udev issues - revisited

udev issues - revisited

[Jim Bodkikns (Dakotacom)]

   Sorry, I have been out of town on business. Thank you, for accusing me of 
being a troll. (Whoever that was)

   This is not a request for assistance. It was what I said it was. A 
comment. Pointless, because I assumed that nothing would come of it. Why? If 
you had intended to provide any form of transitional support (given the 
paradigmatic shift involved here from devfs to udev etc) you would have done 
it by now. I have been writing software a long time and long ago lost any 
immediate interest in wanting to control all aspects of any software. As a 
result, Linux, while nice - is just another OS. Sorry,. I use it - 
appreciate it. (I'm not a brand fan either for that matter). I have simple 
expectations. Software that is released for general use should function more 
or less as advertised. Until it does, it shouldnt be released. (Just my 
opinion) Does udev/hotplug function more or less as advertised? Sure, but 
not everywhere. If I install it from the start, my experience is that it 
functions correctly. That isnt what I am doing. I am attempting to upgrade 
existing systems. I have had nothing but trouble with the transitiont. 
(Including system that wont boot)  I understand that you may think that as 
long as the underlying technology functions, its not your issue. Some-one of 
you, I dont know who, made or influenced the transition from devfs etc to 
udev etc.Its that transition that makes an otherwise stable and functional 
developement system questionable. I read that this was the place to post 
comments. I did that. A new technology that results in the changes present 
here should be accompanied with a means a effecting the change. It doesnt. 
This isnt optional. And if doesnt work. (Transition, the underlying code 
would work if I could get a system that actually ran).

   I am very appreciative of the hard work done by so many in the linux 
community. But with reguards to a mandatory shift, I suggest that it needs 
to either work or be rigorously supported. (Not me, several levels above me) 
I am not a wild MS fan, but frankly - this would not have occured with MS. 
They would have addressed the 1) issues of the technology itself 2)  the 
issues of implementing the changes on existing systems. That is 
professional. (My customers expect as much from me). That hasnt happened 
here. (I'm upset about it because of the consequences that no choice of mine 
has resulted in)

   Sorry if this is rambling. I am exhausted. I hadnt planned a reply, and 
shouldnt reply as tired as I am. But I believe the point needed to be made 
that simply providing an installable new technology is only part of a 
deliverable solution. Issues of implementing this technology on existing 
platforms must also be considered for it to be (IMO) deliverable. That just 
isnt happening in this case. Some would consider it too much to ask to 
simply apt-get the appropriate kernel and expect it to work. Sorry, I expect 
just that. (Sorry again but I am too busy to assume the responsibility of 
trying to determine why it wont work. I just need to use it. If I have 
time - later perhaps - I might wonder how all this works. Not now). For now, 
support appears to be needed (at a level above me - with the packager - 
debian) to solve the issue of transitioning to your new technology.

   This is just a comment. If this offends anyone, sorry. I dont care which 
I use as long as it works. (Udev works if I could just get there)

Jim
P.S.
I am not a troll. It was either join this list to post my comment or email 
the developers at personal emails - the latter would have been 
inappropriate. Frankly, I dont spend enough time idling online to qualify as 
a troll.l :) (I dont have the time at the moment).




-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems?  Stop!  Download the new AJAX search engine that makes
searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid\x103432&bid#0486&dat\x121642
_______________________________________________
Linux-hotplug-devel mailing list  http://linux-hotplug.sourceforge.net
Linux-hotplug-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel

RE: udev issues - revisited

[Robert Kennedy]

Jim,

I understand your frustration.  But I suspect that some of your frustration 
may not be with udev but with the actual Linux distribution in question.  As 
I understand it, you do not need to use udev with the 2.6 kernels.  You can 
still use devfs and the old hotplug daemon.  But some distros may force you 
to switch to a udev with its built in hotplugging support.  Often the 
distros do not realize that the switchj to udev may not be painless for all 
of its users.  (In my view, this is a problem with the distro).

So far the transition from devfs to udev has been pretty painless for me.  
The only linux computers that have had problems are computers that use 
usb.usermaps (e.g. usb.usermaps are used by the hotplug daemon to load 
firmware via the fxload utility into some external USB devices during 
hotplugging).   udev does not use usermaps.  When you switch to udev, you 
need to convert the usermaps to udev rules.  It is not hard to do so.  For 
example, for a linux computer with an older Backpack USB CD writer, I wrote 
a simple script to convert the usb.usermap to udev rules.  Now this linux 
box works great.

In fact I find that hotplugging uder udev is much faster than hotplugging 
under the old hotplug daemon.  And in my experience, cold plugging is also 
better (but some may disagree with me on this point).

