HN user

bdonlan

1,723 karma
Posts0
Comments528
View on HN
No posts found.

Fusible traces aren’t uncommon - but they would effectively destroy the device if current limits are exceeded, and they certainly would be if the power supply is non-isolated, so it wouldn’t actually be a solution to the firmware update problem.

The usual solution here is an optical coupling of some kind - optocouplers (a box with a LED, photodiode, and air gap between them) are very common for this purpose, and are an inexpensive and effective option for digital signaling across an isolation boundary.

In this case they’ve basically built a custom optocoupler out of discrete components, which is a bit unusual, but makes sense when you consider the risk of corrosion due to water ingress at the diagnostic ports, and the need to access it while - for example - a dishwasher cycle is running.

It was originally a switching/signalling waypoint, and later started seasonal passenger service for skiiers in 1932, before later switching to year-round service. Apparently it used to be a popular hiking destination as well, but with the establishment of more convenient rail and road routes became less popular in the 80s.

In this case, the problem was the seed was low entropy, so if you generated a bunch of random numbers with a single initialization of the RNG seed you would get unique values. It's a tricky scenario to test for if you don't know of the failure mode...

Japan has a good solution for this - most of the larger train stations (and many of the medium-sized ones as well) have "coin lockers" (which, despite the name, these days often take transit IC cards instead of coins) you can rent to store your stuff until you're ready to head home. In larger tourist hubs like Akihabara there'll also be a ton of lockers in the vicinity of the station run by third parties as well.

The state _could_ attempt to negotiate with bondholders to repurchase the bonds - probably at a significant premium, however. Of course, if one of the bondholders refuses to sell, there wouldn't be much they can do... (I wonder if Disney holds any of those bonds for this very reason?)

Much harder for a large company than a small one, actually. The coordination overhead to get a bunch of disparate teams in a large company to rearchitect the fundamental structure of the service should not be overlooked.

If you are sending a message to a person in another country, you are consenting to that communication traveling to the location of that person. See article 6:

1. Processing shall be lawful only if and to the extent that at least one of the following applies:

(a) the data subject has given consent to the processing of his or her personal data for one or more specific purposes;

(b) processing is necessary for the performance of a contract to which the data subject is party or in order to take steps at the request of the data subject prior to entering into a contract

[.. other permissible purposes snipped ..]

Generally, the GDPR issues with sending to another country boil down to whether the EU accepts that the other country's government will allow the company protect the data in ways compatible with the GDPR. When sending that data to another company happens for reasons incidental to the permissible purpose of your data (eg EU-to-EU data processed in the US) this becomes an issue, as you've not consented to that risk. If you are specifically requesting that your data be sent to, say, the US, however, then processing that data in the US becomes necessary and thus much more justified.

From a technical perspective, you can certainly silo your data and transfer only as needed. This is however way more complicated, as you need to now deal with the fact that you're potentially performing joins across high-latency datastores, where you might need to be careful about what query data you're sending across the wire, and where your different silos need to apply access controls against each other. If you didn't engineer for this from the start, it's a big shift in architecture.

The reality is you don't really have to transfer anything out of EU in order to keep your service running.

Practically speaking, running FB in a way that doesn't transfer anything out of the EU would involve either:

1. Siloing off the EU facebook, with no contact with the US side

or

2. Building a federated facebook, which transfers across e.g. only the timeline entries US friends are granted view access to

The former would not be well-accepted, as it would cut off communication from e.g. international relatives, and would be a rather large project to launch. The latter would be an even bigger rearchitecture, which would likely take, at a minimum, several years to complete, since it's unlikely this was ever anticipated as being a possibility when FB was originally created.

So, I sympathize with them - while in the long term they might be able to find a solution, in the short-to-medium term, FB would have no choice to stop operating.

The signal will often go green (or in the case of this line, ATC will show that the train is cleared to proceed) prior to the actual departure time. This is because the trains are not allowed to close their doors until they have a departure signal, so this has to happen well before the actual departure time. Depending on the station configuration and signalling technology the signal clearing may simply occur automatically once the line ahead is clear, so it's up to the train driver to ensure that they don't depart too early.

The semantics of "time X at point Y" are different from "time X in time zone Z". Sometimes the former may indeed be what you want, but it's uncommon for the user to be providing precise location information for a point in time far in the future. If you're told "time X, Eastern standard time", then the semantically correct thing to do is to not guess a point in space, but instead preserve the time zone as provided.

x86-64 does not have an addressing mode that uses the 8-bit register alias; in particular, the base and index registers are always either 32-bit (in 32-bit mode) or 64-bit (in 64-bit mode). As such, you need to zero-extend AL to a 64-bit register before using it in an offset addressing mode (or use XLATB).

For more information on supported addressing modes, see the manual: https://www.intel.com/content/www/us/en/develop/download/int... (specifically volume 1, section 3.7.5)

There's a mechanical door handle on the interior.

If you run out of batteries completely and need to get in the procedure is:

1. Open a small panel on the front and apply 12V to the emergency cables to unlock the frunk (this only works if the car is powered off as I understand)

2. Jumpstart the car's low voltage battery, located under a panel under the frunk

3. Unlock and open the door normally

4. Attach an EVSE charger to get the HV back online as well