bionsystem

I have a previous experience with it. I agree with most points. Jobs can be downloaded as xml config and thus kept/versioned. But the rest is valid. I just don't want to manage gitlab, we already have it at corp level, just can't use it right now in preprod/prod and I need something which will be either throwaway or kept just for very specific tasks that shouldn't move much in the long run.

Yeah I was thinking of using it for us actually. Connects to everything, lots of plugins, etc. I wonder what the hate is from, they are all pretty bad aren't they ?

Will test forgejo's CI first as we'll use the repo anyway, but if it ain't for me, it's going to be jenkins I assume.

Overall it works well and I like the defaults, the work done is remarkable, and it's been a huge relief considering the shitshow that's Windows 11, and even an improvement from Windows 10 which I enjoyed for years, but it lacks a bit of polish I feel, depending on what you use it for. I don't blame anybody it's really hard work to maintain something like that and a lot of things are nice, but here are some annoyances :

- It doesn't shut down properly most of the time, I have to cut the power ; which I do anyway to go to sleep but sometimes I forget after I use it in the morning before going to work, and it stays with a black screen and the fan running all day

- There are a lot of updates, a few Gbs per week, and I have to type my password several times a week (even when logged in), I can't find how to change that

- Sometimes after an update I'll lose an icon or two, or some settings like scroll speed, etc ; not a huge deal but forces me to google around to get the setting back

- Lots of apps are in flatpacks or snaps, I could try some other repos or maybe nix/guix/pkgsrc but I would lose the appstore anyway so I might as well look around for something else

- Some things seem painful to setup, nvidia drivers, incus/lxc, zfs on root... NVidia was the most important and I managed to make it work well now but didn't bother with the rest

Funny I've been poking with the latest ISO last night in a VM. ZFS on root with mirroring and boot environment is seamless, which to me is a huge enabler for a rolling release with fast update cycle, so I want to try it deeper. Currently on fedora kde spin which has a lot of quircks, with Cosmic coming out soon I'll probably switch.

I agree. We are going as far as being asked to release our public app on self-hosted kube cluster in 9 months, with no kube experience and nobody with a CKA in a 2.5 person ops team. "Just do it it's easy" is the name of the game now, if you fail you're bad, if you offer stability and respect delivery dates you are out-fashioned, and the discussion comes back every week and every warning and concern is ignored.

I remember a long time ago one of our client was a bank, they had 2 datacenters with a LACP router, SPARC machines, Solaris, VxFS, Sybase, Java app. They survived 20 years with app, OS and hardware upgrades and 0 second of downtime. And I get lectured by a 3 years old developer that I should know better.

If it runs on a mainframe you would lose both the chess and the boxing.

It does happen that the engine doesn't immediately see that a line is best, but that's getting very rare those days. It was funny in certain positions a few years back to see the engine "change its mind" including in older games where some grandmaster found a line that was particularly brilliant, completely counter-intuitive even for an engine, AND correct.

But mostly what happens is that a move isn't so good, but it isn't so bad either, and as the computer will tell you it is sub-optimal, a human won't be able to refute it in finite time and his practical (as opposed to theoretical) chances are reduced. One great recent example of that is Pentala Harikrishna's recent queen sacrifice in the world cup, amazing conception of a move that the computer say is borderline incorrect, but leads to such complications and a very uncomfortable position for his opponent that it was practically a great choice.

At this point, why isn't Amazon shipping us products that they think we should buy ? After all we can always send them back and get a refund if we don't want them.

Is 2026 the year of freebsd on the desktop ?

I used emacs at school some 15 years ago and I remember it being pretty seemless, I had an OCaml repl for one course and a 68000 emulator with memory inspection for another, and gdb integrated for C ; I do NOT remember hours of configuring that, maybe put some files at the right places and that was it. Switched to vim due to work (that was what's installed on remote machines), kept it for years because of the ubiquity.

More recently in a new gig I'm finally able to install stuff on my machine (with homebrew) and not just work remotely, wanted to revisit my choice between (neo)vim and emacs again, but I guess muscle memory is too strong and still chose the former, although trying emacs I can tell that it is maybe even better polished now with the package manager and everything. Turns out neovim has the same with lazyvim, mason, etc. Just a bit more friction sometimes maybe.

My main pain point right now is the lack of tooling for devops/sre in general. Yes we have LSPs for ansible, groovy, terraform... But they do not cover the entirety of plugins and modules that can be used, and I'm not aware of good tools for testing and debugging. Yes there is teamcity but that needs a license and I can't have that at work apparently. I don't think it is at the editor level though, just the ecosystem is lacking.

I "tested" the handbook recently (I think on FreeBSD 14 when it came out) and I can attest that the experience was flawless. It is even surprising that the right way to use it, is to follow a documentation and apply what it says, versus the Linux way which looks a lot more like "google your way through multiple different ways of doing the thing until you find the one that works".

