MaxGanzII

AWS hype everything.

What they emit is the dance routine of a sugar coated cheerleader squad.

I am of the view everything which is not a strength is obfuscated.

I have zero faith, confidence and trust is all information AWS emits.

I approach press releases and the docs on the basis that they cover up the actual implementation, and so my task is to find out what is actually going on under the hood, so I can actually make sense of what's been provided and operate it correctly (or avoid it completely, as it may be!)

I have to say, speaking for Redshift, I significantly agree with the critique.

There are too many simple problems which should have been caught in testing, and the problem which have over time been found absolutely imply unprofessional, even amateurish software development standards.

For example, recently, the format of the version string was changed. This broke a lot of existing software, which had hard coded parsing - SQL Alchemy stopped working - so did AWS's own JDBC driver.

This on the face of it indicates the RS test suite does not include any connections over JDBC.

It then turned out the version string had anyway been inaccurate for months, because RS had moved from GCC 3.4.2 (I think it was) to 7.3. But the version string kept reporting the old numbers.

I can't even begin to describe how many issues - flat factual errors, and profoundly meaningful gobble-de-gook - in the official docs.

The whole thing just feels too much like amateur hour.

In my view, AWS docs (and support and so on) obfuscate everything which is not a strength; they cannot be trusted, not ever.

Redshift is a sorted relational database and as such it provides the performance necessary to handle Big Data when and only operated correctly, which is to say, when the sorting orders of the tables are correct for the queries being issued on those tables.

If you start talking as the quotes here do about "extremely complex SQL" it is going to be utterly impractical to operate Redshift correctly; it's impossibly complicated.

When operated incorrectly, Redshift has the same performance profile as good old Postgres, except you have the cluster rather than a single node.

The cluster is an expensive way to get 10x or 100x performance, but if you have Big Data, you need a LOT more performance than this - you need sorting, with the staggering efficiency gains it brings - but sorting only works when operated correctly.

The docs from AWS here are, in my view, utterly misleading and are dangerous to readers.

The minimum cost for serverless seems quite high - 14.4 USD per hour.

The advantage is the flexibility to easily change compute resource. A disadvantage is that your data is now in S3 or something very like it, and this I think alters the characteristics of write performance, for the cluster; I've not yet looked into this, but it's on the list.

You absolutely should beware of falling into the trap of imagining that serverless simply gives you flexible compute and that's the only change to behaviour.

AWS in their press releases and docs are relentlessly positive - anything which is not a strength is obfuscated - so only actual experimentation and investigation throws light on what you're really getting.

I hold a very dim view of all information AWS emits, by whatever channels.

My experience with support/account managers is that they always tell you "yes, Redshift can do this", and the and the only way to actually get a "no" out of them is to already know Redshift cannot do something, and to explain to them why.

They won't deny reality, but you would never have got that answer from them in any other way.

I suspect the problem is the training AWS give its staff. The material they are taught is relentlessly positive and I suspect AWS staff actually have no idea what Redshift is no good for.

(Indeed, if you read the official docs for RS, which I strongly advise you never to do, you will come out the other end under the impression there is literally nothing Redshift cannot do; the docs describe everything using positive terms only.)

I would say though that this is what I'm arguing.

The planet can't cope now - the environment is falling apart right now, already - and the mid-range estimate there's another two or three billion people to come over the next few decades.

There are estimates in excess of this, there are estimates lower. Estimates which have population declines see slow declines only.

I've not seen any real understanding of why the fertility rate is dropping, particularly in first world countries. That's a concern; what happens if the factors causing this to happen are transient?

That's part of why the estimates are only estimates, of course.

I may be wrong, but I think the world population is growing more quickly than you portray, and will be for some decades yet, and we currently have far too many people for how we conduct ourselves as a species.

I may also be wrong to think it, but I would expect if you halved the world population, assuming it was done equally across the world, you would indeed halve human carbon emissions; half as many people, half as many homes, cars, power stations, etc.

CO₂ is not the only factor driving ecological damage and indeed, the mass extinction event currently in progress.

There are simply too many humans for the way in which we as a species currently behave.

Either we need single parent families, globally, for a few generations, to get numbers down to a sustainable level, or, we have to convince/force 7 billion people to accept and adhere to major changes in how they live their lives.

CO₂ is a basically easy problem, compared to this, if it is possible to take the route of geo-engineering, because, as you say, it's not necessary to change how people behave; but I think no matter what, there is a need to change how people behave, or billions die from famine and economic failure, induced by ecological collapse.

I don't think people will change (indeed, there will be large numbers of people vocally against change), I think Governments will at best effectively do nothing "(we're building lots of renewable! but we're also building lots of gas at the same time!"), or more likely make things worse ("we'll phase out coal by 2070"), so my expectation is human suffering on a scale never seen before in all human history.

