Posts by Doctor Syntax 40485 publicly visible posts • joined 16 Jun 2014
"Usually those rules were associated with reasons, so an expert system could not only make a decision but also support that with the underlying reasons."
The shortcoming about that is that is can only make the decisions it was given reasoned rules to make.
A real expert, OTOH, will have a degree of understanding that helps them, when confronted with a novel situation, to undertake new reasoning and at least suggest what the best decision might be.
When I use a piece of software, I expect the right to ENJOY the use of it without interference (so no "time-restricted trials" or whatever have you). I expect the right to STUDY its operation etc.
Even amongst FOSS S/W users (and I include myself) you are exceptional and even more so amongst software users in general. Most of us (almost all, actually) don't include studying software as one of its uses.
Most users have no understanding of any computer language nor should they need to. Very few will have fluency of all the languages in the stacks in a typical Linux distro and I doubt anyone at all has sufficient understanding of the minutiae of the all the sub-systems to avoid inflicting damage if they stray outside their competence* - hence the frequent advice not to try writing your own cryptography.
* Remember the great Debian random number fiasco.
"But don't forget the Red Hat paradox... Despite the massive contribution of unpaid developers, large FOSS projects often need some source of income, and paid support is often the solution."
No paradox. What you're not realising is that by and large is somebody paying the developers and that somebody is often Red Hat. In other cases it's often people with some other financial interest such as Intel who expect to sell the processors the S/W will run on.
The actual principle behind the funding is that a group of businesses who have something to gain from a project find it more effective to get together, do their bit at their own expense and share it rather than each of them try do the whole thing themselves with huge and pointless duplication of effort.
"Do you believe rich people buy a house, car or a boat because they've seen it on Facebook?"
Maybe not houses and cars but the advertisers presumably believe that they'll respond to adverts and as rich people buy more stuff in general and more expensive brands that's why they'll pay to advertise at them. Whether the ads justify the money spent is another matter but remember Facebook is part of the advertising industry and the advertising industry doesn't sell houses, cars and boats. It doesn't sell phones or trainers. It doesn't sell soap powder. It sells advertising. Only advertising. And all it has to do is to persuade advertisers to buy advertising. What people buy because they see it on Facebook is irrelevant.
"They seem to have a monopoly, which is why they get away paying no tax as they know people will keep buying stuff from them."
As written that's a non sequitur. You need to turn it around. The reason why they (a) have a monopoly and (b) pay little or no tax is because they keep investing in building the business. That means that the operating profit gets swallowed up so there's no profit on the bottom line and the result of the investment makes it impossible for others to compete. At some point, however, they'll have to stop growing. Any growth curve that looks exponential is really the start of a sigmoidal curve.
The trick is going to be to decide when to stop eating the operating profit and start paying dividends. If they invest in facilities they don't need once they start to saturate the market they'll be throwing away cash. That happens. Right at the start of my working life I found I was in the middle of a small but well-known in its field business going bankrupt because it had done just that and the investment wasn't even on the stuff that made its profits.
They're not unique in investing all the available income for years before they start paying divies; I remember the same things being said about Microsoft.
"LOL believers in the current system thinks that continuous economic growth is sustainable."
LOL believers who equate money - which is a tool to trade what you do for what you want - with continuous economic growth.
Of course continuous economic growth isn't sustainable. But then neither is a society running on barter at anything above subsistence level.
"And in cases like the fortune 1000 most of those stockholders are either other corporations or they are banks or they are investment firms."
Do you have pension investments? Life insurance? Any other sort of saving scheme? If so then very likely some of the money that went into those companies' shares is yours.
I was taught a long time ago that money is a means of exchange. The alternative is barter. Is that what you're suggesting? Otherwise if I feel hungry I have to knock off one of the lambs in the next field. The farmer won't mind, will he? After all they will literally have no value to him.
Let me offer you a small example: the Panel on KDE. That's the bar (usually) at the bottom of the screen with the start button at one end, the system tray at the other and between them minimised windows any useful bits you want to put there.
In KDE3 this could be set to auto-hide and then to reappear when the cursor went to a prearranged place. This could be anywhere along the edge or just in the corner.
If you selected the edge the ballistics of a mouse cursor would ensure you hit it whenever you went for a control at the bottom of a widow sitting at the bottom of the screen. You then had to wait a moment for the panel to unhide and rehide.
