Posts by Liam Proven

Small but mighty, 9Front's 'Humanbiologics' is here for the truly curious

Re: Applications

> BTW is there any history available about the naming

I liked this HN comment:

«

Research Unix11 = Plan9

Research Unix12 = Inferno

»

That sounds about right to me.

https://en.wikipedia.org/wiki/Research_Unix

So it's not because it came after Unix V8. There was a V8, V9 and V10, which started to incorporate networking. We (not me!) covered their release:

https://www.theregister.com/2017/03/30/old_unix_source_code_opened_for_study/

I think it's simply a tribute to the Ed Wood film.

http://catb.org/~esr/writings/taoup/html/plan9.html

Re: As I wrote about something else a month or so ago ...

[Author here]

> OSX/Android/ChromeOS isn't Unix

So, a few things about that.

№ 1: Always look at the dateline.

«

My Unix is a general purpose operating system

January 23, 2013

»

A lot has changed in a decade -- very nearly 11 years.

Macs outsold by Chromebooks. Linux on successful mainstream games consoles running Windows games. Microsoft offering a Linux environment in Windows, Linux distros of its own, apps as well as programming tools for Linux, etc.

№ 2:

Quoting from it...

«

where the line is between Unix and non-Unix is ultimately one of feel (and varies from person to person).

»

Well, I think he's wrong on this. Android is a general-purpose OS used by over _a billion people_ and for many, including my own wife, it's their sole computer.

Equally iOS: I know people doing everything on their iPhone and/or iPad for a decade or more now. My own startup was killed off by the iPad.

It doesn't matter what some Unix beardies think at hallmarks, like piping between or plain text files or whatever.

Linux *is* Unix now.

I argued this at the start of the year and I stand by it:

https://www.theregister.com/2023/01/17/unix_is_dead/

Everything else is a rounding error except Apple's XNU and it's massively smaller than Linux. Maybe 50% of units but 10% or less of total volumes.

Linux > Apple XNU > xBSD > all commercial Unix.

Re: sdf.org have a boot camp for Plan9 newcomers

I have a Pi 5 too and I haven't had a chance to fire it up yet... but it is currently over 1000km from me.

I don't think Plan 9 supports it _yet_. But a Pi 4 or 400 should work.

Re: Boat/Car?

I wasn't specifically thinking of one particular vehicle.

But, shortly before it sank, I took a tour of the Albert Dock in one of these:

https://www.liverpoolecho.co.uk/news/nostalgia/fateful-sinking-liverpools-iconic-yellow-27121187

Re: Applications

[Author here]

> This may have changed, it was a while ago,

No, you are right, and it hasn't.

But I have a Cunning Plan™ for that, and I am working on presenting it...

Re: As I wrote about something else a month or so ago ...

[Author here]

> But it's not for everybody

Neither was Unix until Mac OS X. Linux *definitely* wasn't even back then. Ubuntu made it easier, Android improved on that, and ChromeOS brought some of that to the laptop market.

Unless someone somewhere puts the effort in, it never will be.

Microsoft opens sources ThreadX under MIT license

Re: Before the flood...

I don't know if you are the same AC who posted about ThreadX on the Pi last time. I linked to that comment in TFA but just in case, this one:

https://forums.theregister.com/post/reply/4735439

Any road up. I very definitely didn't mean to imply anything nefarious, or that MS had any influence over the Pi or anything of the kind.

But I am not in favour of proprietary BLOBs in general and firmware that isn't really firm. I've written about the problem before, here:

https://www.theregister.com/2022/03/01/proprietary_firmware_blobs/

If the upstream OS is now FOSS then all the Pi Foundation needs to do is get permission from Broadcom and it could release the source code to the Pi's version. There are a lot of people who'd like that and there have been 2 efforts to create such a thing from scratch already. One is linked in the article and the other is a fork:

https://github.com/librerpi/

If nothing else, now that newer versions are available as FOSS it's possible that the Pi 1/2/3/4 firmware could be updated, maybe bringing new features. Or other companies could write their own, or port existing firmware such as coreboot.

I don't see any vast potential but more to the point I don't see any real downside to this.

