The Game What I Wrote

I can't remember if I've blathered on about this before - I don't think I have, but my memory isn't what it was - so let's do it again (if I did it before...I'm confusing myself now).

Back in 1987, when I was 15 (even thinking of that makes me feel old (which I am I suppose)) I wrote an adventure game for the ZX Spectrum called "Homicide Hotel".  It was done with a piece of software called the Professional Adventure Writer (PAW), so I can't claim to have written it from scratch with machine code (at the time my knowledge of Z80 - the machine language of the Spectrum - was pretty much limited to programming really bad side-scrolling shoot-'em-ups) but PAW had a language of its own that was quite complex.  So I was quite proud of my creation - as it all worked as I wanted it to, had a plot, characters and some in-jokes and everything - and I thought it was so good that I sent a copy to CRASH (the leading Spectrum-related magazine at the time; I didn't bother sending it to their main rival - Sinclair User - as they tended to concentrate on arcade games and ignored adventures, in the main part) in an attempt to get a mention and potentially sell it.

I didn't actually expect it to be reviewed, but it was, the review appearing in issue 44, dated September 1987, the cover of which (artwork by Oliver Frey - unrelated to my game, but it's such an artefact of the time that it's irresistible not to include it) is here:

At the time, the adventure game reviews were handled by a guy called Derek Brewster, who really knew his stuff; he'd coded some of the best early adventure games (Kentilla was probably the one that made the most impact) for the Spectrum and had been reviewing for Crash pretty much since its inception.  It's not like I saw him as a god or anything but I - and most others involved in that whole "scene" - definitely respected his opinion.  A good review from Brewster was kind of a stamp of authority, or authenticity, or something.

So you can't imagine my surprise and excitement when I got a complimentary copy of the magazine in the post (prior to its on-sale date, so this would have been early August 1987, around the time I turned 16, anyway), with a little paper slip saying "you are mentioned on page 69*" or something like that.  And lo and behold, there it was:

I seriously couldn't believe it!  84%!  I kept having to read it over and over again to assure myself that it had happened (incidentally, I've still got that copy of CRASH, 30 years on, and I imagine I'll keep it forever for nostalgia reasons).

The market for mail-order adventure games wasn't exactly big, but I got about 50-odd orders (and some surprisingly complimentary letters) from that review.  It doesn't sound much (and indeed it isn't much, in the grand scheme of things) but for a boy just turned 16 it was a big deal.  I probably made about £1.50 for each copy sold after the cost of the cassettes (I used C15s) and postage were taken into account.  This was Big Money to me then.

A year or so later, the market for Spectrum adventures had pretty much dried up - very few new ones were being produced - so when a guy called Tony Collins, who ran a software company called The Guild (which specialised in adventure games) got in touch and offered to take it on and include it in his stable of adventures, I naturally said "yes please!".  He paid me a "royalty" of 40p per copy sold and it went on to sell about another 50-odd copies before that whole thing faded away.

Looking back on it now, it was a bit of a rubbish game really (I think it's still playable on PCs through the World of Spectrum archive, so if you really want, you can find out for yourself just how rubbish it was).  It was far too difficult for a start - I've long forgotten how to complete it myself - but fucking hell, it was fun.  I particularly enjoyed writing the location descriptions and seeing how much extraneous stuff - silly in-jokes and such - I could pack into it.  Ah, nostalgia.


* Although being a young innocent at the time - and that it was undoubtedly a pure coincidence - the happenstance of the review being on page 69 didn't strike me as funny at the time, but it does now.  I am easily amused.

ZX Open World, Part 4

Sorry for the delay...

Excuses: had a birthday, which was all pretty cool, pissed about doing other stuff, then thought "ah shit", need to finish this thing, or at least reach some sort of (probably unsatisfactory) conclusion.

But anyway, excuses out of the way, my favourite open world game on the Spectrum - not to say the best, I lost touch with commercial Spectrum games in 1988 or so - was by quite a distance, Mercenary:

Crash-landing on Targ is imminent...

