keyboard very unresponsive / often missing keypresses

Have you accidentially enabled the 'assistive technology'.

Something like what you have described can happen if you type too quickly, and the keyboard setting requires key to be held down in order to register.

To disable, look for the option under 'assistive' or 'keyboard' dialog, under the System menu.

see bug #63024 and bug #63137

To Paul: the assistive technology is disbled (hope, we mean the same - in German "Barrierefreiheit")

To Alessandro: it seams, that bug report #63024 points to a similar keyboard behavior.

I have played with the "Key Repeat Value" on the Keyboard Settings but the issue still appears.

I have also checked:
* Kernel boot parameters
** acpi=off
** acpi=noirq
* killed some tasks responsible for Harddisk activity every 4-8 seconds
* disabled the Touchpad

Thanx to all!

Frank

I can confirm this bug!

I have the same hardware: ACER Extensa 3000 Notebook with Dapper running.
I also played arround with various settings but nothing worked.

This bug is really annoying and makes (K)ubuntu unusable for writing longer documents!!!

I also can confirm this bug.
Hardware is an ACER Aspire 1520 Notebook with Edgy running now.
Also #bug 63024 is the other way round I confirmed.
Either keystrokes are lost oder repeated too often.

I reported bug 63024.
When "keyboard repetition"(? Herhaaltoetsen in Duch) is enabled, I have too many keypresses. When it is disabled, I lose keypresses.

This bug impedes my basic human-pc I/O, making it virtually unuseable. I have to doublecheck every letter I write. Most annoying are passwords: you don't see what you type...

I am not using a notebook but a diy desktop.

Thanx to all for feedback! I still fight with this issue and like Amedee it is a quite unsatified situation. Unfortunately, I have no experience using this bug tracking tool. How can I trigger this bug report and change the "Needs info" status?

If anybody is monitoring that bug report who needs more information than please ask for information!

Again, thanx to all...

Colleagues of mine using Ubuntu creating SSH Certificates with multiple-part-passwords shared by different users... they ask me about curious issues on password sessions and they are quite happy as I told them from that issue!

I did a bit more troubleshooting.
When I boot with a "recovery mode" kernel, for example:

kernel /vmlinuz-2.6.17-10-386 root=/dev/md1 ro single

the problem does not appear.

When I boot with a fresh compiled plain vanilla kernel, straight from kernel.org, the problem is also there. This rules out any debian- or ubuntu-specific patches to the kernel. So it's either a standard kernel problem, or it's something completely unrelated.

I took a deep breath, made a good backup, and upgraded to Edgy. I don't want to jump to conclusions, but it appears that the keyboard issue is gone. This message is the longest text that I have typed in about two months (!) without double and triple checking every keypress. It took me about 2 minutes (including looking up that line in /boot/grub/menu.lst) when before it would have taken me over 15 minutes.

@amedee: Thanks a lot for your feedback! Do you think we talk about a problem which is allready solved by newer kernel versions (used by Edgy) ?

@Frank: on the contrary, I am now convinced the problem is not kernel related at all!

Let's compare 3 kernels:

Ubuntu Dapper kernel: 2.6.15-27 --> problem
vanilla kernel.org kernel: 2.6.18 --> problem
Ubuntu Edgy kernel: 2.6.17-10 --> no problem

As you can see, the kernel.org kernel is even more recent than the Edgy kernel, but the problem still occurs. On the other hand, the Edgy kernel is not as recent, but the problem is gone.

The simplest explanation is that the problem is completely unrelated to the kernel, but that one of the many, many other packages is involved. And when I upgraded Dapper to Edgy, a faecesload of packages got replaced with a newer version.

i am seeing this on a Powerbook g4 - this occurs occasionally and I have to press each key for about a second for it to register.

extremely annoying

Edgy 6.10

note that this bug affects the on-screen keyboard as well

The keyboard malfunction reported in the above posts concerning the PowerBook was due to having activated SlowKeys by holding down Shift for more than 8s.

Suggest that the SlowKeys activation dialog should be placed above all other windows on the current desktop, and in any case should not be inhibiting keypress events from the keyboard.