As an SRE I did most of my work from windows as almost all of my jobs required it (only one provided a working Ubuntu laptop). Mostly used it only for firefox and PuTTY. At home it was W10 + WSL or virtualbox but mostly used it for games initially and then browsing.

BUT as from W11 which I tried for 11 minutes, I switched to Fedora and didn't look back. Everything felt buggy, start menu taking forever, buttons switching places, apps wouldn't start but then randomly would, and of course ads. I know there are debloat scripts around but didn't have the patience, I felt if I'm going to try a new OS I might as well try Linux before buying a Windows license.

Now Linux has its issues, the occasional icon disappears, sometimes shutdown doesn't entirely shut down the machine and I have to power the machine down (electrically from my outlet), sometimes updates don't go through, and of course I miss excel and a couple other things (although I was far from a power user and OnlyOffice fits my needs). But it's still better and with KDE there is almost no config to get a similar "feel" as W10. Amusingly I actually use MORE apps now that I trust installing new stuff more.

That said I would probably pay a premium to dual boot occasionally on a supported Windows OS without bloat, probably it would be my default boot even.

I don't quite understand the point, why would anybody use S3 then ?

I believe redox is doing the same (the everything as an URI part)

I'm quite fluent in english and "unless" is also weird to me, for some reason, "when" in ansible is also very weird, I don't know why. Back to the original post, I would never work for my (French) government again if they adopt something like rouille ; just reading the README felt very weird.

Right, I was focused on the other links (89 and 100) on mobile.

Impressive. Would be great to have the list linked in the article.

I remember an article here, maybe a year ago, where somebody showed some database concepts from bash examples (like "write your db in bash"), but I can't find it anywhere, does anybody have it ?

I saw this at an airport. Took the same plane twice, one year apart, in between they had replaced the terminal by a web UI. First trip it took 15 seconds from the hostess (well into her 50s) to find my booking and print my pass. Second trip (on the web UI), it took 4 hostesses to team up for something that felt like 5 good minutes to do the same thing.

Sorry I'm exhausted and lazy.nvim is the package manager, but I meant lazyvim which is the distribution. Within this distribution I'm pretty sure it's which-key that provides a popup. If I type <leader> it pops with suggestion (and a little icon in front of each indicating if the next key has sub-commands or not).

Neovim with lazy.nvim has that by default (delay included).

Think of it as types. All of the following are the same thing (declare p as an int* type). It's important for the end :

  int * p;
  int *p;
  int* p;

Now remember that the type is a memory address. I'm sure it is semantically wrong for whatever reason somebody will explain but it helps to think about it. So you can do :

  int my_number = 6;
  int* p = &my_number;

Both sides of the "=" are the same type (int* is an address, and &my_number is also an address, the one of my_number).

Now p is a pointer (or an int* or an address), and *p is... an int ! So this is totally valid :

  printf("%d\n", *p)

and for anything else than int you need to malloc that so you will see a lot of :

  my_struct* s = malloc(sizeof(my_struct);

which makes sense because malloc returns an address (the beginning address of the content of s ; yet again somebody will tell me I'm wrong to call it "the content of s" so sorry for that).

  my_struct* // is the type of s, it is an address
  my_struct // is the type of *s (some custom type of size sizeof(my_struct))

The way a lot of teachers teach it is plain trash. I was raking terrible grades in OCaml and C before a student showed me what I needed to know in 15 minutes, and then I would kill it in the remaining exams. Same thing happened to my 68000 course. It didn't happen with every teacher but still, some people really need to get some better pedagogy.

Non-ops teams in big corp / public services are often fed up by their own IT (slowness, incompetence) and want to deploy, budget lines for recruiting devops/sysops is sometimes different than software/platform licenses, even if the solution is sub-optimal.

Disclaimer I'm an SRE/sysops and went through a bunch of those projects where sometimes entire department see (often shadow) IT from contractors as the only solution moving forward.

That's what I want to do as well, Pis running a vpn and some wake-on-lan scripts for a couple of old boxes (I will probably run SmartOS on those). Anything to share about your setup ? (blogs/gits that you wrote or followed)

I also got some cloud credits from my employer, but a bit paranoïd about putting my data there (although most of it isn't sensitive).

I wonder why people always assume that the strategy would be algorithmic or systematic. How about global macro, long/short equity, or even plain long only done well ? Actually studying markets and assets fundamentally, and finding asymmetric bets ? There are plenty of people that have done that successfully over really long periods of time, I doubt markets are perfectly efficient just because some academics claim so, especially for bets with strong convexity.

Nowadays you can run your VMs inside LXC, SmartOS also run them inside zones by default. I wonder if the same exploits could be used accross the container layer of both technologies or if it would protect from leaks.

So, it competes more with S3/minio than NFS it seems ?

Free to chose which pedophile you prefer.

No, I got multiple cases of managers with speech impediment, I also saw a blind tech lead and a deaf one (who also happened to have a small speech impediment).

I also saw a bunch of C-levels being total sociopaths but that's another story :)