jabr

You’ve been able to do download/upload progress using the Streams API with fetch for more than seven years now. https://developer.mozilla.org/en-US/docs/Web/API/Streams_API

State replication is currently based on corrosion: https://github.com/superfly/corrosion

In fact could you not just cron the cli deployment command on the nodes and get an effective poor man's declarative layer

In the project discord, a user recently experimented with a custom setup that sounds very similar to what you describe.

In fact, a big part of uncloud’s appeal to me is that it also provides powerful building blocks for more complex, custom systems like this, not just the streamlined workflow for simpler, standard cases.

I haven't tried it, but EPMD with DNS discovery should work just fine, and should be similar to this NATS example: https://github.com/psviderski/uncloud-recipes/blob/main/nats...

Basically just configure it with `{service-name}.internal` to find other instances of the service.

There is a `--no-install` flag on both `uc machine init` and `uc machine add` that skips that `curl | bash` install step.

You need to prepare the machine some other way first then, but it's just installing docker and the uncloud service.

I use the `--no-install` option with my own cluster, as I have my own pre-provisioning process that includes some additional setup beyond the docker/uncloud elements.

Do you have an example of what you mean? I'm not entirely clear on your question.

The `compose.yaml` spec for services let's you specify which machines to deploy it on, so you could target the database and web app to the same machine (or subset of machines).

There is also an internal DNS for service discovery and it supports a `nearest.` prefix, which will preferentially use instances of a service running on the same machine. For example, I run a globally replicated NATS service and then connect to it from other services using the `nearest.nats.internal` address to connect to the machine-local NATS node.

There isn't a cli function for adding a connection (independently of adding a new machine/node) yet, but they are in a simple config file (`~/.config/uncloud/config.yaml`) that you can copy or easily create manually for now. It looks like this:

    current_context: default
    contexts:
      default:
        connections:
          - ssh: admin@192.168.0.10
            ssh_key_file: ~/.ssh/uncloud
          - ssh: admin@192.168.0.11
            ssh_key_file: ~/.ssh/uncloud
          - ssh: administrator@93.x.x.x
            ssh_key_file: ~/.ssh/uncloud
          - ssh: sysadmin@65.x.x.x
            ssh_key_file: ~/.ssh/uncloud

I can only speak for myself, but I considered a few options, including "simple k8s" like [Skate](https://skateco.github.io/), and ultimately decided to build on uncloud.

It was as much personal "taste" than anything, and I would describe the choice as similar to preferring JSON over XML.

For whatever reason, kubernetes just irritates me. I find it unpleasant to use. And I don't think I'm unique in that regard.

How does this compare to https://www.mooncake.dev/pgmooncake? It seems there are several projects like this now, with each taking a slightly different approach optimized for different use cases?

DuckPGQ is an interesting option, but unfortunately, that project hasn't been touched in a few months and does not currently work with the latest version of DuckDB.

A couple companies using Kuzu in products are talking about joining efforts on a community fork, including Gitlab and Kineviz. Possible future home of that work: https://github.com/Kineviz/bighorn

My best guess is the company was acqui-hired and will soon be working on implementing Kuzu's tech in a different database owned by the acquirer.

My _hope_ is that it was some IP issue with the University of Waterloo and a new company will appear shortly and pretty much pick up where they left off, but that's probably just wishful thinking on my part.

AT&T

* https://imada.sdu.dk/u/kslarsen/dm546/Material/IntelnATT.htm

* https://en.wikipedia.org/wiki/X86_assembly_language

"HTTPie CLI — modern, user-friendly command-line HTTP client for the API era. JSON support, colors, sessions, downloads, plugins & more."

https://github.com/httpie/cli

http://dl.acm.org/citation.cfm?id=139478

http://www.cs.ucsb.edu/~ravenben/research/CS252/252Paper.pdf

I have an old implementation (in C) of paged, deterministic skip lists here:

https://github.com/jabr/olio/blob/master/skiplist.c

https://github.com/jabr/olio/blob/master/skiplist.h

It looks like the WebRTC Object API might happen for IE13, but no support for the full API is planned.

It looks like IE12 should have the Audio API and getUserMedia, though. That lets you do some interesting things with audio, at least.

It varies by feature, but due to IE's long release cycle, it's generally about a year behind Firefox and Chrome on most things.

While basic sites work, some things just aren't possible to do in IE yet. For example, http://www.wavesurfer.fm/ works in every recent browser other than IE 11.

You do realize that he was born and raised in the late USSR, right?

You are seriously defending US police actions that made a former Soviet citizen worry about what would happen next?!

No. We expect better, I guess.

You probably should, too.