I just updated to edgy and the keyboard works well again (Acer Extensa 3000 Laptop)!
Does anyone know which difference between edgy and drapper caused this problem?
Or will it stay one of these miracle ... ?

This annoying error is still in my kubuntu edgy.
I'm going to drive crazy......

Yes, I can confirm this. I have first installed Kubuntu on my notebook and the keyboard issues were stronger and more frequently than in Ubuntu.

This keyboard bug is very misterious and I have no idea what I can do to narrow down that problem? Maybe updating to Edgy...;-)

@Erik: Thank you for your response/feedback. If I do understand your message in the right manner than something called the SlowKeys-Feature is responsible for that crazy keybord issue, isn't it? What do you mean with "SlowKeys aktivation dialog" and how can I disable that feature/behavior?

@Tjengo: Yes, this is a very unsatisfied situation. But, how the heck can you modify the status of that bug report to get more drive... ;-)

Additional notes:
http://packages.ubuntu.com/edgy/x11/xkbset (=SlowKeys ??)

In https://launchpad.net/bugs/63024 I wrote a solution which at least worked for me. Can anybody confirm the connection between spurios keyboard events and the e100 ethernet driver module?

@MrWulf: this was a hint I was looking for. I uninstalled the ndiswrapper-package und build it on my own. for me the keyboard seems to work fine. A little bit more testing is still needed but so far... thank you

On Mon, November 6, 2006 22:54, WrMulf said:
> In https://launchpad.net/bugs/63024 I wrote a solution which at least
> worked for me. Can anybody confirm the connection between spurios
> keyboard events and the e100 ethernet driver module?

Since I have an entirely different ethernet driver, I cannot confirm nor
deny a connection.

lspci | grep Ethernet
0000:00:0a.0 Ethernet controller: Marvell Technology Group Ltd. 88E8001
Gigabit Ethernet Controller (rev 13)

My solution was upgrading to Edgy.

--
Amedee

Updating to Ubuntu 6.10 (Edgy) works for me. But:

Regarding my Bug Requests I have talked with different people still fight with that keyboard problem. In a current case the user has upgrated to Ubuntu 6.10 but the issues still there. The user works on a desktop system with SiS chipset.

I suggest to consolidate the requests refered to that kind of keyboard problems and give them more drive?

See also: https://launchpad.net/bugs/63024.

That bug report does not reflect my problem 100% but the status is "confirmed" (Importance: medium).

Hi,

I already posted my confirmation of this bug in 63024, but what kind of information are needed?

Cheers,
Benny

[Expired for Ubuntu because there has been no activity for 60 days.]

There are some other Problem Records in that Database refering same or similar issues. As I have recommenden in the past a responsible person should consolidate that records - nothing happens.

If we asked, what information do you need - nothing happens.

This issue is still be there - hope the problem will be fixed in the future, because that issue is really embarrassing for Ubuntu and Kubuntu.

I have had a lot of problems with M$ before I have changed - but never I had a problem with the stupid keyboard!

Thanks.

See also: https://launchpad.net/bugs/63024.

On Fri, September 21, 2007 22:06, Launchpad Janitor wrote:
> [Expired for Ubuntu because there has been no activity for 60 days.]
>
> --
> keyboard very unresponsive / often missing keypresses
> https://bugs.launchpad.net/bugs/63147
> You received this bug notification because you are a direct subscriber
> of the bug.
>
>
It is not very considerate (see: Ubuntu Code of Conduct) to auto-expire a
bug because of lack of activity, when the lack of activity is because no
developer or bugfixer is reviewing the bug report. We, the bug reporters,
are willing to give information. We will give you all the logfiles you
want. But only if you tell us which log files. Or do you want us to upload
a complete image of our system?

Likewise, it would not be very considerate of us, the bug reporters, to do
an auto-bump of bug reports every time we think it is taking too long. We
still believe that it is not productive to interrupt more knowledgeable
people when there is not yet news to report and we do not have anything
new to add.

Just my 0.02€.

--
Amedee