I didn't care that it was a port from an MSX game, nor that the Commodore 64 version was better in many ways, nor even that when the Amiga version came along, it knocked them all into a cocked hat. Mercenary for the Spectrum was - I think - as close to my ideal open world could be at the time. Basically, you crash-land on a planet (Targ - or was the city called Targ? - I forget) which is currently experiencing some sort of civil war (between the Palyars - the goods ones - and the Mechanoids - the baddies - but this isn't immediately apparent). When you land, you're pretty much on your own to do as you please (it definitely helps that you crash-land there's a flying craft nearby - walking around this place takes a long, long time).

The City of Targ becomes visible as
you plummet helplessly...

Handily, a craft is immediately
available on landing

The logical thing to do is to buy the craft that's available when you crash-land. That way you can get yourself exploring from the air, which is far faster than walking around. But you can walk around if you want. You might come across a land-based craft that'll get you around a lot faster, but your line of sight is far more limited than it would be from the air.

Point is, anything went, pretty much.

As can be seen above, the city was broadly laid out in a grid format, although not all co-ordinates contained anything of interest (not on the ground, anyway, although there might be something hovering up above...)

It's Milton Keynes...
in bizarro world

This is a view flying over the city. It's difficult to describe how amazingly well all the flying craft were (once you'd got the hang of them). There was a real feeling of weight and momentum and speed; there was a genuine physics engine working in there.

Of course, this was a 48K Spectrum game (41.5K usable blah blah blah) so by necessity, it was sparse. Boy, was it sparse. Vector graphics to the max.

But that added to the atmosphere if you ask me. Mercenary had an atmosphere all of its own.

Here's a bridge (pictured left). It's a bit like the Humber Bridge, but it only spans a road. Or does it?

Here's a...actually I can't remember what the fuck that was (pictured right). I'm sure it added to the mystery, though.

And there was plenty of mystery, for sure. Mysterious satellite installations! Mysterious forests!

Mysterious circus big tops which were something else but nevertheless looked like circus big tops!

And you could of course blow all this shit up if you wanted (given a suitably-equipped craft). What more is there to like?

The hangar at 09-06. Some nice stuff in there.

Well...here's something. It's a hangar from the outside, but it has an elevator to a subterranean level.

There's a good few of these dotted around the city. Some of them contain useful stuff but some of them just do their best to kill you.

'Course, there's no way of winning without going into the subterranean complexes, as they contain all the stuff you might need to finish the game (plus all the cool stuff).

For a start, you'll find new flying craft with far better capabilities than the one you start with. And you will ultimately need one, as the starting craft is nowhere near capable of getting to the places you need to get to, as it simply can't make the altitude.

Let's take a trip...

This is descending into a hangar and not knowing what the fuck is going on, incidentally. Not that that's a bad thing; in fact, in this game, it was the only way to figure out - gradually - how best to play.

Turns out the triangular doors seen here need a triangular key to get through, but that's not obvious from the outset.

Course, there's all kinds of weird shit down there, it's not all enormous blue rooms with triangular doors.

No, quite the contrary, the area to explore underground was - in terms of mapping data - probably bigger than the above-surface stuff; there were quite a few of these underground complexes, each with different things in them, f'rinstance:

Yellow corridors!

Red rooms!

Blue corridors with lethal
spider webs in them!

Mysterious purple doors!

  

It all looks so basic and primitive now, but the simplicity of the graphics added to it all somehow...as I mentioned before, the atmosphere generated within Mercenary was unique.  I genuinely loved this game. I felt that this was the future of gaming; the basis being that game landscapes could become increasingly vast whilst the detail could be generated depending on the player's point of view.

There's an interesting interview with David Aubrey-Jones (from Crash edition 44, possibly more on which anon) here and sometimes I wonder what people like him are doing these days.

I bet whatever it is, it's not as exciting as what he was doing in 1987.

ZX Open World, Part 3(a)

Before embarking...

