I really hate it whenever I see someone calling Apple fans fanboys or attacking Apple products as useless junk that only sells because it's fashionable.

Every hater is 100% as ignorant and wrong as any fanatically-loyal fanboy who won't consider anything else.

Let me try to explain why it's toxic.

If someone/some group are not willing to make the effort to see why a very successful product/family/brand is successful, then it prevents them from learning any lessons from that success. That means that the outgroup is unlikely to ever challenge the success.

In life it is always good to ask why. If this thing is so big, why? If people love it so much, why?

I use a cheap Chinese Android phone. It's my 3rd. I also have a cheap Chinese Android tablet that I almost never use. But last time I bought a phone, I had a Planet Computers Gemini on order, and I didn't want two new ChiPhones, so I bought a used iPhone. This was a calculated decision: the new model iPhones were out and dropped features I wanted. This meant the previous model was now quite cheap.

I still have that iPhone. It's a 6S+. It's the last model I'd want: it has a headphone socket and a physical home button. I like those. It's still updated and last week I put the latest iOS on it.

It allowed me to judge the 2020s iOS ecosystem. It's good. Most of the things I disliked about iOS 6 (the previous iPhone model I had) have been fixed now. Most of the apps can be replaced or customised. It's much more open than it was. The performance is good, the form factor is good, way better than my iPhone 4 was.

I don't use iPhones because I value things like expansion slots, multiple SIMs, standard ports and standard charging cables, and a customisable OS. I don't really use tablets at all.

But my main home desktop computer is an iMac. I am an expert Windows user and maintainer with 35 years' of experience with the platform. I am also a fairly expert Linux user and maintainer with 27 years' experience. I am a full-time Linux professional and have been for nearing a decade... because I am a long-term Windows expert and that is why I choose not to use it any more.

My iMac (2015 Retina 27") is the most gorgeous computer I've ever owned. It looks good, it's a joy to use, it is near silent and trouble-free to a degree that any Windows computer can only aspire to be. I don't need expansion slots and so on: I want the vendor to make a good choice, integrate it well and for it to just work and keep just working, and it does.

It is slim, unobtrusive for a large machine, silent, and the picture (and sound) quality is astounding.

I chose it because I have extensive knowledge of building, specifying, benchmarking, reviewing, fixing, supporting, networking, deploying, and recycling old PCs. It is over 3 decades of expert knowledge of PCs and Windows that is why I spent my own money on a Mac.

So every time someone calls Mac owners fanboys, I know they know less than me and therefore I feel entirely entitled to dump on their ignorance from a great height.

I do not use iDevices. I also do not use Apple laptops. I don't like their keyboards, I don't like their pointing devices, I don't like their hard-to-repair designs. I use old Thinkpads, like most experienced geeks.

But I know why people love them, and if one wishes to pronounce edicts about Apple kit, you had better bloody well know your stuff.

I do not recommend them for everyone. Each person has their own needs and should learn and judge appropriately. But I also do not condemn them out of hand.

I have put in an awful lot of Windows boxes over the years. I have lost large potential jobs when I recommended Windows solutions to Mac houses, because it was the best tool for the job. I have also refused large jobs from people who wanted, say, Windows Server or Exchange Server when it *wasn't* the right tool for the job.

It was my job to assess this stuff.

Which equips me well to know that every single time someone decries Apple stuff, that means that they haven't done the work I have. They don't know and they can't bothered to learn.

A short extract of Neal Stephenson's seminal essay has been doing the rounds on HackerNews.


OK, fine, so let's go with it.

Since my impression is that HN people are [a] xNix fans [b] often quite young therefore [c] have little exposure to other OSes, let me try to unpack what Stephenson was getting at, in context.

The Hole Hawg is a dangerous and overpowered tool for most non-professionals. It is big and heavy. It can take on big tough jobs with ease, but its size and its brute power mean that it is not suitable for precision work. It has relatively few safety features, so that if used inexpertly, it will hurt its operator.

DIY stores are full of smaller, much less powerful tools. This is for good reasons:

• because for non-professional users, those smaller, less-powerful tools are much safer. A company which sells a tool to untrained users which tends to maim or kill them will go out of business.


• because smaller, less-powerful tools are better for smaller jobs, that a non-professional might undertake, such as hanging a picture, or putting up some shelves.


• professionals know to use the right tool for the job. Surgeons do not operate with chainsaws (even though they were invented for surgery). Carpenters do not use axes.

The Hole Hawg, as described, is a clumsy tool that needs things attached to it in order to be used, and even then, you need to know the right way or it will hurt you.

Compare with a domestic drill with a pistol grip that is ready to use out of its case. Modern ones are cordless, increasing their convenience.

One is a tool for someone building a house; the other is a better tool for someone living in that house.

That's the drill part.

Now, let's discuss the OSes talked about in the rest of the 1999 piece from which that's a clipping [PDF].

There are:

• Linux, before KDE, with no free complete desktop environments yet;


• Windows, meaning Windows 98SE or NT 4;


• Classic MacOS – version 9;


• BeOS.

Stephenson points out that Linux is as powerful as any of them, cheaper, but slower, ugly and unfriendly.

He points out that MacOS 9 is as pretty, friendly, and comprehensible as OSes get, but it doesn't multitask well, it is not very stable, and when a program crashes, your entire computer probably goes with it.

He points out that Windows is overpriced, performs poorly, and is not the best option for anyone – but that everyone runs it and most people just conform with what the mainstream does.

He praises BeOS very highly, which was 100% justified at the time: it was faster than anything else, by a large margin. It has superb multimedia support and integration, better than anything else at the time. It was standards-compliant but not held back by it. For its time, it has a supermodern OS, eliminating tonnes of legacy cruft.

But it didn't have many apps so it was mainly for people in narrow niches, such as music production or maybe video editing.

It was manifestly the future, though. But we're living in the future and it wasn't. This was 23 years ago, nearly a quarter of a century, before KDE and GNOME, before Windows XP, before Mac OS X. You need to know that.

