You are going to give them ideas…

Ironically, reinstall the whole system, make sure to add some CrowdStrike, SolarWinds, and Ivanti for security and management though…

All well and good when ssh activity is anchored in a human doing interactive stuff, but not as helpful when there’s a lot of headless automation that has to get from point a to point b.

Problem they had was that ssh doesn’t really have any way to enforce details of how the client key manifests and behaves. They could ship out the authentication devices after the security team trusted the public key, but that was more than they would have been willing to deal with.

Rotating the passphrase in the key wouldn’t do any good anyway. If an attacker got a hold of your encrypted key to start guessing the passphrase, that instance of the key will never know that another copy has a passphrase change.

Meanwhile, my company has systems insisting on expiring ssh keys after 90 days…

Yeah, that localization was one of the reasons that was so well loved. The PSP redid all that and while still enjoyable, it wasn’t quite as compelling as the WD effort.

This article is an example where statistical confidence doesn’t help. The model has lots of data so it likely has high confidence, but it didn’t have any understanding of the nature of the relation in the data.

I recently did an application where we indicated the confidence of the output of the model. For some scenarios, the high confidence output had even more mistakes than the low confidence output

As I said, I’ve dealt with logging where the variable length text was kept as plain text, with external metadata/index as binary. You have best of both worlds here. Plus it’s easier to have very predictable entry alignment, as the messy variable data is kept outside the binary file, and the binary file can have more fixed record sizes. You may have some duplicate data (e.g. the text file has a text version of a timestamp duplicated with the metadata binary timestamp), but overall not too bad.

I still have weird glitches where applications don’t seem to update on screen (chrome and firefox, both natively doing wayland).

Lack of any solution for programmatic geometry interaction. This one has been afflicted with ‘perfect is enemy of good’, as the X way of allowing manual coordinates be specified is seen as potentially too limiting (reconciling geometry with scaling, non-traditional displays), so they do nothing instead of proposing an alternative.

The different security choices also curtail functionality. Great, better security for input, uh oh, less flexibility in input solutions. The ‘share your screen’ was a mess for a long time (and might be for some others still). Good the share your screen has a better security model, but frustrating when it happened.

Inconsistent experience between Wayland implementations. Since Wayland is a reference rather than a singular server, Plasma, Gnome, and others can act a little different. Like one supporting server side decorations and another being so philosophically opposed to the concept that they refuse to cater to it. While a compositing window manager effectively owned much of the hard work even in X, the X behavior between compositors were fairly consistent.

I’ve been using Plasma as a Wayland compositor after many failed attempts, and it still has papercuts.

Thing is that they could have preserved the textual nature and had some sort of external metadata to facilitate the ‘fanciness’. I have worked in other logging systems that did that, with the ability to consume the plaintext logs in an ‘old fashioned’ way but a utility being able to do all the nice filtering, search, and special event marking that journalctl provides without compromising the existence of the plain text.

Uh… Sounds like it’s not really system’s fault, your setup is just terrible.

I don’t know his specific issue, but the general behavior of systemd going completely nuts when something is a bit ‘off’ in some fashion that is supremely confusing. Sure, there’s a ‘mistake’, but good luck figuring out what that mistake is. It’s just systemd code tends to be awfully picky in obscure ways.

Then when someone comes along with a change to tolerate or at least provide a more informative error when some “mistake” has been made is frequently met with “no, there’s no sane world where a user should be in that position, so we aren’t going to help them out of that” or “that application does not comply with standard X”, where X is some standard the application developer would have no reason to know exists, and is just something the systemd guys latched onto.

See the magical privilege escalation where a user beginning with a number got auto-privileges, and Pottering fought fixing it because “usernames should never begin with a number anyway”.

I’m fully onboard with the “mean people are offended” smokescreen when they produce bad product that also is very visibly “progressive”. It also works because a lot of people do fixate on that when it’s the least of the problems in a reboot/remake.

