Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
dom0
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
47 ms
·
541.
▲
by
dom0
9y ago
I don't think you are; this story condensed into a 150 second clip would fit great into sketch comedy (Mitchell and Webb?).
542.
▲
by
dom0
9y ago
m4 is orders of magnitude slower (notice the different unit). If you write the Python script using bytes (not Unicode), then Python 3 is faster than Python 2, at least for me. Python 3 is 20 % slower than GNU yes and Python 2 35 % slower th
543.
▲
by
dom0
9y ago
Hot snot gets icky with time, rather use a conformal coating; they're available in spray-form as well (e.g. CRC Urethan or Plastik 70). That being said FR4 is a really tough material and it's quite difficult to pry SMD parts off.
544.
▲
by
dom0
9y ago
That is, unless you're on Windows.
545.
▲
by
dom0
9y ago
Signed overflow is IB, not UB.
546.
▲
by
dom0
9y ago
> It's worth comparing to other languages. There are known ways to segfault PHP and most other scripting languages, but those bugs aren't considered security problems unless they meet all three criteria above. With Python it&#x
547.
▲
by
dom0
9y ago
I'd argue that this depends a lot how the code is structured. Back in the day of Win32 and MFC it was not unusual at all to very tightly integrate logic and the UI, often even outsourcing parts of state machines to GUI controls etc.; s
548.
▲
by
dom0
9y ago
> Qt is an easy sell on fresh projects, but switching a large existing codebase is huge. There have been a lot of major projects ported to Qt over the years, both commercial and open source (e.g. Wireshark, Subsurface).
549.
▲
by
dom0
9y ago
(from linked resources) > Keccak won the SHA-3 competition, and became the FIPS 202 standard on August 5, 2015. All the other contenders lost. I don't think many professionals would agree with this view of cryptographic competitions
550.
▲
by
dom0
9y ago
AMD and Intel have agreements (mainly due to x64).
551.
▲
by
dom0
9y ago
The structure of a problem specifies it; an unstructured problem is less well-specified, and that makes it intuitively more difficult to approach. E.g. summation of a column of decimal numbers is a highly structured problem that's very
552.
▲
by
dom0
9y ago
At least some of these devices don't have that option.
553.
▲
by
dom0
9y ago
Python has had closures for a long time, which make great ad-hoc factories that can capture state from upstream functions without making state visible along the way.
554.
▲
by
dom0
9y ago
Well, pure DI also has practical downsides. E.g. anything that wants to log something has to be able to access the logging configuration — so it needs to be passed into everything.
555.
▲
by
dom0
9y ago
Technology isn't intrinsically good or evil. It's how it's used, like the death ray.
556.
▲
by
dom0
9y ago
RSTDISABLE means that no in-circuit programming is possible any more and for re-flashing a parallel HV programmer is required; it's quite annoying for tinkering.
557.
▲
by
dom0
9y ago
Does Rust support FMV (function multi-versioning)?
558.
▲
by
dom0
9y ago
const does not imply pure. If applied to a member function, it makes this const; if applied to types it makes types const. Purity is a different concept. The closest match would probably be constexpr.
559.
▲
by
dom0
9y ago
(attic) borg check --repository-only does not transfer any data except informational logs. This is CRC32 only [1]. Borg 1.1 beta has borg check --verify-data, which does full decryption & full MAC + ID checks -- by downloading all data.
560.
▲
by
dom0
9y ago
"bup - I can't believe no-one had this idea earlier!"
561.
▲
by
dom0
9y ago
Yes, that's right, concurrent addition of the same chunks would generally mean that some work is wasted; so concurrent long running jobs would not synchronize well in this model, and lock-free performs clearly better there. The only
562.
▲
by
dom0
9y ago
The verdict of the "open source competition" in Duplicacy's README is not entirely accurate. Exclusive locking in the sync'd approach is just the easiest implementation, not the sole possibility. I can't speak for o
563.
▲
by
dom0
9y ago
Not related at all. (Duplicacy != Duplicity) Duplicity is a pretty straight good old-fashioned incremental backup program. Duplicacy on the other hand is hash-based deduplication (BorgBackup / Attic, Restic etc. are some others). The d
564.
▲
by
dom0
9y ago
There is the latent question of how this works with B&W printers which are still frequently found in offices. I assume that all of these have a similar technique, just sans the colour. I did notice some scattered black dots on a Kyocera
565.
▲
by
dom0
9y ago
In a Person of Interest episode an ex-Stasi officer was amazed by surveillance cameras, "These small cameras ... they're extraordinary. The Stasi would have killed for this technology!" Fun-fact: The Stasi installed Caesium-b
566.
▲
by
dom0
9y ago
I heard that before, I think it's a tactic to bully people into buying more ink they don't need. Ensuring tracking dots is maybe just an unfortunate side effect .
567.
▲
by
dom0
9y ago
No printer controls their motors with >600 dpi resolution. Inkjet printers have print heads with many nozzles; the motors do the rough positioning / slide the head over the paper, the nozzles do the hard work. In laser printers a mo
568.
▲
by
dom0
9y ago
MICs are usually mentioned in the same sentence that also explains why you can't print nor scan money. The big Wikipedias have long articles on it with many sample images: https://de.wikipedia.org/wiki/Machine_Iden
569.
▲
by
dom0
9y ago
Imgur very much does, with a visible loss of quality.
570.
▲
by
dom0
9y ago
macOS does not seem like an OS that has been optimized much at all in regards to the CPU. Also, Apple dragged macOS from PPC to x86; switching x86 implementations is no big deal for the rest of the world, it surely should not be a challenge
More ›