I haven't said that for a while, so I thought it was a good thing for getting this long-dormant blog going again.
What's my particular reason for saying it on this occasion? Internet Explorer and Cookies.
We've written some software that uses cookies to identify users who have logged in, and I wanted to actually look at the cookies to see exactly what it was doing.
How? Well, IE 8, at least, doesn't make that it obvious how you do that, but after a brief burst of Googling, I found the answer: go to the Tools menu, select "Internet Options", and in the "Browsing History" panel, select "Settings". You get a new window, entitled "Temporary Internet Files and History Settings", from which you select "View files". Not exactly obvious, but OK. And there, in amongst 99 assorted files, were a bunch of 10 with names that started cookie:. Super.
Except that none of those cookie files had been modified for weeks, and none of them were obviously related to the site I was working on.
Puzzled, I looked again at the "Temporary Internet Files and History Settings". One of the settings, "Current Location" was the name of a folder:
C:\Users\Robert\AppData\Local\Microsoft\Windows\TemporaryInternetFiles\
Presumably, I thought, the name of the folder where my cookies (and other "Temporary Internet Files") were being stored. And when I looked in that folder, I indeed saw the same 99 files that Internet Explorer had shown me. And no sign of the cookie I actually wanted.
So I prodded about, and after a while tripped over another folder:
C:\Users\Robert\AppData\Roaming\Microsoft\Windows\Cookies
Where I found 13 files that looked like cookies: 9 corresponding to the ones I had seen in The Other Place, plus a couple more that didn't correspond to anything, but that weren't the one I was looking for either.
But there was also a folder, named Low. And lo! In there, I found 98 further Cookie files! Including the one corresponding to the cookie I was looking for.
So when I looked in IE, it told me I had 9 cookies stored in one place.
But it was actually using cookies stored somewhere else, where it had almost 100!
I have long hated the way Microsoft, and Windows in particular, try to protect naive users (or, alternative, avoid putting off non-technical buyers) by hiding complexity, in such a way that turns the whole system into black magic.
But this surpasses their normal level of obfuscation. To explicitly say IE has 9 things stored in one place when it is actually using 98 things stored somewhere quite different, is not just "hiding complexity" -- it is confusing, and misleading users (perhaps dangerously).
This kind of thing just makes it more likely for people to find computers incomprehensible, and think they are beyond them.
And I think that is a nasty thing to do.
P.S. And I've just noticed that when I told IE to delete cookies, it deleted almost all of them from the "Other Place". Strangely, it left behind four that came from microsoft.com.
Saturday, 21 May 2011
Tuesday, 24 August 2010
Standards Of Literacy
I realise I may sound a bit like Lynn Truss, but to my mind, "Say your family name" and "Say 'your family name'" are two different instructions.
They are different because one has quotes around the words "your family name", and the other does not. The quotes are there because they make a difference. And in this context, that difference is to indicate that their content should be taken literally.
So while the first ("Say your family name") is asking me to say "Inder", the second ("Say 'your family name'") is asking me to say the three words "your family name".
But this distinction seems to be lost on the people who design BT's bills related paperwork. For years, they have instructed customers to "quote 'your account number' on all correspondence".
What do they think the quotes round "your account number" are for? What effect do they think they are having on the meaning of the sentence?
It is not as if this were a letter sent out by a junior secretary in a hurry. This is in the payment instructions that have gone out with every bill for years now: someone would have been picked for the job of designing the forms, and expected to spend time getting them right.
But it is worse than that. Given the associated printing costs, and the high visibility of the documents, I cannot believe those instructions were not vetted by at least three layers of management. But they were approved. Presumably without anyone having enough confidence in their own level of literacy to query the purpose of these stray quotes.
And in fact, it is even worse than that. These bills must have been seen by practically every manager in the company (well, except those that get their phone service from one of BT's competitors). Surely some of them must have noticed something amiss. But none of them has the confidence and concern for their employer's image to have done anything about it. Do they all think those quotes are needed? Or look sophisticated? Or that so few people will notice that it isn't worth getting them removed?
Oh dear, oh dear...
Sunday, 28 February 2010
Curling
I tried curling yesterday. And although I was continually aware I was flapping about on ice (and indeed did plant my backside on it once), that didn't stop me enjoying myself.
It's much more of a team game than I'd expected.
At one level, it is very similar to bowls: there is a target at the far end of the playing space; players take it in turns to launch a "projectile" towards it; and the side/team that end up closest to the target score points. In both, obviously, the skill of the player launching the projectile is in finely judging the weight and line of the launch. And, if the curler starts their stone spinning as they launch it, it will tend to travel on a curved path: it will curl. More so as it slows down. Just like the bias on a bowling bowl.