What Unix people interpret as praise here is in fact criticism.

That Unix is very unfriendly and can easily hurt its user. (Think `rm -rf /` here.)

That Unix has a great deal of raw power but maybe more than most people need.

That Unix is, frankly, kinda ugly, and only someone who doesn't care about appearances would choose it.

That something of this brute power is not suitable for fine precision work. (Which it still mostly isn't -- Mac OS X is Unix, tuned and polished, and that's what the creative pros use now.)

Here's a response from 17 years ago.

In fact, there are two free versions: one for Classic MacOS, made freeware when WordPerfect discontinued Mac support, and a native Linux version, for which Corel offered a free, fully-working, demo version.

But there is a catch – of course: they're both very old and hard to run on a modern computer. I'm here to tell you how to get them and how to install and run them.

WordPerfect came to totally dominate the DOS wordprocessor market, crushing pretty much all competition before it, and even today, some people consider it to be the ultimate word-processor ever created.

Indeed the author of that piece maintains a fan site that will tell you how to download and run WordPerfect for DOS on various modern computers,  if you have a legal copy of it. And, of course, if you run Windows, then the program is still very much alive and well and you can buy it from Corel Corp.

Sadly, the DOS version has never been made freeware. It still works – I have it running under PC-DOS 7.1 on an old Core 2 Duo Thinkpad, and it's blindingly fast. It also works fine on dosemu. It is still winning new fans today. Even the cut-down LetterPerfect still cost money. The closest thing to a free version is the plain-text-only WordPerfect Editor.

Edit: I do not know if Corel operates a policy like Microsoft, where owning a new version allows you run any older version. It may be worth asking.

But WordPerfect was not, originally, a DOS or a PC program. It was originally developed for a Data General minicomputer, and only later ported to the PC. In its heyday, it also ran on classic MacOS, the Amiga, the Atari ST and more. I recall installing a text-only native Unix version on SCO Xenix 386 for a customer. In theory, this could run on Linux using iBCS2 compatibility.

When Mac OS X loomed on the horizon, WordPerfect Corporation discontinued the Mac version – but when they did so, they made the last ever release, 3.5e, freeware.


(Image source.)

Of course, this is not a great deal of use unless you have a Mac that can still run Classic – which today means a PowerPC Mac with Mac OS X 10.4 or earlier. However, hope springs eternal: there is a free emulator called SheepShaver that can emulate classic MacOS on Intel-based Macs, and the WPDOS site has a downloadable, ready-to-use instance of the emulator all set up with MacOS 9 and WordPerfect for Mac.

To be legal, of course, you will need to own a copy of MacOS 9 – that, sadly, isn't free. Efforts are afoot to get it to run natively on some of the later PowerMac G4 machines on which Apple disabled booting the classic OS. I must try this on my Mac mini G4 and iBook G4.

The non-Windows version of WordPerfect that lived the longest, though, was the Linux edition. Corel was very keen on Linux. It had its own Linux distro, Corel LinuxOS, which had a very smooth modified KDE and was the first distro to offer graphical screen-resolution setting. Corel made its own ARM-based Linux desktop, the NetWinder, as reviewed in LinuxJournal.

And of course it made WordPerfect available for Linux.

Edit: Sadly, though, Microsoft intervened, as it is wont to do. The programs in WordPerfect Office originally came from different vendors. Some reviews suggested that the slightly different looks and feels of the different apps would be a problem, compared to the more uniform look and feel of MS Office. (The Microsoft apps in Office 4 were very different from one another. Office 95 and Office 97 had a lot of effort put in to make them more alike, and not much new functionality.)

Corel was persuaded to license the MS Office look-and-feel – the button bars and designs – and the macro language (Visual BASIC for Applications) and incorporate them into WordPerfect Office.

But the deal had a cost above the considerable financial one: Corel had to discontinue all its Linux efforts. So it sold off Corel LinuxOS, which became Xandros. It sold its NetWinder hardware, which became independent. It killed off its native Linux app, and ended development of WordPerfect Office for Linux, which was a port of the then-current Windows version using Winelib. In fact, Corel contributed quite a lot of code to the WINE Project at this time in order to bring WINE up to a level where it could completely and stably support all of WordPerfect Office.

I'm not sure if the text-only WordPerfect for Unix ever had a native Linux version – I didn't see it if it did – but a full graphical version of WordPerfect 8 was included with Corel LinuxOS and also sold at retail. Corel offered both a free edition with fewer bundled fonts, as well as a paid version.

This is still out there – although most of its mirrors are long gone, the Linux Documentation Project has it. It's not trivial to install a 20-year-old program on a modern distro, but luckily, help is at hand. The XWP8Users site has offered some guidance for many years, but I confess I never got it to work except by installing a very old version of Linux in a VM. For instance, it's easy enough to get it running on Ubuntu 8.04 or 8.10 – Corel LinuxOS was a Debian-derivative, and so is Ubuntu.

The problem is that even in these days of containers for everything, Ubuntu 8 is older than anything supports. Linux containers came along rather later than 2008. In fact, in 2011 I predicted that containers were going to be the Next Big Thing. (I was right, too.)

So I've not been able to find any easy way to create an Ubuntu 8.04 container on modern Ubuntu. If anyone knows, or is up for the challenge, do please get in touch!

But the "Ex WP8 Users" site folk have not been idle, and a few months ago, they released a big update to their installation instructions. Now, there's a script, and all you need to do is download the script, grab the WordPerfect 8.0 Downloadable Personal Edition (DPE), put them in a folder together and run the script, and voilá. I tried it on Ubuntu 20.04 and it works a treat so long as I run it as root. I have not seen any reports from anyone else about this, so it might be just my installation.

Read about it and get the script here.

Edit:

For more info, read the WordPerfect for Linux FAQ. This includes instructions on adding new fonts, fixing the MS Word import filter and some other useful info.

From the discussion on Hackernews and the FAQ, I should note that there are terms and conditions attached to the free  WP 8.0 DPE. It is only free for personal, non-commercial use, and some people interpret Corel's licence as meaning that although it was a free download, it is not redistributable. This means that if you did not obtain it from Corel's own Linux site (taken down in 2003) or from an authorised re-distributor (such as bundled with SUSE Linux up to 6.1 and early versions of Mandrake Linux, and the "WordPerfect for Linux Bible" hardcopy book, and a few resellers) then it is not properly licensed.

I dispute this: as multiple vendors did re-distribute it and Corel took no action, I consider it fair play. I also very much doubt that anyone will use this in a commercial setting in 2021.

If you are interested in the more complete WordPerfect 8.1, I note that it was included in Corel LinuxOS Deluxe Edition and that this is readily downloaded today, for example from the Internet Archive or from ArchiveOS. However, unless you bought a licence to this, this is not freeware and does not include a licence for use today.



(Image source.)

Postscript

If you really want a free full-function word-processor for DOS, which runs very well under DOSemu on Linux, I suggest Microsoft Word 5.5. MS made this freeware at the turn of the century as a free Y2K update for all previous versions of Word for DOS.

How to get it: Microsoft Word for DOS — it’s FREE

Sadly, MS didn't make the last ever version of Word for DOS free. It only got one more major release, Word 6 for DOS. This has the same menu layout and the same file format as Word 6 for Windows and Word 6 for Mac, and also Word 95 in Office 95 (for Win95 and NT4). It's a little more pleasant to use, but it's not freeware — although if you own a later version of Word, the licence covers previous versions too.

Here is a comparison of the two: Microsoft Word 5.5 And 6.0 In-depth DOS Review With Pics

I have several RasPis lying around the place. I sold my π2 when I got a π3, but then that languished largely unused for several years, after the fun interlude of getting it running RiscOS in an old ZX Spectrum case.

Then I bought myself a π3+ in a passive-cooling heatsink/case for Yule 2018, which did get used for some testing at work, and since then, has also been gathering dust. I am sure this is the fate of many a π.

The sad thing about the RasPi is that it's a bit underpowered. Not unreasonable for a £30 computer. The π1 was a single rather gutless ARM6 core. The π2 at least had 4 cores, but still weedy ones. The π3 had faster cores and wifi, but all still only have 1GB of non-upgradable RAM. They're not really up to running a full Linux desktop. What's worse, the Ethernet and wifi are USB devices, sharing the single USB2 bus with any external storage – badly throttling the bandwidth for server stuff. The π3+ is a bit less gutless but all the other limitations apply – and it needs more power and some form of cooling.

But then a chap on FesseBouc offered an official π touchscreen, used and a bit cheaper than new. That gave me an idea. I listen to a lot of BBC 6music – I am right now, in fact – but it needs a computer. Czech radio seems to mainly play a lot of bland pop which isn't my thing, and of course I can't understand a useful amount of Czech yet. It's at about the level of my Swedish in 1993 or so: if I listen intently and concentrate very hard, I may be able to work out the subject being discussed, but not follow the discussion.

But I don't want to leave a laptop on 24×7 and I definitely don't want a big computer with a separate screen, keyboard and mouse doing it. What I want is something the size of a radio but which can connect to wifi and stream music to simple old-fashioned wired speakers, without listening to me. I most definitely do not want a spy basestation for a dot-com listening to my home, thank you.

So I bought the touchscreen, connected it to my old π3, powered them both off a couple of old phone chargers, bunged in a spare µSD card, and started playing with software. I know where I am with software.

First I tried OSMC. It worked, detected and used the touchscreen, and could connect to my wifi... but it doesn't directly support streaming audio, as far as I can tell, and I could not work out how to install add-ins, nor how to update the underlying Linux.

I had a look at LibreElec but it looked very similar. While I don't really want the bloat of an entire general-purpose Linux distro, I just want this to work, and I had 8GB to play with, which is plenty.

So next I tried XBian. This is a cut-down Debian, running on Btrfs, which boots straight to Kodi. Kodi used to be called XBox Media Centre, and that's where I first met it – I softmodded an old original black XBox that my friend Dop gave me and put XBMC on it. It streamed movies off my server and played DVDs through my TV set, which is all I needed.

XBian felt a lot more familiar. It has a settings page through which I could update the underlying OS. It worked with the touchscreen out of the box. It has a UI for connecting to wifi. It too didn't include streaming Internet radio support, but it had a working add-ons browser, in which I found both BBC iPlayer and Internet Radio extensions.

Soon I was in business. It connected to wifi, it was operable with the touchscreen, connected to some old Altec Lansing speakers I had lying around. So I bought a case from Mironet, my friendly local electronics store. (There is a veritable Aladdin's Cave even closer to my office, GM electronic – but I'm afraid they're not very friendly. Sort of the opposite, in fact.)

I assembled the touchscreen and π3 into my case, and hit a problem. Only one available opening for a µUSB lead, but the screen needs its own. Some Googling later, it emerges than you can power the touchscreen from the π's GPIO pins, but I don't have the cables.

So off to GME it was, and some tricky negotiations later, I bought a strip of a dozen jumper cables. Three of them got me in business, but since it was £1 for all of them, I can't really complain about the wastage.

So now, there's a little compact unit in my bedroom which plays the radio whenever I want, on the power usage of a lightbulb. No fans, no extra cooling, nothing. I've had to use my single Official Raspberry Pi PSU brick, as all my phone chargers gave me the lightning-bolt icon undervoltage warning.

This emboldened me for Project 2.

Some years ago, Morgan's had a cheap offer on 2TB hard disks. I bought all their remaining stock, 5 mismatched drives. One went into an external case for my Mac mini and later died. The other four were in a box, pending installation into my old HP Microserver G1, which currently has 4×300GB drives in it, in a Linux software RAID controlled by Ubuntu. (Thanks to hobnobs!) However, this only has 2GB of RAM, and I figured that wasn't enough for a 5TB RAID. I may have accidentally killed it trying to fit more RAM, and the job of troubleshooting and fixing it has been waiting for, um, a couple of years now.

Meanwhile, the iMac's 1TB Fusion Drive was at 97.5% full and I don't have any drives big enough to back up everything on it.

I slowly and reluctantly conceded to myself that it might be quicker and easier to build a new server than fix and upgrade the old one.

The Raspberry Pi 4 is quite a different beast. Apart from a beefier 64-bit ARM7 quad-core, it has 2GB and 4GB RAM options, and it has much faster I/O. Its wifi and Ethernet are directly attached to the CPU, not on the USB bus, and it has 2 of those: the old USB2 bus (480Mb/s) and a new, separate 5Gb/s USB3 bus. This is useful power. It can also drive dual monitors via twin µHDMI ports.

But the π4 runs quite hot. The Flirc case my π3+ is in is only meant for home theatre stuff. A laden π4 needs something beefier, and sadly, my local mail-order electronics place, Alza, doesn't offer anything that appealed. I found the Maouii case on Amazon Germany and that fit the bill. (It also gave me a good excuse to buy the entire Luna trilogy by Ian McDonald in order to qualify for free shipping.)

So, from Alza I ordered a 4GB π4 and 4 USB3 desktop drive cases. From Mall CZ I ordered a USB3 hub with a fairly healthy 2.5A power output, thinking this would be enough to power a headless π4. USB C cables and µSD cards I have, and I figured all the USB 3 cables would come with the enclosures, which they did. In these quarantine lockdown times, the companies deliver to electronically-controlled mailboxes in shopping malls and so on, where you enter a code and pick up your package without ever interacting with a potentially-infectious human being.

It was all with me within days.

Now, I place some trust in those techies that I know who are more skilled and experienced than I, especially if they are jaded, cynical ones. File systems are one of the few significant differentiating factors between modern Linux server distros. Unfortunately, a few years ago, the kernel maintainers refused to integrate EVMS and picked the far simpler LVM instead. This has left something of a gap, with enterprise UNIXes still having more sophisticated storage tech than Linux. On the upside, though, this is driving differentiation.

SUSE favours Btrfs, although there's less enthusiasm outside the company. It is stable, but even now, you're recommended not to try to repair a Btrfs filesystem, and it can't give a reliable answer to the 'df' command – in other words, the basic question "how much free space have I got left?"

I love SUSE's YaST admin tool, and for other server stuff, especially on x86, I would probably recommend it, but it's not ideal for what I wanted in this role. Its support for the π4 is a bit preliminary so far, too.

Red Hat has officially deprecated Btrfs, but that left it with the problem that LVM with filesystems placed on top is a complex solution which still leaves something lacking, so with its typical galloping NIH syndrome, it is in the process of inventing an entirely new disk management layer, Stratis. Stratis integrates SGI's tried-and-tested, now-FOSS XFS filesystem with LVM into a unified disk management system.

Yeah, no thanks. Not just yet. I am not fond of Fedora, anyway. No stable or LTS versions (because that's RHEL's raison d'etre). CentOS is a different beast, and also not really my thing. And Fedora is also a bit more bleeding-edge than I like. I do not consider Fedora a server OS; it's more of a rolling tech testbed for RHEL.

Despite some dissenting opinions, the prevailing opinion seems to be that Sun's ZFS is the current state of the art. Ubuntu has decided to go with ZFS, although its license is incompatible with the Linux kernel's GPL. Ubuntu is, to be honest, my preferred distro for desktop stuff, and I've run it on πs before. It works well – better than Fedora, which like Debian eschews non-free drivers completely. It doesn't have Raspian's hardware acceleration but then everyone uses Raspbian on the π so it's an obvious target.

So, Ubuntu Server. Modern versions include ZFS built-in.

I tested this in a VM. Ubuntu Server 18.04 on its own ext4 boot drive... then add a bunch of 20GB drives to the VM... then tell it to create a RAIDZ. One very short time later, it has not only partitioned my drives, created an array, and formatted it, it's also created a mount point and mounted the new array on it. In seconds.

This is quite impressive and far more automatic than the many manual steps involved in doing this with the old Linux built-in 'mdraid' subsystem, as used in my old home server.

Conveniently – it was totally unplanned – by the time all my π4 bits were here, a new Ubuntu LTS was out, 20.04.

I installed all my drives into their new enclosures, plugged them one-by-one into my one of iMac's USB3 ports, and checked that they registered as 2TB drives. They did. Result. Oh, and yes, the cables were in the boxes. USB3 cables are entertainingly fat with shielding, but 5Gb/s is not to be sniffed at.

So, I put my new π4 in its case, put the latest Ubuntu Server on a µSD card – and hit a problem. I can't connect a display. I only have one HDMI monitor and nothing that will connect to a π4's micro-HDMI ports. And I don't really want to try to set this all up headless.

So off to Alza's actual physical shop I trogged to buy a µHDMI to HDMI convertor. Purchasing under quarantine is tricky, so it took a while, but I got it.

Fired up the π4 and it ran fine. No undervoltage warning running off the hub. So I hooked up all the drives, and sure enough, all were visible to the 'lsusb' command.

I referred to various howtos. Hmm. Apparently, you need to put partition records on them. Odd; I thought ZFS subsumed partitioning. Oh, well. I put an empty GUID disklabel on each drive. Then I added them to a RAIDZ, ZFS' equivalent of a RAID5 array.

Well, it wasn't as quick as in an VM, but only a minute or so of heavy disk activity later, the array is created, formatted, its mountpoint created and it's online. This is quite impressive stuff.



Then came the usual joys of Linux' fairly poor subsystem integration: Samba is a separate, different program, Samba user accounts are not Linux user accounts so passwords are different. Mounted filesystems inherit the permissions of their mountpoint. Macs still favour the old Apple protocol, so you need Netatalk as well. It, of course, doesn't integrate with Samba. NFS has two alternatives, and neither, of course, integrate with either Samba or Netatalk. There are good reasons NT caught on, which Apple successfully imitated and even exceeded in Mac OS X – and the Linux world remains as blindly indifferent to them as it has for a quarter of a century.

But some hours of swearing later, it all works. I can connect from Windows, Linux or Mac. It's all passively-cooled so it runs almost completely silently. It does need five power sockets, which is a snag, and there's a bit of cable spaghetti, but for an outlay of about £150 I have a running server which can sustain write speeds of about a gigabyte per second to the array.

I've put my old friend Webmin on it for a friendly web GUI.


So there you are.

While the π3 is a little bit underpowered, for a touchscreen Internet radio, it's great, and I'm very pleased with the result.

But the π4 is very different. It's a thoroughly capable little machine, perfectly usable as a general-purpose desktop PC, or as a server with quite decent bandwidth.

No, the setup has not been a beginner-friendly process. Apparently OpenMediaVault has versions for some single-board computers, including the π3, but not for the π4 yet. I am sure wider support will come.

But overall I'm impressed with how easy this was, without vast expert knowledge, and I'm delighted with the result. I will keep you posted on how it works longer-term.

A lot of my speculations concern the future of new, alternative operating systems which could escape from old-fashioned, sometimes ill-conceived models and languages.

But I do spend some time thinking about what is happening with Linux, with FOSS Unix in general, and especially with container technologies, something I deal with in my current and recent day-jobs more and more.

One answer to legacy nastiness for years now has been to virtualise it. Today, that's changing to "containerise it".

There is a ton of cruft in Linux and in the BSDs and so on which nobody is ever going to fix. It's too hard, it would break too much stuff... but most of all, there is no commercial pressure to do it, so it's not going to happen.

I can certainly see potentialities. There are parallels that run quite deep.

For instance, consider a few unrelated technologies:

- FreeBSD jails and Solaris Zones. Start here.

They indirectly evolved into LXC, the container mechanism in the Linux kernel which gets relatively little attention. (Docker has critical mass, systemd namespaces are trendier in some niches, CRIO is gaining a little bit of traction.)

Docker now means Linux containers are a known thing, already widely-used with money being poured into their R&D.

Joyent, a company with some vision, saw a chance here. It took Illumos, the FOSS fork of Solaris, and revived and modernised some long-dead Sun code: lxrun, the Linux runtime for Solaris. Joyent SmartOS is therefore a tiny Solaris derivative -- it runs entirely from RAM, booted off a USB stick, but can efficiently scale to hundreds of CPU cores and many terabytes of RAM -- which can natively run Docker Linux containers.

You don't need to run a hypervisor. (It is a hypervisor, if you want that.) You don't need to partition the machine. You don't even need a single copy of Linux on it. You have a rack of x86-64 boxes running SmartOS, and you can throw tens of thousands of Docker containers at them.

It gives capacities and scalability that only IBM mainframes can approach.

Now, if one small company can do this with some long-unmaintained code, then consider what else could be done with it.

 - Want more resilient hosts for long-lived containers? Put some work into Minix 3 until it can efficiently run Linux containers. A proper fully-modular-all-the-way-down microkernel which can detect when its constituent in-memory services fail and restart them. It can in principle even undergo binary version upgrades, piecemeal, on a running system. This is stuff Linux vendors can't even dream of. It would, for a start, make quite a lot of the Spectre and Meltdown vulnerabilities moot, because there's no shared kernel memory space.

Unlike Darwin and xnu, it's a proper microkernel -- no huge in-kernel servers for anything here. (Don't eve try to claim WinNT is a microkernel or I will slap you.) Unlike the GNU HURD, it's here, it works, it's being very widely used for real workloads. And it's 100% FOSS.

 - Want a flexible cluster host which can migrate containers around a globe-spanning virtual datacenter?

Put some work into Plan 9's APE, its Linux runtime. Again, make it capable of running Linux containers. To Plan 9 they'd just be processes and it was built to efficiently fling them around a network.

I have looked into container-hosting Linux distros for several different dayjobs. I can't give details, but they scare me. One I've tried has a min spec of 8GB of RAM and 40GB of disk per cluster node, and a minimum of 3-4 nodes.

This is not small efficient tech. But it could be; SmartOS shows that.

 - Hell, more down to earth -- many old Linux hands are deserting to FreeBSD in disgust over systemd. FreeBSD already has containers and a quite current Linux runtime, the Linuxulator. It would be relatively easy to put them together and have FreeBSD host Linux containers, but the sort of people who dislike systemd also dislike containers.

Not everything would run under containers, sure, no. But they're suitable for far bigger workloads than is generally expected. You can migrate a whole complex Linux server into a container -- P2V migration as was once common when moving to hypervisors. I've talked to people doing it.

Ubuntu LXD is specifically intended for this, because Ubuntu isn't certified for SAP, only SUSE is, so Ubuntu wants to be able to run SLE userlands. Ditto some RHEL-only stuff.

But what if it doesn't work with containers at all?

Well, as parallels...

[1] A lot of Win32 stuff got abandoned with the move to WinXP. People liked the new OS enough that stuff that didn't work got left behind.

[2] Apple formalised this with Carbon after the NeXT acquisition. The MacOS APIs were not clean and suitable for a pre-emptive multitasking OS. So Apple stripped them out and said "if you use this subset, your app can be ported. If you don't, it can't."

Over the next few years, the old OS was forcibly phased out -- there is a generation of late-era gigahertz-class G4 and G5 PowerMac that refuses to boot classic MacOS. Apple tweaked the firmware to prevent it. You _had_ to run OS X on them, and although versions >= 10.4 could run a Classic MacOS VM, not everything worked in a VM.

So the developers had to migrate. And they did, because although it was a lot of work, they wanted to keep selling software.

It worked so well that in the end the migration from PowerPC to Intel was less painful than the one from classic MacOS to OS X.

