fadesibert

Goldman took the same approach with instrument pricing ~30 years ago. I recall long discussions about "Node Purpling" in my ~13 year tenure there.

Computer Science has evolved, and AFAICT this is not a graph approach, but things like differentiation are computationally expensive, and therefore you want to minimize the number of times you do it to as close to the theoretical minimum.

Edit: Related HN discussion https://news.ycombinator.com/item?id=36006737

I was sad to see this, but a natural conclusion for a Google Product (whose history of non-search products is lamentable).

Google Fiber / GFiber did a lot to both fulfill the promise of Universal Broadband AND challenge perception of what was possible with home broadband.

If you are waiting until undergraduate level to take ethics, it's far too late to matter anyways.

Doubly so for "business ethics" classes which became à la mode in the post-Enron era. They attempt to teach fundamental ethics, when at most it should be a very thin layer on top of a well founded internal moral framework and well-accepted ethical standards inculcated from day 1 of kindergarten.

Morals are taught 0-9 [0], Ethics perhaps slightly later as it requires more complex thought processes.

[0] https://familiesforlife.sg/pages/fflparticle/Young-Children-...

I moved to Asia from London / NY.

I found this draconian policy jarring at first (never a drug user, but casual cocaine / pot use was everywhere in both London and NY, and the usual cocktail of whatever was fashionable too).

You get used to these policies pretty quickly, and in exchange there are no (visible) drug users and no (visible) homelessness; I don't think in the West we are willing to sacrifice the freedom to do these things, or impose the death penalty for importing drugs (we have abolished it for nearly every other crime apart from murder in most jurisdictions).

I say that not making a value judgement (I cherish and in some cases miss western freedoms, and believe we do all too little to defend them at home), rather observing from nearly 40 years in western society and <12 months in the East.

It's worth remembering that much of Asia went through terrible drug addiction epidemics in the 20th century [0], and they decided to take drastic action, which probably took 25 years to fully bear fruit.

I also don't believe this policy, in isolation, is the whole answer. Asia (and particularly Singapore) focuses on society, community and other values which attenuate the factors which lead to, and are exacerbated by, drug use (violence, theft, vagrancy, unemployment, under-employment).

You give up a lot of freedom, but you get order in return. For some of us, that is acceptable. For others, this is not (and that is ultimately a matter for voters in each polity).

[0] https://www.unodc.org/unodc/en/data-and-analysis/bulletin/bu...

And of course, peculation means misappropriating or embezzling funds. Again, given crypto (and certain notorious crypto exchanges), even more appropriate.

A quick inspection of the article suggests there's a difference of intent.

<snip> Datoviz is a relatively low-level visualization library. It focuses on rendering visual primitives like points, lines, images, and meshes — efficiently and interactively.

Unlike libraries such as Matplotlib, Datoviz does not provide high-level plotting functions like plt.plot(), plt.scatter(), or plt.imshow(). Its goal is not to replace plotting libraries, but to serve as a powerful rendering backend for scientific graphics. </snip>

I'm not sure how much that's creating outsized income for the founder...

There are between 98 (2022 annual report number) and 120 (ZoomInfo) and 133 (LinkedIn number). German filings are notoriously opaque vs Europe or UK.

So that's 637k EUR / 120 employees (although the payroll number jumps around between 450 and ~640 - weird, but who knows, # of employees shifting around or some paid quarterly or on commission?).

That's around 5,300 EUR / month per employee, or 64k / year. Germans notoriously don't work on the cheap - so unlikely that everyone else is working below market to line the CEO's pockets.

That said - they are still a profit seeking enterprise (another commenter noted that they aren't gGMBH - but also they set up a Feeder fund in January - https://www.sec.gov/Archives/edgar/data/1999332/000199933224...)

Which presumably CAN be profit seeking.

So yeah - it doesn't invalidate their mission - if you're into that - but it's not 100% of what it says on the tin.

Also - monthly financial statements may be a German thing (sorry, I actually quite like Germany and Germans - just German company law is quite cumbersome) - but annual statements would give a clearer and more transparent picture.

This is a really fun app. I took a painting class for a few years in an effort to exercise the other side of the brain - and this does what I failed to do with tens of hours of effort with GIMP, and no amount of prompt magic could do with $LLM.

I trust your tip jar gets and stays full - bringing this back to life is a wonderful contribution!

I liked the idea of tako - a slightly slimmed down, less clunky, version of Xonsh. However, it requires significant patching in order to work with modern python (I tried with 3.12, but I believe some of the breaking changes occur in 3.9+).

Specifically, there are a number of imports from collections which appear to no longer work, and I had to install a fairly old version of collections to get Set and MutableSequence (which `tako/takoshell/tools.py` subclasses).

`hz.mit.edu` is not a git forge, so I don't think I can submit a PR without e-mailing the author.

Brit / American checking in and agreeing. My first startup was a B2B SaaS and hiring in the UK was fantastic - the arbitrage was just silly. Experienced software developers (10+ years) @ GBP 70k / year - and that was close to non-finance full-market pay. The same people were averaging $250k in NYC / SF.

And yet, the UK hires were often better off after all expenses than the US hires.

Largely due to housing being slightly cheaper (other posters have pointed out, London is on par with SF / NY - the big difference being London expands, NYC and SF are both "islands" - yes SF is a peninsula, but commuting up 280 or 101 is not a pleasant experienced).

Also, even offering private healthcare (BUPA) - the UK hires were cheaper. I'm in my late 30s and reasonably healthy - my all-in, gold-plated UK policy was GBP 2k / year - I was at $2,000 / month in the US.

*However* - salaries in the UK are unsustainably low.

Three reasons: [1] BOMAD - The Bank of Mom and Dad (parents paying / lending the deposit for a house so the mortgage is at a low rate) is effectively exhausted. This means that current entrants into the housing market are either renting (which is nearly as expensive as NYC, especially after the inflationary / interest rate jump), or saving to "buy" a house (I enclose in quotes because at a 95% mortgage you don't own much of your house). [2] Professional salaries outside of finance are way too low. My fiancee works in a highly skilled, professional field and her salary in 2024 was, in nominal terms the same as my starting salary in NYC 17 years ago working for a large investment bank IN THE BACK OFFICE - where salaries were decidedly blue-collar. My unproven hypothesis is that the UK professional world is still largely geared towards those with alternative assets, private incomes (especially high-prestige non-professional jobs, especially around politics). This makes it impossible to compete with US venture backed startups, even post-ZIRP, because the offer is always going to be better. And yet that private-income driven base has largely been eroded through capital gains, inheritance tax and general downward social mobility (or, perhaps, less doom-and-gloom - averaging towards the center. The difference in wealth and income between the upper-middle class and the lower-middle class has narrowed significantly). [3] There has been over the last 5-7 years significant negative messaging and tax policy against economic success. A confiscatory top-tax band, an erosion of a "job perks" friendly tax regime and a political climate that is very anti-success, even prior to the labour govt (largely started at the same time, though perhaps not by, Theresa May's 2015 speech and focus on "Just about managing").

VC in the UK is hard, largely because the majority (though by no means all) VCs are focused on aping mid-market pension managers. Their ambition is limited to businesses that already work (and yet anything transformative by definition does not work yet) - and are interested mostly in post-revenue companies with linear or lightly superlinear growth.

This, IMNSHO, is largely caused by the fact that, given state expenditure and the corp and personal tax burden, there simply isn't enough capital for US style VC - the portfolio approach requires capital to absorb failures. Most VCs here cannot afford failure.

The closest we get is the EIS / SEIS tax policy, which allows the offsetting of losses in failed businesses (by the equivalent of Accredited Investors) - as well as a friendly Cap Gains treatment of successes. But these are largely made as common stock investments by individuals - and limited to a very small scale.