I was in grad school with Peretz for a few years, and I can assure you that he was neither drunk nor a smart-ass. And I have no doubt that his narrative is entirely credible.

(And to be clear, we're talking about a guy that has a PhD from one of the top five biophysics programs on Earth.)

He is earnest and sincere, and I can imagine that the officer mistook him for someone he is not that night, but I don't think any of his actions come close to justifying how he was treated.

Further, it was my experience while living near his current neighborhood that the police were as dangerous (if not more so) than any of the myriad criminals operating there. I'm not anti-police, but I am not a fan of the cops that covered those areas while I lived there.

He wasn't told bluntly. They said they no longer needed him. He said he would like to stay, and then he was tackled, cuffed, and sent off for a bit of solitary.

The police overreacted. And it sounds like one cop in particular overreacted and then triggered a cascade of nonsense.

This kind of thing is not rare, but it's not common, either. It happens nightly, but it's unlikely to happen to you.

A good rule of thumb is to not be anywhere near a scene that has an inbound emergency response. There will be police, and if they have nothing else to do, they might harass you for fake reasons.

But it all depends on where you are going. In most places, this is not a common problem. In certain areas of larger cities, the police can be a bit over zealous.

Have you ever lived in that neighborhood of San Francisco?

99.9% of my interactions with the police have been entirely professional and exactly what I would expect.

The 0.1% occurred in the Tenderloin, Mission, and SOMA (neighborhoods of SF). A SFPD beat cop, at night, in one of those areas... Yeah, no.

I've had more pleasant and rational interactions while being mugged in the Mission than I ever had with cops on the Mission beat.

You apparently missed the part where he did all of that, and then ended up naked in solitary for 12 hours.

/.@./ is pretty much the best you can do.

And even the dots are a bit iffy.

I responded to the sibling comment a while ago, and I'd welcome you joining that thread.

But for a quick synopsis of my response to your statement:

1. Billions of people use computers with a variety of operating systems every day; 2. Thousands of people touch their first computer ever, every day, and very, very few run Windows; 3. Millions of people use computers every day and are still deeply confused when the start button moves.

Regardless, the point I was responding to was that Microsoft had some sort of special advantage over the run-of-the-mill enterprise soul-suckers like IBM or Oracle. They might; but they won't for long.

Ultimately, I don't see how any reasonable person could disagree with simply making note of the fact that Microsoft no longer has the ability to change the course of computing: enterprise or consumer.

I was arguing your point; specifically, I don't think their "major footholds" have sufficient strategic value to push Microsoft into a class of its own.

And to be clear, my "just another enterprise software company" category includes IBM and Oracle, as well. And I imagine Microsoft is still on the top of that list, but I don't think they have a radically stronger position. It is a very profitable position for now, but that wasn't what the original comment was about.

The question is whether they can again "control the future of computing" or just continue to make a lot of money on roughly the status quo (i.e. simply be "yet another enterprise software company"). I can't speak for the original commenter, but I suspect their emphasis was on "enterprise software company with no apparent mission except to make money" rather than "yet another", so that line could probably be replaced with something like "yet another boring, immensely profitable company".

On that question, I think you can make a good argument that some of their footholds are actually a handicap now. For example, Windows and Office make so much money that the company becomes conservative, unwilling to do anything that might significantly disrupt them.

They haven't released an OS without the Windows name in about 20 years now. Even with their mobile OS attempts, it always has Windows in the name. Even the original XBox promotion had a bit of "it's based on Windows" push. And on that line of thought, why hasn't there been an XBox Phone? The smartphone market is largely consumer right now, so why not try a different brand approach?

Yes, Microsoft dominates in the enterprise space right now, but they are losing badly in the consumer space (which will eventually break into the enterprise position) and the cloud space (which is the enterprise position of new companies). What do you see them doing to change the course now?

On a "consumer tech" tangent, IE 11 will finally implement WebGL. Direct3D gave them a unique hold on desktop gaming, paired with a really strong console gaming position. WebGL finally forced them to submit. I expect there are a few divisions that are wickedly pissed off now, and I totally understand why: Direct3D is now in its deathbed.

An utterly dominant, gaming tech position undermined and eliminated in the span of maybe five years.

In the enterprise market, I'd give them a longer timeline, but it's hard to imagine a moribund, hydrogen blimp of that size managing to see (much less avoid) the fireball collapse that's coming.

My point is that operating systems are largely commodified now, and that Windows' strategic value (for both Microsoft and the corporations using it) only holds in very homogenous environments.

That is the state of most corporations today, but at best, I would consider Windows a rapidly depreciating asset (in a strategic sense -- I'm sure it will continue to be a major source of cash for a while).

In a single day, I routinely use up to four different operating systems and none of them are Windows.