wfaler

A couple of things: * It's mostly hobbyists running their instances, some literally in their basement. * Most scaling glitches have usually been addressed within a few days. The comparison to early-day Twitter is fair, maybe people have literally not expected such a fast scaling route. * I don't think anyone in their wildest fever-dreams expected Elon Musk to burn down Twitter this fast, thus causing a mass-migration.

The UK royal family has pretty wide control over what can and cannot be said about them in the media.

There is for instance, many stories about royal improprieties that have been covered up in recent years, in exchange for “juicy” Harry & Meghan stories being circulated in their place. This is probably at the core of the schism within the royal family. I can read about it, because I’m not in the UK. Bet it’s not very well covered in the UK though.

It’d be great if some sort of self-professed free-speech absolutist could just take over Twitter moderation.. Oh, wait.

Ok, the browser I buy and agree.

But outside web-browsers, I'm not sure it is anything anyone wants - like you imply, in China it's probably handy, because it is a reliable route into all the services that are blessed by the CCP, which means you avoid running into firewalls & thought police.

"Everyone else has it (China)" doesn't make me want one.

Not sure he lied. I think he fooled himself into thinking that moderating a social media platform can be done on a rose-tinted ideological basis. Then reality hit, and the fact is, there are so many edge-cases that being ideologically consistent is almost impossible.

Moderation is hard. And it matters, both to keep audiences & advertisers. Besides, First Amendment-based moderation is not even possible for an international company. Most of Europe have strict laws on holocaust-denial. Thailand & the UK have laws banning speech offensive to their royal families.

..etc, etc..

Forgive me for being stupid, but why would any consumer want an "everything app"? It's not exactly a compelling need I've ever heard anyone express.

Yes, multiple things can be combined, but if it works, probably just coincidence.

Not sure it’s long term the most valuable thing I built, but definitely short-term most profitable:

Built an airline pricing system as the sole developer in 3 months in the early 2000’s. When demoed during late stages of development, it received pre-order guaranteed sales from airlines of $60mn for the next 12 months.

I was paid a paltry $500/day for the contract, and got my marching orders when it was done.

I'm wondering if this erratic behaviour is Musk having some sort of breakdown over how things are unravelling?

* Currently that revenue is cratering and things are not going well. Bankrupcy is a real risk with the high debt servicing cost for the take-over. * If the value is going quickly in the direction of $0, banks will call the loans. * If banks call the loans, he has to sell Tesla stock to pay the $44bn. * If Musk starts to sell Tesla stock at that scale, the value of Tesla could quickly crater in the current climate. It's the last bubble-stock of this cycle, still 80% way to go down, if valued like its auto industry peers.

The outcome of that could be that even selling his entire holding of Tesla, Musk could end up broke, and potentially still owe money from the Twitter take-over.

It would make the wealth destruction of FTX look like a Sunday stroll in the park.

Honestly, I think any "smart money" managers who invested in this should be ashamed. It's absolutely an indictment of their intelligence.

The basics of due diligence is: * How is their record keeping of board decisions & financials? * What board decisions and indemnities exist in those records that may have conflicts of interest or claims from other third parties? (FTX didn't even have a board) * etc..

Furthermore, probably should raise every red flag ever for a reasonably intelligent person if some 20-somethings with a few years of experience claim they have the secret sauce to trade successfully under every possible market regime. It falls on its own implausibility and only reeks of hubris.

To understand most market regimes likely to occur, you likely need people in their late 50'ies onboard, or even older (there's a reason Buffett, Soros et al come out relatively unscathed out of most crises, while everyone else bleeds out).

This is a result of stupidity meeting easy money.

I know someone personally who won the lottery twice. The exceptionally and undeservingly lucky exist.

I moved to Mastodon. 1 day later I have a pretty nice feed and good amount of followers. Many of them old familiar now Twitter-refugees.

Why yes. The point is AWS is charging people through the nose for bandwidth.

Switch the bandwidth past Cloudfront, and you still pay bandwidth costs that no one else charges.

It's no wonder Cloudflare and a bunch of other providers have latched onto this in their positioning to AWS.

AWS, GCP et al have already commoditized DB management. These are pretty well-defined steps, there are no good reason why they should require you to pay someone $200k specifically for those.

