titzer

46 ms·

781.
by titzer 3y ago
Capability security is great, but it isn't native to the Linux kernel. It must be implemented on top. WALI allows that capability security model to be implemented in a Wasm module, rather than built into the Wasm engine. Thus a bug in
782.
by titzer 3y ago
This is not a competing standard, it's a complementary, lower-level interface that can be used to implement (and sandbox) the WASI implementation. It's defense-in-depth and layering: WASI moves up a layer, and not every engine nee
783.
by titzer 3y ago
This is the most common misconception of our work. Please read the paper. WALI doesn't have to be exposed directly to programs, but can be used to implement a more restrictive sandbox as a Wasm module a layer above. For example, WASI c
784.
by titzer 3y ago
> It's all microcode these days. I know we're splitting hairs here, but really the only difference between microcoded instructions and "native" instructions is that the instruction decoder directly produces u-ops ver
785.
by titzer 3y ago
Testing, fuzzing, and verification are all pretty hot SE topics now, at least in my department at CMU. All of those have some element of program analysis, both static and dynamic, so there's potentially deep PL stuff, which is core CS.
786.
by titzer 3y ago
If you're specifically referring to CMU's compilers course, feel free to follow up with me offline.
787.
by titzer 3y ago
CMU constantly reevaluates its MSE program with input from many different angles. I've participated here and I think we're trying hard to balance important foundational knowledge with practical skills of the day. I don't thin
788.
by titzer 3y ago
I agree with you! It's hard to change curricula because there are so many competing interests. CS is an evolving field and things like machine learning have burst onto the stage, clamoring for attention. There is also an age-old debate
789.
by titzer 3y ago
We do teach these things, they are just not core CS topics, but rather in other areas, relegated to electives like a software engineering course. At CMU we have entire Master's program for software engineering and an entire PhD program
790.
by titzer 3y ago
Yeah, and ads on Google search have the teeniest, tiniest little "ad" chip on them, a long progression of making ads more in-your-face and less well-distinguished. In my estimation, given the context around AI-generated content an
791.
by titzer 3y ago
I say ban all the plastic. If even 1% of is disposed of improperly , you get a deluge of plastic litter that accumulates in the environment. I've picked up hundreds of bags of this stuff, personally. "Properly disposed of" b
792.
by titzer 3y ago
Note also that paper filters are recyclable and compostable.
793.
by titzer 3y ago
Nine times out of ten, the person saying this will turn around and complain about all the "political hacks" running things, referring to political appointees with no experience or background in the area of government they are task
794.
by titzer 3y ago
> But gagging a company from disclosing untargeted or semi-targeted surveillance, especially if it involves American citizens, seems like it should be unconstitutional on free speech grounds. I see you have not read the Patriot Act, an O
795.
by titzer 3y ago
> So the thing people actually care about is the risk of bad outcome, not the actual data itself. It's more than that. Privacy affects the psychological context for daily activity and alters behavior, sometimes subtly, sometimes ove
796.
by titzer 3y ago
The proper reason to give for privacy is: I don't need to have a reason for privacy; you're the one that needs to justify violating it.
797.
by titzer 3y ago
Steal a loaf of bread, do hard time. Steal $5 here and there millions of times over from unsuspecting customers with completely made-up junk fees, and it's an FTC fine...after years of haggling in court. The magic of indirection!
798.
by titzer 3y ago
Sure, you can still have safe code with irreducible loops. Safety is orthogonal. There is a proposal to add "multi-loop" ( https://gist.github.com/conrad-watt/6a620cb8b7d8f0191296e3eb... ) to Wasm, but it hasn&
799.
by titzer 3y ago
It's not just that it's more restrictive, it's that it forces the producer to do work for the verifier. Structured control flow follows a stack discipline and also pre-declares all labels, so the consumer (verifier) can use t
800.
by titzer 3y ago
Irreducible loops (loops with more than one way to enter) are not representable by structured control flow. It's possible to model with with state variables (e.g. loop over switch). Incidentally if irreducible control flow is modeled t
801.
by titzer 3y ago
Structured control flow is more efficient to verify, more compact, and disallows irreducible loops by design. Also note that "br" "br_if" and "br_table" are all just text format mnemonics. You can think of the
802.
by titzer 3y ago
That's impressively fast. Well done!
803.
by titzer 3y ago
Thanks for the reference. I think at this point I kind of want to just build my own UI to playback GP files. I looked into the guts of TuxGuitar and was pretty put off by how much code there is. It's kind of a hot mess. Really I want t
804.
by titzer 3y ago
I hate what ultimate guitar has become. Before it evolved into the current ad-laden, walled-garden rat trap that it is now, it basically was just a search engine for tabs. What amazes me is that so many hobbyists made Guitar Pro tabs in t
805.
by titzer 3y ago
When there is such a huge scale difference between the entity that causes harm and the person/group harmed, it just doesn't register. E.g. if you wanted down the sidewalk and inadvertently stepped on a cockroach because you were t
806.
by titzer 3y ago
> the point of the article is that you don't have to pay for the stuff you don't need. I kind of don't get this attitude, TBH. A bad register allocation decision could cost 5% in the wrong circumstance, yet no one is sugg
807.
by titzer 3y ago
Thanks for the reference. Several of the examples given could benefit from loop versioning and/or iteration space partitioning, i.e. the compiler splitting a loop that it cannot prove is always in-bounds to an in-bounds part and a poss
808.
by titzer 3y ago
> we've got safety at no cost No one ever said that safety was zero-cost. The article mentions bounds checks, but gives no measurements. Have they ever measured? Most data I've seen is that bounds checks cost less than 1% in re
809.
by titzer 3y ago
> Are we talking about safety against hackers? The large majority of C++ applications are non-public facing. That is like saying that most screws aren't load bearing, so it's fine if they fail 30% of the time. The way engineeri
810.
by titzer 3y ago
> 2. Microsoft can quite seamlessly pick up the ball and commercialize GPTs like no tomorrow and without restraint. And while there are lots of bad things to say about microsoft, reliable operations and support is something I trust them

More ›