Philip Storry's Website

Intel really struck it big when IBM's PC design used their 8088 as it's CPU.
The Intel 8088 was a budget version of the Intel 8086 - the money was saved by only giving the 8088 an 8-bit data bus instead of a 16- bit one.

So here's the thing...

IBM's replacement for the IBM PC - the IBM PC AT - used an Intel 80286.
A few years later, Compaq famously beat IBM to the punch when it came to shipping the first PCs with the next generation of processors - the Intel 80386.
And then we had the 80486, and then came the Pentium - a sudden name change which threw everyone, because they were expecting an 80586.
(US courts said you can't trademark a number, so Intel stopped using them.)

So we can easily see a fairly decent progression in Intel's numbering scheme. But isn't there a hole?

Shouldn't there be an Intel 80186?

Yes. And there was one. It just wasn't used in IBM PCs.

The Intel 80186 was an odd beast. It was designed to be binary compatible with the 8086, but at the same time it included a lot of features that were usually part of the supporting chipsets on a computer.
That made it more suitable for use in embedded computers rather than in IBM PCs, especially as some of those extra features actively clashed with the standards expected in an IBM compatible machine - so most computer manufacturers just skipped it and waited for the 80286.

A few plucky companies did use the 80186 in an "IBM Compatible", and one of them was where I first saw this chip. It was an RM Nimbus, a common machine in the education sector of the UK back in the late 80's.
And it was gloriously almost-compatible. It could do EGA graphics! Kind of. If you remembered to tell it to be, by running the "setmode" command first...

I still run across people who don’t believe that the 80186 existed. And before the internet, it was kind of difficult to prove that it did, unless you happened to have an example handy. So the debates could get a little circular...
Interestingly, although the 80186 isn’t remembered by the public, it evidently sold well. Intel stopped making the 80286 in the early 90’s, but manufactured their last 80186 in 2007 - less than a decade ago.

25 years is a pretty good run for a processor. Although perhaps not a good enough reason to waste neurons on knowing about it...

An impressive outturn from the SMWS this month, with plenty of older (and therefore more expensive) bottlings.

G3.4 Pride of Bengal
27 years old, distilled 29th November 1984, 57.5% abv, refill ex-bourbon hogshead
A light nose of vanilla, butteriness, PVA glue, cornflake dust and a hint of ginger. The body is very sweet, with naan bread and peanuts complimenting a massive vanilla note. The finish is fiery and sees a return of the ginger note from the nose.
Water doesn’t bring out much more - perhaps a little cumin on the body. Holding this whisky on the tongue for a few seconds gives an astonishingly satisfying mouth feel - truly a highlight of the day.
5/5

58.14 Ye olde sweetie shoppe
23 years old, distilled 5th June 1989, 57.8% abv, refill ex-bourbon barrel
The nose has midget gems, berries and a touch of oak. The body adds peaches and fruit fondant centres from a box of chocolates, and the finish brings in golden syrup and oak.
After water,the nose has a touch of old dusty books, and in the body there is a much larger presence of oak and berries, as we as sponge cake in the body.
A very decent dram, nicely balanced and dangerously drinkable.
4/5

41.56 Rabbit in soured cream source
24 years old, distilled 28th August 1988, 51.9% abv, refill ex-bourbon hogshead
A very light nose - sweet vanilla and oats, but not much else. The body is similarly clean and straightforward. Custard Creams, a hint of Bourbon biscuits, and a floral sweetness - honeysuckle? The finish is oak and corn - unexpectedly heavier than the body and nose.
After water, you can spend a little while marvelling at the heavy whorling in your glass, before returning to the flavours. The nose and body now have a slightly grassy flavours added, and the finish is slightly less large.
A fine dram for a summer's day.
4/5

