This is very nice---it catalogs a number of standard continuation tricks in a manner clear enough (I think) to enlighten those who barely know what continuations are in the first place. As a bonus, the devil/angel/milestone idea is new to
23.
by mgreenbe 17y ago
Or any of the other soft/partial/gradual typing work that's been ongoing since the 80s. Check out Jeremy Siek's papers on gradual typing for some of the latest. Also Tobin-Hochstadt and Felleisen's work on Typed Scheme. On the one hand, t
24.
by mgreenbe 17y ago
A neat trick, but also one of the better arguments for more principled macro systems. These macros are fragile, and the error messages for misuse must be inscrutable. So: cool, but I hope I never have to use it.
25.
by mgreenbe 17y ago
Settling on language-wide naming conventions reduces cognitive dissonance. Naming conventions don't supplant reading library docs, but they support it. In any case, you have to pick something , and some people will always be dissatisfied
26.
by mgreenbe 17y ago
On the one hand, the physicist counterpart sets a high conversational bar. On the other hand, even in academia the "computer science is fixing computers" misconception and the lack of awareness of theoretical aspects to computer science ar
27.
by mgreenbe 17y ago
An excellent article! I found the idea of "mind creationist" intriguing. While I'm not a creationist, I think I have mind creationist leanings....which now I feel obligated to question.
28.
by mgreenbe 17y ago
I must say, I'm really impressed. My first search, a curveball, was "gong bao ji ding". Found, and a good recipe, too! A follow-up search for "kung pao chicken" gave a longer list, but all on topic. A search for "chocolate tart" yielded
29.
by mgreenbe 17y ago
Reminds me of Cavafy: http://www.poetry-chaikhana.com/C/CavafyConsta/GreatYes.htm Though I think he understands the trade-offs better than the OP. And, conspicuously, the decisions described mostly rest with and affect the decision maker
30.
by mgreenbe 17y ago
PHP, welcome to the early 80s. Who knows, soon you might have real datatypes! Does the Zend engine do TCO? When I search, Google thinks I mean "Total Cost of Ownership". All I can find is http://devzone.zend.com/article/1235 .

More ›