I learnt 6502 assembly on the BBC Micro (by Acorn). The assembler embedded in BASIC made that a cheap and convenient option. Still use that skill from time to time. And the 6502 is still manufactured, by the way. Still useful as a fast, reliable embedded device.

1. Congratulations! It must have hurt, but you did it! 2. Bookmarks. They've been around for a while, but they let you save tabs for later visits and close them in the meantime. Arranged into folders, you can easily find them in the future and open a whole group in one go. Takes all the fear out of closing tabs, I find. Mostly. Still had 13 when I opened Firefox this morning :sigh: Incidentally, one of my major reasons for using FF – decent managment of bookmarks. Chrome has long sucked at this, not even having a sidebar until recently, and that still doesn't have a keyboard shortcut to open it (though inexplicably there's one to close it again). I've got over 2000 bookmarks collected since the early nineties, so I may have a problem too...

Thinking, like horseback riding, will be reserved for the few.

100 years? It's already happened. We British go horse riding. To quote Michael McIntyre, only Americans need it spelling out which part of the horse they're supposed to sit on.

I don't think this is spaghetti code; it's just long and complicated. Spaghetti jumps all over the place and would obscure the actual logic far more than this example does (which is completely hypothetical. Not real world at all, by the way.) Last year I tried a challenge -- to rewrite a hangman game from the late 70's, which was written in an early dialect of BASIC with GOTOs and GOSUBs all over the place, into a modern language. The code was very short, barely a page of A4 when printed out. But to understand this very simple program, I had to print it, and draw lines all over it to work out the program flow, which took over an hour. Imagine if it went to more than two sides of paper! I found a line that was unreachable and would never be executed. Even the person who wrote it wasn't aware. Just a bit of structure, named procedures etc. simplified and clarified it enormously. But I suspect that what the OP is talking about is something a bit different from that. The code may have had a lot of IFs and branches, but writing it in a clear way with meaningful names, rather than abstracting it all to classes, may have been the right call in this instance. But do a reality check: give it to a colleague and see how long it takes them to figure out the flow and logic.

The rule hoary old IT professionals use (born of long and sometimes painful experience) is: You need at least three copies of anything you don't want to lose. That is, your working copy plus at least two backups sets. A set will usually have at least the last three backups on it, in case the most recent, or its preceding one, turn out to be faulty. Sometimes errors occur right in the middle of a backup, rendering it useless. Of the two backups, one should be local (SSD, HDD, NAS, whatever) and one off-site in case of disasters like fire or a very thorough burglar. You can achieve off-site by having two (preferably three) hard drives, keeping one at, say, the office or nearby relatives', and rotating them using the grandfather-father-son principle. Why three? Well, say you only have two and you bring the HDD from the office home on an evening to do the next backup, intending to take the other drive to the office in the morning. For a night, both drives are in the same place and the worst happens... Or your off-site backup could use a cloud backup service. Note that OneDrive, Google Drive etc are not backup services: they are synchronisation services. If you accidentally delete a file or it gets corrupted and you subsequently need to restore it, you will find that it has already gone from your cloud drive, and worse, the change has been replicated to all the other computers you had linked to the same account. OneDrive does give you a 30 day window when you can reclaim files from the online recycle bin, but that is all. It is not an archive. Microsoft's terminology gets confusing here, describing the service correctly as a synchronisation service most of the time, then having a tab named Backup. Bad! That said, it is good for wholesale restoration of files to their last known state after a disaster and has saved many a person's bacon. So the short answer to your either/or question is 'both'! The choice between HDD and SSD is less important than having multiple tested backups in place and keeping them up to date. You choice will hinge on cost, capacity and longevity: nothing lasts for ever, you can drop an external drive at any time, so I rely on two local and one cloud solution for my personal files. DVDs deteriorate over time as well: I have many over 10 years old that cannot be read any more, so check and copy them every few years. Don't forget that you also need system backups on a less frequent cadence, which can be local (on the grounds that they are rather large for uploading to the clo

Well, if I'm writing a very short procedure purely as a way to check it works or compare two methods that have occurred to me, and am probably doing it in a test rig, then yes, I will use short variable names for speed. But I do it properly when I rewrite the code into the actual program. One thing I will admit to: if I write a short utility function, one or two lines... you know, to return a filename without the extension, or the _n_th word in a string or something... I will often use astr, x etc. And counters are i. Always. Unless there is a nested loop, in which case one of them is j. Unless the code is long and complicated, in which case one is outer and the other inner. Should I hang my head in shame? [Please specify precise angle]

I tried to install 4DOS on my Win10 box the other week so I could use aliases. Should I just die in shame?

Have you seen Mark of RetroClinic's DataCentre (RetroClinic DataCentre - USB on your Beeb[^]) for the BBC Micro? The DataCentre is discontinued, but I expect he would be willing to discuss any questions you have.
I seem to recall that one of the big issues was the Beeb's 16-bit address bus, which meant only accessing a fraction of, say, an SD card's capacity, but I have no idea what the solution was.