...on what looks like it's going to turn into a very long post about another game, I just wanted to drop in the brief story of the Game That Never Was (except for later, when it was overhauled and was released, but let's not dwell on that for now), The Last Ninja.

It did well on other platforms and spawned sequels, but the Spectrum version was doomed to failure; it seemed to be imminent all the way through 1987 and 1988 but never actually appeared, which was a shame, as when it did eventually appear for the superannuated Spectrum (as Last Ninja 2, some two and a half years late) it looked damn good:

These screens are from Last Ninja 2, I think, or maybe another of the versions that was eventually completed for the Spectrum. It doesn't exactly look state of the art now, but back in 1987, when I still believed this thing was going to appear, I was pretty excited.

It looked like a proper cross between Saboteur and Turbo Esprit, with proper solid-looking graphics, fighting, exploring, etc. which was pretty much my dream game in 1987. But by the time it eventually appeared, I'd long moved on.

ZX Open World Part 3

Isometric 3D wasn't the only game in town, of course.

There were a number of other games that came out on the Spectrum during the "isometric wars" that can't be ignored, either because they were direct attempts at open world games, or contained elements or ideas that would later prove essential for future advances. Funnily enough, the programmers or teams writing games with non-isometric 3D views tended to produce very unusual, individualistic (and sometimes downright odd) projects.

Gyron: DANGER - GIGANTIC
BALLS ON THE LOOSE

Gyron: IMMINENT LARGE BALL
ENGULFMENT DANGER

One that came out in late 1985 was Gyron. I can't say I ever really fully understood what the hell was going on in this one. You were in a series of mazes, which had giant balls rolling around in them, and towers that kept shooting at you.

It was apparently a great game once you understood what was happening, but I never got that far, I'm afraid, so I can't report in any detail, and I'm not prepared to play through it on an emulator to find out. Life, I fear, is too short.

Tau Ceti: Four views of the game, there

The other game I always group with Gyron is Tau Ceti (and as it turns out, the latter was directly inspired by the former, so there's a nice thing). Tau Ceti was actually a great game and given the excessive delays that were affecting the Spectrum port of Elite (as mentioned in part 1), it sort of stole its thunder. Elite appeared on the Spectrum at much the same time, but Tau Ceti was more immediate, had better graphics, was faster, etc.

Elite had its fans but the Spectrum version could never hope to replicate the massive success of the versions on other platforms; it had simply been beaten to the market and no longer really had a USP. In fact, there was a comparitively vast space-based exploring/ collecting game with vector graphics that had been out over six months already, in (Z80 programming pioneer) David Webb's Starion.

The Sentinel: Absorb energy and ascend, or something

The next obvious game to mention is The Sentinel (early 1987). It was the first that I can recall to attempt full-screen "solid" (OK, solid/shaded) graphics to create a game world and it still managed to be vast, with ten thousand large landscapes (the screens shown here each show only a tiny part of a landscape) to get through.

Also I'm pretty sure that it must have been one of the first tries at procedural generation for its 3D landscapes - if I remember correctly, which I probably don't - the graphics were all generated from algorithms depending on your position and direction of view. As I understand it, the program could render the view of any point on a landscape from any other point, i.e. as close to the way modern procedural games are done.

I think.



All that said, it was more of a puzzle or strategy game than an arcade explorer and as your character teleported, there wasn't the freedom of movement that a true open world game would allow. To be honest, it's just as well; whilst the game could render the graphics at a reasonable speed, jump-cuts for character movement were essential to keep the pace up.

Driller: all singing, all dancing, but
unfortunately verr-r-r-rry sloo-o-o-ow

The final game I think fits into this section is Driller, shown here at something like five times normal speed in a clever animated GIF wot I made and then forgot to take the logo off.

You can imagine what it was like at the regular Spectrum speed, but if you've no imagination then I can tell you - it was fucking slow.

That wasn't really all that important; although complete freedom of movement was allowed, this was another game that leaned towards the strategic, so its being slow was often quite useful in allowing you more time to work out what on earth to do.

This was autumn 1987, which was pretty late on in the Spectrum's lifetime, and Driller (plus its sequels) probably best represented the limits of what the machine could do with an solid 3D, arcade style open world game. However, it isn't (in my opinion) the best 48K Spectrum open world game that was produced during its normal lifespan, another game that appeared in that autumn takes that title for me.

But before going on to that, there are two other games that really have to be mentioned - given the excessive length of this never-ending piece, it would be rude not to - as they encapsulated the essential elements that would prove to be so influential in one particular genre of future open world game. I'm talking GTA and its clones, so you know what's coming - FIGHTING and DRIVING.

Saboteur: killing had never been so much fun

Turbo Esprit: this is the real proto-GTA 3, trust me

Coincidentally, both Turbo Esprit and Saboteur appeared at much the same time, around Xmas of 1986. I say coincidentally, because they ended up being two of the handful of games that I would regularly play long after I'd moved on from the Spectrum and had got fancy computers like the Amiga and PC (I had a Spectrum set up until the mid-1990s for programming anyway, so it wasn't a totally weird thing to do).

I mention these two in particular because the similarities to GTA and its successors are so obvious. Saboteur introduced the tone of amoral violence and Turbo Esprit the drive-anywhere freedom - including the freedom to run pedestrians over - that are essential parts of any modern urban open world game.

And in part 4, I'll finally get round to saying what I think was the best effort at fitting an open world game onto the Spectrum. It was a game that first appeared for the Commodore 64, but the Spectrum port was just as good. I still think it's an all-time classic that plays well today.

ZX Open World Part 2

March of the Clones...

Alien 8: Knight Lore in space

So, we're into 1985 and the era of the Knight Lore clones (as opposed to the Marble Madness clones). Ultimate's next game was the space-based Alien 8, which was really just Knight Lore with different graphics and a few gameplay tweaks, but at least they were ripping themselves off with the concept.

Everyone was getting in on the act and some of the quickest off the mark were the budget software houses (Knight Lore and Alien 8 were sold at £9.95, whereas the budget houses tended to be £1.99 or £2.50). There were two that stand out for me (even though both were obviously plagiaristic, they were sufficiently odd to remember: Firebird's Cylu and Chimera.

Chimera: more solid, more weird

Cylu: alarming play area shrinkage

To be fair, I don't recall much about the gameplay of either of these; both were just large maze games with collecting tasks, really. Neither set the world alight, but it was quite impressive and novel that the then state-of-the-art in Spectrum gaming could be ripped off and repackaged so cheaply and quickly.

Molecule Man: £1.99 bought quite
a lot back in the summer of  1986

And probably the best of the cheapo ripoffs was Mastertronic's £1.99 Molecule Man, which, if a bit late to the party (there were genuinely better ripoff isometric 3D games by this point) it was quite unbelievable value for money.

Not only was the game itself pretty good, the map was vast and it also included a level designer program along with the main game, so you could write your own versions, and that in itself was very much a novelty for this sort of game, never mind one selling at a quarter of the price of most games at the time.

Head Over Heels: packed a lot
 of format into 48K

Sweevo's World: The Monty
Python of the isometric 3D world

The fixed-view isometric 3D games probably reached their peak with games such as Head Over Heels and Sweevo's World, which added new elements to the template (for instance, in Head Over Heels, you had two controllable characters with unique abilities and often both were needed to get past a room; Sweevo's World deserves a mention for simply being the clone that was simply the most fun to play).

Quazatron: Paradroid, Spectrum-style

Oh, and not quite finally (as I've got a bit bogged down in the whole isometric thing) I must mention one of my all-time favourite Spectrum games, Quazatron, which was basically a port of the Commodore 64 game Paradroid (itself one of my favourite C64 games).

It was a great example of what each machine was good at; Paradroid couldn't have been reproduced on the Spectrum and been as good as the 64 version (no hardware sprites, for a start), so it was done in a style the Spectrum did well.

Using an open-view 3D perspective actually made the game feel quite different - less tension, but more scope for action. Both are classics even now, I think.