So maybe Linux workloads that won't work in containers will just go away, replaced by ones that will -- and apps that play nice in a container don't care what distro they're on, and that means that they will run on top of SmartOS and FreeBSD and maybe in time Minix 3 or Plan 9.

And so we'll get that newer, cleaner, reworked Unix after all, but not by any incremental process, by a quite dramatic big-bang approach.

And if there comes a point when it's desirable to run these alternative OSes for some users, because they provide useful features in nice handy easy ways, well, maybe they'll gain traction.

And if that happened, then maybe some people will investigate native ports instead of containerised Linux versions, and gain some edge, and suddenly the Unix world will be blown wide open again.

Might happen. Might not. It's not what I am really interested in, TBH. But it's possible -- existing products, shipping for a few years, show that.

I have been meaning to try Arch Linux for years.

As a former RPM user, once I finally made the switch to Ubuntu, more or less exactly 10y ago, well, since then, I have become so wedded to APT that I hesitate with non-APT distros.

My spare system on this machine is Crunchbang, which I like a lot, but is a bit too Spartan in its simplicity for me. Crunchbang is based on the stable version of Debian, which gives it one big advantage on my 2007-era built-for-Windows-Vista hardware: it uses a version of X.org so old that the ATI fglrx drivers for my Radeon HD 3470 GPU still work, which they haven't done on Ubuntu for 2 years now.

But there was a spare partition or 2 waiting. I tried Elementary -- very pretty, but the Mac OS X-ness is just skin-deep; it's GNOME 3, very simplified. No ta. Deepin is too slow and doesn't really offer anything I want -- again, it's a modification of GNOME 3, albeit an interesting one. Same goes for Zorin-OS. I've tried Bodhi before -- it's interesting, but not really pretty to my eyes. (Its Enlightenment desktop is all about eye-candy; as a desktop, it's just another Windows Explorer rip-off. If it shipped with a theme that made it look like one of those shiny floaty spinny movie-computer UIs, I might go for it, but it doesn't, it's all lairy glare that only a teenage metalhead could love.) Fedora won't even install; my partitioning is too complex for its installer to understand. SUSE is a bit bloaty for my tastes, and I don't like KDE (or GNOME 3), which also rules out PCLinuxOS and Deepin.

So Arch was the next logical candidate...

I've been a bit sheepish since an Imaginary Internet Friend, Ric Moore, tried it with considerable success a month or two ago. (As I write, he's in hospital having a foot amputated. I've been thinking of him tonight & I hope he's doing well.)

So I have finally done it. Downloaded it, burned it to a CD -- yes, it's that small -- installed it on one of my spare partitions and I am in business.

After a bit of effort and Googling, I found a simple walkthrough, used it, got installed -- and then discovered that Muktware only tells you about KDE, and assumes you'll use that and nothing else. I don't care for KDE in its modern versions, so I went with Xfce.

Getting a DM working was non-trivial but now I have LXDM -- the 3rd I tried -- and it works. I have an XFCE4 desktop with the "goodies" extras, Firefox, a working Internet connection via Ethernet, and not much else.

It does feel very quick, though, I must give it that. Very snappy. I guess now begins the process of hunting down all the other apps that I use until I've replicated all my basic toolset.

The install was a bit fiddly, much more manual than anything I've done since the mid-1990s, but actually, it all went on very smoothly, considering that it's a lot of hand-entered commands which actually do not seem to depend much on your particular config.

Facebook readers may have noted my post yesterday, when I mentioned that I was trying to resurrect an old notebook with a dead screen by using a screenreader. I commented:

"Just spent an hour trying to update a fresh install of Windows XP SP3 on a PC with no screen, using speech alone. Haven't felt so lost since 1988. It's currently on 100 of 125, though, which is a sort of success..."

Well, I've spent a little more time on it today.

According to http://update.microsoft.com I now have all essential updates installed. I'm not feeling brave enough to tackle the optional updates just yet - I'm still terrible at navigating web pages.

I've also managed to install MS Security Essentials, and currently, Ninite claims to be installing Opera, OpenOffice and a FOSS PDF reader.

It's a very chastening experience. I am a dab hand with driving Windows without a mouse - I learned on Windows 2.0 in the days when my employers didn't own a PC mouse. But much of the XP and Windows apps' UI is either inaccessible by keyboard, unreadable or just unlabelled.

For instance, stepping through the icons in the notification area, I get "icon... icon... NVDA... icon... Automatic updates... clock." Selecting each icon and opening it is the only way to find out what it's the icon for. One gives the wireless network connection info, for instance, but some lazy-ass Microsoft programmer forgot to give it a text label.

The entire UI of the MS Security Essentials consists of the following: "home... update... options... scan... exit." That's it. No legible text at all. I can open Task Manager and move between the tabs, but there's no way to sort the list of tasks to find what is hogging the system. That needs a mouse-click.

Progress bars are unreadable, but NVDA makes a series of rising beeps to tell you that something's happening. It's hard to tell how far you've got, though. The mandatory Windows Genuine Authentication installer stops at about 80%, every time, even after 3 reboots. I gave up and used a third-party WGA killer app to nuke it into oblivion.

And I've compared notes with ednun on this. Ubuntu seems to be about the best Linux for accessibility, with an integrated screenreader, Orca - but it can read considerably less than NVDA can. Windows does seem to be the best option.

It's quite scary. Certainly I'm nowhere near being able to post status updates from a screenless PC.

(Weird font changes courtesy of the LJ rich-text edit control. Sorry about that.)

I have spent a lot of time and effort this year on learning my way around the current generation of Windows Server OSs, and the end result is that I've learned that I really profoundly dislike them.

Personally, I found the server admin tools in NT 3 and 4 to be quite good, fairly clean and simple and logical - partly because it was built on LAN Manager, which was IBM-designed, with a lot of experience behind it.

