Back To Home - Archive.org (CLEO) - Archive.org (VIGLEN Utilities)

I picked this up for really more than it was worth but I saw the labels and figured I'd take a gamble on it because it looked very unique and I wanted to make sure it got preserved. It is in fact unique, but sadly it's missing the hardware it needs to properly show off this very niche item. For one it needs a BBC Micro which is a computer I do not own, but that system's very well emulated even if I have some gripes with the specifics of BeebEm. But the real problem is that it's intended to be a diagnostics package for external hardware. I do not own whatever ADC module this thing was using, and I have no real desire to reverse engineer such a thing. Not that it really matters because I definitely do not have the device it was intended to test.

Because that device is a 737.

This is CLEO, the British Airways Control Loading EvaluatOr, licensed to/by Britannia Airways. This is, as far as I can tell, a system to verify that human control inputs are mapped to the correct output levels, which is simultaneously not interesting at all and yet fascinating just for seeing what else computers got used for. The lot included seven disks: two versions of the actual software (dated March 13 and July 12 of 1985), a diagnostics disk (no date), two source disks for CLEO (both dated July 12), one source disk for the diagnostics (also July 12), and one named disk that just contained a copy of VIGLEN Utilities 3.05 that I'm not going to cover here other than this incredibly cursed "40 track" format:

Yes, that's just the first 40 tracks of an 80 track disk, that seemingly used to be a PC 360KB 40 track format. I'm amazed that it read correctly. There's an existing archive of this online, though it doesn't seem very common.

Anyway, on to the main attraction, CLEO itself.

The two versions are very similar, seemingly differing only in splash screen and one specific test that isn't present in the March edition:

(March first)

It looks like the date is internally just a string, thus the "X". The pre-filled date on the March edition suggests this was in use up through 1990 at least, but probably even later than that as systems like this tend to last way longer than you'd think they should. Once you enter a date (or don't) you're presented with three choices:

Help is, well, help text:

Which other than the connect/disconnect order really doesn't seem that helpful, honestly. Very "now draw the rest of the owl" on step 8 there. Not to mention you had to figure out how to boot to the disk to even get here so telling the user to shift-break is sort of redundant. Maybe the intent was the guy who Knew Computers would print this out for the likely less-computer-savvy people doing the actual testing.

Library is a submenu, with the July edition having a new option compared to March:

(Yes, I entered a date of "L" by accident trying to open the library...)

Drive gives you the exact same menu.

As far as I can tell, the intent is you'd pull up the library chart to see what the response curve is supposed to look like, and then compare to the real one to make sure your control system is behaving correctly. The "drive" version is a blank chart and there's a flickering / character that I assume is where it would draw the chart if I had the input rig this thing requires.

(Library left, drive right)

Here's some other charts from the July version for the hell of it:

That last one implies this was connected to a simulator of some kind, but I don't know the specifics there.

Without being able to simulate the inputs that's all we can do here, so onto the diagnostics. You're greeted with a splash screen that asks for a date and then "Timer constant in centiseconds" which is not a unit I'm used to seeing. I assume if you had access to this software you'd know what the "timer constant" was. Sample hold time, maybe.

Yes, the second question is rendered vertically above the first. I guess they added that later? It looks terrible.

Once you specify that you get an even more spartan menu:

"SIM DECODE" just seems to read some ID value, I guess because the idea is you'd have one computer connected to several test rigs. It seems wild to me that that's the place where you'd save money by only buying one; even at aviation prices surely the off-the-shelf computer was the cheapest part of this setup.

ADC1-3 are raw ADC reads, probably for calibration or functionality testing (this is the diagnostics disk after all). Other than 2 and 3 being +/-10V instead these are all the same. Using voltage for analog inputs rather than current is offending my professional experience, but I'll forgive them for current input sensors you could connect to a BBC Micro probably not existing when this was designed.

And finally "CLEO check" gives you what I assume is just an immediate measure of the input values:

While the source disks read without issue, they appear to be incomplete, with some listings referring to files that seem to not exist. Oddly one of the source disks appears to be the exact same content on both sides, so maybe somebody just screwed up archiving the source back in the 80s, though it's equally likely that whoever got their hands on this just didn't get every original disk. Some searching led me to BasList to dump the tokenized BASIC source files to text from Windows to avoid having to read it from within BeebEm, but seemingly CLEO is occasionally using tokens it doesn't understand as there's a fair number of lines that came out like "370 ?&FCFC=0:?&FCFD=0: REM set analogue switches to normal". I don't know enough about the BBC Micro to understand what's going on there. Presumably that specific one is a variable or I/O location.

I've archived CLEO as-is, flux and all. You can check it out yourself at Archive.org (see above). You need to turn off BeebEm's defaut write protect to use the "Run" disks. The VIGLEN Utilities disk has been uploaded separately at Archive.org (see above) since it has no relation other than having been part of the same lot.

One other technical note for people who didn't see my ramblings elsewhere a week or two ago: the CLEO disks are weird. They got recorded where both sides are written as if the disk was used by a dual-head drive rather than a "flippy" one (this is correct for a BBC Micro as far as I know), but both sides claim to be head 0, which is not normal, at least for systems I'm familiar with. Greaseweazle refused to decode these as a dual-side disk, so I had to read them as two single-sided flux rips, rename the second side's files to lie to Greaseweazle that it was also "head 0", decode both sides into .ssd files, then reassemble them back into a dual-sided disk image by interleaving the two .ssd files into one .dsd file. I have no idea if this is typical for this sytem or if these disks are just odd, but if you want to read the flux for some reason keep in mind you'll need to recreate that dance. It just makes sense that the one time I get a set of weird floppies with no read errors it throws this ridiculous curveball at me instead.

That about covers it for CLEO. It's weird, it's boring, and I adore it. If nothing else I think it's fascinating that the BBC Micro got chosen as the system to run this rather than anything more "professional", but it's possible this was started early enough that the PC's market dominance wasn't obvious yet. Or maybe somebody just sold the idea to BA on the back of "it's a british computer". But I want to celebrate the boring uses of computers just as much as the flashy ones, because things like this are just neat and the average person is never going to encounter anything like this on their own.