Re: an oldy
He's the superior officer of Major Error.
42030 publicly visible posts • joined 16 Jun 2014
I was thinking more generally.
Certainly as regards error messages long numerals or alphanumerics are seldom going to be reported correctly. I do wonder how such error messages come to be generated.
Does some development manager tell a team "You can have errot numbers 00000333387000000 to 00000333387999999" while someone else has 00000333388000000 to 00000333388999999?
Are they some sort of mapping? Is the code file given number 00323334, the functions within it numbers starting 0000 and the errors within the functions also given numbers starting 0000 with the error reported as the concatenation?
A better approach would be to use a What 3 Words approach and map them to a short phrase which might be memorable enough for the user to remember and easy to map back to where the error was found. "Out of cheese" might well be a better way.
But there are other issues. Take the situation where the user goes to close the application with unsaved work. There are instances in which the user may reasonably choose to do this, one being that they've made such a bollocks that the easiest thing is just to quit and start again and a prompt which confuses the user into the wrong choice is not helpful.
Another, mentioned here recently is the GIMP prompt about unsaved work. The real issue there is that GIMP's idea of saving is saving work in its own format even if the user used it to edit a JPEG and pnly wants to save the resulting JPEG. In this case the "Quit without saving" option is quite often the one to choose as the user has already opted for the Overwrite or Export options which are the only ones to save as a JPEG. A bit of thought about that would maybe have steered the developers into a unified save function in which .jpg and .xcf were equally valid choices.
Getting that short message unambiguous can be crucial. A notice at a level crossing saying "Wait here while light flash" sounds OK doesn't it? In Yorkshire dialect "while" and "till" can have opposite meanings than expected. "Wait here while lights are flashing" is better.
What sort of morons code these websites that say "Enter card number - no spaces"
A slightly better class of morons than the morons who simply say "Enter card number" but don't tell you they don't accept spaces or do but treat them as valid characters and truncate the entered string at 16 characters.
This is a consequence of using the wrong early implementation of Pascal. UCSD Pascal would have taken you back into the editor at the point where the error was detected. After growing up on batch runs of FORTRAN* this was a mind-blowing revaluation. Even more mind-blowing was the fact that it would also do this for assembler.
* For added fun once the compiler had been thrown off track by the first error it would consider most following lines as being in error even if they weren't so there would be pages of error messages of which only the first was necessarily true - although some of the others might have been.
"FOSS) Fix it myself or hire a programmer at a competitive rate to fix it."
Can't remember having to fix anything. It must be pushing 20 years since I wrote a line of C and that's because I was being paid. This stuff Just Works these days.
"Proprietary) Beg the monopoly supplier to fix it. Buy the next version and hope the fix is included and new breakages aren't."
Well, if you want to beg Microsoft to fix things, good luck. Personally, I'll stick with the FOSS route.
Edit: I'll just throw this in. One thing I did like about Windows back in the day was the cardfile program. It's one thing they dropped along the way. I don't think it can be made to run at all in W10. And that is one thing I have just written for myself - using Lazarus. If I had any reason to I could port it to Windows.
Added functionality doesn't get there accidentally. Somebody has to write it and, bugs apart, that's likely to be deliberate. This is a little too elaborate for a bug. Unless whoever coded it was sneaking something in for their own gratification we must assume they were told to. Telling them to do so must surely imply that usage was at least considered.
Where then is the scope for accident? Possibly exposing it at this stage? Hmmm...
What's the betting that these "firms" will go bankrupt
Which is why I advocate a PAYG (for the callers) of compensating the callee. Dial 1461 or similar and receive as small fee for taking the call, greater if you're TPS registered, taken from the caller's account. The going bankrupt before paying up trick becomes a matter for the telco's credit controllers who won't let big bills rack up.
Yes, it would require some monitoring to stop the bright spark who tries to get money from anyone who calls him regardless of who it was.
An yes, it would take some work and expense upfront by the telcos to set up the required systems. In fact, if OFCOM started proposing this I think the telcos would manage to solve this so far insoluble problem fairly quickly.
"But I really don't understand why one would develop a compiler for such an inferior language. Just emulate your AS400"
Why lose performance emulating something instead of compiling to native code? The COBOL code carries out the business that pays the bills for the company. There's nothing inferior about that.
The finger pointing issue is a more general problem than utilities provision. I know of 4 instances in recent years of retaining wall failure and these seem to run on for years. One has left a house teetering on the brink of a landslip - and a closed road - for well over a year. In each case it seems to be the local council arguing responsibility with insurers.
We need legislation to ensure that, in the event of stand-offs between service providers (in the widest sense of the term) the individual has the right, after a reasonable lapse of time, to put the parties on, say, a week's notice to finally resolve the issue after which the costs of rectification and any additional costs, loss of service, etc, will be split equally between them. Providing those additional costs are made sufficient (maybe throw in some statutory compensation) to make attempts to game the system too expensive, the parties should have sufficient incentive to resolve things quickly. Perhaps as an additional incentive the extra costs should come out of the management bonuses rather than shareholders' funds.
As to the suddenly dumb "smart" devices - well, if you will buy such things..
If they've been made redundant the present tense doesn't apply: they've left and it wasn't their decision hence going to competitors wasn't their purpose. If the former employer then starts to throw their weight around they're just making a case to be taken to an employment tribunal and subsequently to the cleaners. However I suspect your use of the term "General Counsel" means you're from parts where there might be less protection for the victims.
"applications move to probabilistic computing models, in which processing and decision making is based on inferences drawn from large piles of data"
Back when I was doing these things it never occurred to me to decide where to address an order based on inferences drawn from large piles of data. I took the simple route of dispatching it to the address given with the order. This might be wild supposition on my part but I think that's still how it's done today and that that's how business collects money from customers. That raises the question of what extra value is added to a business by making inference and whether that added value, if it exists, justifies the expense of making the inferences.