Marcel Kilgus writes:
<>
Im assuming that you were answering two different mails here. Forget the QPC
'hole' that got me going and lets look at path depth for SMSQ/E in general:
> Unfortunately directories have to be read "raw", meaning that the
> format is limited to 36 characters. If one were to overcome this, one
> would probably have to create a few new "real" directory traps. After
> those are established and used in all the applications one could then
> think about extending them to allow more characters.
That would be one way forward, and I did have that in mind as a possibility.
> But I can already hear the whining "but I can't adapt my application
> to use the new traps as then it wouldn't be QDOS compatible anymore",
> so I probably just don't bother. Not trying to discourage anybody
> else, of course, it's just my view of things.
There are a number of different defenses to this argument:
1) Ignor Qdos as such a DDD neednt apply to low-end devices such as
floppies
The recent questionnair should be able to answer the question: What
percentage of QLers use both Qdos AND hard disks [HDD] (a small percentage I
would think) Minerva and emulator users can upgrade the OS. Hardware Qdos
users with HDD would have to migrate to new hardware - or stagnate.
2) We could use a method that could be added on, eg
a) utility Things
b) trap #3 (extendible even in Qdos)
c) a new "System Services" trap, eg adopt trap #[5..15]
d) Other ;)
I could go on, but that would bore eveyone sick. Ive been pondering this
question ever since I realised the full horror of the current implementation
(back in '87 or therabouts).
Per
_______________________________________________
QL-Users Mailing List
http://www.q-v-d.demon.co.uk/smsqe.htm
