raggi

37 ms·

421.
by raggi 3y ago
glibc, I’m too lazy to lookup if it removes the getenv lookup with posix build flags.
422.
by raggi 3y ago
there are lots of programs that access `*environ` directly, so while this might be good, it wouldn’t solve all classes of the problem. There are also uses out there which are performance sensitive (and often just as if not more unsafe, such
423.
by raggi 3y ago
Correct. libc is largely thread unsafe.
424.
by raggi 3y ago
getaddrinfo is far too ossified to be changed
425.
by raggi 3y ago
It’s neither in the headline or in the article. The question was about setenv, not getenv. It is best to avoid calling setenv in a threaded program. Some programs do it to make space for rewriting argv with large strings (freeing space from
426.
by raggi 3y ago
The language extensions feature is designed to provide these kinds of facilities is it not?
427.
by raggi 3y ago
Can we just get wgsl compute good enough and over the line instead, and do away with these moats?
428.
by raggi 3y ago
In purely practical terms the old system interfaces are sufficiently problematic that for any workload with necessarily smaller buffers than tens of kb, most implementations will get stuck being syscall bound first. Spectre really didn’t he
429.
by raggi 3y ago
Yup, the ffi cost is extremely painful. I work on a Go product that is performance sensitive and this exact problem is a constant source of aggravation
430.
by raggi 3y ago
every single comment extrapolating a trival example is a strawman
431.
by raggi 3y ago
you're right, the api for ordering a pizza will probably stay the same. the cooking process won't though. stuffed crust? add some stuff in the middle. square? add some stuff in the middle. deep dish? add some stuff in the middle.
432.
by raggi 3y ago
If you never have to write any tests, perhaps this is ok.
433.
by raggi 3y ago
We have an article that tries to provide a detailed answer to this question! https://tailscale.com/compare/openvpn/
434.
by raggi 3y ago
Tailscalar here: Tailscale servers never see your traffic in plain text. In the integration with Mullvad in particular, WireGuard connections are always direct from your device to the selected Mullvad exit node.
435.
by raggi 3y ago
This doc kinda needs to say "TCP" somewhere, as it's very focused on TCP concerns - which is useful, people are mostly using TCP. The default UDP tunings are awfully low and as such are notably missing.
436.
by raggi 3y ago
This largely depends on how the application is written and in particular what if any non-POSIX interfaces it uses. If you are looking to hit line rates with UDP, or looking to head well above ~1-10gbps with TCP, you're fast headed into
437.
by raggi 3y ago
This has been surprisingly quiet. I’d like to know more about the scope of the leak, was it just to a log file? An OSS repo?
438.
by raggi 3y ago
This step is what I was expecting you to talk about, and it has some tricky subtleties to get right, which is why I looked for it in the whitepaper. A trivial problem with a naive implementation is being able to perform presence proofs usin
439.
by raggi 3y ago
Based on the blog you referenced up thread: I upload a large document to your drive product from my workstation. I go to search on my phone. My phone needs to download the content in order to index it. My phone downloads the content from yo
440.
by raggi 3y ago
To put this in context, the trivial example of a user with a 15gb account, say you happened to be using s3 for storage. They buy a new phone, that costs you ~$1.50 that month, or 50% of your revenue at current pricing. They buy a new iPad a
441.
by raggi 3y ago
So if I have 15gb of email, I have to process all of that on every client bootstrap? How do you plan on scaling your service with respect to this problem? once you have a non-trivial user base with non-trivial data volumes this is likely to
442.
by raggi 3y ago
Without producing confirmation side channels and so on? Doesn’t seem that trivial- at least it needs more thought than assumption of simplicity
443.
by raggi 3y ago
The marketing site says there’s full text search. The paper doesn’t cover this. How is this being solved here?
444.
by raggi 3y ago
Signal doesn’t have a web client. Most of the stores it is distributed through have fairly strong resistance to compel orders.
445.
by raggi 3y ago
I don't buy (as sold) the argument that we need associative and commutative merges, and I believe that this constraint when applied by the software makes resolving merge requirements harder on the user, rather than easier. It's pu
446.
by raggi 3y ago
The intersection of probabilistic optimization and timing based side channels is a gift that will never ever fully go away. Everything _really fast_ which is approximately any very mature systems gear has probabilistic optimization in it no
447.
by raggi 3y ago
I do not, this is not my area of expertise. Cross integrations are ongoing, for example the model presented in https://acp.copernicus.org/articles/11/11707/2011/acp-11-117... (among others) was cross int
448.
by raggi 3y ago
https://www.cesm.ucar.edu/community-projects/lens2
449.
by raggi 3y ago
> If the Sulfur emissions policies were known for >5 years and discussed for >10 years why aren’t they part of mainstream climate models? Why do you believe that they are not?
450.
by raggi 3y ago
I use custom elements directly in a bunch of stuff I had roll and I find the composition and reuse advantage nice. I tend to not bother with template elements, they don’t help much. I also don’t bother with shadow dom or element styles, as

More ›