37.54 A contradictory dram
27 years old, distilled 24th April 1985, 52.4% abv, refill ex-bourbon hogshead
An excellent nose - slight hints of oak and almost burnt toast, and lots of perfumed notes from scented soaps. Subtly floral. The body is less subtle, quite substantial, and brings in tannins and a hint of linament. The finish is much woodier, bringing back that toast note yet having a floral note too - violets?
After water, the nose is sweeter, with buttercups and a suggestion of gelatin. The body has wood - camphor wood? - and the finish is also woody, but still retains those floral hints.
A very satisfactory dram, which brought a little more light into a dull overcast day.
4/5

24.124 Close to the edge of extreme
23 years old, distilled 22nd December 1988, 50.8% abv, refill ex-sherry butt
Furniture polish and cedar wood on the nose, which also hints at Terry's Chocolate Orange. The body is robust in flavour, but quite thin in mouth-feel. It's full of toffee, oranges, dark chocolate, tobacco and Arabic spices. The finish is toffee and treacle.
I didn't find that water did much for this dram - a little coffee, but otherwise very similar to its unwatered experience.
An excellent example of sherry maturation, from the distillery that made its reputation on sherry casks.
3.5/5 (would have been 4 if the mouth-feel was less thin)

44.57 Sea salted caramel pebbles
22 years old, distilled 13th August 1990, 52.9% abv, refill ex-bourbon hogshead
Toffee, caramel, grapefruit and melon on the nose. The body has salted caramels, hints of lemon and custard. The finish is like an assortment of penny sweets.
After water, the toffee becomes a bit more dominant, but this is still a very sweet, light dram.
A fine aperitif dram.
3/5

29.129 Leather and lime in a smoky room
22 years old, distilled 9th November 1989, 55.1% abv, refill ex-sherry butt
Smoke and embers of the nose, with a background hint of leather and mahogany. The body has plenty of smoke, with a slightly nutty background - walnuts? The finish is big, smoky, and has a hint of oranges.
After water, the nose brings in a bit more mahogany, but is still dominated by smoke and ash. The body is a little sweeter, bringing peanuts alongside the smoke. The finish remains unremittingly smoky, with the oranges having given way to walnut.
A fine finishing dram, which provides a palate coating barrage of pleasantly complimented smoke. If only the law permitted me to smoke a cigar with this...
3.5/5

The IBM PC was powered by an Intel 8088 processor.

That processor could address 1Mb of memory. Yes, a whole luxurious megabyte! Trust me, that was a big figure in 1981.

However, you didn't get the whole megabyte. Oh no. Things like video cards and the built-in BASIC need some RAM to use, so IBM reserved the upper 384Kb of RAM for that purpose. That gave the machine 640Kb of RAM to play with.

In 1981, that was actually pretty amazing. Remember, the first IBM PC (Model 5150) could only have a maximum of 256Kb fitted into it. The minimum RAM to be fitted was 64Kb. So at the time, these probably seemed like fairly generous limits...

So, the IBM PC was cloned, and became effectively an industry of its own. Its originally generous architecture lasted some 21 years longer than I think IBM would have wanted...

So some fairly nasty hacks had to be performed to get more RAM under MS-DOS.

The first was Expanded Memory. That required a special memory card, and a driver to be loaded. It was also a horrifically ugly bank-switching affair.

For those that aren't aware, "bank switching" meant that you didn't get access to all the memory you had on that card. You got access to the memory in 16Kb pages - and only a maximum of four of them. So basically you got 64Kb at a time.

And yet this was better than no memory, so people persevered.

Then, with the Intel 80286, we got Extended Memory. That allowed flat access - no bank switching - to all the memory above 1Mb.

(Provided the processor was in protected mode, that is.)

And it gets no better. There is also the High Memory Area, which is a best described as a godawful hack.
Technically described, it's the first 64Kb (almost) of the Extended Memory area. Due to some trickery involving - believe it or not - the keyboard handling interrupt, that small sliver of memory at the start of the Extended Memory area can be accessed by real mode programs.

Truly, the High Memory Area is a hack which makes Expanded Memory look graceful by comparison.

Basically, RAM on an IBM (Compatible) PC was a mess. An unruly, damnable mess. And we wrestled with it using the tools that MS-DOS provided (HIMEM.SYS, EMM386.EXE, DOS=HIGH) or with third party programs (QEMM, 386MAX).