> TFA seems to imply that ThreadX was something that MS found themselves with, but didn't really need, a

Does it? Because I wrote TFA and I did not mean that or think that or intend to imply it. Which doesn't mean I didn't do it by accident, of course.

No, MS wanted it -- I don't know why or what for -- bought the company, but the founder and original programmer up and left and started a rival company which now offers a more sophisticated rival product.

FWIW ThreadX isn't his first; that was Nucleus, some 4Y before ThreadX came out.

Although MS claims 12Bn licences to "just" 3Bn for PX5, the latter sounds quite healthy to me.

If it found that it was a played-out market with no more money to be made, making it FOSS seems like a good plan.

[Author here]

> a little bit more history of this RTOS before blaming Borkzilla for burying it ?

Um. It's not buried it -- it's just dug it up and let it go.

And I gave what history I could...

Revival of Medley/Interlisp: Elegant weapon for a more civilized age sharpened up again

Re: "there was a war between two rival factions over the right way to build computers"

[Author here]

> Was it really a war?

That was by way of a hook for a future feature-length article which should appear over the holiday season. I have a lot of notes which need to be bashed into shape.

Re: Obligatory XKCD

[Author here]

> ...Of course, I think that this might have straight up been referenced by Liam with the headline.

It 100% was, yes.

And of course that leads us to...

https://xkcd.com/224/

(I especially like "My god -- it's full of `car`s!")

Re: You /what/ Liam?

> I'm disappointed that your editorial staff aren't sufficiently on the ball to have picked up the lisp/list inconsistency.

Don't worry -- they are, too. ;-)

Re: lore

> I meant the fact that the notation is constructed from S expressions. Just like LISP.

Huh! How about that?

I could insert some waffle about decomposition to abstract syntax trees here but I'd only be trying to cover up that I didn't know about that.

Thanks!

Re: You /what/ Liam?

> Interlisp or Interlist? Parehelion... who the heck were they?

Whoops. My bad. I have pointed these out to TPTB and they will be fixed forthwith.

> And if LISP really was so great, how come it was almost totally ignored when the DoD was looking for a foundation for Ada?

I find Ada quite readable, myself.

Whereas I recently had strips torn off me on Mastodon for calling Lisp unreadable -- by Piers Cawley, renowned (retired) Perl guru, who ought to know about unreadable languages and who has clearly not read any of my tech blog, ever.

I think things like RPN, prefix and postfix notation, Forth and Postscript and Lisp, are readable to a certain kind of mind, but will remain forever opaque to most people. And programmers are just people. While C notation is easy enough to construct and Java can be thrown together by untrained non-domain-experts and will work fine, in a language designed for safety, a simple, clear, Pascal-like syntax was probably the right choice to make.

Re: lore

[Author here]

> No mention of Web Assembly, Liam?

I'm not aware of an effort to build Maiko on WASM but I suppose it is possible, and someone somewhere would doubtless like that.

Will anybody save Linux on Itanium? Absolutely not

Re: There are many VLIW systems

FWIW the MIPS claim was widespread at the time. Wikipedia mentions it and it came up on the Reg as well:

https://www.theregister.com/2010/06/22/tilera_quanta_server_stratton/

Re: for well over ten years

[Author here]

I find your post rather hard to follow TBH and I am not sure what you are trying to say here, but you seem to have your information a bit muddled.

NT 3.x and 4 ran on Alpha, yes, but only in 32-bit mode. There were no 64-bit builds of NT.

An unreleased copy of a prototype 64-bit Windows 2000 was recently uncovered and I wrote about it:

https://www.theregister.com/2023/05/19/first_64bit_windows/

The Reg wrote about the forthcoming x86-64 XP nearly 20Y ago:

https://www.theregister.com/2004/02/04/microsoft_teases_users_with_64bit/

I also looked at x86-64 XP recently:

https://www.theregister.com/2023/07/24/dangerous_pleasures_win_xp_in_23/

That story carefully tries to explain the differences between the two different 64-bit editions of XP, which it seems to me that you are muddling up.

Re: Liam Proven is spectacular!

