ckastner

For larger enterprises with complex org charts, I consider an agenda one of two invaluable tools for meetings, the other being post-meeting minutes including a RACI chart.

If someone can't think up a few bullet points for a meeting in advance, that person has not prepared for their meeting, and will waste some of the participant's time.

Not creating (one-line) minutes that most importantly include decisions that have been made in the meeting is also the perfect set-up for wasting other people's time.

I see this similar to the effort of writing a good commit message.

Why on earth would people believe his veiled threats to abandon Tesla.

That's where most of his net worth is. Nobody would stand to lose more from Musk leaving than Musk himself.

Because it suggests a possible explanation why non-skilled consumers were not the immediate concern of this legislation.

> Why can't I as a consumer view a service manual and buy a part?

Have you ever helped a family member with their computer problems, only to notice that they often start as something benign, and the mess you are dealing with is from the family member attempting to fix that benign problem themselves?

Outside of tech, how many people do you know who actually read manuals instead of just going at it?

Not only did they have a reason, they discussed the reason and the proposed change with upstream.

> Tesla’s current valuation is pretty much thanks to him.

And nobody has a larger interest in Tesla's valuation them Musk himself, the largest shareholder.

He can't walk and any threats to that effect are theatrics. Shareholders would be idiots to re-approve the $55bn.

An important fact that I only recently learned about this vulnerability is that the change that introduced it was not some rushed act: the maintainer raised the problem they were seeing on the OpenSSL mailing list, proposed a change to fix the problem with a request for feedback, and got some (including from upstream).

The result was a terrible vulnerability, but it seems more of a case of spectacularly bad luck of everyone not spotting the issue.

On Android, I use Keepass2Android [1]. It is open source, is updated frequently enough, and it's compatible with my Yubikey-backed databases.

Yubikey Challenge-Response for KeePassXC needs its own setting though, as KeePassXC made some design decision that is incompatible with KeePass, IIRC.

[1] https://github.com/PhilippC/keepass2android/

I love this. Texts appear both coherent yet utterly random:

> Satan's heat-image can sometimes be seen here and it has recently been proven that all of the earth's insects were born just inside the metal door.

You wouldn't be able to use the PyTorch as built and provided by Ubuntu, as in 'apt-get install python3-torch' [1], as that version currently only has CPU support.

You could still use a version provided by a third party, for example directly from pytorch.org, or a containerized version provided by ROCm upstream [2].

[1] https://packages.ubuntu.com/mantic/python3-torch

[2] https://hub.docker.com/r/rocm/pytorch

As of right now, the most recent versions would probably be Debian 'testing' (our next release-in-development). But eventually, this shouldn't matter.

As with all other Debian development, we target the 'unstable' distribution for packaging work (similar to git HEAD). Packages eventually migrate from 'unstable' to the 'testing' distribution once they pass certain checks. And when Debian is ready for a new release, 'testing' will basically be tagged as '13 (trixie)'.

Ubuntu periodically syncs their packages from Debian 'unstable', so eventually they will have they same packages that Debian has.

Either way, we realize that running bleeding-edge distributions is probably not for everyone, so we intend to provide backports for both Debian and Ubuntu. Hence why I said earlier that eventually, it shouldn't matter. With "eventually" being sometime in the next few months.

(Site note: our own PyTorch does not yet ROCm support, we're still working on a few dependencies.)

> I have a 7900 XT, and all my attempts at getting ROCm to work on different Linuxes seem to have failed at different points.

Debian formed a ROCm Team a while ago and we are working hard on getting this to work out of the box, as in: apt-get install and you're done.

For various reasons, our initial focus was on RDNA2 and CDNA2 and earlier, mainly because we are building a CI network in which we test all of our packages (and their reverse dependencies) on actual cards, and we had to bootstrap the network infrastructure first. CI is central to the Debian ecosystem but all our official infrastructure cannot deal with the requirements for specific hardware attached to a machine -- that was never a factor, until now.

AMD has generously supported us with hardware donations, among which are RDNA3 cards. These are already in our possession, and it's just a question of person-time until these are integrated into our CI. We are already spec'ing the systems, though.

> `apt remove` will often leave other things lying around.

If you mean configuration files, then this is by design.

`apt purge` removes those as well.

For Microsoft -- probably great, as they can now also get the people driving this.

This would have been a hostile move prior to the events that unfolded, but thanks to OpenAI's blunder, not only is this not a hostile move, it is a very prudent move from a risk management perspective. Forced Microsoft's hand, and what not.

I'm starting to think that Christmas came early for Microsoft. What looked like a terrible situation surrounding their $10bn investment turned into a hire of key players in the area, and OpenAI might even need to go so far as to get acquired my Microsoft to survive.

(My assumption being that given the absolute chaos displayed over the past 72 hours, interest in building something with OpenAI ChatGPT could have plummeted, as opposed to, say, building something with Azure OpenAI, or Claude 2.)

> I remember attending a company-wide all hands in 2016

Times change, though. A lot of companies back then thought remote work would negatively impact productivity.

Then Covid happened, and mandatory remote work demonstrated that it actually can work (for the right company, etc.).

Are you at liberty to say how high the frequency gets in connection with these models?

I assume the latency is comparably much higher but also wouldn't be surprised if microseconds generally aren't a problem, eg because the patterns detected are on a much larger scale.