I agree with you that it would be so much better if there was better 
documentation and tools to make the transition.   I also think that some of 
the Linux distros do not know what needs to be done to make the transition 
from devfs to udev easy and painless for their users.  If they did, perhaps 
they could give more guidance and perhaps provide some tools.

There is a pretty good HOWTO on how to write udev rules.  And there is a 
blog (or two) that provides some good info on udev.  And of course there is 
the man page.   But a lot of what I have learned comes from reading posts on 
this mailing list.

Rob


>From: "Jim Bodkikns (Dakotacom)" <JimBodkins@yahoo.com>
>To: <linux-hotplug-devel@lists.sourceforge.net>
>Subject: udev issues - revisited
>Date: Tue, 24 Jan 2006 01:59:41 -0700
>
>   Sorry, I have been out of town on business. Thank you, for accusing me 
>of being a troll. (Whoever that was)
>
>   This is not a request for assistance. It was what I said it was. A 
>comment. Pointless, because I assumed that nothing would come of it. Why? 
>If you had intended to provide any form of transitional support (given the 
>paradigmatic shift involved here from devfs to udev etc) you would have 
>done it by now. I have been writing software a long time and long ago lost 
>any immediate interest in wanting to control all aspects of any software. 
>As a result, Linux, while nice - is just another OS. Sorry,. I use it - 
>appreciate it. (I'm not a brand fan either for that matter). I have simple 
>expectations. Software that is released for general use should function 
>more or less as advertised. Until it does, it shouldnt be released. (Just 
>my opinion) Does udev/hotplug function more or less as advertised? Sure, 
>but not everywhere. If I install it from the start, my experience is that 
>it functions correctly. That isnt what I am doing. I am attempting to 
>upgrade existing systems. I have had nothing but trouble with the 
>transitiont. (Including system that wont boot)  I understand that you may 
>think that as long as the underlying technology functions, its not your 
>issue. Some-one of you, I dont know who, made or influenced the transition 
>from devfs etc to udev etc.Its that transition that makes an otherwise 
>stable and functional developement system questionable. I read that this 
>was the place to post comments. I did that. A new technology that results 
>in the changes present here should be accompanied with a means a effecting 
>the change. It doesnt. This isnt optional. And if doesnt work. (Transition, 
>the underlying code would work if I could get a system that actually ran).
>
>   I am very appreciative of the hard work done by so many in the linux 
>community. But with reguards to a mandatory shift, I suggest that it needs 
>to either work or be rigorously supported. (Not me, several levels above 
>me) I am not a wild MS fan, but frankly - this would not have occured with 
>MS. They would have addressed the 1) issues of the technology itself 2)  
>the issues of implementing the changes on existing systems. That is 
>professional. (My customers expect as much from me). That hasnt happened 
>here. (I'm upset about it because of the consequences that no choice of 
>mine has resulted in)
>
>   Sorry if this is rambling. I am exhausted. I hadnt planned a reply, and 
>shouldnt reply as tired as I am. But I believe the point needed to be made 
>that simply providing an installable new technology is only part of a 
>deliverable solution. Issues of implementing this technology on existing 
>platforms must also be considered for it to be (IMO) deliverable. That just 
>isnt happening in this case. Some would consider it too much to ask to 
>simply apt-get the appropriate kernel and expect it to work. Sorry, I 
>expect just that. (Sorry again but I am too busy to assume the 
>responsibility of trying to determine why it wont work. I just need to use 
>it. If I have time - later perhaps - I might wonder how all this works. Not 
>now). For now, support appears to be needed (at a level above me - with the 
>packager - debian) to solve the issue of transitioning to your new 
>technology.
>
>   This is just a comment. If this offends anyone, sorry. I dont care which 
>I use as long as it works. (Udev works if I could just get there)
>
>Jim
>P.S.
>I am not a troll. It was either join this list to post my comment or email 
>the developers at personal emails - the latter would have been 
>inappropriate. Frankly, I dont spend enough time idling online to qualify 
>as a troll.l :) (I dont have the time at the moment).
>
>
>
>
>-------------------------------------------------------
>This SF.net email is sponsored by: Splunk Inc. Do you grep through log 
>files
>for problems?  Stop!  Download the new AJAX search engine that makes
>searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
>http://sel.as-us.falkag.net/sel?cmd=lnk&kid\x103432&bid#0486&dat\x121642
>_______________________________________________
>Linux-hotplug-devel mailing list  http://linux-hotplug.sourceforge.net
>Linux-hotplug-devel@lists.sourceforge.net
>https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel




-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems?  Stop!  Download the new AJAX search engine that makes
searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid\x103432&bid#0486&dat\x121642
_______________________________________________
Linux-hotplug-devel mailing list  http://linux-hotplug.sourceforge.net
Linux-hotplug-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel