Philip Meyer;505354 Wrote:
> I meant if there were more than one file. If there's only one, then use
> it, if there's more than one, and none match the names that the scanner
> would usually look for, then don't arbitrarily pick any of them.
+1.
I want to see where errors are made and don't want to have an 'arbitary
random' image.
Whatever algorithm you take - alphabetical, resolution, filesize - from
the users point of view you take a random image which *could* be the
right - but most of the time plain wrong even in the simplest
scenarios:
- simple scenario: back.jpg + front.jpg - image
alphabetical: takes 'back.jpg' => wrong image
by size/resolution: takes 'back' (as the back image is bigger) =>
wrong image
of course these names change - but when choosing by alphabet in the
end this boils down to a 50:50 chance to get either one.
- advanced scenario: back + front + booklet
alphabetical: random, as we don't know how booklets typically are
named
by size/resolution: takes booklet (its biggest, because 2-sided) =>
wrong
- more advanced scenario: back + front in different resolution
alphabetical: random (as the size is somewhere in the name - always
in a different spot for each album downloaded somewhere)
-100x100front.jpg, front100X100jpg, front50x50
by size/resolution: random
--
bluegaspode
1x SB-Controller+Receiver (Duet), 1xSB-Boom. 1xSB-Radio
Server (7.4.1) running on SheevaPlug (Ubuntu) with attached Western
Digital MyBook Essential.
Secondary 7.4 Server on Debianized Buffalo Linkstation LS-CHL.
------------------------------------------------------------------------
bluegaspode's Profile: http://forums.slimdevices.com/member.php?userid=31651
View this thread: http://forums.slimdevices.com/showthread.php?t=71768
_______________________________________________
beta mailing list
beta@lists.slimdevices.com
http://lists.slimdevices.com/mailman/listinfo/beta
