News for hackers. Stuff that matters.

chrischapman

Ask HN: Does Starship mean that data centers will soon be in orbit?

So, I've just been reading this HN link (https://news.ycombinator.com/item?id=38227463) and it got me wondering about near-term profitable use cases for Starship.Would it be possible to move data centers into space? Would it be cost effective? It would be better for the planet from a climate change point of view but how practical is it? How many servers could you fit on board Startship? Maybe...

( Read More... | 8 comments | 2 points | on HN )
Re:11/17/23 starship test 2 on Saturday November 11, 2023 @11:57AM (#38232010)
Maybe move all data centers into geostationary orbit. They would be solar powered with direct communication between each orbiting center and transmitters/receiver stations on the planet surface. With that sort of lifting capacity, it wouldn't take too long to move most data center servers into orbit. Would it be cost effective? Maybe not. But I reckon there would be some climate change benefit by moving all that heat out into space.
Re:Server-Driven UI – Everything Old Is New Again on Friday March 04, 2022 @04:53AM (#30552620)
> thus alleviating the need for the developer to release a new version through the app store review process

I'm not an native app developer but this seems like a win, right? I wonder what app store owners will make of it?

> All users see the latest version of the UI at the same time on both iOS and Android

Seems very much like what you can do with a typical browser application. In the long run, I wonder if this is detrimental to the native app ecosystem. Why not just build for the browser?

Re:Against an increasingly user-hostile web (2017) on Tuesday February 15, 2022 @03:42PM (#30352402)
> it seems rather clear to me that all these things cannot and will not be resolved on the implementation level

Absolutely! This is why we need good legislation. Tech issues will never be solved by tech builders. All tech measures invariably have a tech counter-measure. We've been round this loop so many times over the last two decades. It's a Sisyphean nightmare.

> It is structurally impossible to design your way out of capitalism

This is what good legislation can do for us. The only problem is it takes a lot longer to introduce good legislation than it takes to build bad tech. I'm optimistic that we're approaching the end of the bad tech era.

> Short of economic changes that are completely outside the particular problem domain that these kinds of posts tackle, there is nothing at all to be done.

I reckon legislators in the EU would disagree. They are doing a lot and it's starting to work.

Re:Against an increasingly user-hostile web (2017) on Tuesday February 15, 2022 @03:12PM (#30351988)
> surgeon general moment

I get the impression that the GDPR is ushering in the surgeon general moment that you mention. It's slowly starting to improve the health of the web. The latest rulings may even finally kill the cookie banner. Hallelujah.

Not sure if you're being humorous but just in case, it means the 'making of a sigh' as in 'generating something new' rather than a 'group of people'. As a joke, it works for my generation. The older I get the more I sigh and these days it has a lot to do with cookie banners. But that's another matter.
Re:In defence of the boring web on Sunday January 30, 2022 @12:29PM (#30139081)
Nice read. Firm supporter of Web Pi (3.1415). When it comes to building for the web today, I'm always amazed that "so much can be done with so little" and yet the default is the opposite - "so much is needed to deliver so little" - so irrational! Where did we go wrong? I wonder what Web Euler (2.71828) would have looked like?
Re:If it were my home on Sunday January 30, 2022 @08:55AM (#30137209)
Upvote for 'Oh Goodie' and the humour in general. But had to reflect on the sad reality of the following being in green!

> spend 99.58% less money on health care

That is apparently a 'benefit' of living in the Democratic Republic of the Congo. When compared to:

> die 23.88 years sooner

perhaps the health care statistic should have been in red?

Disagree. The key information is no longer than an average privacy policy (and frankly, less complex) and whereas there are millions of privacy policies (which no one ever reads), there is only one GDPR. Give it one hour of close reading and you'll see just how important and useful it is. It's worth the effort because you only need to do it once.
This kind of 'leakage' is key to personal profiling. If this ruling becomes widespread, profiling may prove much harder to do (which is a good thing).

Edit: clarification

Re:Nuanced communication usually doesn't work at scale on Sunday January 30, 2022 @06:15AM (#30135949)
I really like that quote. It seems like it would work just as well for programmers too: "I didn't have time to write a 100 SLOC, so I wrote 1000."
Re:The old internet shows signs of quietly coming back on Wednesday January 26, 2022 @05:24AM (#30083789)
> And, a high percentage of websites load as slowly as a sick turtle awaking from a long slumber.

I love that the full page makes just 2 requests (1 html file and one favicon). 20KB transferred. 1 second to load. No JavaScript, no cookies, no dependencies at all. This, to me, is the essence of 'respect your visitor'. Kudos.

