Mail Archives: djgpp/2009/12/02/07:45:12
"Rugxulo" <rugxulo AT gmail DOT com> wrote in message
news:51b83e93-a4cc-49b3-b177-b4bd51b6b8b1 AT l13g2000yqb DOT googlegroups DOT com...
> > Of course, the Henrik Haftmann LFN driver, or Jason Hood variants,
should
> > probably be used instead of Bill Currie's lfn.zip.
>
>... (To be honest, I almost like StarLFN's
> idea better as it doesn't directly access the hard drive, uses
> LONGNAME.DAT, hence should be patent free, heh. But it has some very
> small incompatibilities with various DOSes by default. I never pushed
> myself to mess with it too much.)
>
Prior to HH's LFN driver, I was partial to Chris Jones' LFNDOS. I did a bit
of work updating it, fixing bugs, and I merged in some patches he posted
elsewhere.
I had also considered creating a hack-job "cut-n-paste" LFN TSR from Odi's
(Ortwin Glueck) LFN tools, since they seem to work _really_ well with LFN's.
But, I didn't want to track down MSVC 1.52c (or whatever...) and attempt to
then port to DJGPP. I haven't looked at his code, but I still suspect that
either he or the C compiler was using some additional undocumented
features... They are very fast with LFN's - i.e., no "bog" down with large
amount of LFN data. His tools seems to do things behinds the scenes with
LFN's without accessing the disk. (a ramdisk? accessing internal DOS
structures? undocumented DOS calls? ...) I've found LCOPY to be useful.
LREN is on occasion also. The 179 version is on Sourceforge. I have a few
of the older versions too.
Rod Pemberton
- Raw text -