Hacker Newsnew | past | comments | ask | show | jobs | submit | corey_moncure's commentslogin

For 楷書 type fonts this may be true, but you ought to know there’s more to it, don’t you?


> For 楷書 type fonts this may be true, but you ought to know there’s more to it, don’t you?

With all due respect, this is the type of comment that really makes online discourse so exhausting.

Yes, I know! Unless you put up two pages of disclaimers and footnotes there's always someone who comes out of the woodwork and "um ackshually"ies you. It was only supposed to be a quick and dirty comment talking about the topic in general, and not an in-depth, ten page treatise on the subject.

If you have something to add to what I wrote, then please do so, but heckling random people who put up their comment up in good faith is not helping anyone.


Decorticate posturing of the hands


Javascript put me off a career in IT for 20 years, and even today I still avoid it like radioactive waste.


If I understand it correctly, the effect of experts is a weighted sum of the individual calculation of each token meeting each expert, where experts to be met by a token are selected on an individual basis. Since a sum is commutative, though, it should be possible to send a large batch of tokens copied to multiple GPUs, where experts are streamed into VRAM, partitioned across GPUs. Then the bottleneck is your PCI-E bandwidth. With 2 GPUs at Gen 4 x16, you should have 60 GB/s of TX bandwidth, allowing you to upload a half precision quant of DeepSeek (about 360 GB) in about 6 seconds.

  1 GPU  -  30 GB/s TX - 12 seconds
  2 GPUs -  60 GB/s TX - 6 seconds
  4 GPUs - 120 GB/s TX - 3 seconds
Then you just optimize your batch size to match the compute time to the upload time of each GPU. The expert calculation results can be retrieved from the GPUs and summed up.


Its venting gas and yawing again. None of that deploy and re-entry stuff is gonna happen.


This is a tall tale told by one of his sons to pump up the mystique (=> market value) of the unpublished work at the time of his father's death. Common sense and technical analysis tell a different story.

By all means,there's no way anyone, even the great Master himself, could have undertaken a project of 14 fugues with various subjects, with the final one (14, not 15) being a quadruple fugue that combines four of them together, without being certain of what the subjects were and that ultimately they could indeed be combined.

Bach scholars have analyzed the paper and the device used to draw staff lines to show that a completed version likely existed, now lost.


Yes inside the DB where it cannot be debugged or optimized


If you're putting advanced feature support into a db engine , you're probably also putting in semi-competent debugging support (at least i'd hope so).

But again, at that point you're really just moving the surface rather than addressing the issues.


We definitely don’t need any such “feature”. If you want to live in a safety bubble you are free to do so. Kindly respect the freedom of the rest of us as well. Have a nice day!


Then you can come up with your own AI, on your datacenters. You are free to do so, so far.


Ars has been like this ever since Conde Nast bought them. It was a great publication a long time ago.


Pure ideological drivel. Anti technology.


Eric Berger used to be a major standout from them, and has traditionally never delved into the partisan nonsense. I've recommended him, on this account alone, countless times. He was a diamond surrounded by a pile of crap.

It's extremely out of character and the writing on this exact topic is somewhat contradicted by other articles he himself wrote on it, like this [1] one. I have a suspicion he's being pressured to increase engagement/subscriptions, and it's trivial to do that in the age of such extreme political derangement.

Countdown to Substack.

[1] - https://arstechnica.com/space/2025/03/butch-and-suni-send-mi...


In the linked article he stated:

“For what it is worth, all of the reporting done by Ars over the last nine months suggests the decision to return Wilmore and Williams this spring was driven by technical reasons and NASA's needs on board the International Space Station, rather than because of politics.”


Read the article I linked above, also written by Eric Berger! Butch confirmed Elon's claim that he made an offer to bring them home last year was accurate. So we know that that offer was made, and must have been rejected. What we don't know is why.

But the real thing here is that Eric Berger has written extensively about the political games NASA plays or (depending on one's perspective) is compelled to play. They support a lot of fundamentally flawed programs that are either a complete waste of money or will never go anywhere, like SLS/Artemis. The reason they do this is because, in NASA's case, their funding depends on it and, in the astronauts' case, because they'll never fly again if they don't play ball.

So you don't simply take things NASA says at face value - again something he has written about ad nauseum. His understanding of the industry, and the games surrounding it, are part of what made (and hopefully in the future continues to make) him such an excellent reporter. But this latest article just dumps all of that and is written with all the insight and worldview of somebody on /r/politics.


This is the main quote from the Willmore interview you linked, that seems most relevant:

"I can only say that Mr. Musk, what he says, is absolutely factual. We have no information on that, though, whatsoever; what was offered, what was not offered; who it was offered to, how that process went. That's information that we simply don't have. So I believe him. I don't know all those details, and I don't think any of us really can give you the answer that maybe that you would be hoping for."

