Story about keyboard?

[micky]

Not sure if this is hardware or software, Posted to alt.comp.hardware
and microsoft.public.windowsxp.general

Right now the problem is not bothering me, but I wanted to tell you
all about it. Both because I'm curious and I expect the problem will
be back.

I have a Micorsoft multimedia keyboard. Dob't know the model.
Details about the keyboard at the bottom** I'm runing WinXP-Pro-SP3
on a Dell something.

I've been using it four years or more, and it might have been used, or
new old stock when I got it, and in the last 6 months Cntl-V stopped
working often. Yet I never had any trouble with plain V or with any
other Cntl combination. Isn't that weird in itself?
Cntl-C for example, would copy in text, but it was the text
I saved the previous time I used Cntl- V

Does Cntl-V use the V key switch in a different way from plain V? I
wouoldn't think so but this electronics stuff often ends up being
amazingly complicated.



Second weirdness: II only noticed this when copying a url from the
location bar at the top of a firefox page, and maybe from the body of
the page. When I copied text from within Agent (the newsreader) or
Eudora or a text editor, I don't remember ever having trouble. Isn't
that weird?

When I would have trouble, I would go back to the webpage and do
Cntl-V three or four times in a row, and almost always when I did
cntl-C after that, it had the right text and copied it in. I figured
the V key was wearing out, but wondered why V itself always worked.


Then I thought I was overwhelmed with open tabs in Firefox, so I
started keeping my own list of the ones that mattered, and closing
them. Over 4 days I got down from maybe 200 to about 40. On about
the 4th day of my doing this, Firefox came out with Version 12. All
of a sudden response time from the browswer was much faster, and
Cntl-V stopped giving me problems. Isn't that weird?

I tried Internet Explorer a little bit and it was much faster too, so
it probably wasn't the FF version.


So once in a great while now, Cntl-V doesn't work, like once in the
last two weeks, but for months it gave me trouble every day.

The computer has been in the basement since last August, and it's
still there. . If I can trust a 10 dolllar thermometer/hygrometer, the
humidity has been 60 to 75 percent. Other than that, the basement is
quite dry. Water spilled on the floor evaporates or soaks into the
cement quickly. No mold. I think the humidity has been in this
range all year. but I'll admit when the humidity was low outside, the
hygrometer was in the kitchen, not the basement.

Thought you'd like this story, and thanks for any help.


**I don't know if it matters but for the record, the model number is
no longer visible but I'm been using it for 3 or 4 years and I bought
it at a hamfest. It looked new, but maybe the user washed his hands
more than I do. It's computer cream colored and has 16 buttons
across the top and 3 more above the numeric pad.

 

[micky]

Yes. Ctrl-V (or Ctrl-anything, for that matter), sends out a different code
sequence from a microcontroller than simply pressing a letter key (assuming
both it, and the Ctrl key itself, are working).

I was unclear. Yes, I was sure about what you say, but it still
uses the V key switch in the same way, doesn't it? There isn't a
second set of contracts in the V key that works when Cntl is pressed,
and a first set of contacts that works when only V is pressed.

That would account for much of the weirdness, but I wouldn't think
they would make it that way.

 

[micky]

Each key has one set of contacts. But the resultant keyboard codes (numeric
values) coming out when key(s) are pressed will end up being different,
depending on whether or not the Ctrl, or Shift, or Alt, key were pressed at
the same time

Of course.

or not (so that the software can look for that sequence or
unique value and act appropriately.

I'm assuming you can't get Ctrl-V to work right in any application. (If

I know my first post was longer than maybe it should have been, but
no. I don't remember every having the problem using Cntl-V in Agent,
or Eudora, (or a text editor, though I don't use them much), and I
only noticed it happening in Firefox. But even there only some of
the time, and even when it happened in FF it was cureable by going
Cntl-V several times in a row.

that's not the case then the issue lies with the respective application (but
I bet it is the case). For example, you could try Ctrl-C and Ctrl-V in
different apps, like Notepad, Wordpad, and Word, and if you get the same
problem, it sounds like the problem is with the keyboard controller or a
related microcontroller.

Well, it's barely given me any trouble at all since what? Maybe since
a month ago when I got rid of 160 open tabs, or a month ago when I
stopped using so much RAM (although surely I always had enough for a
100 or 200 character url.) I don't know what, but now it works too
well to do tests like you suggest.

That's why this started out just as a story. As the subject line
shows.

Then I got greedy and asked for advice

If it starts up again, I'll keep real records and post back.


BTW, with FF, I think one can close a window and then reopen it from
the Frequently Saved Windows list (which is good for 5 minutes at
least) and when one reopens it, it doesn't actually fill in the tab
data until one goes to the tab. A good way to keep RAM usage low.

 

[SteveH]

Bill in Co brought next idea :

Well, who knows, maybe somebody else can chip in and shed some light.

Yeah, he's bonkers.

