[BUG] 2.6.12-rc1/rc2 mouse0 became mouse1

[BUG] 2.6.12-rc1/rc2 mouse0 became mouse1

[Ali Akcaagac]

Hello,

This doesn't sound right to me. After upgrading from 2.6.11.5/6 to
2.6.12-rc1/rc2 I detected that my mouse didn't operate anymore when
loading up XOrg, I realized that /dev/input/mouse0 (which worked for
years) had to be changed to /dev/input/mouse1. Anyone care to explain
why this had to be changed or what the intention behind this was ?

greetings,

Ali Akcaagac

Re: [BUG] 2.6.12-rc1/rc2 mouse0 became mouse1

[Dmitry Torokhov]

Hi,

On Apr 8, 2005 6:58 AM, Ali Akcaagac <aliakc@web.de> wrote:
> Hello,
> 
> This doesn't sound right to me. After upgrading from 2.6.11.5/6 to
> 2.6.12-rc1/rc2 I detected that my mouse didn't operate anymore when
> loading up XOrg, I realized that /dev/input/mouse0 (which worked for
> years) had to be changed to /dev/input/mouse1. Anyone care to explain
> why this had to be changed or what the intention behind this was ?
> 

Input devices names are not guaranteed to be stable, they get them in
order of their registration and therefore can change based on config,
order in which modules are loaded and changes in other modules. In
this case shift was caused by scroll support recently added to atkbd.

Ideally hotplug system should notify interested parties about current
layout, but for programs that do not support dynamic setup changes
please use mice multiplexor device (/dev/input/mice) that gives access
to data from _all_ mouse-like devices in the system.

Hope that helps.

-- 
Dmitry

Re: [BUG] 2.6.12-rc1/rc2 mouse0 became mouse1

[Ali Akcaagac]

On Fri, 2005-04-08 at 09:13 -0500, Dmitry Torokhov wrote:
> Input devices names are not guaranteed to be stable, they get them in
> order of their registration and therefore can change based on config,
> order in which modules are loaded and changes in other modules. In
> this case shift was caused by scroll support recently added to atkbd.
> 
> Ideally hotplug system should notify interested parties about current
> layout, but for programs that do not support dynamic setup changes
> please use mice multiplexor device (/dev/input/mice) that gives access
> to data from _all_ mouse-like devices in the system.
> 
> Hope that helps.

I see, thank you... Good to know this wasn't a bug or something.

Re: [BUG] 2.6.12-rc1/rc2 mouse0 became mouse1

[Pavel Machek]

Hi!

> This doesn't sound right to me. After upgrading from 2.6.11.5/6 to
> 2.6.12-rc1/rc2 I detected that my mouse didn't operate anymore when
> loading up XOrg, I realized that /dev/input/mouse0 (which worked for
> years) had to be changed to /dev/input/mouse1. Anyone care to explain
> why this had to be changed or what the intention behind this was ?
> 

This is a bug where we assume keyboard has scrollwheel, and therefore
we pretend its keyboard and a bit of mouse?

Did we stop generating zero mouse movements during typing?
				Pavel
-- 
64 bytes from 195.113.31.123: icmp_seq=28 ttl=51 time=448769.1 ms