Tweetovi

	Rohan Padhye @moarbugs	20 h
	Odgovor korisniku/ci @koronkebitch @heathercmiller i 2 ostali I used to do this too! I think I initially shrugged it off as a possibly Americanized spelling of "depreciate". It wasn't an unreasonable assumption, given that this was around the time when I would constantly break my CSS while misspelling "color".
	Prikaz razgovora ·

	Rohan Padhye @moarbugs	5. velj
	Odgovor korisniku/ci @clegoues Tangentially, this reminds me of The Independent's genuine confusion about Whatsapp's group chat limit of 256 users, and the (unsatisfactory?) resolution thereof. pic.twitter.com/ImuoiNHZVe
	Prikaz razgovora ·

	Rohan Padhye @moarbugs	4. velj
	Odgovor korisniku/ci @krismicinski @DrPaulRalph I am not necessarily saying that *current* abstracts meet this bar, but I feel like judging relevance should require reading far less than ~12 pages. That said, I've read many more accepted papers than ones under review, so perhaps I am lacking counter-examples.
	Prikaz razgovora ·

	Rohan Padhye @moarbugs	4. velj
	Odgovor korisniku/ci @krismicinski @DrPaulRalph Right, but then that's a direct fail on evidence to back claims. Completely different reasoning there.
	Prikaz razgovora ·

	Rohan Padhye @moarbugs	4. velj
	Odgovor korisniku/ci @krismicinski @DrPaulRalph Here's a claim: Relevance and interestingness can be judged from the abstract alone. For a reviewer, downloading the paper PDF = checking a box that says "I agree that this paper is worth reading in full".
	Prikaz razgovora ·

	Rohan Padhye @moarbugs	4. velj
	Odgovor korisniku/ci @krismicinski @DrPaulRalph @v0max Of course the PC should arbitrate! To clarify: I meant that reviewers need not be the direct beneficiaries of the research. The hard task of the reviewer is to judge whether this work may be interesting and relevant to *someone* in the community (often: a mythical "user").
	Prikaz razgovora ·

	Rohan Padhye @moarbugs	3. velj
	Odgovor korisniku/ci @krismicinski @DrPaulRalph @v0max Right. I do think there is a burden of proof on authors to convince reviewers that a paper is relevant to and may benefit some subset of the publication's readership community. But the reviewers need not be in this subset.
	Prikaz razgovora ·

	Rohan Padhye @moarbugs	3. velj
	Odgovor korisniku/ci @krismicinski @DrPaulRalph @v0max Interesting and relevant to whom?
	Prikaz razgovora ·

	Rohan Padhye @moarbugs	31. sij
	Odgovor korisniku/ci @vj_chidambaram Oh wait, I just saw the following reply. I can totally relate to this. twitter.com/danielbittman/…
	Prikaz razgovora ·

	Rohan Padhye @moarbugs	31. sij
	Odgovor korisniku/ci @vj_chidambaram Is it possible to be somewhere in the middle? If I now consciously try to think how I think, then yes, there are words in my head. But I don't remember thinking in words for the last couple of hours before I read this tweet.
	Prikaz razgovora ·

	Rohan Padhye @moarbugs	31. sij
	Odgovor korisniku/ci @jayashree2912 Thank *you* for teaching me something new! I did not know anything about filesystem crash-consistency before this, but now I really want to go fuzz all the things!
	Prikaz razgovora ·

	Rohan Padhye @moarbugs	31. sij
	Odgovor korisniku/ci @moarbugs CrashMonkey and Ace are available here: github.com/utsaslab/crash…
	Prikaz razgovora ·

	Rohan Padhye @moarbugs	31. sij
	An *excellent* talk on automated testing of filesystem crash consistency. If you are into fuzzing/testing but don't know anything about filesystems: this talk explains the problem in depth, shows why it's *very* hard, and presents an automated system that finds real bugs! twitter.com/jayashree2912/…
	Prikaži podatke ·

	Rohan Padhye @moarbugs	30. sij
	Odgovor korisniku/ci @basus If your IDE is Web-based and you want something lightweight, you can checkout Ace: ace.c9.io. I've used Ace to implement syntax highlighting and error marking on chocopy.org, but it appears to be more programmable than that.
	Prikaz razgovora ·

	Rohan Padhye @moarbugs	30. sij
	Odgovor korisniku/ci @basus Oh, I read "custom language" instead of "custom IDE". This may not actually be that relevant if you want to build your own UI.
	Prikaz razgovora ·

	Rohan Padhye @moarbugs	30. sij
	Odgovor korisniku/ci @basus If you write a spec in the Language Server Protocol (microsoft.github.io/language-serve…), you could potentially hook it up to multiple editors: langserver.org
	Prikaz razgovora ·

	Rohan Padhye proslijedio/la je tweet
	M. Umar B. Niazi @umar_n11	29. sij
	Odgovor korisniku/ci @SigProcessing "For instance, we store E=mc^2 for Einstein's face."
	Prikaz razgovora ·

	Rohan Padhye proslijedio/la je tweet
	ShriramKrishnamurthi @ShriramKMurthi	29. sij
	And this is why we need basic research. Someone is out there today is coming up with answers to questions you'll have tomorrow. twitter.com/mariagoneviral…
	Prikaži podatke ·

	Rohan Padhye @moarbugs	27. sij
	Odgovor korisniku/ci @krismicinski @peter_a_goodman Also makes describing dynamic analysis algorithms in a paper quite straightforward.
	Prikaz razgovora ·

	Rohan Padhye @moarbugs	27. sij
	Odgovor korisniku/ci @krismicinski @peter_a_goodman This is actually a common representation used in dynamic analysis. Local variables are just fields of stack-frame objects! The call stack is just a linked list with the saved frame ptr as the "next" link. Programs mutate a graph of objects rooted at the stack pointer.
	Prikaz razgovora ·