 

[Ken Blake, MVP]

Not sure if this is hardware or software, Posted to alt.comp.hardware
and microsoft.public.windowsxp.general

Please do not send the same message separately to more than one
newsgroup (called multiposting). Doing so just fragments the thread,
so someone who answers in one newsgroup doesn't always get to see
answers from others in another newsgroup. And for those who read all
the newsgroups the message is multiposted to, they see the message
multiple times instead of once (they would see it only once if you
correctly crossposted instead). This wastes everyone's time, and gets
you poorer help than you should get.

If you must send the same message to more than one newsgroup, please
do so by crossposting -- sending a single message simultaneously to
multiple newsgroups (but only to a *few* related newsgroups).
Ken Blake, Microsoft MVP

 

[micky]

Bill in Co brought next idea :

Yeah, he's bonkers.

Hehehe. Just got back from the grocery. I put in my phone
number, because I refuse to carry a bogus card and I just put in some
phone number every time, and it asked "Is this the correct phone
number? No or Yes". I pressed the Yes part of the screen 10 times,
light and hard, and it never worked. The cashier had to do it.

A minute later I had put in my credit card and it asked "Is this
amount correct? No or Yes." And the Yes worked just fine on the
first try.

Not the fist time this has happened.

 

[Brian Cryer]

Not sure if this is hardware or software, Posted to alt.comp.hardware
and microsoft.public.windowsxp.general

Right now the problem is not bothering me, but I wanted to tell you
all about it. Both because I'm curious and I expect the problem will
be back.

I have a Micorsoft multimedia keyboard. Dob't know the model.
Details about the keyboard at the bottom** I'm runing WinXP-Pro-SP3
on a Dell something.

I've been using it four years or more, and it might have been used, or
new old stock when I got it, and in the last 6 months Cntl-V stopped
working often. Yet I never had any trouble with plain V or with any
other Cntl combination. Isn't that weird in itself?

When diagnosting a problem with a keyboard the easiest thing is to swap the
keyboard for another one and then see whether the problem has gone away. If
you don't have another keyboard then can you borrow one from a friend or
from work, failing that ps2 and usb keybaords are quite cheap (although
wirless ones cost a lot more).

If the problem goes away with a new keyboard then bin the old one, if it
doesn't then at least you know its not a hardware problem - although I'm not
sure where you would go from there.

 

[Brian Cryer]

Please do not send the same message separately to more than one
newsgroup (called multiposting). Doing so just fragments the thread,
so someone who answers in one newsgroup doesn't always get to see
answers from others in another newsgroup. And for those who read all
the newsgroups the message is multiposted to, they see the message
multiple times instead of once (they would see it only once if you
correctly crossposted instead). This wastes everyone's time, and gets
you poorer help than you should get.

If you must send the same message to more than one newsgroup, please
do so by crossposting -- sending a single message simultaneously to
multiple newsgroups (but only to a *few* related newsgroups).
Ken Blake, Microsoft MVP

Whilst I agree with everything you said ... the OP was cross-posting not
multi-posting, so he was doing it right.
http://everything.explained.at/Crossposting/

 

[micky]

When diagnosting a problem with a keyboard the easiest thing is to swap the
keyboard for another one and then see whether the problem has gone away. If

Darn. I know that but didn't think of it. Another senior moment?

If it starts up again, and I bet it will eventually, I'll do that.

Thanks and thanks for your other post. And thanks Bill.

 

[Paul]

In message <[email protected]>, micky

I wouldn't have thought so; obviously (and as you had some difficulty
conveying that you realised), the _controller_ produces different codes
if another key is pressed at the same time, but I doubt there's more
than one set of contacts.

Actually, the approach has changed with time.

Back when I built my own keyboard from scratch, the controller
used a matrix for most keys, but the modifier keys were
separate. And the modifier keys, actually "modified" something
at the hardware level. So the shift and control keys, used
separate wires and were monitored separately. The matrix is
scanned to detect ordinary key closures. That keyboard design
was so old and crusty, I think the chip may have been NMOS.

+ --- Ctrl ---- Old --- rows--------- | | |
+ --- Shift --- Controller --- columns --+ | | |
| | | |
+---+-+-+

I was checking a datasheet for a USB keyboard controller,
and *all* the keys on that, were on the matrix side of the
chip, including the control key. So it looks like this
at the hardware level.

New --- rows--------- | | |
Controller --- columns --+ | | |
| | | |
+---+-+-+

That means that *something* has to keep track of
closures of shift and control. It might be done
inside the keyboard chip, or it could be tracked
somewhere else.

The problem with such an approach (put everything on
the scanning matrix), is avoiding ghosting, and
implementing N-key rollover. On my home-made
keyboard, I used a diode per key, which made
my keyboard N-key rollover (ghost free). I've not
noticed that in retail keyboards - I've not seen
a keyboard PCB, featuring a diode at each key crossing.

http://en.wikipedia.org/wiki/N-key_rollover

Another possible difference, is my homemade keyboard
only did "key down" codes. There were no "key up" codes.
I understand keyboards now, may do both, not really sure.
The datasheet I was looking at, did not clarify that issue.
The datasheet showed hex codes for what happens when a key
is pressed, but didn't offer any more details than that.
You'd think if it was doing key up and key down, there'd
be two tables of values, rather than just one table.
Maybe there's a byte pre-pended that says whether its
keyup or keydown, but that wasn't clear.