Was about to say, the last few times I have bought something, Amazon was actually the more expensive choice. Once we looked at them to buy some grocery type products and they were just absolutely horribly expensive compared to any local grocer.

I’ll say one problem is that for a number of items, there’s a technicality in the supply chain that exempts stuff from the price match. I don’t know about Microcenter, but have seen it in other contexts.

For example “Oh, Amazon is selling a 120 pack, but we only carry 125 packs, so it’s not equivalent”. Or in the most egregious, “You have the price for model number 762LAZ, but we stock 762LWM”, and you search and find out those two model numbers are absolutely identical, but “AZ” models come in a box with an Amazon logo printed on it.

I’ll say that when people notice the white character is recasted as black, it generally means the source material was absurdly popular and any follow up is likely to be pretty meh. The live action disney adaptations. of their biggest animated properties have been generally bad.

Rinse and repeat for almost any reboot/remake of some iconic movie or show. The chances of getting it at least as right the second time around are slim. Even slimmer than bolted on sequels that generally do poorly even with the benefit of the original creative teams at the helm.

They could have preserved the race of every character and it still would have sucked.

YT can’t really serve more ads

They say to hold their beer and watch this…

A bit more of a direct comparison would be IRC to, say, Matrix. Last year I see an article announcing Matrix user count and it was more than all the internet users combined in 1997. This is a near-nothing number in modern internet scale, not even 4% of Facebook userbase, but I’d say that Matrix is about as close as I can conceive of “IRC-like” mindset applied with more modern principles in play. Yes you have billions in more popular social networking and communication networks, but there remains many millions of people’s worth of “internet” that resembles the 90s in some structural ways, which is how many people we had on the internet total in the 90s.

One huge difference is of course that no longer does a wider populace see those folks as potential pathfinders for others to join, but their own little weird niche not playing the same way as everyone else, with no advantage that they can understand in play.

I would say internet “in kind”, not necesarily verbatim the content from back then. I think if someone inventoried the subset of the internet that was “like the good old days” more or less that it would probably match in scale 1997 internet or so, or be larger. Styles may change and content, but the general spirit and approaches persist, just as a now minority in a much bigger sea of crap that came to join it.

Yeah, and the ol’ “slashdot effect” is hardly a concern anymore because things have gotten so much more capable as slashdot didn’t grow.

I’m sitting at a laptop with 8-way 2.3 ghz, 32GB of RAM, a way faster NVME storage than any datacenter array would deliver in that era with a gigabit internet connection from my house. Way outclassing any hosting demands from the 90s for the most severe “slashdotting” that slashdot ever could inflict back then.

To deal with ‘modern internet scale’, you have to resort to more resources, but to keep up with the ‘like 90s subset’, little old rasberry pis can even keep pace.

To a large degree, the same internet that used to be, still is.

Keep in mind that in the era they are nostalgic for, the internet involved roughly 4% of the world’s population. As big in the public conciousness was, it was a relatively small thing.

For example, most people see Lemmy as pretty small and much slower content coming at you than reddit. However Lemmy is still way bigger than what a mid 90s experience with the internet would be. I can still connect to play BBS Door games and there’s barely anyone there, but there were barely any people there back then either. The “old” internet is still there, it’s just small compared to the vast majority of the internet that came about later.

Some things are gone, but replaced. For example Geocities now has neocities, which is niche by today’s standards, but wouldn’t be shocked if neocities technically is bigger than geocities ever was in absolute terms.

Some things are gone and won’t come back. The late 2000s saw a really nice and stable all-you-can-watch streaming experience from Netflix, and their success brought about maddening licensing deals where material randomly appears, moves, and disappears and where a lot of material demands more to “rent” than buying an actual Blu Ray disc of it would cost (have gone back to buying discs as of late because it’s cheaper than streaming).

From what I’ve heard Intel on top of doing more of the work ready to go, they also tend to pay for more development expense associated with products based on their stuff.

With AMD, less work is done for you and you are kind of on your own. This is why partners tend to be more enthusiastic about Intel offerings.