#include <hallo.h> * Marco d'Itri [Mon, Jan 08 2007, 05:34:47PM]: > On Jan 08, Eduard Bloch <[EMAIL PROTECTED]> wrote: > > > My impression is that it always takes the X string part from cdromX and > > appends it to cdrw, resulting in no "cdrw" symlink if "cdrom" has no > > writting capabilities. If I am right then... uhm, it sucks and should be > > changed ASAP. A setup like mine (first drive a pure reader and the > > second is the writter) is not that uncommon. > Yes, it works this way (see /lib/udev/write_cd_rules). > I think this is the least confusing scheme *and* other distributions > work this way too so I am not sure it's a good idea to change it. > But feel free to try to persuade me...
Okay, think yourself: whom does that symlink serve? The user, the user's convinience. And the programs that do expect some functionality represented by that device name. That is why it is called /dev/cdrom and not /dev/cdrom0 because it needs to be used by humans and human-oriented programs. Agreed? And what do we have now? Imagine, the system has a CD-RW device. And where is a reliable link representing this functionality? It is MISSING. Not there. So why do you want that I do not add that obvious behaviour to wodim but refuse to add a udev solution for that? > BTW, now I see that it's a removable device. See #395962 about how to > get really persistent names. I'd like to fix this for etch, but I am > really busy with my day job right now and I have not even been able to > follow up this bug. I will try tonight. Btw, do you have an idea why the z25... file has not been created at the first time after reboot? Eduard. -- * Tolimar ist sooooo doooooooof. Der Befehl zum installieren eines einzelnen Debian-Paketes lautet nicht »rm«. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
