In the UK the IBM PC was a very rare beast outside of a blue chip company, they were just too darned expensive.

In January 1984 the IBM PC XT was the only version on offer, using an 8088 or 8086, the PC AT had not even been announced. The tape only version with built in BASIC on ROM would probably set you back ~£1500, if you were lucky, with MGA mono-graphics maybe. Then you had to add the monitor for a couple of hundred more.

Most people and small companies just couldn't afford that sort of money. It was almost the price of a small car!

Even at £400, the Sinclair QL and BBC Model B were way out of the price range of the average person, being close to half a month's wages, if not more. This is the main reason that the ZX Spectrum sold so well.

Actually, there was a QL derivative sort of available, the Q40 and Q60 machines running SMSQ/E, the QDOS derivative.

Out of production currently though.

Having said that, there is still a small but active QL hardware and software development community.

Check out the QL-Users mailing list archives.

Hey, don't knock the Microdrives, they worked well enough.

Only now that the foam rubber pads under the tape on the cartridges are failing is the data being lost from them. Thankfully someone's found a solution to this and has got some high-quality felt pads made up as replacements. (See: RWAP Software for more details.)

It only took a year until 3.5" floppy disk drives and interface were available, if you had the cash. Remember, back then the drives cost a fortune and the disks themselves a significant amount of the weekly wage packet.

That's an urban myth and only slightly true of the early ZX Spectrum+ cases.

Erm... so how many machines costing less than £2000 had a 68000 processor and fully pre-emptive multitasking in January 1984? None.

The Macintosh, which you're possibly talking about, was announce two weeks after the QL and was half an order of magnitude more expensive.

This isn't to say that the QL was perfect, which it very much wasn't. Some of the design decisions, such as using the keyboard processor for serial receive and sound as well as the keyboard were nuts.

Tony Tebby's OS and SuperBASIC were way ahead of their time though.

So, is the quality of the output equivalent or has it suffered due to compromises due to the speed increase?

Probably is... as long as you don't show any nipples!

You have to remember that enterprise level storage isn't a single set of drives holding the data, it's a hierarchy of different technologies depending upon the speed of data access required. Since SSDs arrived they've been used at the highest access rate end of the spectrum, essentially using their low latency for caching filesystem metadata. I can see that now they are starting to replace the small, high speed drives at the front end entirely. However, it's going to be some time before they can even begin to replace the storage in the second tier and certainly not in the third tier storage where access time isn't an issue but reliable, "cheap" and large drives are required. Of course, beyond this tier you generally get on to massive robotic tape libraries anyway, so SSDs will never in the foreseeable future trickle down to here.

<quote>

<quote><p>I built a model to calculate the fuel consumption of locomotives on 24 routes crossing the nation. on each route, i had a record every tenth of a mile that calculated instantaneous speed, acceleration, and power. rolled it all up to aggregate fuel economy, horsepower, etc. metrics. more than 10^6 records. power user, bitch.</p></quote>

<p>Model building like that is probably better done in R anyway</p></quote>

Indeed, especially as Excel doesn't use IEEE standard floating point maths and doesn't round the results of arithmetic operations reliably and loses precision. It's also slow as it doesn't use the floating point unit of the processor.

Actually, other than a very few packages, I've not had any problems installing "generic" RPM packages on Mandriva or Mageia.

The biggest problem comes with dependencies which have different names, in which case you manually install the correct packages before forcefully installing the RPM with --ignore-deps.

Working in a scientific environment there are sometimes you just have to do this as the RPMs are only available for RHEL and nothing else.

It seems mostly to be a "me too" bragging right. MacOS has launchd, Solaris has svc.configd so someone thought that Linux needed one too.

On the whole it's also trying to boot marginally quicker, but not necessarily correctly. i.e. play fast and loose.

Let's face it, does it really matter if a server or desktop takes 20 seconds rather 30 seconds to boot if the machines going to have an uptime for several weeks?

Wouldn't it be better that it is guaranteed to be running correctly after 30 seconds rather than having services try to start up before the rest of the system is ready for them and failing?

Well, obviously I'm not hip and trendy enough and think that shiny-shiny is no substitute for correctly working.

The reason for the fork was the Mandriva fired all their French developers, moved production to a cheaper country and then totally broke the distribution (Mandriva 2011.0).

The original programmers took the Mandriva 2010.x distribution, forked it, updated it and made the Mageia (mage-ee-ah) 1 distribution, which actually worked.

Mageia 2 moved to systemd (*spit*) but generally didn't break backwards compatibility. I've been running the pre-release version of Mageia 3 on a server for the last month or so (because the chipset needed a newer kernel than previous releases had) and it's been very stable.

Subsequently, Mandriva's management have had a small rethink and are now basing their server distribution upon Mageia (because it actually works).

Of all the Linux distributions I've found the Mandrake/Mandriva/Mageia family to be the least primitive and actually work, both in a scientific computing desktop role and a server roll. They're generally hassle free and the update and upgrade system practically flawless.

Actually, they've not abandoned it, it's merely the thrusters are simulating the aircraft-like ability. If you look at the rolls video you see that the craft can do some very un-aircraft-like turns if the player wants them to.

It's actually a sequel and not a re-implementation.

Of course, it's still pre-alpha quality, so you can't judge the final game from the Skunk Works developed initial test system.