But the interesting difference is that that, once the stone is on its way, the other team members can influence its course. Sweeping the ice in front of the stone reduces the friction, and thus makes it slide further, which allows the other members of the team (as directed by their captain) to adjust the effective strength of the launch. And the sweeping also affects the extent to which the stone curls, so that sweeping (or not) at different points along the stone's travel fine tunes the aim of the stone on its way towards its target.
Fascinating.
It's much more of a team game than I'd expected.
At one level, it is very similar to bowls: there is a target at the far end of the playing space; players take it in turns to launch a "projectile" towards it; and the side/team that end up closest to the target score points. In both, obviously, the skill of the player launching the projectile is in finely judging the weight and line of the launch. And, if the curler starts their stone spinning as they launch it, it will tend to travel on a curved path: it will curl. More so as it slows down. Just like the bias on a bowling bowl.
But the interesting difference is that that, once the stone is on its way, the other team members can influence its course. Sweeping the ice in front of the stone reduces the friction, and thus makes it slide further, which allows the other members of the team (as directed by their captain) to adjust the effective strength of the launch. And the sweeping also affects the extent to which the stone curls, so that sweeping (or not) at different points along the stone's travel fine tunes the aim of the stone on its way towards its target.
Fascinating.
Tuesday, 1 December 2009
Windows 7: One week, two bugs
Well, I've been dabbling with Windows 7 for just on a week now, and I've found two bugs.
First, it has the same "auto hide task bar" feature that Windows XP. Except that in XP, the task bar hides automatically. And in Windows 7, it usually doesn't. Don't know why. Sometimes it does. But, it seems, after a while it gets tired, or bored, and stops hiding.
Great.
Second, it has unnecessary and un-documented interaction between features.
Specifically, Windows 7 introduces two new features: Aero Snap and Aero Shake.
Aero Shake is a brand new feature. If you use the mouse to start dragging a window by its header bar, and "shake" the window, it will minimise every other window, leaving the one being dragged and shaken the only one open on the desktop. So you can see it clearly and, I guess, get at other things on the desk top at the same time. Shake it again, and the others go back the way they were. Kinda neat, I guess.
Aero Snap is a feature that, when you drag a window into the top or bottom of the screen, maximises it, while if you drag it into the side of the screen, it "half maximises" it --- that is expands it to fill half the screen. Now, XP has something like this, too, but I've never understood it. All I've noticed is that when I try to move a window, it sometimes maximises itself. Which is, to a very good approximation, never what I want.
So the Windows 7 mechanism is an improvement, in that I understand what it does. And, moreover, it only does the resize when you release the mouse button. But it does an animation the moment the mouse hits the edge of the screen, so you get warning of what is going happen, and can prevent it by dragging the mouse/window back away from the edge of the screen.
But, best of all, it can be disabled! There is a registry key (hey, it is windows, after all), and Google will find you any number of people telling you how to set it.
But there is also a tick box, which appears on several screens within the "Control Panel" hierarchy, that controls it. The box is labelled thus:
"Prevent windows from being automatically arranged when moved to the edge of the screen"
A 14-word description that does not rely on terminology, like "Aero Snap", that may be unfamiliar and off-putting to the novice user. Probably the result of much attention to usability, and no doubt the subject to extensive user testing and evaluation. Just the sort of label that one would hold up as an example of how to do interface design, if it only described what the tick box did.
But unfortunately, it doesn't: the tick box doesn't just disable "Aero Snap". It also disables "Aero Shake" --- a facility which has nothing whatsoever to do with dragging windows to the edge of the screen.
So the 14-word description of the tick box is clear, friendly, and wrong.
Thanks, guys...
First, it has the same "auto hide task bar" feature that Windows XP. Except that in XP, the task bar hides automatically. And in Windows 7, it usually doesn't. Don't know why. Sometimes it does. But, it seems, after a while it gets tired, or bored, and stops hiding.
Great.
Second, it has unnecessary and un-documented interaction between features.
Specifically, Windows 7 introduces two new features: Aero Snap and Aero Shake.
Aero Shake is a brand new feature. If you use the mouse to start dragging a window by its header bar, and "shake" the window, it will minimise every other window, leaving the one being dragged and shaken the only one open on the desktop. So you can see it clearly and, I guess, get at other things on the desk top at the same time. Shake it again, and the others go back the way they were. Kinda neat, I guess.
Aero Snap is a feature that, when you drag a window into the top or bottom of the screen, maximises it, while if you drag it into the side of the screen, it "half maximises" it --- that is expands it to fill half the screen. Now, XP has something like this, too, but I've never understood it. All I've noticed is that when I try to move a window, it sometimes maximises itself. Which is, to a very good approximation, never what I want.
