Posts by Amblyopius

Packets DO cost more to move one way than the other

In the last 30 years or so that I've been dealing with this we've never reached that magical point where egress and ingress for anyone was balanced (and I would generally consider that impossible anyway). If you are in services, you scale on egress. If you're dealing with end users, you scale on ingress. If the pipe does both, you'll quickly know which one is the most demanding and yes that makes the other direction essentially "free". Given the services they host, these providers are spending on egress so that's what they charge customers for.

As to multi-cloud: mostly in the software. Hardly any of it is truly an infrastructure issue. With some abstraction layers it's easy enough but like it has been for decades: write crappy software and then expect the infrastructure people to magically make it resilient, portable ... Works just fine if you don't take that route and it's not up to the Cloud Provider to make that decision for you, they are not developing your software.

Maybe AWS has a point? Market seems to be misunderstood indeed.

As usual it's all semantics, what can be defined as "reliving a memory"?

Why bother with brains if saving/replaying memories just requires you to record/replay signals? Memories are fairly inaccurate remnants of what we perceived and they are not really "stored". If I need to remember a concert I went to, I listen/view a recording of it. In order to do the same on a person-by-person level you need the ability to record the signals going towards the brains and you leave the interpretations of it up to the brain. Bonus is of course you can discover endless amounts of things in those "memories" that you can not discover in actual memories.

The ability to develop recording/replay devices to record these signals is going to outpace the development of interpretation of what is actually stored/processed in the brain for quite some time and the quality of the memories will be better.

There are obviously things you can not record such as "what were your thoughts at the time" but we generally strive to remember what we saw, heard, tasted ... so why not just record it before the interpretation is done and replay it after. It makes much more sense.

Competing products? Well, you could of course just equip everyone with cameras and audio recording devices, do a reconstruction of the world based on that and then allow people to relive in VR. Quite a few privacy issues of course, given it allows for after the facts eavesdropping on any conversation but comes with the advantage that you can change what you do rather than just relive.

Endless possibilities, just none that really record brains. The detailed recording of brains for the sake of memories is fairly pointless really.

4th May post is in the wayback machine

Click the link and go to previous snapshot and there it is.

Once you're there I'd also read the other snapshotted post called "Extraordinary Claims Require Extraordinary Proof".

"Specialist" Dan is a bit clueless

His whole FTTP vs FTTC thing is definitely only half of the story. See VMO2 in that story. Not exactly FTTP/FTTC related is it? Now I can immediately name 2 countries in Europe that do a lot better than the UK regardless of FTTP roll out: Belgium and Netherlands. In both cases cable TV is in over 90% of the houses and cable internet has been what has pushed up average bandwidth for 2 decades already and at a far more significant rate than FTTP.

Who here really knows RMS anyway?

Like most people I've only seen the most public things and have no idea how he is the rest of the time. As a result I wouldn't be able to defend/vilify him. The fact that he has done contributions in the field of software is undeniable but I'm not sure how that supposedly means he can do whatever he wants in society and the workplace.

For example Thomas Bushnell said in 2019:

"RMS’s loss of MIT privileges and leadership of the FSF are the appropriate responses to a pattern of decades of poor behavior. It does not matter if they are appropriate responses to a single email thread, because they are the right thing in the total situation."

Which would be a clear indication that dragging up some posts is not really relevant as apparently people close to him saw a pattern of misbehaviour that got halted way too late (after decades) and where for a long time RMS was protected.

There's also pushback against Bushnell. For example Thomas Lord claimed Bushnell had an axe to grind but on the flip side another person who had worked with RMS side-to-side (Giuseppe Attardi) agreed with Bushnell and Bushnell himself simply suggested talking to other people (including women) who worked with RMS to form an opinion.

Regardless of what side you're on it would be interesting if anyone commenting can actually indicate in how far they really know Stallman as he seems to be mainly a tool to wield while advocating for certain beliefs.

Just ask the jury to get out their phones and ask them how many cores each one has. Your options are:

a) they don't know

b) they know but the cores are not compliant with the definition pushed upon them

c) they run out to sue their phone manufacturer for false advertising