Thank you very much. :-)

The core image that I was riffing on was from a Neal Asher novel I read some years back, which had a *splendidly* macabre concept called "reefs" -- reified individuals: uploaded copies of human minds, running in silicon, occupying their original bodies re-animated with technology keeping the dead meat body up and walking around.

Re: Raymond Chen reads TheRegister...

[Author here]

It could be and it would be most gratifying if so, but it wasn't an exclusive of mine. :-)

Re: i860

I think the tech went into the Pentium *Pro*, the P6, not the P5.

Re: Liam Proven - Reply Icon [Author here]

Gosh. Thank you!

[Author here]

> Akin to a dodgy kebab shop store room?

Please imagine Willow, from the 'Bad Willow' episode of Buffy the Vampire Slayer, saying simply:

"Ew."

(I did love that episode especially. But then I am, originally, English. And Englishmen are famous for loving the sound of leather upon Willow.)

Re: There are many VLIW systems

[Author here]

> Many, many VLIWs exist (and ship annually.)

I will freely admit that I'd never heard of the Kalray, and while I did know about Tilera, I thought they only made massively-multicore MIPS chips. I'd never heard of the TileGX before.

These would seem to be forms of DSP, though, and not what I'd call a general-purpose processor: something that runs a conventional OS and can be used to power a desktop, laptop or server.

POWER, SPARC, MIPS, Alpha, ARM, x86 of course, and yes, Crusoe, or MC680x0, and so on... those I'd call general purpose processors. A DSP I would not. Is that unfair or a nonstandard distinction?

It's not an area I keep up with. From a very quick and cursory search'n'skim, it does indeed seem to be something of a stronghold of VLIW designs. I am amazed, but rather pleased, to discover this.

[Author here]

> As for "hundreds " - the number is in the thousands, even now.

OK, I can accept that. :-D A decimal place out? Close enough for government work.

It's a tiny tiny market and it always was, but I am happy to accede that it could be 10× bigger than my back-of-a-fag-packet estimate.

[Author here]

> Itanium has been a dead man walking for well over ten years at this stage.

I would strongly dispute that.

I need to see some evidence of it ever even standing up, let alone walking.

It's not a dead man walking at all. It's a sort of coffin-shaped tub full of decaying giblets immersed in formaldehyde, but with a lot of life-support tubing going in and out of a small number of cadavers, keeping a few bits of them inflating and deflating in some horrible parody of metabolism.

That's the point of the summary I included: if anyone really cared, they'd have been maintaining it already. They weren't, which is why it's gone. They are 100% not going to do it independently out of tree either, so it is not going to happen.

There are definitely people running Itanium in production, yes. Maybe as many as hundreds of them. But they're running OpenVMS or maybe HP-UX and they'll keep it limping along until it can be emulated.

The sales numbers were much much worse than generally realised. The Intel and HP charts looked good if you don't know how to read graphs... but if you did, you looked at the scales on the axes and saw that while x86 numbers were in millions of units shipped, the Itanium charts were unscaled. They were total sales: thousands of units per year.

Very broadly speaking, HP scammed Intel into making a doomed failure of a successor to PA-RISC while at the same time ditching its Arm devices and killing Alpha, and HP got away with it.

It's dead, Jim. It was never really alive but it's definitely dead...

And it never walked. It never even shambled with arms outstretched moaning DAAAATAAA CEEENTEEERRRRSSS...

[Author here]

> Texas Instruments had a VLIW family of DSP processors

That's an interesting point, but a DSP isn't a general-purpose thing of course -- as GPUs are not today.

(As an aside, a reader also wrote in to note that the Russian Elbrus processors apparently have a VLIW design, but then they are not mainstream, either.)

You don't seem to hear so much about dedicated DSPs any more, but related kit still exists, like Sophie Wilson's Firepath processors at Broadcom.

I think, like Intel's doomed "Native Signal Processing" initiative, in the end it just moved into the main CPU core, possibly in some generalised form like SIMD instructions, that is good enough for mainstream general-purpose computers.