I think the natural evolution is that both smaller cloud providers AND open source catches up, providing either managed services that are equivalent, or software that effectively makes your DB and other services feel like a managed service.

Depends: If you work remotely, using VS Code Liveshare or IntelliJ CodeTogether (depending on tooling) works a charm. In office environments, I once worked in a place where each workstation had two sets of mice and keyboards connected.

Main thing about pairing is communication: one person "drives", while both people talk through what they are doing/thinking. The non-typing partner can be a "map reader" of sorts, thinking ahead. It really will be different from person to person.

The benefit of remote setups with two people working the same code in their own editor is one person can run ahead and write the tests, while the other writes the implementation code (and both communicate what they're doing).

What either setup has had in my experience is constant communication between the parties.

This is a decent article as well: https://gds.blog.gov.uk/2018/02/06/how-to-pair-program-effec...

Firstly, you got your _first_ swe job in August. Companies that hire juniors for first jobs shouldn't expect to have fully formed contributors from the gate. It's probably a year or two before you should be expected to be a reliable contributor.

I know with todays inflation in "senior" titles, it comes earlier, but personally, I have low trust in anyone who hasn't done the job for 10+ years.

As more concrete advice: Own the fact that you are green & new. Ask the "dumb" questions. If you are unsure on how to start something, ask "what would be a sensible, concrete first outcome on the way?". Ask, ask, ask.

Also, as a junior, it is immensely valuable to pair on problems and code with more senior members. So if it's not a thing, maybe suggest if you could try pair programming from time to time in the team. It reduces bugs and errors even for very senior people.

Finally, mistakes happen. Sure, try to avoid them, but don't freak out when they happen, just find a solution to correct it. Every mistake you make is a mistake you are less likely to make again (but some of them, you will make again, and that's fine). If you are able to do expensive mistakes on production systems, I'd suggest it's more of a process problem than a people one.

Autoscaling groups solve the problem of ensuring you have the desired number of server nodes running. They do not in anyway solve optimal placement of jobs. It only works “optimally” if you are running just a single service, or services that can all run fine on a single node, in which case you don’t need a scheduler to begin with. The Autoscaling group in this case only provides you with scaling & redundancy, not scheduling capabilities.

How did you arrive at that conclusion? The post claims nothing of the sort.

Yes, by cutting out the pieces that make people do cargo cult Agile of buying a JIRA license, standing around a board explaining what you did yesterday, and shipping once in a blue moon (AKA "Scrum", "Enterprise Agile", or "Scaled Agile Framework" etc)

Nope, nope, nope.

If it was just about registering a company somewhere else and your home tax authorities had no jurisdiction, every company in the world, small or big would be incorporated in tax havens like the British Virgin Islands only.

If a company is controlled from the UK, you can bet your bottom dollar that the UK tax authorities have jurisdiction. Only a very small minority, mostly less well-developed countries allow you to incorporate elsewhere without having to pay tax and report where you are.

Corporate tax residence internationally tends not to be a matter of where a company is registered, but from where it is effectively controlled. This is a matter of fact, not a matter of paperwork, so you can't get around it by just appointing a buddy to sign papers who lives in the right country.

Not only that, incorporating elsewhere opens a world of complexity, pain and double taxation. A simple example: A US LLC with more than one owner is usually taxed as a partnership - eg, US tax is due at individual marginal rates. The UK on the other hand treats US LLCs as "opaque", which means they'd want to tax it as a corporation (there are legal cases to this fact). So a UK resident LLC owner could end up having to pay up to 39.6% tax in the US, then 20% AGAIN on the very same income in the UK, before having extracted a single cent from the company. Then if they extract money from it, there's an additional up to 38.1% dividend tax, for a whopping 90% marginal tax rate.

Now, double taxation treaties _may_ come to the rescue, but I wouldn't bet on it being without a fight through the courts to prove they apply.

Case of "I'm going to pick the sentence I like the least out of its context and argue against that"?

It's pretty easy to be "10x" (or infinity) someone who clearly doesn't know what they are doing.

However, assuming some minimum baseline of having a clue, the multiplier is more likely to be due to environment.