Fairlight: atmospheric, detailed, great graphics
and loads of stuff to play with

Another I should mention is Fairlight, which whilst obviously borrowing liberally from Ultimate's ideas and template, took a different approach to the look of the graphics, at least. Plus it allowed more interaction with non-player characters than had been seen before, but most of all you could move pretty much any object in the game, stack things up to ridiculous degrees of wobbliness, and so on. More freedom, in other words; this was as close as a graphical adventure had come to emulating the open world aspects of the text adventure in terms of scope.

It's a shame that the programmer got ripped off, he could've done more great things, I'm sure. It was a bit of a Wild West world back then though, there was a lot of that going on.

Part 3 follows, at some point...

ZX Open World Part 1 (revised)

More retro-memories.

I was reading about No Man's Sky and Outerra Anteworld the other day and - apart from being blown away by the sheer audacity of scale - it brought home to me how out of touch I've got with gaming. This thing looks absolutely incredible. It looks like it could be the game that I've been dreaming of since about 1985, but vastly larger and more complex than I ever thought could be done.

Outerra Anteworld: one view from one angle of
one tiny part of a planet, all apparently procedurally-generated

Basically, the concept behind my dream game, as I envisaged it in 1985, was the creation of a virtual world which could be explored and examined at every level. It would have lots of objects in it, with different sizes, weights and properties, so a proper physics engine would be essential for it all to interact logically. So, if you were on foot, outside a house, you would be able to go into the house (assuming an unlocked door) and explore its rooms. If the living room had a TV in it, you would be able to turn it on and change channels; you would be able to open all the drawers in the kitchen and examine things found inside the drawers, and so on.

You'd be able to pick things up and carry them around, but again it would all have to be logical - you could conceivably carry one or two televisions, but not ten; however, you would be able to load ten televisions into a car, or onto a cart, and move them that way.

That was about it really, there wouldn't necessarily be any objective beyond exploring and finding things (actually the way I describe it above, "stealing things" would probably be more apt). The point wasn't about making a specific type of game anyway, it was about simulating a world in which any and all games could then subsequently take place, possibly at the same time.

Probably because I was young and technology seemed to be moving so fast, and because there were games already in existence that contained the basic elements of my dream game, I was convinced that it would happen...oh, definitely within ten years. I wasn't daft enough to think it could be done on a Spectrum, but I was daft enough to think that I'd be playing some sort of version of it on whatever future super-Spectrums we'd be using in 1995 (ahem).

OK, that didn't quite turn out the way I'd envisaged, but there were some really interesting games released for the Spectrum that can be seen as moving towards this grand concept I had - either by making programming advances that would be necessary, or by containing some of the elements required - even if they inevitably would fall short of the concept in my head...how could they not? A world in 48K was only ever going to be a very limited world.

This post might end up being very long; I was going to try to do a brief history of open world games on the ZX Spectrum, but when I came to think about it, there are too many strands to include and still keep it relatively brief, so I'm just going to concentrate on the style of arcade-style controlled open world game that we're familiar with now (the real open world games in early computing were all text adventures and that's a whole other subject).

In 1983, when I got my ZX Spectrum, the closest thing to an open world game was probably something like Ant Attack. Although being set in a walled city - so more of a closed world game, really - it had the ethos of an open world game, in that you could follow the objectives of the game if you wanted, but crucially you could just run around and explore and have fun if you wanted to.

Hey honey, bet you're glad we came to Giant Crazy
Killer Ant World on our honeymoon, huh?

It was actually a great game too, primitive though it looks now (once you were used to the finicky rotate-move forward control system).

Plus - in an era when it was assumed that girls just didn't play games - you could play as either a male or female character. That was really going against the grain in those days - it was unusual to be given a choice of character at all - and on the odd occasion a game offered it, the choices were generally things like warrior, knight, warlock, wizard, serf, that sort of thing, i.e. all male.

That said, it was only a gimmick here, it didn't actually change the gameplay. The only change was cosmetic - a few judiciously-placed pixels on the character you controlled - but the principle was novel.