I reckon this is the eventual fate of discrete GPUs. They are already ridiculously large and hot and expensive, and as cryptocurrencies thankfully fade away, now banks of racks of the things are being used to run glorified neural network simulations in datacentres to power these allegedly-intelligent chatbots all the technologically-illiterate-but-excitable currently love.

I reckon Apple has the right plan here. Make a smaller simplified version that you can built right into your SOC where it will be much more closely coupled to your main processor and memory, and as long as the drivers are good, that will do 80% of what 80% of your users need, for 20% of the cost and a lot less than 20% of the electricity and heat output.

If and when Intel make core-integrated GPUs that are good enough for most people most of the time, the ludicrous ugly kludge of laptops with discrete GPUs and some ugly switching system will just go away and become a faintly-remembered historical embarrasment.

Rocky Linux and Oracle Unbreakable Linux also hit 9.3

Re: RHELatives

[Author here]

> If I could upvote articles, I would for that wonderful word.

*Big grin* Oh good!

Spread the word. Tell all your friend. Let's make this a thing. ;-)

Those of you with more than one friend probably don't talk about operating systems over dinner, but tell them anyway.

Linux Deepin's big brother claims it's hit three million installs

Re: Huh? Chinese troops in WW2 surrendering to their allies?

Ahh, I defer on which war. It was only a faintly-remembered ancedote from a book.

[Author here]

> Unisys?

You do understand that this is a Chinese distro from a Chinese software company that aims at the Chinese market, right?

The company is called 统信软件 and the product is called 统一操作系统.

That's its name. Names don't translate. My name is Liam. It *means* something like "battle-helm-love" or something, from _Wiljahelmaz_, but that's not what it denotes. It's just a name now.

The company's name can be *transliterated* as _Tǒng xìn ruǎnjiàn_ and *translated* as Tongxin Software. Note those are two different processes and two different results. Neither is "true" or "right", because its name is really 统信软件, and no, I can't read or pronounce that either.

The name translates as something like "Unified OS" or "Unity OS", and the company as something akin to "Unity Software Systems" but the "unity" part of the name also means "faith". Translating to and from Chinese is a bit of a game because it doesn't really match one-to-one to Indo-european language concepts.

The linguist Paul Linebarger, known to the world better as SF writer "Cordwainer Smith", wrote a phrase for Chinese troops to shout when they lost to American troops at the end of WW2 which he felt was his greatest work. It was printed on leaflets that were dropped over enemy lines.

It told them to march holding their weapon above they heads while shouting "love, virtue, humanity! Love, virtue, humanity!"

As it happens in Mandarin, that sounds to an English speaker a bit like

"Ai! Suh! Ren! Dah!"

He saved tens, maybe hundreds of thousands of lives.

The point? It's hard to translate Chinese to English and English to Chinese.

Google Translate spotted a chance to turn a phrase like "computer program systems united in faith" to "uni-sys".

That does not mean anything to do with Burroughs or Sperry, no.

Canonical intros Microcloud: Simple, free, on-prem Linux clustering

[Author here]

You're both right.

Also see:

https://twitter.com/lproven/status/1724417893847241177

«

Proven's first law: "if it's got the word 'blockchain' in it, it's bollocks."

Candidate second law: "anything claimed to be 'artificial intelligence' is neither artificial nor intelligent."

(The former because training a neural network isn't artifice.)

»

Re: Unix doesn't have anything like this

[Author here]

> the Tru64 cluster-enabled AdvFS file system

Now FOSS:

https://en.wikipedia.org/wiki/AdvFS

Note the comments in there about how it never really delivered on its promise, and wasn't very usable until v4, which is the version in the last-ever release of Tru64, v5.x.

Also note that HP laid off the team and we covered it ~20y ago:

https://www.theregister.com/2004/12/02/hp_ends_tru64/

RHEL and Alma Linux 9.3 arrive – one is free, one merely free of charge

Re: Advanced Kernel

[Author here]

> Alma's adoption of such a kernel would be a major blow.

It would, but I can't see it happening. I think there's no love lost between Alma and Rocky, and I suspect Alma would go _far_ out of their way to avoid working with CIQ.

[Author here]

Thanks!