Since Windows 2000 Server, the new basis, Active Directory, is very similar that of Exchange. Much of the admin revolves around things like Group Policies and a ton of proprietary extensions on top of DNS. The result is a myriad of separate management consoles, all a bit different, most of them quite limited, not really following Windows GUI guidelines because they're not true Windows apps, they're snap-ins to the limited MS Management Console. Just like Exchange Server, there are tons and tons of dialog boxes with 20 or 30 or more tabs each, and both the parent console and many of the dialogs containing trees with a dozen+ layers of hierarchy.

It's an insanely complicated mess.

The main upshot of Microsoft's attempts to make Windows Server into something that can run a large, geographically-dispersed multi-site network is that the company has successfully brought the complexity of managing an unknown Unix server to Windows.

On Unix you have an unknown but large number of text files in an unknown but large number of directories, which use a wide variety of different syntaxes, and which have a wide variety of different permissions on them. These control an unknown but large number of daemons from multiple authors and vendors which provide your servers' various services.

Your mission is to memorise all the possible daemons, their config files' names, locations and syntaxes, and use low-level editing tools from the 1960s and 1970s to manage them. The boon is that you can bring your own editors, it all is easily remotely manageable over multiple terminal sessions, and that components can in many cases be substituted one for another in a somewhat plug-and-play fashion. And if you're lucky enough to be on a FOSS Unix, there are no licensing issues.

These days, the Modern way to do this is to slap another layer of tools over the top, and use a management daemon to manage all those daemons for you, and quite possibly a monitoring daemon to check that the management daemon is doing its job, and a deployment daemon to build the boxes and install the service, management and monitoring daemons.

On Windows, it's all behind a GUI and now Windows by default has pretty good support for nestable remote GUIs. Instead of a myriad of different daemons and config files, you have little or no access to config files. You have to use an awkward and slightly broken GUI to access config settings hidden away in multiple Registry-like objects or databases or XML files, mostly you know or care not where. Instead of editing text files in your preferred editor, you must use a set of slightly-broken irritatingly-nonstandard and all-subtly-different GUIs to manipulate vast hierarchical trees of settings, many of which overlap - so settings deep in one tree will affect or override or be overridden by settings deep in another tree. Or, deep in one tree there will be a whole group of objects which you must manipulate individually, which will affect something else depending on the settings of another different group of objects elsewhere.

Occasionally, at some anonymous coder's whim, you might have to write some scripts in a proprietary language.

When you upgrade the system, the entire overall tree of trees and set of sets will change unpredictably, requiring years of testing to eliminate as many as possible of the interactions.

But at least in most installs it will all be MS tools running on MS OSs - the result of MS' monopoly over some two decades being a virtual software monoculture.

But of course often you will have downversion apps running on newer servers, or a mix of app and server OS versions, so some machines are running 2000, some 2003, some 2008 and some 2008R2, and apps could span a decade or more's worth of generations.

And these days, it's anyone's guess if the machine you're controlling is real or a VM - and depending on which hypervisor, you'll be managing the VMs with totally different proprietary toolsets.

If you do have third-party tools on the servers, they will either snap-into the MS management tools, adding a whole ton of new trees and sets to memorise your way around, or they will completely ignore it and offer a totally different GUI - typically one simplified to idiot level, such as a enterprise-level backup solution I supported in the spring which has wizards to schedule anything from backups to verifies to restores, but which contains no option anywhere to eject a tape. It appears to assume that you're using a robot library which handles that automatically.

Without a library, tape ejection from an actual drive attached to the server, required a server reboot.

But this being Windows, almost any random change to a setting anywhere might require a reboot. So, for instance, Windows Terminal Services runs on the same baseline Windows edition, meaning automatic security patch installation - meaning all users get prompted to reboot the server, although they shouldn't have privileges to actually do so, and the poor old sysadmins, probably in a building miles away or on a different continent, can't find a single time to do so when it won't inconvenience someone.

This, I believe, is progress. Yay.

After a decade of this, MS has now decided, of course, that it was wrong all along and that actually a shell and a command line is better. The snag is that it's not learned the concomitant lessons of terseness (like Unix) or of flexible abbreviation (like VMS DCL), or of cross-command standadisation and homogeneity (although to be fair, Unix never learned that, either. "Those who do not know VMS are doomed to reinvent it, poorly," perhaps.) But then, long-term MS users expect the rug to be pulled from under them every time a new generation ships, so they will probably learn that in time.

The sad thing about the proliferation of complexity in server systems, for me, is that it's all happened before, a generation or two ago, but the 20-something-year-olds building and using this stuff don't know their history. Santayana applies.

The last time around, it was Netware 4.

Netware 3 was relatively simple, clean and efficient. It couldn't do everything Netware 2 could do, but it was relatively streamlined, blisteringly fast and did what it did terribly well.

So Novell threw away all that with Netware 4, which was bigger, slower, and added a non-negotiable ton of extra complexity aimed at big corporations running dozens of servers across dozens of sites - in the form of NDS, the Netware Directory Services. Just the ticket if you are running the network the size of Enron or Lehman Brothers, but a world of pain for the poor self-taught saps running single servers of millions of small businesses. They all hated it, and consequently deserted Netware in droves. Most went to NT4; Linux wasn't really there yet in 1996.

Now, MS has done exactly the same to them.

When Windows 2000 came around, Linux was ready - but the tiny handful of actual grown-up integrated server distros (such as eSmith, later SME Server) have never really caught on. Instead, there are self-assembly kits and each sysadmin builds their own. It's how it's always been done, why change?

I had hoped that Mac OS X Server might counteract this. It looked the The Right Thing To Do: a selection of the best FOSS server apps, on a regrettably-proprietary but solid base, with some excellent simple admin tools on top, and all the config moved into nice standard network-distributable XML files.

But Apple has dropped the server ball somewhere along the line. Possibly it's not Apple's fault but the deep instinctual conservatism of network and server admins, who would tend to regard such sweeping changes with fear and loathing.

Who knows.