All to gain a few extra bytes in that precious 640Kb area that PCs were originally so generously given.

It was no wonder that when Windows 95 came along, so many people were happy to be free of these kinds of ugly hacks!

And I remember all of this and more. All of it, in these days of technological luxury, just so many wasted neurons...

It's that time of year - resolutions time.

Why?

I'm quite serious. And this isn't an attempt to shirk anything.

(I'm making no resolutions this year because I'm more of a fan of kaizen. So when I needed to start running, I didn't wait - I just started. I'm simply wondering about why we make resolutions at New Year.)

Despite preferring continual improvement, I can see that having a time to take stock of things and set new goals is positive. I just don't see why it should be January 1st.

My first problem is that it seems a very arbitrary point. There may be some psychological attachments of importance for some but that's about the only thing to recommend it.

After all it's just a point on the calendar. And that calendar has been tinkered with so much to try to make it align with the Earth's orbital oddities that our New Year is now almost two weeks after the Winter Solstice, which is both the origin of the New Year and a more logical candidate, as the Winter Solstice is a physically measurable phenomenon which heralds a real change - days growing longer after it.

But having thought about it, I don't think that the Winter Solstice is the best candidate either.

The second problem I have with New Year's Day as a point for reflection and goal setting is its frequency. It is, unrelentingly so, yearly. (The clue is in the name.)

If you're employed in an organisation of any moderate size or larger, you'll probably have Annual Reviews. And if they're anything like mine, half the goals that were set a year ago are then ignored in that review, because "things changed" and "new priorities appeared".
(And you should have been meeting with your boss to cover that and amend the goals every four to six weeks, but that didn't happen because both of you were busy working. On things that had changed, and new priorities that had appeared.)

So I think that yearly is too long. And that brings us back to Solstices, which occur twice yearly.
That's a much nicer schedule. But I don't like using the Solstices, because they mark the longest and shortest days - it's no good planning activities that may rely on conditions outdoors then, because it's either too late or too early.

So the only sensible options, it seems to me, are the Equinoxes.

At an Equinox, you know that the next six months (approximately) are going to be either better indoors or better outdoors. With a Solstice, you're at the peak - if your resolution requires outdoors activity, then you're either at the best or worst time for that RIGHT NOW.

Which is probably not the best time for a resolution.

Whereas with an Equinox, you're at a point of balance. Days will either grow shorter or longer, and you'll find out how the activity works with your schedule for both as you'll experience both lengthening and shortening of days during the next six months.

And the Solstices are also celebrations in themselves - even if not absolutely celebrated these days. But we all take summer holidays, or visit family for Christmas - making the Solstices a busy time at which reflection might not be easy.

And yes, I appreciate that not all resolutions rely upon the weather and the length of day. But many do. And we subconsciously change our behaviour during the longer or shorter days - if people wanted to stay indoors during summer, pubs wouldn't have gardens and terraces. Let's face facts - we are swayed by what it's like outdoors, and that may well affect how easy it is to accomplish your resolutions.

The more I think about it, the more I think that anyone taking up Resolutions should make them Equinoxal. Six months, easier to plan around the big events of the year, a better idea of the weather... They seem ideal.

Which is probably why the idea will never catch on. But I'll try to use it as a review point myself, to help my kaizen along...

Tonight and tomorrow are the Internet's busiest time of year.

In the same way that at New Year the mobile phone networks get a free stress test with millions of messages of goodwill, the Internet is about to get a massive stress test as people open their Christmas presents.

Because when's the last time you had a shiny new computer that didn't have to download updates almost immediately after being turned on for the first time?

Not to mention registration, apps being downloaded, and so forth.

And all that on top of the extra traffic generated by a holiday.

Yes, tomorrow is a difficult day for the Internet.

So have some patience, and spare a thought for those unfortunate souls who are working behind the scenes to keep it all running.

(And all those folks working in the emergency services, armed forces, utilities, health care and social care, and other essential roles.)

It is, after all, Christmas.