43,B8,00,00, Row 16 column7 Normal Key: <Ctrl-R>
42,81,00,00, Row 16 column1 Normal Key: <Ctrl-L>

00,00,00,3D, Row 3 column7 Normal Key: <C> <--- Something tracks
00,00,00,0A, Row 4 column 7 Normal Key: <V> shift state... No
small c or small v.

43,CD,00,00, Row 15 column7 Normal Key: <Shift-R>
43,94,01,00, Row 15 column4 Normal Key: <Shift-L>
00,00,00,1A, Row 2 column4 Normal Key: <Caps> <--- Caps Lock?

I've always found tracing the path from keyboard to screen
on a computer, to be pretty torturous. Even before sitting
in this chair, I've tried to research this topic for other
computer systems (eons ago), and come up empty. I could never
get all the details in one place. If I knew of a good
tutorial, I'd offer it.

Paul

 

[Char Jackson]

I also found no difference when I went from 1 GB to 2 GB of RAM, apparently
because I haven't yet tapped it out, even with some (limited) video
processing work. I tried checking the PF on a few such occasions and never
got close to needing it, so apparently you really have to be doing something
"VERY HEAVY" to need that 2 GB. Like perhaps multitasking a bunch of big
apps, which I never do.

I usually just single task - I rarely run multiple applications all at
once. (Probably for two reasons: 1) I don't tend to multitask all that
well, anyways, and 2) it's just generally safer.

You sound a little like one of my regular customers, a friend of mine
who is about 75-80 years old. He called me over to help him with a
problem he was having, and it went something like this.

When I got there, he was doing something in Photoshop. Nothing
important, so he exited without saving. Then he (re)launched Photoshop
to load the image he wanted to ask me about. He couldn't find it, so
he exited Photoshop and launched Windows Explorer. Using that, he only
found a copy that had already been modified, so he exited Windows
Explorer and launched Outlook Express. In OE, he found the email that
had the original image as an attachment, and saved it to his desktop.
He exited OE and launched Windows Explorer. He navigated to his
Desktop (not directly, but in Win Explorer) and renamed the newly
saved image. He exited Win Explorer and launched Photoshop. He loaded
the image, we discussed it a bit, and he then edited and saved the
image to his desktop under a new name. He exited Photoshop and
launched OE. In OE, he created a new email and attached the modified
image, then he sent it. He exited OE and launched Win Explorer, where
he navigated to his Desktop again (not sure why he couldn't just click
on the images on his desktop, but he has his own ways), then he
deleted both images from his desktop. He exited Win Explorer, then he
double clicked on his desktop's Recycle Bin. Once that window opened,
he selected the menu item to empty the trash. At no time did he have
more than one application open.

I wanted to strangle him, metaphorically speaking.

 

[Char Jackson]

I don't think I'm quite that bad.

Yeah, that was pretty extreme, but I've spent a lot of time with him
and that's just the way he prefers to work.

Nonetheless, if I'm (say) trying to restore an audio or video file, that,
and that alone gets all my attention. I'm not going to be emailing, or
running Skype, or working on a Word document, or what have you, while I'm
doing that. Works a lot better that way. (I'm not one of these newbies who
has to always be connected, and twittering, or texting, or Googling, or
Facebooking, or what have you, lest I feel "bored" (aka: newage short
attention span), or feel I'm out of the loop, so to speak. (I sometimes
feel sorry for these folks). You know, the ones that also text and drive,
and gab on the cell phone while driving, etc, etc.

And - I always know (or can readily where my documents in any app are
actually being saved (if sometimes even having to run Agent Ransack and
checking the file times . And if anything crashes, only one app is
affected, and the likelihood of that is diminished since I'm basically only
running one app. (and just monitoring files thru windows explorer while any
of this is going on doesn't count)

I normally have enough things running that they require a second row
of Taskbar icons. Once in awhile they spill over into a third row, but
not usually.

 

[Char Jackson]

In message <[email protected]>, Char Jackson

You sound a little like one of my regular customers, a friend of mine [frustrating tale deleted]
I wanted to strangle him, metaphorically speaking.

Talking of frustrating habits: I find similar frustration with people
who don't know about the enter, let alone tab key. My team leader at
work will type something into a box, then mouse up to the go key next to
the box - when enter would do fine! This is in several applications.
People who don't know the tab key fill in a web form by filling in a
box, then using the mouse to get to the next box, and finally use the
mouse to activate "go" or "submit" or whatever, when again enter would
work in the vast majority of situations.

It is these people I want to strangle ... (-:

I agree, but programmers have to do their part, too. Fortunately, they
almost always pay attention to tab order, but quite a few years ago I
used to use a file renaming program called TheRename 1.4, (current
version seems to be 2.16, although I no longer use it). Pressing the
tab key would take you on a crazy convoluted tour of the semi-busy
interface, with no rhyme or reason. With that one program, I learned
not to use tab, but of course enter always worked.

How about people who have a problem and assure you they've "tried
everything". Or that they've tried everything and "nothing worked". No
mention of exactly what was tried or any error messages encountered.