Which brings me to my final point - the SAFE note is not only not ubiquitous here, it's rare. Even pre-seed investments are either common stock or (more rarely) convertible notes. This requires a level of diligence (even on small tickets) that make capital formation incredibly burdensome.

There's absolutely a path to resolving this - but the UK first has to make a political and cultural decision to embrace startup-led GDP growth, which is has not yet made.

A lot of discussion focuses on the supply side (not unexpected given the article).

The demand side is where we can take individual / small group action. On the one hand, you have CA (and most of the US) phasing out incandescent light bulbs (great - lowers baseline demand) - but you also have an order of magnitude increase in compute (model training anyone?) as well as a push towards electrification of cars, household appliances, heating etc. All fine - but that energy has to be moved (unless you're generating locally with Solar Panels).

Grid scale batteries and similar ancillary services do a pretty good job of levelling this out - making the market more "liquid" (to borrow a stock market term) - but they take 3-7y to bring up depending on the usual infra variables.

<shill> We think batteries to absorb household demand are part of the solution. [www.energyapplied.com]

If you can control when these batteries act in unison (and, as with anything, the devil is in the details) - and deploy them densely enough to absorb forecasted demand spikes (another comment ITT mentioned LMP markets - astute) - then you complement grid scale storage with something you can roll out in weeks to cover a high stress area, not years </shill>

Another comment mentioned "differently regulated" - and that's right - in order to participate in any of the supply and demand programs on the grid, you need to do a lot of things a certain way. The more you can encapsulate that for the user (don't have a household participate in Demand Response, do it through Nest or something similar and have 1mm households participate) - gets a lot more achievable.

In a funny way, power markets are probably the US's last real free-trade market. My first few years deep in the back office of an investment bank, there were 31 NYSE floor traders, when I left 13 years later in 2020, there were 2, and most volume is program trading / etfs / electronic trading. Ditto fixed income (bonds and bond like instruments), foreign exchange (though these latter two are slower).

Financial Markets inherently move towards being vanilla and scalable because it's cheaper - and this has been one area where the race to the bottom hasn't entirely harmed the consumer (there once was a time when you had to pay a broker a commission on each stock trade. Etrade / Robinhood anyone?).

Power still has a long way to go before it's there - but it will _probably_ be a good thing when it does, and given that it touches hard infrastructure and nobody's forgetting Enron any time soon - it _probably_ will end up better off for the consumer.

BOFH is still a periodic fixture on TheRegister: https://www.theregister.com/offbeat/bofh/

BTW - the fictitious calendar is very often used in bond repayments - e.g. https://www.lawinsider.com/dictionary/bond-payment-dates

It has historically been used to allow for bonds to have uniform repayments across maturities (when the principal of a bond becomes due) - useful in the context of a bank wanting to hedge risks such as interest rate risk, repayment risk (if the bond is callable it can be paid back early, which means the bondholders have cash yielding ~0% instead of whatever the bond yield was)

Mortgages, though not traditional bonds (in that it's not a sophisticated company borrowing) are, depending on the jurisdiction, treated as bonds because they get packaged up and sold as bonds - the underlying bit of financial engineering that, after serious perversion, gave us the 2008 credit crisis

Trivial pedantry - but the phrase is more commonly _requiescat in pace_

pacem would be the accusative[0] vs the ablative[1] - the difference being him being peaceful in rest, vs his rest being peaceful.

Nonetheless, I echo your sentiment.

[0] https://en.wikipedia.org/wiki/Accusative_case

[1] https://en.wikipedia.org/wiki/Ablative_case

While I certainly don't encourage over-policing or the proliferation of "rules" - I subscribe to Churchill's view on regulations [0] - I'm not sure I can entirely fault the police in the above quoted passage - "The driver turned violent."