Re:Get to Know the New Topics API for Privacy Sandbox on Tuesday January 25, 2022 @09:36AM (#30072201)
> More importantly, topics are thoughtfully curated to exclude sensitive categories, such as gender or race.

'Thoughtfully curated' by whom? The user or Google? This seems like it's less about privacy and more about advertising. I wonder if it would have been more honest to call it Advertising Sandbox?

Re:Bayesian A/B Testing Considered Harmful on Sunday January 16, 2022 @12:28PM (#29957914)
> Again, all I am arguing is that if you use a sequential test, you should analyze it accordingly.

I hate to be picky and negative but this doesn't seem like a justification for the use of the word 'harmful' in the title. The phrase 'considered harmful' has been used quite a bit lately and I get that it attracts attention but a title like 'Bayesian A/B Testing Considered Harmful' sounds like a group of authoritative individuals has collectively decided that it is indeed, harmful (and I was expecting a list of the reasons why). If it's just one persons opinion, wouldn't it be more accurate to say 'Why I consider Bayesian A/B Testing harmful'.

Re:Do success stories cause false beliefs about success? on Sunday January 16, 2022 @09:03AM (#29956036)
Wow. This is so useful. I'm always left a bit speechless when I come across something really useful like this. It should be a first stop for all would be entrepreneurs. Thanks.
Re:Shenanigans on Microsoft Feedback Hub on Sunday January 16, 2022 @08:41AM (#29955894)
Thanks. That was a really good explanation and covered all the bases. Makes you wonder if there's a business case for implementing a one-off payment mechanism for response time. For instance, 'free' means "we're listening but probably won't respond" while '$1' means "we're listening and will respond within a week". Just a thought.
Re:Shenanigans on Microsoft Feedback Hub on Sunday January 16, 2022 @06:05AM (#29955010)
Yep. Kind of like "we value your privacy" means "we don't give a shit about your privacy". I guess the only way to uncover a platitude is to test it. The problem with "We are listening to your feedback" is that "listening" is hard to disprove. It doesn't require any action - they're just "listening". It sounds good but doesn't mean anything useful. If, instead, it was "We will respond to your feedback within 48 hours" - well, that's testable. That sounds like a real, honest, useful commitment. Is it just too expensive to do that? Even for a billion dollar business?
Re:Death Clock on Saturday January 08, 2022 @11:21AM (#29853100)
Excellent idea. If they tied it to a fitness calculator it would show your life extending when you exercise and shrinking when you don't. Spend more time on the couch and watch your life get shorter. Go for a long walk and watch it grow longer. Seems like a pretty good motivator. Don't show me how many steps I've taken. Just show me how much life I have left. I would buy one!
Re:Open-source tests of web browser privacy on Saturday January 08, 2022 @09:05AM (#29851773)
This reminds me so much of the ACID tests[1] from the late 90's, early 2000s. I wonder if it will have a similar effect, i.e. to drive people away from Chrome in the same way it eventually drove people away from IE.

[1] https://en.wikipedia.org/wiki/Acid1

Re:A unified theory of VC suckage (2005) on Monday January 03, 2022 @05:21AM (#29779227)
Wow - he would have been 14 years old. Also, the second comment is from Dave Winer.
Re:All of Wittgenstein is now public domain on Sunday January 02, 2022 @04:26AM (#29767191)
Nope. You probably just had JavaScript disabled (which would be the sensible thing to do). Or maybe they haven't got round to it yet and in a few weeks all of what you mention will be implemented along with 48 tracking cookies, 4MB of JavaScript downloads (of which only 16KB is ever used) and an incomprehensible cookie banner referencing 641 privacy policies which no one will ever read!

The things we do, eh? I wonder what Wittgenstein would have made of it all.

I'm just seeing arrays everywhere. 1, 2, and 3 seems wrong to me. I reckon it should either be 1, 2, 3 or 1, 2 and 3. When a comma is followed directly by 'and' (or by 'or'), one of them just seems superfluous.

On a side note, it makes you wonder if lawyers should take a course in programming. I wonder if these two arrays would have clarified matters:

The Intent of the Contract: no overtime for the following: [ canning, processing, preserving, freezing, drying, marketing, storing, packing for shipment, distribution ] This would have included truck drivers because it separates packing from distribution.

The Ruling: No overtime for the following: [ canning, processing, preserving, freezing, drying, marketing, storing, packing for [ shipment, distribution ] ] This excludes truck drivers because it puts distribution with packing and truck drivers don't pack! The power of arrays.

Older »