|
Kristopher Micinski
@
krismicinski
Syracuse, NY
|
|
Asst. CS prof, @EngineeringSU. Programming languages and security. Always looking for students (email me!). Formerly @haverfordedu
|
|
|
29.473
Tweetovi
|
1.134
Pratim
|
2.314
Osobe koje vas prate
|
| Tweetovi |
| Kristopher Micinski proslijedio/la je tweet | ||
|
Otto The Orange
@TheOttoOrange
|
30 min |
|
Fuzzy socks ☑️
Hot chocolate ☑️
The BEST sledding hill EVER☑️☑️☑️
Stay cozy today friends!! pic.twitter.com/tnhDTyJbSf
|
||
|
||
|
|
Kristopher Micinski
@krismicinski
|
3 min |
|
🤣🤣🤣 genuinely wondering if this tweet is sarcasm
|
||
|
|
||
|
|
Kristopher Micinski
@krismicinski
|
4 min |
|
looking forward to see what you come up with!
|
||
|
|
||
| Kristopher Micinski proslijedio/la je tweet | ||
|
Adam T. Geller
@atom_geller
|
7 h |
|
Bad PL joke time: "Why did the chicken cross the road?"
I can't tell you the answer because I don't know how to typeset derivations in Twitter.
|
||
|
|
||
| Kristopher Micinski proslijedio/la je tweet | ||
|
Andrew Myers
@AndrewCMyers
|
51 min |
|
Just as the culture of "ship insecure code and patch it later" has gotten thoroughly established, people are shipping bazillions of IoT devices that won't be patched ever.
|
||
|
|
||
|
|
Kristopher Micinski
@krismicinski
|
5 min |
|
.. so this was a teaching thread originally ... ;-)?
|
||
|
|
||
| Kristopher Micinski proslijedio/la je tweet | ||
|
Rolf Rolles
@RolfRolles
|
12 h |
|
IMO the further the work is from current practice, the less sense it makes to focus on performance. E.g. King's original 1975 work on symbolic execution was impractical before advances in constraint solving in the late 90s/early 00s. Sometimes practice has to catch up to theory.
|
||
|
|
||
|
|
Kristopher Micinski
@krismicinski
|
1 h |
|
I have always wondered about this..
|
||
|
|
||
|
|
Kristopher Micinski
@krismicinski
|
9 h |
|
also: seen this? pages.cs.wisc.edu/~loris/papers/…
|
||
|
|
||
|
|
Kristopher Micinski
@krismicinski
|
12 h |
|
I broadly agree with you tough
|
||
|
|
||
|
|
Kristopher Micinski
@krismicinski
|
12 h |
|
But most analyses today have algorithmic walls: uniform k-CFA is exptime complete!
|
||
|
|
||
| Kristopher Micinski proslijedio/la je tweet | ||
|
Adrian Herrera
@0xadr1an
|
12 h |
|
I think Peter O'Hearn's work with Facebook's Infer is another counter-example, where scalability was important from the get-go
cacm.acm.org/magazines/2019…
|
||
|
|
||
|
|
Kristopher Micinski
@krismicinski
|
13 h |
|
i guess now it's time to train another generation to sass.
|
||
|
|
||
|
|
Kristopher Micinski
@krismicinski
|
13 h |
|
hey hey that's my job!
|
||
|
|
||
|
|
Kristopher Micinski
@krismicinski
|
13 h |
|
if you want a counterexample: dl.acm.org/citation.cfm?i…
|
||
|
|
||
|
|
Kristopher Micinski
@krismicinski
|
13 h |
|
Static analysis research is not driven nearly experimentally enough. So many papers propose cute ideas that do not prove effective when done at scale. No wonder, it's really hard, but still, I wonder things will change when we can rapidly test analyses.
|
||
|
|
||
|
|
Kristopher Micinski
@krismicinski
|
16 h |
|
yes, the testing paper is what I wanted!!
|
||
|
|
||
|
|
Kristopher Micinski
@krismicinski
|
16 h |
|
that could be great. If you can remember which, it could be really cool.
|
||
|
|
||
|
|
Kristopher Micinski
@krismicinski
|
16 h |
|
I'm basically using AI to mean SA here, but yeah, I think less of a focus on technical properties of abstract domains would be good if possible.
|
||
|
|
||
|
|
Kristopher Micinski
@krismicinski
|
16 h |
|
giving this to someone who uses a lot of LLVM, so technically interesting abstract domains are not really what I'm looking for. More like, papers that use abstract interpretation to do things like instrument programs for security, etc..
|
||
|
|
||