[sane-devel] [lsb-discuss] LSB DDK: Support for scanner drivers/SANE into LSB3.2?

[Till Kamppeter]

Wichmann, Mats D wrote:
> Just comments on this piece, not on the bigger question of whether
> SANE support should be added to LSB and when (when mostly depends
> on who has time to do the work).
>

I have already asked on the SANE mailing list whether the SANE 
developers would provide documentation and testing scripts, but did not 
get an answer yet.

> Hardware specifics were always considered outside the scope of
> LSB as it was purely focused on defining a target for user-level
> applications.  With the new focus on "hardware" in the form of
> printers and perhaps now scanners we may need to think a bit
> about that.
> 

The LSB DDK is generally intended to allow hardware manufacturers to 
supply drivers in a distribution-independent way and so to make it 
easier for them to support Linux. It covers only user-space drivers, not 
kernel modules. Having the printing and scanning infrastructure in the 
LSB is also important for user-level applications, as they can have 
printing and scanning functionality then (most common user applications 
have printing functionality, and many also scanner access, like 
OpenOffice.org or the GIMP).

> getpass has been explained as one of those "don't use this" 
> interfaces as the password is stored in clear text in the 
> process's address space.  The glibc documentation contains 
> one alternative implementation.  It should be possible to
> have password prompting without using getpass specifically.
> 

So we should file bugs against CUPS and SANE that they use this unsecure 
function and should replace it?

    Till