Mail Archives: djgpp/2000/05/08/20:11:34
Hello.
Damian Yerrick wrote:
>
> On Mon, 8 May 2000 11:11:35 +0300 (IDT), Eli Zaretskii
> <eliz AT is DOT elta DOT co DOT il> wrote:
> >
> >Sorry, I don't really see how did you arrive at that conclusion.
> >
> >The addition of LFN functionality when Windows starts is analogous to
> >the additional functionality in, say, keyboard remapping that you get
> >on Unix when you start X. I fail to see how does this reveal anything
> >(good or bad) about the design.
>
> That LFN support is tied to having Windows resident in RAM.
Really that's got nothing to do with the design of LFNs themselves or the
API itself. Microsoft chose not to _implement_ the LFN API under DOS.
That's a bad decision IMHO, but then they don't really care about DOS
anymore.
It could be argued that the method used to store LFNs in the FAT is ugly,
but it works (most of the time).
> Product tying is thought to be bad and to result in bloatware;
I'm not sure that this follows. Surely if you tie two products together,
you can reduce bloat, since they can be more tightly coupled? I think the
runaway development process might have more to do with bloat.
> hence the *n?x philosophy of "one tool for one job".
I think we're all aware of how easy it is to combine multiple Unix-ish
tools to get the job done. ;) After all, we're using lots of GNU tools
here!
Bye,
--
Richard Dawe
richdawe AT bigfoot DOT com ICQ 47595498 http://www.bigfoot.com/~richdawe/
- Raw text -