rathel

From what I was able to gather, the onus is on the SBF for misappropriating FTX customers' funds. On the Alameda side it does not seem like a typical scam in the sense of "get money and run with it". "Just" a desperation out of illiquidity due to a credit crunch. Keep in mind that SBF was a 90% owner in Alameda anyway.

I am however positive she should have stayed out of crypto. Everybody in this space becomes at least morally gray at some critical point, even if they started well-intentioned.

But regardless of CE's intentions, you're posing a philosophical question: should we evaluate worldviews for what they are, or by who is possessing them. I lean towards the first one.

Internships? Yeah, nah. TSMC doesn't hand out technology NDAs to interns, which makes them kinda useless.

All in all, the IC design field is a decade or two behind software dev in terms of ergonomics. They're not going to attract a lot of talent, if the tooling remains as kludgy and unreliable as it is. What if GCC or Clang crashed on you once in a while "just because"? That's the reality of IC design flow.

On the flip side, the world at large has just realized the importance of chips, which makes the outlook mildly positive.

Yes, BUT... Each new technology node (excluding shrinks) already disrupts the flow in many breaking ways. Seemingly small changes, new DRC requirements, the need to model more side effects all cooperate to make porting to a new node all but easy.

Rest assured that one of the companies you mentioned has deployed a catalog of reusable software components not long ago. It's already quite populated. But the funny thing is, you need VP approval to submit your modules.

I agree that EDA companies are unnecessarily siloed and dated in many places.

Weren't it for the moat of hundreds of PhD's working on the algorithms, it'd have been disrupted to hell by startups.

Same with PostgreSQL:

https://www.paulnorman.ca/blog/2016/05/improve-your-st-geoha...

I wonder what will come of it. Portugal had Chipidea which was somewhat successful, to be later acquired by MIPS and sold shortly after to Synopsys.

Cadence and Synopsys use it.

Sadly, the "cocktail party test" part is so true. How am I supposed to be proud of my work between that and "this one asshole customer" from Japan? (Tier 1's from USA are demanding, but reasonable).

We are woefully underappreciated industry. While we're B2B vendors, it's no rent-seeking sales-driven crap. It's actual innovation 1-3 years ahead what's in consumers' hands.

However, I think it is slowly beginning to change - with news about "chip sovereignty" initiatives and AI helping design SoC's going mainstream. People now realize how crucial ICs are when car factories close.

Just designing on their node as a foundry customer.

Intel did offer foundry services, at least 3 years ago or so, albeit designing stuff for that node required putting everything in separated environment, and simulating in separate compute farms. Even TSMC which was and is far ahead does not demand this level of paranoia, only documentation is put in secure isolated environment, other stuff being managed in the more reasonable way of unix groups, which are, well, good enough.

Do you realize that most locations where tech companies would seek talent have much better employee protection? No at-will employment, guaranteed annual and parental leave.

Why aren't such dense polynomials used more often?

Is it that we trade off something in return for less short-term correlation (maybe more long-term correlation)?

At work I had to take a course on US export control. The restrictions they bully everyone into are pretty nazi. Likewise with SWIFT. As evidenced by TFA it's always regular citizens that suffer. Compare this with EU sanctions that are targeted to particular companies and individuals.

As Moore's law slows down, this time for real, there have been a few unorthodox ideas lately. 3DIC, or Cerebras' whole-wafer CPU. This looks like a thing worthy pursuing. With free money and big corps racing for a market first (disruptors trying not to be disrupted), who knows.

I wonder how much in terms of power-performance-area can be gained from going asynchronous.

Corner reflectors are crazy effective for such application, as referenced in another comment thread here.

The whole digital ASIC synthesis flow revolves around timing closure i.e. satisfying setup and hold at max possible clock speed.

I guess we may even have to invent a new HDL to pursue asynchronous paradigm efficiently.

Not every form of cultural misunderstanding is racism. Why dilute the word? What does it even mean in this context?

I think, that this "art of" while exaggerated as you point out, is thrown around not entirely without reason. It's the first association that comes to writers' mind when you see how Japanese can elevate the most mundane (in Western understanding) stuff to the next level - think polishing mud balls, senko hanabi, kintsugi and so on.

Zuck never asks why, he just collects whatever he can.

People make jokes he's a robot or something, but I think for him company is everything. All the money but no respect earned, and his "status" ends at $10m house and whoever wants to use him for their goals. Hence the paranoid need to have FB stay on top, otherwise he'll end up at the dumpster bin of history. Collect all the data, copy all the competitors.

Compare with Elon Musk who is also kinda asshole, but has a vision and charm.

Holy. Fuck. You weren't kidding. This is blazing fast. Thanks for the tips, definitely bookmarking this as a reference.

Nice tool, just in time I sought an alternative to Google Alerts, because they filter out most stuff and are biased towards bigger/news portals.

For now something is not quite right, whatever I search for, I only get results for Twitter and when I uncheck Twitter, I get nothing.

It'd be also cool to search by language (or would site:pl work to filter by TLD?)

> The dilemma these companies have is wanting to keep those things proprietary and OSM not really allowing for bundling proprietary and open data

Provided you attribute OSM data, you can slice across horizontal layers and country borders. [1]

> I'd argue Here maps is a lot less attractive these days then it was when Nokia bought Navteq (which became Here later). It's still alright for car navigation but that's just one niche application of maps these days.

The silver lining here is that limiting scope lowers the effort needed. Much of OSM edits go beyond road network.

[1] https://wiki.osmfoundation.org/wiki/Licence/Community_Guidel...

Maybe she could instead unfuck Google Maps that became a UX trainwreck ever since she had left? [1]

[1] https://news.ycombinator.com/item?id=20238332

And often the damage is already done. Cf. AARD code which was only in Windows 3.1 beta

Skills are not a real barrier. Data quality and license are. When you look at official datasets, you find ways to screw up you never thought were possible

People in USA still complain about the botched TIGER import [1].

It may sound flippant, but I'd rather not have such end-to-end tutorial geared towards imports be made. It'd lower the barrier too much. People who self-taught GIS (it's possible) can be trusted to be marginally more responsible (not fully, though).

[1] TL;DR the OSM road networks in USA was initially imported from TIGER dataset. However the importing mappers didn't know TIGER was being updated, state by state. Hence, the quality of imported data is not great in many places.

You may have a look at http://hdyc.neis-one.org/?Stereo to try and find metrics relevant to you (example profile, to retain privacy profiles are hidden behind an OSM account login-wall, unless said user links to that page in their OSM profile site, which this example user did)

One great metric is mapping days (distinct calendar days in which the user was active) which sort of abstracts the fact that different people tend to have a different idea how much work warrants a single changeset.