That has to be a nominee for doublespeak quote of the year -- and that's not a low bar this year.


I think the most likely scenario is that a message made its way up the grapevine to him along the lines of 'We're working with Elon to look at getting you guys brought back home on a Dragon.', nothing more, nothing less. NASA awarded a ~$270k contract to SpaceX on July 14th called "Special Study for Emergency Response." NASA claimed that study had nothing to do with Starliner, but that was probably a typical administrative lie of the sort Eric Berger has regularly pointed out (until this article...):

"NASA said this study was not directly related to Starliner's problems, but two sources told Ars it really was. Although the study entailed work on flying more than four crew members home on Crew Dragon—a scenario related to Frank Rubio and the Soyuz MS-22 leaks—it also allowed SpaceX to study flying Dragon home with six passengers, a regular crew complement in addition to Wilmore and Williams."

Nobody knows the details of exactly what was offered, or why it was rejected, besides Elon and whoever he was talking to in the previous administration. But at this point I think one cannot reasonable argue that no offer was made. And we know because of what happened that it was rejected. So the only question remaining is why.

[1] - https://arstechnica.com/space/2024/08/yes-nasa-really-could-...


That interview doesn't really say much. It is mostly an carefully worded, polite but empty statement made after earlier extemporaneous comments by Suni and Butch seemingly contradicted Musk. The headline claim is basically just Butch saying he has no insight into the decision process but he trusts Musk is telling the truth.

In contrast other articles Eric wrote talking to people who where involved in the decision of how and when to return Butch and Suni, those officials clearly state that the decision was made for technical and programmatic reasons, not political pressure.


I 100% agree he's phrasing things 'politically', but think about how you might also be reading what you want to read. For instance the section where you [reasonably] claim he's contradicting Musk (by claiming the decision was not political) was also not only phrased politically, but even came with a sort of disclaimer starting with "From my standpoint." There were no such disclaimers when stating that Musk made some sort of an offer to return the astronauts.

And again something you can't discount here is that Eric himself has written extensively about NASA frequently carrying out/endorsing poor decisions (SLS/Artemis being the low hanging fruit there) owing much more to political pressure than pragmatic decisions about the best direction for progress. Here [1] is one example, including an interview with a former high level NASA insider (30 years experience, up to deputy administrator) openly and casually talking about such.

It's not a secret whatsoever that NASA is under constant and significant political pressure. It's just a part of the game. And in this case you had a situation where the guy, who had basically become public enemy #2 (from the previous administration's POV), was going to be spearheading a high visibility rescue of a launch that should never have been approved in the first place - undoubtedly while blasting it all to his tens of millions of followers. To imagine this would not have provoked some behind the scenes 'management' just seems unthinkable to me.

[1] - https://arstechnica.com/science/2022/08/former-nasa-official...


> And again something you can't discount here is that Eric himself has written extensively about NASA frequently carrying out/endorsing poor decisions (SLS/Artemis being the low hanging fruit there) owing much more to political pressure than pragmatic decisions about the best direction for progress.

The reason he was able to write about this was that he had sources within NASA that would tell him the inside story behind the decisions, and how much politics influenced them. Now his sources are telling him that this decision was not political. I don't have any reason to trust his previous sources but not his current ones (especially when many are the same).


No it wasn't. It's based on visible logic, as everybody knows this, well at least everybody within the 'space domain.' People outside of the 'space domain' don't realize how absurdly dysfunctional things like the SLS or Artemis are, and generally have a completely erroneous impression of NASA.

If you want to see this in action search for pretty much any article on SLS or Artemis by him. The one I offered with the former NASA administrator was to clarify to people who might want to claim he was just speculating or whatever. NASA is the posterboy for making bad decisions under political pressure, and not just in contemporary times...


How are you possibly parsing "We have no information on that, though, whatsoever; what was offered, what was not offered; who it was offered to, how that process went" as "I can confirm that Musk made that offer and it was declined"?

That is a HUGE disclaimer. "no information on that, though, whatsoever" - it doesn't get any bigger than that.


> Read the article I linked above, also written by Eric Berger! Butch confirmed Elon's claim that he made an offer to bring them home last year was accurate. So we know that that offer was made, and must have been rejected. What we don't know is why.

Read his goddamn quote you're citing. The part where he says "We have no information on that, though, whatsoever; what was offered, what was not offered; who it was offered to, how that process went".

Which part of "I believe him despite having no information" do you think is "confirming"?


When you have the owner of the company which provides your space flight lying about spaceflight and nasa, calling the commander of the space station an idiot, and the president making total lies up about the entire process, you can't be apolitical.


Take a picture of your energy screen and post it.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: