I don't really know that this is avoidable without buckets of work and probably legal issues on behalf of core's (or anyone's) engineers- it's really just something that plagues hardware in general.
Hell, lots of sensors/etc these days are running fairly complicated software that's totally opaque.
I don't think _anyone_ who's buying the new pebble watches is to some degree not interested in software, and probably pretty interested in open-source community work. It's a wildly niche userbase, and this sort of thing is going to put crazy pressure on Eric and co, I imagine.
Still keeping my preorder, but damn dude this kinda sucks.
> restart required a hardware security module (HSM) smart card.
Out of curiosity, does anyone know why? My guess would be the PW DB would be encrypted with some token generated from this card.
I've had lots of "I have a secret and the server needs it" type problems but I've never been very happy with my solutions- smart cards seem like potentially an elegant solution.
This article highlights exactly why a HSM may be potentially elegant, but also really really dependant on embedding the process for using it in your operational processes (which would include performing that operation regularly to ensure it still works and that knowledge of its use is retained).
For a 'best effort' hosted internal service, this is not a good choice.
Figure 3 from the report- that's an Adafruit sensor module on a 3d printed bit of plastic with a teensy-brand microcontroller just sitting in there! Actually, the entire electronics enclosure appears printed.
Very funny to see in what I assume is a million-dollar product.
This is a camera outside the sub, and the electronics are inside sealed enclosure. If that had any leak for water to get in some conformal coating isn't going to save you, it will get pancaked.
Of course random arduino module and teensy used for product is amateur hour, even for low volume production. They must have crazy margins on that camera and producing custom board is very cheap.
>They must have crazy margins on that camera and producing custom board is very cheap.
It's expensive in time and expertise to do a custom board, and to debug a custom board. All to what, save 20$ on a bom which might not even be 1% of the profit per unit?
Far more efficient to just ship the dev board. They could have perhaps picked a better dev setup to start with, but if it looks stupid and it works...
Point isn't to save BOM cost, margins are high already and they can eat into them.
Unless they are making literally less than 10 of those, custom board will be easier to manufacture than that mess of dev boards and more reliable than random wires and headers all over the place. Plus they can spend money where it makes sense, using better regulators etc.
Only if you happen to have the relevant skill set.
Speaking as someone capable of designing the mechanical hardware and who is broadly electrically savvy but who is most definitely not an embedded engineer: I could bang out a few hundred hacked together dev boards in week, but doing a custom board would take me a few months. Starting with reading 'Prototyping PCBs for Dummies'.
Also it's actually simpler than it might seem, especially if you are not doing anything high-power or high-frequency (which, again, if it's a bunch of breakout boards connected over 0.1inch headers they clearly aren't).
Watch a few YT videos, copy-paste the reference schematics for all those boards, delete what you don't use and you are almost done :)
It's a bit silly IMO, but USB PD EPR _can_ support 24v and 48v- for charging laptops, I believe. The day I see a server rack with a pair of USBCs plugged into it is a far day off I hope though.
Didn't know, thanks for the info... I still have to see a power supply capable of delivering those voltages on USB-C though. All the ones I've seen can output 5V, 12V and 19V.
The dude tests basically anything and everything, and gives them generally pretty reasonable goes at a proper test. Sure, it's far off from an industry standard in a lot of cases, but unlike the industry (generally) he does side by side comparisons.
I've learned from watching vintage computer repair videos on youtube that touching chips on a running system and looking for a hot one is a common troubleshooting technique.
Fun fact if you take a board out of a Burroughs B6700/B6800 that you think is bad (has a bad chip) if you drop it flat on the floor, chips side up from 36" (about belt height) 9 times out of 10 the lid on the bad chip will fly off. Turns out the heat from a failing chip heats enough to release the glue.
It's a bit of a convolution, but let's say we've got some iterator of unknown size and we want to have a prime each iteration.
We can do the following in python, which saves us from having to pre-compute the primes _or_ know how long `some_iterator` is.
primes = gen_primes()
for item, prime in zip(some_iterator, primes):
Please god no. So much info is already getting dumped to discord and nowhere else making searching or archiving impossible, we don't need more of that.
Discord no doubt keeps every conversation ever typed on their servers. they might toss out something huge like a movie file but text compresses -very- well.
Sending something to discord is like sending it to /dev/null except that it is also diverted into the mailbox of Big Brother, with you having only very limited access to it on their whim.
Every time I see tailscale do something really neat I'm always a little disappointed to find out they still offer only the three auth schemes- and I really don't want to tie my networking to google/github/ms. On top of the various tinfoil hat reasons, I know a variety of people who have had these accounts terminated out of the blue, and it throwing out my networking stack would be insanely aggravating.
If you're reading tailscale, I will pay you actual real dollars per month to offer a different not-tied-to-a-megacorp authentication scheme. Till then, guess I've got headscale.
Well god damn there it is! Three days fresh, even! Thanks!
Looks like a fair lot of work to get it configured, but few good things come entirely free. Wonder if there's enough people that could get together for a communal one...?
Got to the end of that post and thought: definitely don't want to self host that!
Are there good options for an IdP that has good data policies that are easy to wire in with tailscale? I'm not opposed to paying for it. I wonder if Zoho can do this for me, I'm very happy paying them $12/yr for email.
Question about the docs, it mentions that "The WebFinger endpoint must be hosted at the domain of the email address provided during setup". Would it be possible to support a subdomain?
Also, a small ask: could the webfinger request that's sent include the `rel` and a well-known user resource params, for the situations where there's already a webfinger implementation there that isn't 100% under dev control which requires these params
like
GET /.well-known/webfinger?
resource=tailscale-webfinger%3A%40mydomain.com&
rel=http%3A%2F%2Fopenid.net%2Fspecs%2Fconnect%2F1.0%2Fissuer
HTTP/1.1
Host: mydomain.com
lastly, is this request resent at every auth event?
Are there really Microsoft accounts that were terminated out of the blue? I always had the feeling they acted a bit more responsibly around that than Google.
Hell, lots of sensors/etc these days are running fairly complicated software that's totally opaque.