Hot zombie on zombie action amidst the colour clash

There was a follow-up the following year called Zombie Zombie, based on the same game engine (by Sandy White, a real pioneer of this sort of thing), but it could never hope to have the same impact as Ant Attack. After all, it was 1984 now and callow youths were not so easily impressed. It did have a helicopter in it though, which you could get into and fly in order to lead zombies off tall buildings and so on, thus introducing another key open world concept - modes of transport - you could run around on foot, then get in the helicopter, fly it, and crucially, leave it anywhere to come back to and fly again.

Elite for the Spectrum - too little, too late

If Ant Attack was a distant predecessor of the GTA-style format, then its analogue in space was Elite. But Elite was a Commodore 64 and BBC Micro (of all things) game! I only had a Spectrum so I never really played it (by the time it eventually appeared for the Spectrum, it was late 1985 and things had moved on apace).

Again though, it was important in that - as far as I can recall - it was one of the first games to use the concept of procedurally generating a large reproducible 3D game space with defined objects and characters to explore and interact with; a standard now, but hard to do in 48K.

Lords of Midnight - over 31,000 distinct views from any
 angle...all of them the same (joking, it was great)

Though it doesn't really fit with my theme, I should really mention Mike Singleton's Lords of Midnight and Doomdark's Revenge here, as he was the king of the vast procedurally-generated landscape game on the Spectrum.

I can't go into this in much more detail because I always found his stuff a bit impenetrable, but others swore by it. For me, I could never get over the impression that most of the locations were so similar as to be impossible to tell apart, so I never really got the sensation of exploration which I think is key to this sort of game. There was to be a final part of the trilogy called Eye Of The Moon, but I don't think it ever appeared (possibly it ended up becoming Midwinter).



So that was pretty much the scale of it towards the end of 1984.

Then Ultimate released Knight Lore.

There were other, bigger games, and isometric 3D games weren't new, but it was the feel of solidity to the graphics that was the breakthrough here. The addition of (very rudimentary) physics made it feel like a real world, with movable and immovable objects, which was pretty good for the end of 1984. And while not huge (128 rooms rings a bell) it was large enough to get lost in. We were getting there, or so it seemed, but then everything got a bit confused.

The huge success of Knight Lore meant that seemingly every software house then went on a mad rush to get their own clone of it out. Coincidentally, Marble Madness was big in the arcades at the time, and a lot of software houses were also working on their own ripoffs of that. Cue an absolute slew of isometric 3D games, some good, some bad, some bloody awful.

Spindizzy: not at all like
Marble Madness, honest guv

Gyroscope: even less like that
 Marble Madness game, honest guv

It didn't help that - as it turned out (although not revealed until years later) - Ultimate had actually had Knight Lore ready to go a full year earlier, but held it back to sell more of their other games. Well, you could hardly blame them. They were so far ahead of the opposition that it would have been madness to put Knight Lore out ahead of Sabre Wulf.

In fact, Knight Lore was pretty much the extent of what Ultimate's two main programmers (Chris and Tim Stamper) thought they could get out of the Spectrum and they were already planning to move on (which they did, forming Rare Ltd. and making shitloads of money). Ultimate did continue to put out games for the Spectrum, but other programmers were responsible and the quality definitely dipped.


Part 2 soon, with any luck...

ZX Spectrum fun

When you think about it, the 48K Spectrum was a fucking amazing piece of hardware for its time.

I got mine in late 1983 as a combined birthday and Xmas present and I think it was £129.99, but it might have been reduced to £99.99 by then. Now I know that nowadays you could probably get them cobbled together in a Chinese sweatshop for about ten pence each, but at the time it was unbelievable value for money. 48 kilobytes RAM FFS! OK, only 41.5K was usable, but that was still more than the C64 (38K I think), which cost twice the price.

Anyway, I did loads and loads of programming in Spectrum BASIC - which is still an awesome programming language by the way - and have been trying to remember all the tricks I used to save on memory.  I mean, 41.5K seemed loads at the time, but it's very quickly filled, especially given the weird quirks of Spectrum Basic.