Humans are kākāpō; we're over-reproducing, having as they did no meaningful natural predators to keep us in check, and sooner or later, that leads to ecological collapse.

(The kākāpō went through several cycles of this and evolved to reproduce very, very slowly, and so came into balance with their environment, and then were very nearly almost completely wiped out when humans arrived.)

Famine.

Consistent crop and fishery failures induced by ecological collapse and massive environmental change, and attendent economic failure.

The world has billions living on one or two dollars per day. These people barely survive now; what happens when food shortages cause prices to strongly rise, across whole belts of countries at a time, and they stay high because crop failures become normal, and where the loss of wealth generation, from the loss of crops, hammers the economy?

Bear in mind farmers cannot just "change crops". Each crop requires a particular set of skills, technologies and experience, and these are extremely poor people, both financially and in other forms of capital, such as knowledge resources. These are the billions of people without mobile phones.

In poorer countries - populous countries - the production of crops is the main production of wealth for the mass of the population.

For those people, with significant, systemic, sustained food price rises, there comes a point where you have had so little food, for so long, you either die from it directly, or you become ill and can't fight that illness off, and it kills you.

The Paris agreement has not been kept to.

This agreement will not be kept to.

What will happen is climate change will continue, perhaps accelerating a bit more slowly than it would otherwise have done, becoming more and more destructive and harmful, until the situation is so up-front right-now deadly serious that action is finally taken, by which time action will not matter, because of the huge lag between cutting emissions and reduction in the rate of climate change.

What will be needed at that point is geo-engineering, on a global scale. The technology for this does not exist, it will need to be invented, and then cost-reduced, and then deployed on a global scale on a point-blank timescale - like going from zero to manufacturing every single car in the world in five or ten years. I have no hope this will actually work, and it's madness for civilization as we know it, and to avoid the deaths of billions, to depend upon it.

Steam is a trap though, no?

You buy software, but if you leave, you lose it all.

Is that correct? there's no way to retain what you've bought?

> It's done to help their user acquisition. It uploads your contacts to match against other contact lists and let you know who's on Signal.

I may be wrong, but I think this functionality existed prior to the server-side state effort. I recall when people in my contact list joined Signal, I was notified.

However, these days I do not keep contacts in the phone contact list. It's too big and juicy a target.

> And besides, a 4-6 digit PIN would be a terribly insecure way to "encrypt" anything server-side :)

Very much so. That does seem odd.

I think we may be talking about different PINs.

I am not talking about the PIN you would have to enter when starting Signal, to get into Signal.

I Googled a bit and found an approachable blog post from the time this all happened, here;

https://blog.cryptographyengineering.com/2020/07/10/a-few-th...

This has refreshed my memory of events.

In short, Signal wanted to store what had been purely client-side information (contact lists, for example) on their server, but - in principle at least - in a form Signal could not access.

The PIN in question is used to provide access to that information.

> Server state comment aside, it seems your main complaint is about a pop-up PIN entry UI that can be opted out of?

The dialog to force the user to set the server-side PIN disabled the app. You either had to do it, or stop using Signal. There was no opt-out.

I had a look at the app now. I found the settings you mentioned. It's not clear to me from what I see there is this if an app-locking PIN, a SIM protection PIN, or a server-side state PIN, or all three rolled into one.

In any event, at the time it happened, the presented dialog was full-screen and could not be dimissed; even if there had been options to disable this (and there were not prior to the full-screen dialog - I looked, in an effort to dismiss the permanent partial-screen dialog) you could not get to them, because it was a full-screen dialog which you could not dismiss; you could not get to the app, and so could not get to settings.

The only option was to stop using Signal or provide a PIN so your client-side state could be stored server-side.

Signal was superb for a long time, and then received a hefty chunk of funding, and, although I may be wrong, has declined since then, and in fact jumped the shark about a year ago.

They attempted ever more forcefully to make users to set a PIN to protect server-side state; it started with a dialog at the bottom of the screen, obscuring about 20% of the user list, which could not be dimissed, and then after a few weeks progressed to the a full page dialog, which could not be dismissed - rendering the app unusable.

All you saw upon starting was the full page dialog demanding you set a PIN to continue using Signal.

I did not want any server-side state, and so did not set a PIN, and stopped using Signal. After a few weeks, the full-page dialog went away, and I found I could use Signal again.

Signal actually blocked usage of the app to force users to adopt unwanted new functionality. It's hard to imagine any app doing well with such mis-management.

I opened a thread discussing the problem on their support/public discussion forum, which was deleted. I also at first opened a bug report on Git, before I understood it was all intentional, this was also deleted.

Since this experience, I've regarded Signal as on the way out, but it's still the best there is right now.

Whales have brains which physically are much, much larger than a human brain; size does not directly equate to intelligence. Note also brains are extremely expensive, physiologically, to maintain. It may simply be we are evolving to be more efficient.