> Semiconductor in general is going bust, too much supply. All semiconductors companies invested more because of Covid (and the sudden huge demand).

Semiconductors are cyclical in general, and most sources that I trust believe that the bottom is already behind us. Many of the big players have already begun guiding revenue growth again, though not as massive as Nvidia's windwall, obviously.

Granted, the merit of some of the awards is questionable, but I think e.g.: Nash's work on game theory or Kahneman/Tvservky's on the pyschology of judgment definitely fall under the umbrella of "science".

The Debian ROCm Team [1] has made quite a bit of progress in getting the ROCm stack into the official Archive.

Most components are already packaged, the next big target is adding support to the PyTorch package.

Many of the packages are older versions; this is because getting broad coverage was prioritized. The other next big target that is currently being worked on is getting full ROCm 5.7 support.

I fully expect Debian 13 (trixie) to come with full ROCm support out-of-the-box, and as a consequence, also derivatives to have support (Ubuntu above all). In fact, there will almost certainly be backports of ROCm 5.7 to Debian 12 (bookworm) within the next few months, so one will be able to just

  $ sudo apt-get install pytorch-rocm

One current obstacle is infrastructure: the Debian build and CI infrastructures (both hardware and software) were not designed with GPUs in mind. This is also being worked on.

Edit: forgot to say that the CI infra that the Team is setting up here tests all of these packages on consumer cards, too. So while there may not be official support for most of these, upstream tests passing on the cards within the infra should be a good indication for practical support.

[1] https://salsa.debian.org/rocm-team/

> What does a credit rating even mean for a country that prints reserve currency?

The credit rating affects borrowing cost (how much interest you have to pay). When you think of the debt ceiling that the US frequently hits, somebody is buying that debt.

And reserve currency or not. Printing your own money dilutes it, contributing to inflation.

Edit:

Forgot to say, a lot of that debt is held by U.S. tax payers and corporations. To see what effect rising interest rates have, you need to look no further than Silicon Valley Bank.

> Six months ago it was a global chip shortage that had supply chains backed up around the planet.

chip != chip. TSMC manufactures hundreds of products, check out slide 4 of the last quarterly results [1].

"AI" (on 5nm or below) is currently only a smaller share of total capacity. 7nm was only around 50% utilization, compared to 110% a year ago.

The recent chip shortage was mostly from 28nm or above. Think automotive, and the like.

The inventory correction happening now is from a strong reduction in smartphone business, and HPC is also down thanks to PC sales declining 10-15% this year.

And there is nothing surprising about this, TSMC has been guiding this inventory correction for more than a year now.

[1] https://investor.tsmc.com/english/encrypt/files/encrypt_file...

> There is no perfect scientific study, if you critique all of them to the point you critique this one, you will dismiss more-or-less all science.

It doesn't have to be perfect, it has to convincingly (to some degree) reject the null hypothesis, as per GP's own words. And that doesn't dismiss more-or-less all science, it more or less sets the lower bar for what constitutes science in the first place.

> it’s the sort of confusion which will inspire many people — especially those who deal with text files all day and get pedantic about this stuff — to satisfy their incredulity and thus be exposed to your program.

Confusion doesn't inspire, it irritates. And to pedantic people, the satisfaction comes from being rightfully irritated.

You're basically advocating in defense of misleading/false advertising.

What I'm missing in this (great) list: products from Shuttle [1]. Which is odd, because a Shuttle XPC seems to have been the author's very first thin client.

Some of the XPC slims are fanless, and the number of ports they come with is terrific.

[1] https://www.shuttle.eu/en/products/slim

> I'm well aware cookie-less analytics does not require consent

Tangent, but cookie-less analytics may also require consent. It's about PII, and cookies are just one mechanism to tie information to a person.

For example: browser fingerprinting doesn't use cookies, but serves the same purpose.

See also: https://en.wikipedia.org/wiki/Temporal_database

This is a dick move par excellence, I'm sorry that you're losing your handle.

But the cherry on top is that three of the handles they offered you (@TheTV, @tvme, @TommiV) seem to belong to someone else already. So I guess the evictions will cascade.

This report will be all about the guidance for the next quarter (and hopefully the next FX), though.

> Simple: by offering you to host a marker on google maps or by even giving you an account, they are entering a contract with you/your corporate entity from which they can't just unilaterally exit without good reason.

Contracts work both ways: now imagine Google, or Facebook, or Twitter stating that you can't close your account without a good reason because you have a contract.

The point is that you don't need a good reason, you just need a previously agreed upon reason.

A cursory examination of the ToS [2] is shows numerous cases in which Google (or the user) can terminate the contract. Whatever happened here, it's all but certain that Google claims that one of these cases is fulfilled, hence they have the right to terminate the contract.

> If you're interested in the finer details and a bit of ranting, read e.g. this post from lawyer Christian Saefken [1] - it's geared towards Twitter (and Facebook, which a friend of mine had success with just the same).

> [1] https://christian-saefken.de/abmahnen-aber-richtig/

That lawyer is claiming a Right to Free Speech. I don't how this strategy can be successful, as Twitter will counter-claim that they it's a private platform they are free to regulate. Sexualized content can be free speech but surely one wouldn't claim that Twitter has the obligation to host it if they don't want to.

If Twitter indeed caved, then they caved simply because they assessed that said Tweet wasn't worth the trouble, not because they were legally obliged to.

[2] https://cloud.google.com/maps-platform/terms