For instance, using a single digit number - such as 0, 1, 7 - used seven bytes of storage space.  But a single character variable only used one byte of space.  So, every time you can use a static variable rather than a number, you save six bytes.  So I would set up static variables for any number that was going to be very commonly used in the code, such as:

10  LET o = 0
20  LET i = 1
30  LET tw = 2
40  LET th = 3

But you could go further than that, because in Spectrum BASIC, PI was a defined in ROM, and that only took one byte of storage. Using the logical functions (one byte), I would instead do:

10  LET i = SGN PI   (five bytes saved!)
20  LET o = NOT i   (five bytes saved!)
30  LET tw = i+i   (four bytes saved!)
40  LET th = INT PI   (five bytes saved!)

and so on.

Actually I wouldn't even have done that, because each new line of code cost bytes (can't remember how many), so I'd have actually stuffed them all into a single line, separated by colons:

10  LET i = SGN PI: LET o = NOT i: LET tw = i+i: LET th=INT PI: ...

What other ones were there? I remember using bits of screen memory as RAM (which you could only do if you weren't displaying anything on it at the time, obviously) and I'm sure there was a way of using the UDG character set (all 168 bytes of it!) amongst other things.

I really should plan these things properly instead of just sitting down and typing; I've forgotten (and probably misremembered) a lot of it. Trigonometric functions, other logical stuff, etc.

Ah well.


[Edit @ 22/09/2016:  This isn't too bad.  It's actually readable and (to me, anyway) quite interesting. It's odd that I remember a lot of these posts as "mad" when they weren't really.]

Must get this place cleared up

I've always had a bit of a "thing" about keeping things.

Not in a weird Secret Hoarder way or anything, I'm not that mad.  But I do hang onto things "just in case" they might once again become useful.  It's not unusual, loads of folk are like this.  And sometimes it's justified, especially with old PCs, because of their largely modular architecture.  In fact, the machine I'm typing this on has definitely got some bits of the old ones in it somewhere - memory modules I think - but in the main it's just clutter.

There must be at least five old TVs up in the top room (most of this stuff hangs out up there), at least as many old monitors (I mean, why the fuck have I still got two - not one, but two - massive old CRT monitors with tiny 14 inch screens?), god knows how many empty boxes and boxes inside boxes (which may contain further stuff).  Then there's all the boxes of audio and video tapes, pieces of old furniture, appliances (lamps, kettles, toasters etc.) that don't work any more and so on.

Obviously the easiest thing to do would be to just get someone to clear it out and take it all away. But there's stuff there that I really do want to hold onto mixed up in it all somewhere.  I mean, I've got a collection of NMEs going back almost continually to about 1990, which apart from being historically fascinating to a nerd like me, would be worth a few hundred quid to someone, going by the prices on Ebay (and far more if sold piecemeal).  I'd be a fool to chuck them out, wouldn't I?  And I'd never chuck my original 48K Spectrum (which still works, 32 years on) or C64. Or the Amiga. Or the Speak & Spell machine etc. etc.

Ah, I'll get to it eventually.  It's a work in progress, let's say.

This behaviour extends as far as data, of course; even when I was programming stuff on the Spectrum and everything had to be stored on audio tape (I can't recall if solid-state storage was available for the Speccy in 1986, but if it was it would've been far too expensive for me), I very rarely taped over anything, because I thought I might need the old versions as backups (come to think of it, this is now regarded as best practise, so well done me).

But now (2015 I think it is) data storage is - if you make best use of cloud storage - effectively free. Even if you want an actual physical storage device, 3 terabyte hard drives only cost about fifty quid these days and as for 500Gb drives, I think they come free in cereal packets these days. That's my justification for keeping absolutely everything, anyway.  Including all my old emails, both sent and received, since 1997. That is a bit odd, isn't it? I don't know anyone else who does that.  But hell, they take up almost no physical space, so why not?  Why, it would be a fool not to follow my lead.