The better option was to select the bottom left corner as the target. This is where the menu button is which is often what you want from the panel anyway. Even if it isn't it's quick enough to go for the corner and then slide along to whatever you want. Very effective and very unobtrusive.
In KDE4 the corner option was done away with. Just that. Annoying panel behaviour unavoidable.
In KDE5....no they haven't seen the light and restored the corner option. They've just added, whether as a bug or by deliberate design, another breakage. The panel now bobs up if a window opens at the bottom of a screen. Not only that but it stays there until the mouse cursor visits and leaves it.
What's worse, if the panel is hidden and a confirmation dialog is invoked, say to confirm binning something, it's invoked minimised on the panel which, just ot be contrary, doesn't unhide. To confirm you have to bring the mouse cursor down to the bottom of the screen, unhide the panel, click on the minimised dialog to restore it and click on it. At least that resolves the unhide by window problem - the only workable solution is to sacrifice screen real estate to a permanently visible panel.
And did I mention that in 5 the background can no longer by a gradient unless you make a gradient as an image and set it as the wallpaper? Or the eye-rattling image as the default wallpaper? Or the bland scribbles that have become the default icon set?
The only rationale I can think if is that they didn't want Microsoft to be the only ones with crap UI design.
To be nit-picky, a 100% CUA-compliant browser ought not call its first menu item "File" but something like "Page" instead.
Oddly enough out of the 14 items on the menu only one is "Open Web Location". The rest wouldn't look out of place on the File menu of any other application. Even Page setup is to configure the printer settings.
"I think the problem with a hamburger menu is that it only works if people have learned that a bunch of horizontal lines in a corner somewhere means a button that can be clicked to show settings or options."
It gets out of hand PDQ. Bing on Waterfox. Waterfox has a hamburger menu top right. Bing has its own hamburger menu just below it.
It's hamburger menus all the way down! I look on them as the graphical equivalent of the acquired speech impediments such as scattering "like" into sentences where it doesn't belong or the introductory "So" which seems to mean "I'm a millennial and I'm starting to speak".
"The main factor is time. If you have time, you can do anything"
It's a matter of how you use that time. Do you use it, and budget for using it, upfront to investigate the requirement or do you use it unbudgetted later trying to make the product fit reality and probably throwing in a few bugs because you're under pressure trying to do things in time you haven't really got?
"1. It's not actually a link, it needs to be copied and pasted.
2. You don't know what it is beforehand so you don't know if it's relevant or not."
Copied and pasted is preferable depending on your level of paranoia. Unless you're in the habit of hovering over a link to check the URL it resolves to you don't know what it is anyway so you don't know whether it's safe or not.
"Does it need a manual? If you need a manual then the design can take a step back. "
Even better, write the manual and call it the design. OK, things get iterated and the draft manual needs to get update. but another gem from TMMM is that the manual is the first thing thing that gets started and the last thing that gets finished.
"Exactly because a hamburger menu does state what it does."
A hamburger menu allows you to choose hamburgers. That's what it does.
If you mean a menu that's hidden behind three lines being called a hamburger menu because that sounds kool the problem isn't with the menu, the problem is with it being hidden.
Right now I'm looking at a browser that has the classic CUA menus. If I were to click on File I know what to expect because, give or take application-related variations File menus do similar things.
If I use a browser with a "hamburger" menu I've no idea what it will do and, more to the point, if I know what I want to do there's no obvious way of finding how to do it unless I try everything in turn until I find it. I want to bookmark somethis? My CUA interface has a menu "Bookmark" - I think I've found it. My "hamburger" browser - no it's not on that menu. Oh, look, it's hidden under an icon of a star. What numpty thought of that?
Because if it doesn't, a Chrome "Now guaranteed with more ads!™" might have a hard time keeping its #1 position.
Especially against a forked Chromium without ads rival. In fact, if you were going to the trouble of forking Chromium why not build the ad-blocker in as part of the browser?
"They're on the backfoot with concerns over privacy, data, consent... the politicians are in there, and they want to legislate. For the blockchain community, this opens the door for you. Go for it."
Really? So privacy legislation requires you to remove a record but your records are being kept by blockchain. Now you find you can't remove the record without breaking the integrity of the chain.