But the current generation of both Unix and Windows server products both look profoundly broken to me. You either need to be a demigod with the patience and deep understanding of an immortal to manage them properly, or just accept the Microsoft way: run with the defaults wherever possible and continually run around patching the worst-broken bits.

The combination of these things is one of the major drivers behind the adoption of cloud services and outsourcing. You move all the nightmare complexity out of your company and your utter dependence on a couple of highly-paid god-geeks, and parcel it off to big specialists with redundant arrays of highly-paid god-geeks. You lose control and real understanding of what's occurring and replace it with SLAs and trust.

Unless or until someone comes along and fixes the FOSS servers, this isn't going to change - it's just going to continue.

Which is why I don't really want to be a techie any more. I'm tired of watching it just spiral downwards into greater and greater complexity.

(Aside: of course, nothing is new under the sun. It was, I believe, my late friend Guy Kewney who made a very plangent comment about this same process when WordPerfect 5 came out. "With WordPerfect 4.2, we've made a good bicycle. Everyone knows it, everyone likes it, everyone says it's a good bicycle. So what we'll do is, we'll put seven more wheels on it."

In time, of course, everyone looked back at WordPerfect 5.1 with great fondness, compared to the Windows version. In time, I'm sure, people will look back at the relative homogeneity of Windows 2003 Server or something with fondness, too. It seems inevitable. I mean, a direct Win32 admin app running on the same machine at the processes it's managing is bound to be smaller, simpler and faster than a decade-older Win64 app running on a remote host...)

In unrelated news, I had to bring up my main fileserver to retrieve the OpenSolaris & PC-BSD ISOs. Alas, its evaluation copy of Windows 2003 SBS has expired, but I get 1h to pull files off it each reboot, apparently.

I am considering trying to install Windows 2008 Server over the top. I don't care about saving my settings, I just don't want to have to backup & completely reformat. Alas, I seem to have lost my ISO of that.

Going looking, I found that W2k8 R2 is out & it is Micros~1's first ever 64-bit onlyOS. I'd missed this one. It's the server version of Windows 7, basically.

And I had no idea if the fairly-late-model Pentium 4 in my HP Proliant was a 64-bit capable one or not. I know it has hyperthreading, but not if it sports 64-bit extensions.

It doesn't have a mouse of its own and I couldn't get Ubuntu's rdesktop client to connect so I could run CPUID on it, so I tried logging in - only to be told that my time was up and be spanked with a BSOD. Thanks, Redmond.

An idea occurred. I could try booting my 64-bit Ubuntu CD. If it worked, it's 64-bit capable; if it doesn't, it's not.

Well, it's not, and Ubuntu helpfully printed a little message to tell me that I needed an x86-64 CPU and it could only find an x86-32 one. No worries; I am limited to original W2K8 Server then. I am sure I'll cope.

On a whim, I tried my copy of 32-bit Ubuntu 9.10, and to my considerable surprise, not only did it boot but it found the RAID controller and happily mounted my NTFS volume. I tried all manner of Linux distros on this last year - Ubuntu 8.04, 9.04, CentOS and SME Server - and none of them could see the RAID5 volume on its Dell-badged ALI MegaRAID card. So at some point late last year, they fixed the driver in the kernel.

Which was nice.

Which leaves me wondering... try to upgrade it to a newer Windows Server, in which I could do with more experience, or stick Ubuntu on it, which will probably be quicker and easier and more use, and won't date-expire on my in 6mths...?

A final caveat to my previous post: there is one thing you probably shouldn't try doing under XP-inna-VM: play games. The VM does sport optional 2D graphics acceleration, although I've spotted a few display glitches, but the copy of Windows in the VM can't get at your shiny whizzy fanheater of a 3D card & any modern 3D game is going to run like crap. For that, I'm afraid, you need to dual-boot into real native Windows.

TinyXP will do that just fine, but remember, you're going to have to find the latest drivers for every bit of kit in your machine. My advice:
- install TinyXP first, in a primary partition on the 1st hard disk.
- leave plenty of space for Linux; put all its partitions in logical drives in an extended partition
- next, after TinyXP is working but before it's got its drivers, install Ubuntu
- now, in Ubuntu, you can carefully peruse the output of

dmesg | less

... and work out what motherboard chipset you have, what graphics, sound, network card(s) &c. your machine is sporting. The best way to identify a motherboard, though, is just to look at it. Use a torch. You'll probably find the makers' name and the model number printed between the expansion slots.

- Using Linux, go download all the relevant Windows drivers from the manufacturers' websites.
- Go to Places | Computer and open your Windows partition. Copy the downloaded drivers into

C:\Documents and Settings\All Users\Desktop

- Then reboot into Windows again and they're all there, ready to install.

This method saves an awful lot of hassle trying to get Windows working if you have no driver disks.

If you install Ubuntu after Windows, it's smart enough to set up dual-boot for you. Install Windows after Ubuntu, it will screw your bootsector and you won't be able to boot Ubuntu any more. Also, Windows likes being in a primary partition, preferably the first, whereas Linux doesn't care.

Oh, and don't waste your time on anything other than Ubuntu. If you are at the level of expertise to have got any useful info from this piece, you probably don't need advice on choosing a distro... but just in case:

- OpenSUSE is huge and its package-management system is frankly a bit past it.
- Fedora is a sort of rolling beta. It never stabilises, it's not supported and there are no official media addons, which are free with Ubuntu.
- Kubuntu is OK if you're a KDE freak but if you don't know the difference between KDE & GNOME, just go for vanilla Ubuntu, which involves a lot less fiddling.
- Mandriva is OK but again its package-management system, like that in SUSE and Fedora, is a decade or so less advanced than the one in Ubuntu.
- Debian is too much like hard work unless you actively enjoy fiddling.
- Gentoo is for boy-racers, the sort of person who drives a 6Y old Vauxhall Nova with a full bodykit and a 150dB sound system. Just don't.
- All the rest are for Linux hackers. You don't want to go there.