Banks hire some very smart people, yet are historically aweful at delivering software. Startups, especially outside SV often have to survive by hiring "scraps" for peanuts, yet run rings around big corps.

If a company forces people to work with their hands tied behind their backs, they shouldn't be surprised results are lacking.

The rule generally applied, which I think is a sensible one, is that you pay tax in a country on profits derived from it IF you have a permanent place of establishment there (branch office, sales office, retail outlets etc). This means a company that has no physical presence in a country but sells products digitally there has no taxes due (except sales tax/VAT in the EU in some cases if over a certain threshold).

The issue in the case of Amazon, Starbucks et all in the UK right now is that these companies DO have a local permanent establishment, but manage to minimize their corporate tax through a mechanism called "transfer pricing", whereby they pay their sister branches in lower tax jurisdictions license fees etc for using brand, IP, software systems etc, thus increasing their tax deductible expenses in the high tax jurisdiction (UK).

What's useful to note is that none of this would be possible without double taxation agreements and the EU, both which successive UK govts and tax authorities have painstakingly and explicitly negotiated and agreed to. If there are no double taxation treaties, taxes for cross-border payments would in fact be quite punitive (withholding taxes etc).

It seems to me the UK govt thought they would benefit from these agreements when they where originally negotiated, but it has not turned out to be the case. The whole "moral outrage" is just the government passing the buck on a ball they themselves dropped.

As a side note, I don't believe moral outrage is conducive to the rule of law. These multinationals follow the letter of the law and have no further obligations. If the government thinks otherwise, they should renegotiate their positions and stop pointing fingers at others for a ball they dropped.

True, but problems are there to be solved. :) We've done a lot of work around data normalization/scrubbing from a multitude of sources as part of a sister project, so I'm fairly confident about this aspect. Curation and classification is another issue, but we have a few ideas.

As for business, you never know, just let me try to get off this Ramen based diet first. ;)

Appreciate your comment. And you're right about decision trees, though they can be useful for simpler problems, such as classifying documents into categories, whereby you have "sub categories" and the parent categories are mutually exclusive.

Decision trees are really only useful for problems where there is mutual exclusion between the different options, so they are definitely no silver bullet.

That's a very good list. I think we're just starting to scratch at the surface of what is possible, which is why I've devoted myself to working in this field!

I'm not familiar with 80legs, but the main idea is to "democratise" access to this technology and make it easily accessible to anyone who wants to build something on it.

The initial few beta releases will probably be aimed at people aiming to build applications themselves by providing them with API's, but hopefully we'll build out the analytics side of things soon enough so it becomes accessible to non-techies as well.

I find the thread starter and the highest ranking comments to be seriously deluded, and here's why: I find that TDD almost 99% of the time speeds me up, lets me iterate and test my thesis much faster than without, and for this very reason, I almost always write my code with good test coverage.

The times I have omitted tests, I have always come to regret it, having to re-write the code from scratch for it to be up to par.

A few reasons for this: - Yes, TDD is belated gratification - the first few cycles of write-deploy-open browser and test are quicker than writing a test. But as your functionality grows, instead of linearly incremental effort to write new test code, your manual regression testing grows exponentially. - TDD actually HELPS dealing with change: when you refactor functionality, you have instant feedback as to what still works and what doesn't. Though features change the whole system and code base rarely do. See previous point. - TDD helps writing minimal, flexible architectures that are adept at change, as systems are de-composed into, well, testable units! - the "prototype" code almost always ends up being the production system. What is easier once that is the case without tests: trying to write tests for code that isn't very testable, rewrite the system, or just live with testing costs that are much higher than that of the competition?

I have actually seen startups slowly die due to the first two points that I raise. But if you think it's still a good idea to skimp on testing for the sake of expedience, good luck to you, you're going to need it..

I'm betting someone in the Vatican is taking note of this..

This is likely due to you being an American citizen or green card holder: American authorities have lately completely disregarded that some people may expatriate for completely legitimate reasons, and American crackdowns and IRS rules makes it exceedingly unattractive for any non-American financial institution to deal with American customers.

The minimal gain opposed to the guaranteed hassle of dealing with American authorities and rules simply isn't worth it for most foreign banks.