I think I have a solution which should please users and advertisers. The advertising industry might have to do a bit of adjustment.
Add a variable that shows the user's current attitude to ads. This variable is used to introduce a weighting into the auction. By default it's zero - the user doesn't care so everything works as at present assuming no adblocker. The user gets all sorts of crap shovelled at him him on behalf of hopeful advertisers. It's a gamble as to whether it does the advertiser any good but the actual probability of that is pretty poor. However if the user sets the value negative it tells the system that the user really doesn't want to see ads and will likely punish any advertiser whose wares get shoved in his face by not buying from them in future. This will weight the auction against showing any ads without actually using an adblocker (precautions need to be taken to avoid gaming the system by arranging for competitors' ads being shown). OTOH if the normally ad-adverse user is looking for something they can set the value positive. At this point the value of showing an ad rises and it becomes worth bidding high.
Under such a scheme the users benefit - they don't get crap shovelled at them all the time but when they're looking for something to buy they get shown useful ads. The advertisers benefit because they're not getting negative value for a lot of their ads and getting good value for the ads they show. The advertising industry has to rethink the way it works. At present the advertising industry only sells one thing. It sells advertising. Not soap, not new shiny, not widgets, just advertising. It needs to change to selling results. Not just results which look good when you can't see what damage you're doing but actual results and price its services on those, not just on ads foisted on the uninterested and unwilling.
"Well, let's do some numbers."
For the advertising industry there's an even more terrifying set of numbers if only the raw data were available. That's the net value to the actual advertisers. The gross value is easy - the marketing department or consultants who sold them the deal can say how much it cost to place the ads, what percentage clicked through, made a purchase and what the value of that was. What they can't really get a handle on is the number of people who might otherwise have made a purchase and were so pissed of by the ads that they went elsewhere and hence they can't get that cost to be added to the cost of their advertising campaign.
I'm surprised that there didn't seem to be any kind of penalty for making pointless frivilous FOI requests that took ages to find information so they could then try and talk their way into selling me a bunch of 20 year old printers that are "proven technology mate".
If it had been me there'd have been a penalty other then being blacklisted.
They'd have been greeted with their pile of requests. "You have half an hour of my time. This request took me x minutes. This took my y minutes.... You've already had your half hour before you arrived. Goodbye."
"Denham has often said that handing out an order to stop data processing would have a greater impact on a firm than a cash payout."
That option has been there since DPA v1.0. Has it ever been used?
AFAICR it looked to me back then (?the 80's!!) that, apart from anything else, it was a trip-wire in that the ICO might not be able to prosecute a criminal offence but it was in their power to impose such a ban and that breaking it did constitute a criminal offence. Of course I may have misunderstood it but it seemed like a crafty bit of legislation.
"thereby betraying that they actually store your password in plain text."
You're assuming that the only way to hash a password is as a single hash.
What they can do is produce a series of hashes of a few select characters and then ask for one of those selections. It has the disadvantage that if all permutations are stored the number of hashes to store expands very quickly as the password length grows. This means that either they have to pre-compute and store a large number of hashes, restrict you to short passwords or only store a small subset of permutations. Guess which is least likely.
I don't suppose they ever did one for the BSA Bantam. MOT was a strip-down and rebuild. David Brown's facilities were used for taking the crankshaft apart & reassembling (the roller bearings tended to go hexagonal). We also had an old paint tin which was the exact size for compressing the spring to take the clutch apart. I was/am short-sighted and dad was long-sighted so he never believed I could read the colours marked on the circuit diagram on the tiny little handbook. Once when he took a bike in for MOT he made a note of where the tyre valves were. When he came to pick it up the bike hadn't been moved - the mechanic knew it was in top condition - as much as a Bantam can be.
Years later I took my MGB head into a garage to get the valve seats re-cut (the blow-back through a carburettor eventually set the air filter on fire). There was a Banty leaning against the wall and it took me ages to work out why it looked different. One of the very early ones with the solid back end.
"While alternatives do exist, they are considerably less capable and productive."
I'm reminded about the story of the bees and aerodynamics. According to aerodynamic theory bees can't fly. Nobody told the bees that so they continue flying quite happily.
Can we have a bee icon please?
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