So the Windows 7 mechanism is an improvement, in that I understand what it does. And, moreover, it only does the resize when you release the mouse button. But it does an animation the moment the mouse hits the edge of the screen, so you get warning of what is going happen, and can prevent it by dragging the mouse/window back away from the edge of the screen.
But, best of all, it can be disabled! There is a registry key (hey, it is windows, after all), and Google will find you any number of people telling you how to set it.
But there is also a tick box, which appears on several screens within the "Control Panel" hierarchy, that controls it. The box is labelled thus:
"Prevent windows from being automatically arranged when moved to the edge of the screen"
A 14-word description that does not rely on terminology, like "Aero Snap", that may be unfamiliar and off-putting to the novice user. Probably the result of much attention to usability, and no doubt the subject to extensive user testing and evaluation. Just the sort of label that one would hold up as an example of how to do interface design, if it only described what the tick box did.
But unfortunately, it doesn't: the tick box doesn't just disable "Aero Snap". It also disables "Aero Shake" --- a facility which has nothing whatsoever to do with dragging windows to the edge of the screen.
So the 14-word description of the tick box is clear, friendly, and wrong.
Thanks, guys...
Sunday, 15 November 2009
Accessible Toilets: Going through the motions
So Scotrail have done it again. Though I suspect the blame is
elsewhere.
Obviously we, as a society, have to take reasonable steps to make our environment and society accessible to those with a disabiltiy. So we have ramps and kneeling busses for people who cannot cope with steps; subtitles and induction loop systems for the deaf; bleeping pedestrian crossings and bobbly ramps for the blind.
I'm sure that many of these things make disabled people's lives better. But sometimes, I get the feeling someone is just going through the motions.
One such example is the toilets on some of the recently refurbished Scotrail diesel railcar units. The toilet in one of the carriages is designed to be particularly friendly to the disabled. It is large enough to accommodate a wheelchair, with a wide powered sliding door, operated by pushing buttons. And for the benefit of the blind, everything inside is labelled in Braille. Including this...
Yes, that's right. Someone has labelled the door status lights in Braille.
I admit it was the superficial absurdity that first caught my attention: if someone is reading the signs in braille hardly needs labelled indicator lights. Or are they supposed to use the heat from the bulb to work out which one is on?
OK, so it's not actually completely pointless: there are no doubt some people who can see well enought to know that there is a light on, but not well enough to read its label.
But Braille labels on indicator lights still feels like some kind of "legislation gone mad" arrangement.
And that's not the limit of the design problem, because the lights, and associated door controls, are located on the wall opposite the door. So our hypothetical blind traveller has to grope his way all the way round the (extra spacious) toilet area in order to find the controls to lock the door.
This, surely, isn't the best we can do for a disabled-friendly toilet.
Off the top of my head: instead of buttons and lights, have a chunky switch or lever that turns (or slides) between "Open" and "Closed" positions. And instead of putting it on the opposite side of the room, put it either on, or near, the door --- where one might expect things to do with the door to be located.
Just adding Braille to the labels on some indicator lights is just going through the motions.
Obviously we, as a society, have to take reasonable steps to make our environment and society accessible to those with a disabiltiy. So we have ramps and kneeling busses for people who cannot cope with steps; subtitles and induction loop systems for the deaf; bleeping pedestrian crossings and bobbly ramps for the blind.
I'm sure that many of these things make disabled people's lives better. But sometimes, I get the feeling someone is just going through the motions.
One such example is the toilets on some of the recently refurbished Scotrail diesel railcar units. The toilet in one of the carriages is designed to be particularly friendly to the disabled. It is large enough to accommodate a wheelchair, with a wide powered sliding door, operated by pushing buttons. And for the benefit of the blind, everything inside is labelled in Braille. Including this...
Yes, that's right. Someone has labelled the door status lights in Braille.
I admit it was the superficial absurdity that first caught my attention: if someone is reading the signs in braille hardly needs labelled indicator lights. Or are they supposed to use the heat from the bulb to work out which one is on?
OK, so it's not actually completely pointless: there are no doubt some people who can see well enought to know that there is a light on, but not well enough to read its label.
But Braille labels on indicator lights still feels like some kind of "legislation gone mad" arrangement.
And that's not the limit of the design problem, because the lights, and associated door controls, are located on the wall opposite the door. So our hypothetical blind traveller has to grope his way all the way round the (extra spacious) toilet area in order to find the controls to lock the door.
This, surely, isn't the best we can do for a disabled-friendly toilet.
Off the top of my head: instead of buttons and lights, have a chunky switch or lever that turns (or slides) between "Open" and "Closed" positions. And instead of putting it on the opposite side of the room, put it either on, or near, the door --- where one might expect things to do with the door to be located.
Just adding Braille to the labels on some indicator lights is just going through the motions.
Saturday, 31 October 2009
Mac Air Fan, Boy!
So I guess I should confess that I've got a Macbook Air.
One of the first generation models, with the smaller disk and the less-capable graphics card, so Lewis's had it at about a third off the price of the current one.
It's nice in many ways, though at times it does tend to get rather warm, and, when the cooling fan ramps up, distinctly audible.
That wouldn't make much difference to me, since I mostly use it for typing text and occasional bursts of web surfing, except that it can be driven to working hard by, say, a flash animation running un-noticed in a corner of an off-the-screen Web page. So I've got into the habit of making sure I avoid such things. But still, sometimes, the machine starts to get warm and "noisy" (well, OK, noticeably not silent :-) for no apparent reason.
When this happened the other day, I took the time to investigate. And discovered, eventually, that part of the sound system (coreaudiod) was using over 10% of the CPU, even though there was nothing that used sound being run at all. Moreover, Mr. Google showed me that other people were also observing this same phenomenon. It seems that it is prone to running away like this under some circumstances. The problem is present in Snow Leopard, too, and can be triggered by "any" application that uses audio --- even a terminal window (can you say "bell"?)
I also found a command to restart the daemon:
sudo launchctl stop com.apple.audio.coreaudiod
I tried it, and the new incarnation of the daemon was as lethargic/inactive as you'd hope for an audio daemon on a machine being silent.
Straight away, the CPU temperature started to fall, as in due course did the fan speed,
So it seems that the 10% load was enough to keep the CPU fully powered up, and thus fully heat producing...
Good, eh!
Robert.
One of the first generation models, with the smaller disk and the less-capable graphics card, so Lewis's had it at about a third off the price of the current one.
It's nice in many ways, though at times it does tend to get rather warm, and, when the cooling fan ramps up, distinctly audible.
That wouldn't make much difference to me, since I mostly use it for typing text and occasional bursts of web surfing, except that it can be driven to working hard by, say, a flash animation running un-noticed in a corner of an off-the-screen Web page. So I've got into the habit of making sure I avoid such things. But still, sometimes, the machine starts to get warm and "noisy" (well, OK, noticeably not silent :-) for no apparent reason.
When this happened the other day, I took the time to investigate. And discovered, eventually, that part of the sound system (coreaudiod) was using over 10% of the CPU, even though there was nothing that used sound being run at all. Moreover, Mr. Google showed me that other people were also observing this same phenomenon. It seems that it is prone to running away like this under some circumstances. The problem is present in Snow Leopard, too, and can be triggered by "any" application that uses audio --- even a terminal window (can you say "bell"?)
I also found a command to restart the daemon:
sudo launchctl stop com.apple.audio.coreaudiod
I tried it, and the new incarnation of the daemon was as lethargic/inactive as you'd hope for an audio daemon on a machine being silent.
Straight away, the CPU temperature started to fall, as in due course did the fan speed,
So it seems that the 10% load was enough to keep the CPU fully powered up, and thus fully heat producing...
Good, eh!
Robert.
Wednesday, 21 October 2009
Windows WTF
So, on an XP machine, I was transferring a file to a files system shared from a remote machine - a "network drive".
I browsed to the directory/folder in Windows Explorer (N.B. Windows Explorer), and noticed that there was an awful lot of accumulated junk in it.
So I selected a handful of files, and right-clicked on them. At which point I got a popup message from Internet Explorer (NB Internet Explorer) saying
Continue with what? I was only trying to bring up a menu!
And why does that, or anything else, involve Internet Explorer?
On the third attempt, out of curiosity, I tried clicking "yes".
The window went away.
Nothing else appeared to happen.
Thanks, Microsoft.
I browsed to the directory/folder in Windows Explorer (N.B. Windows Explorer), and noticed that there was an awful lot of accumulated junk in it.
So I selected a handful of files, and right-clicked on them. At which point I got a popup message from Internet Explorer (NB Internet Explorer) saying
This page has an unspecified potential security risk.
Would you like to continue?"
Continue with what? I was only trying to bring up a menu!
And why does that, or anything else, involve Internet Explorer?
On the third attempt, out of curiosity, I tried clicking "yes".
The window went away.
Nothing else appeared to happen.
Thanks, Microsoft.
Subscribe to:
Posts (Atom)
