Warning, /utilities/kfloppy/TODO is written in an unsupported language. File is not indexed.

 2005-06-07
 
 As I am wirting an answer to an email about the to-do in KFloppy, 
 better have it in the source too.
 
 
 
 The biggest challenge was to support non-legacy (USB, IDE...) floppy drives.
 I have opened a way to it with what has the working title "user-given devices".
 (A better title for the feature would be a good thing too.
 It just mean that instead of selecting "Primary" or "Secondary" as floppy drive,
 you can give a device, like /dev/sda4 )
 
 The "user-given device" is pretty raw currently, allowing nearly no comfort for
 the user. So if you have an external drive (may a ZIP drive), you could try to
 make this mode better.
 
 A little variant would be to support the device like /dev/fd0u2880 for
 formatting. Currently it does not work with Full Format,
 as a "user-given device" is supposed not to be low-level formatable.
 (I do not know any Linux or BSD tool that can low-level format 
 non-legacy drives.)
 
 Another thing that I have very recently discovered is that 
 extended floppy drives will not work in "superfloppy" format 
 (disks without partition table), as neither mkdosfs nor BSD's newfs allow this
 without a special parameter. (I have not checked the other tools used.)
 
 Before I forget it: in the "user-given device" mode, it would be nice if there
 was support from KDE (for example devices:/ ) to know what devices are really
 flopy drives (not that a user roasts an umounted partition of his harddisk.
 
 A much easier task would be perhaps to catch more errors from 
 the external programs so that it is displayed to the user.
 (Also to catch anything error message having the string "error" in it,
 which is currently not done.)
 
 That is I think the list of what users could benefit.
 
 Internally, when the Qt4 switch is done, it would need a review to see how 
 to improve the way how the classes calling the external applications are
 handled.
 
 Of course, when modifying, care must taken not to break the BSD part of the
 code. (Perhaps we need a way on Linux to be able to compile the BSD part,
 as there is nothing BSD specific in the code, only the external BSD programs 
 are specific.)