My email is eritead@dashborg.com. I'm based in London and interested in discussions related to web development. I wrote a book about delay insensitive circuit design whose landing page is at https://www.delayinsensitive.com. I've been teaching myself Go and AWS lately. My github is https://github.com/gueststar.
Another explanation is that it's too radical of an idea even for HN. I
see no reason a judge in a hypothetical anti-trust case against Apple
couldn't give it a choice between ceasing operations and open sourcing
its whole technology stack.
I've heard of people with stereo blindness able to perceive VR in 3D
thanks to the eye tracking. Do you have any idea whether it could
benefit a wearer with intermittent amblyopia ("lazy eye") by
dynamically realigning the images accordingly?
The scanner is an Epson ES-580W, but any model would be adequate that
can scan directly to a network folder by FTP or FTPS over wifi or
ethernet. Some are claimed to support SFTP but I think the copywriter
meant FTPS. They seem to follow the network protocol standards well
enough to work with Linux even without any special drivers. Although
the scanner interacts with my file server, I prefer to run
paperless-ngx on my desktop, so an incron job runs on the file server
to copy the scanned documents to the desktop if the latter is on line
but retain them otherwise, and a startup script on the desktop copies
the retained scanned documents from the file server whenever the
desktop boots.
I bought a high speed sheet-fed scanner to be used with paperless-ngx
based on a recent thread on HN. Before that my filing system for
financial statements and other important documents was verging on
meltdown and causing me considerable anxiety, but now that burden has
been lifted. The scanner is configured to send documents directly to a
home network file server that's always on and will store them until
the next time I boot my desktop rig, so it's nearly frictionless to
drop every piece of mail that comes through the door immediately into
the scanner, which has a box under it, instead of piling it up on my
desk where it used to stay for months. I wish I started this years ago
and I thank everyone who contributed to that thread. Honorable
mention: a printer with liquid ink tanks for £100 with thousands of
pages worth of ink has just replaced a £30 inkjet printer I got ten
years ago that made me cough up another £30 or £40 pounds every time
the cartridges ran out. It's expensive to be poor.
Correct me if I'm wrong about this, but according to one of the
author's cited references, the time needed for metastability to
resolve is theoretically unbounded. Why wouldn't every solution based on
introducing a fixed delay amount to kicking the can down the road?
While they're incentivizing me to learn the ropes and be more
efficient, I'm secretly incentivizing them either to respect my
privacy by accepting cash payments or to compensate me with a discount
for giving up some of my personal information. I wonder which of us
will learn the ropes sooner.
On a more general note, I prefer to tell people that I'm too old for
newfangled stuff like smartphones, because I prefer the reaction of
condescending reassurance to the alternative of hostility evoked by
saying I've made an informed decision not to own one.
Well-known game developer Jonathan Blow has a lot to say about software and technology more generally being in decline due to skills being lost when older generations retire [1]. In my own experience, writing code with pointers in it used to be a skill expected of entry level programmers, and getting the hang of it was something of a rite of passage. Around the time Java started being used as a teaching language, it became best practice to rely on the language's standard libraries instead of rolling one's own linked list or whatever, and I instructed university students accordingly. As things have developed, the design effort put into Rust seems to imply that pointers are now regarded as a grand challenge problem. These days I don't tie my own neck ties any more but wear the clip-on kind because the knots have been developed and tested by experts.
If you think anyone might care about your software in a hundred years,
explain it in writing in a technology-independent way that would make
it easy for anyone to re-implement it by reading your explanation,
along with a reference implementation in pure lisp, lambda calculus,
or the MIX architecture from Knuth's TAOCP. Self-publish your
explanation and reference implementation in a book printed on
acid-free paper, get an ISBN for it, and send the book to the Library
of Congress. Spread the risk by also sending it to other national
depository libraries.
Another advantage to building your own system not mentioned (unless I
missed it) is that pre-built systems might not be available with the
latest and greatest components because the vendors are limited to what
they can reliably source with a fast turnaround, whereas when building
your own you can be more patient and selective. Given the need for the
person who builds a pre-built system to get paid, you'll be getting a
lower spec than you could build yourself and maybe paying the same or
more for it. You might save even more down the line when you want a
component repaired or upgraded and you feel comfortable ordering and
swapping it in yourself instead of worrying you might break something
unless you pay somebody to do it. The only downside I've experienced
is that I'm still too scared to overclock it too much and would have
felt better about getting someone who's good at it to do that for me.
Startup idea: video content moderation service for big social media
companies with moderators given medically supervised Midazolam; no
more high turnover due to PTSD and no counseling needed
FWIW, Barr and Wells in a well respected reference on category theory
[1] use a different definition of a function than the one in this
book. Essentially they avoid insisting on the concrete model of a
function as a set of ordered pairs (which they call its graph), and
would regard two functions having identical graphs as being
nevertheless distinct if they have different co-domains (section
1.2). IMO their definition is better thought out and ultimately less
troublesome when it comes to reasoning about functions. The definition
of a function is always the first thing I check in any book about math
for a rough assessment of whether reading the rest of it might make me
dumber or smarter.
I'm a big fan of asynchronous design, though more as a spectator than
a participant in recent years, and I'm really glad to see a bold new
idea like yours because I was beginning to wonder whether the whole
field had been stagnating. If there's anything else you want to share
publicly about your work I'm all ears. I'm particularly curious
about how you've coped with arbitration or contention in on-chip
packet switched networks, or whether the issue ever came up. BTW,
cloudflare says your web site is not responding as of 20:39 UTC,
January 9.
> we don’t know if anything from archeology is “for ritual use (for real).”
I apologize for being unable resist thinking of a counterexample any
time someone makes a generalization. Would household utensils found in
a sealed burial chamber be a clear case of ritual use for real, or did
dead people in ancient times put them to some practical use we haven't
identified?
Why couldn't a relatively nearby galaxy recede from us faster than a
more distant one due to causes beyond the limits of our observation?
I'm thinking primarily of gravitational attraction to a massive object
past the boundary of our observable universe, but wouldn't dark matter
also imply in any case that all bets are off? What objection could
anyone raise to the idea of a big load of dark matter right next to
some galaxy making it recede faster?
Me too. According to the the way it's always explained to me it should
be impossible for airplanes to fly upside down sustainably because the force of
gravity and the lift should add to each other rather than cancelling,
yet they often seem to do that in air shows.
not to mention the film "I Care a Lot", with an admittedly outlandish plot but a disturbingly plausible dramatization of the legal aspects of guardianship
C gets out of your way and lets you do your job without buying into a
lot leaky abstractions and without going to annual C conferences. C
doesn't argue with you about whether something is a number or an array
of bits. C is blazingly fast compared to most alternatives. C enables
you to detect and handle heap overflow errors without panicking or
crashing (just in case your memory is finite).
On top of everything else, Cloudflare blocks Tor users or makes them
solve captchas, which won't work without javascript and cookies
enabled. It was making some noise at one point about having developed
some sort of privacy-preserving cookies for Tor users, as if that were
ever going to appease anyone other than its established apologists.
I'm not the only HN reader who claims to use a different email alias
for each recipient, but mail user agents require me to remember to set
the return address manually to the right alias on each outgoing
message and risk leaking the wrong one if I forget. Some kind of email
middleware that maintains a mapping from recipients to aliases and
automatically slaps the right alias onto each outgoing message seems
like it should be feasible for a single developer, but I can never
find the motivation to do it myself. Apologies if I'm overlooking some
well known application or service that already does.
Strategically the best thing you can do career-wise is to write
something with multiple chapters on a common theme but with each chapter
understandable separately with a minimal introduction, get each chapter
published as a separate conference paper, and then stick them all
together into a thesis. Then you not only have a publication list all
ready by the time you graduate, but you make it awkward for the
examiners to find fault with your thesis given that it has already
passed peer review. You'll probably have to pick the trendiest topic
you can find even if your interests lie elsewhere, but you'll be in a
better position to vary your research agenda when you're more
established in your career. This advice is the opposite of what I was
given, which is that a Ph.D. is a time to explore your own interests
and then get serious later.
I don't know how scientifically reliable this article is, but it gives
me an idea. Could tsunami avoidance be an ideal use case for a jet
pack? It doesn't need a lot of altitude or more than a few minutes of
flight time, just easy rapid deployment without an airstrip. Some
startup should be all over this. A combined package with an alarm
triggered by an ultrasonic sensor would make a great sales pitch to
well heeled customers living in that region. It might even lead to
beneficial government policy changes by raising awareness of the
danger when everybody who's not rich enough to afford one whinges
about it on facebook.