I tick every single privilege box, some of them a few times over, and would think myself quite lucky if I was only tazed for turning violent with the police. While it's unfortunate that the driver's poor health led to an unintentional heart attack, I don't believe it does the rule of law much good to allow drivers, or anyone, to be violent with the police with impunity.

[0] https://www.forbes.com/quotes/10331/

Some questions come to mind after poking around the catalogue (having worked in an enterprise that ended up with many API catalogues, and now many APIs in startup land, I get the pain point).

- What's the difference between currency and forex? They look to be serving the same case, but in different ways, without reference to one another - challenging the implied claim to "this is the best way to X"

- payments. Stripe/... integration into any moderately saas business is a ball-ache. Current "subscription management" services are woefully, laughably inadequate . If you had a /pay endpoint, where I could offload payments, I would have been quite happy indeed - email. Needs better documentation. When sending an email through this interface, is there impersonation? Or is it "me@mine.com <someemailservice@someintermediary.foo.bar>"

Neat idea - the enterprise survivor in me wishes you success

I just read the article (shamefacedly seeking confirmation bias - I wanted to say "Oracle bad") - but I ended up agreeing with her main points: 1) most of the "third party security researchers" customers were hiring were blindly running tooling and the FP rate was near 100% 2) They had found and were working on 87% of genuine issues 3) (and I don't 100% agree here) - the license agreement forbids decompiling the source code because IP. OK fair point (grudgingly admitted) but I take issue with "a contract freely entered into" - you have to accept the tos/eula if you want to use the product. Not entirely freely entered into...

All in all, I thought it was a balanced and well written post - much better than the usual corporate effluent ( "We are thrilled to announce that we delight our customers in achieving their dreams of democratising toothbrushing")

Depends on your definitions of cheap and large. Some banks are very good at FX in the 5,6,7 figure range. Others aren't.

Fidelity, for example, charges no fee but a 1% spread, which is a "hidden" fee.

FirstRepublic charges .01% (and thats because we are "small" by institution standards ~50k / transaction).

Most banks are badly set up for consumer size transactions, unless they are by default multi currency (Revolut and Starling come to mind).

The traditional banks may be "no-fee" - but some have charged as much as 2%.

Still nothing on the ultimate rip off - the foreign currency ATMs at airports - fee +5+% spread...

Using Linear on an 8 person team. Refugees from JIRA in the corporate world. Absolutely love it - keeps us on track across time zones.

The one piece that's probably missing is a tie in to broader "roadmapping" software - I know what's coming in the next 3 sprints, but we haven't found a way to intuitively tie this to our 12-month roadmap - well automatically anyway.

Excellent article - goes into some detail on the myth of oversubscription (well, inflation of reality)

Consider adding a Crossfilter or something similar that shows how many results exist in the criteria you are narrowing by.

Eg: https://github.com/crossfilter/crossfilter

For a working example - see the flight search results on Kayak

Why bother with the .99? Why not just charge €6 / €5 - the .99 gambit is, IMHO trivially deceptive and - speaking for myself - if I am presented two purchase / subscription choices that are approximately equal, I will usually choose the product that acknowledges the fact that I am capable of rounding in my head.

I fly frequently from London to New York and San Francisco. The flights are quite fast - but here's the killer: in New York and San Francisco (and London if you aren't going near or via Paddington) - the trip from the Airport to my next destination - {hotel,office,client...} Can take 1/4 to 1/2 of the flight time!

New York is especially bad (2.5 hours at rush hour).

Would I pay a premium for speed of flight? Yes

Do I think that there are enough people to form that use case? Yes!

Will it work? No - the benefit obtained through a fast flight is obliterated by 3rd world airport links into the major airport destinations.

Concorde had a hard beginning - the French and Brits cooperating? On research?! (It uses the French spelling...)

But when you couldn't fly a helicopter to Midtown anymore, or into the City... The clock started ticking. (This last I have from dear old Dad who was in Finance in the Concorde era and through its end)