Agreed on both IBM and Oracle counts.

Although at one point I feared that Oracle would buy Apple and I am very glad that never happened. There was zero synergy and Oracle has screwed both the deal and Solaris.

IMHO there was one opportunity here and it failed:

Dump the entire Solaris userland, adopt the GNU one instead, and do a product that subsumed both OpenSolaris and Oracle Linux into one: the CentOS userland, on top of the OpenSolaris kernel. It could run Linux binaries natively -- Joyent SmartOS proved that. Preserve legacy Solaris stuff in branded Zones.

Too late now, and drivers would have probably killed it anyway.

Re: RHEL and Alma Linux 9.3 arrive – one is free, one merely free of charge

[Author here]

> What is the difference?

You know what, I regret that title now. Oh well. Too late.

What I was attempting to allude to is that Alma is free-as-in-freedom whereas RHEL can be obtained without paying only via a free developer's sub, but that was a poor way of putting it.

Blast.

Canonical shows how to use Snaps without the Snap Store

[Author here]

> I avoid it because it's a resource hungry, insecure piece of crap.

So are all modern OSes. This is an inadequate justification.

*If* the app is available as a .deb then by all means use it. I do; I install `deb-get` and install native packages for all the apps I can, and for those that are not in there, like Panwriter and Logseq, I download the appimage.

I described deb-get here:

https://www.theregister.com/2022/10/13/zinc_ubuntu_remix/

I described why all this is being done in "the quest to make Linux bulletproof" and "the future of Linux packaging".

There is a huge amount of R&D going into this. That costs money. You don't think they are doing it for fun, do you?

Re: You can't use gzip/tar?

[Author here]

> I untar Firefox into /opt or /usr/local and it works just fine.

Sure but there are multiple problems with this approach that I've laid out repeatedly.

* Your tarball apps don't get upgraded with the rest of the OS

* If they update themselves then they need to be in a directory writable by that user

* If they're shared by more than 1 user it needs to be globally writable

* The OS doesn't know they're there, so e.g. checks for the presence of a tool will fail.

* There's no desktop integration unless you DIY.

That's off the top of my head. There are a tonne of problems with this approach _which is why distros invented package managers_.

I used Linux before there was automatic dependency resolution and it was hellish and I very much do not want to go back.

[Author here]

> So why do I need(?) another one?

They solve a very specific problem which I have described in some detail already.

I carefully described it last year here:

https://www.theregister.com/2022/11/09/canonical_conference/

«

Canonical started packaging Firefox as a snap in Ubuntu 21.10, but that means that the company now need only maintain a single Firefox snap. Every time the Firefox snap is updated, the same package updates the Firefox version in Ubuntu 21.10, 22.04, 22.10, and those in the foreseeable future too. Each successive Ubuntu release means a modest reduction in Canonical's support burden.

»

Re: OK...So package managers suck....but.....

> They run on any distro that has systemd.

OK, give you that one. :-)

Although the calls are simple enough, and Oliver Grawert told me that there's no specific dependence on systemd itself, just on some of the functionality, so if another init chose to implement that, it's theoretically but perfectly possible.

I suspect that Chimera Linux might be the first to be able to do that.

Re: Trollers must be haters...

I agree, the mounts are a pain.

I wonder if it would be possible to add a "mount only on first run" flag to a snap...?

Re: OK...So package managers suck....but.....

[Author here]

> Snap packages are similarly tied to a specific Debian/Ubuntu release.

No.

Snap is not the native Ubuntu packaging format. Snap is not included in Debian.

.DEB is the native format for Debian and Ubuntu just as .RPM is the native format for Fedora, CentOS, RHEL, openSUSE, Mandriva, Mageia, etc.

Snap is a cross-distro format. Its main rivals are Flatpak and AppImage.

The whole point of Snaps is that they run on any distro, just like RH's Flatpak format.

Re: Apologies, can't seem to find the "Corrections" link

[Author here]

*Much laughter*

I think you are right. I will ping the editors.

[Author here]

They exist, like AWS, like cloud computing, like Kubernetes, to save work for somebody somewhere.

Canonical typically has 5 "current" releases of Ubuntu in support. Like Debian, most of the packages in the distro stay at the same version for the lifetime of the distro.

But Firefox can't. Browsers need to be more current. They are _the_ core tool now and security problems and so on are eternally recurrent. So, Canonical has to maintain Firefox $CURRENT for 5 or 6 distros going back 5-6 years, constantly.

One package that can run on all of them saves the company a lot of work.

Much flows from that. Same goes for RHEL/Fedora with LibreOffice etc.

It's about making stuff easier, for maintainers and for users, at the cost of some disk space.

That, on the face of it, is not a bad thing.

Canonical reveals more details about Ubuntu Core Desktop

Re: Oh, neat

It's very simple and I've explained it, at length, before.

SUSE distros want `/` on Btrfs because that's how they implement snapshots and rollback.

Before you do an update or install a package, the OS calls a tool called "Snapper" to make a pre-installation snapshot. *But* the packaging tool does not account for the space used by that snapshot in its estimation of the space the installation will need.

So, over time, your root fs gradually fills with snapshots. It's not obvious because Btrfs does not reply to the `df` command with real numbers, so any tool using normal Unix commands to check free space gets incorrect info: your file manager, your desktop environment, and your packaging tool.

One day, you do an update. The packaging tool checks for free space. There's enough. It starts. Snapper takes a snapshot. The packaging tool continues, but now, there's much less space because the snapshot is taking some. It fills the partition, but it keeps writing.

Result: filesystem corruption. Reboot and it won't boot because booting writes to the root FS: log files and things. The corruption gets worse. Try to fix it, and *poof* there goes your rootfs.

SUSE's answer is: don't partition the disk; put everything in one big Btrfs volume. Give it enough room and it won't fill up. Don't keep `/home` separate.

To which I say "no" because I've seen it self-destruct many times. The only reason I could recover quickly is that /home _was_ a separate partition and it wasn't Btrfs.

If you don't keep root on Btrfs, you won't see this -- but you lose transactionality. If you don't use snapshots, you won't see this, so Oracle and Fedora see no problem.

Re: (1)Wait...what? and (2) Is it or isn't it?

Oh, thank you very much!

Re: (1)Wait...what? and (2) Is it or isn't it?

OMG I wish!

Thank you. :-)

GhostBSD makes FreeBSD a little less frightening for the Linux loyal

The userland is a free choice, yes... But apart from a few packaging tools and config files, under the hood, most mainstream distros are much of a muchness. They all use the GNU libc, GNU coreutils, systemd, the same handful of desktops, etc.

The differences are the setup program, the packaging tools, and where they keep things on what FS.

There are outliers. Chimera is wonderfully weird. Alpine and Void are different. There are strange little router and firewall ones from CBL Mariner to OpenWRT

ChromeOS is weird. Flatcar is server ChromeOS. Gentoo is odd. Things like Deepin and SteamOS are starting to explore ChromeOS-like tech and that's good.

The immutable ones aren't so strange, it's just the way they are installed and updated that is weird.

Gobo, NixOS and Guix are pretty strange though.

I said distros of *FreeBSD*.

Dragonfly is a whole different OS. So are NetBSD and OpenBSD. They are related but diverging.

Fedora 39 waves goodbye to modularity, but has enough spins to make your head spin

If you want neither snap nor flatpak but want Ubuntu, Zinc is great.

If you want to banish systemd as well, try MX Linux.

Re: I really do not get this...

I wasn't able to get as far as installing it. This is the live boot environment, over which you have no control of session type.

FWIW the GNOME, Xfce & LXqt live ISO versions all worked fine.

Re: I really do not get this...

[Author here]

> Why can't you just install a meta-package to pull in the DE of your choice,

You can. Exactly as you can on Debian, Ubuntu, Arch, whatever.

But having it ready-made is easier if you know what you're doing and what you like, because you can just get it, ready to go and with less effort.

And it's easier if you _don't_ know what you're doing, because you can just try them all in VMs, or go for the mainstream default option, or read a few reviews and pick the one that sounds like you'd like it.

So everyone wins.