Mail Archives: opendos/1997/04/03/16:32:22
On 3 Apr 97 at 13:10, James Lefavour wrote:
> in either case, when searching for topical help (ie. help diskcopy,
> for example) search first for compressed default file, then for
> uncompressed?
Hmmm - I wouldn't support uncompressed; the idea of the compression
is not so much the actual /compression/ as the binding of loads
of little files (as HTML documents tend to be) into single
logical files (like, "This is the OpenDOS Help file" rather
than "These group of help files are for things in OpenDOS, these
ones for...").
Now, for specifying help files... IMHO we should have
single large help volumes, ie:
\Resources\Help\OpenDOS.ZIP
\Resources\Help\My Application.Zip
(Oooh, LFNs!)
Now, normally, these help files are accessed from the
applications, which name themselves to the help applet and
also name a place inside the help file, eg:
HELP "OpenDOS" DISKTOOL.HTM#DiskCopy
The user could do just that manually, but really we
want a simple gateway to OpenDOS help at the command
line (ie: "HELP DiskCopy"), so we should either
call the help engine program something different
("HelpViewer.exe") and provide a HELP.exe for
OpenDOS that converts short names to full locations using
a lookup table, or - my favourite idea - put
an INDEX.TXT as well as an INDEX.HTM in every
help archive, with just such a lookup table in.
Invoking the help viewer with just one argument
prompts a search of it's own personal index cache
("\Resources\Help\INDEX.TXT"). If the key is not
found in there, it scans the help DIR for ZIP files
that aren't mentioned in it's cache, or have more recent
modification dates than those in the cache, in which case
it searches them in turn, adding their indexes to
the cache index, until it finds the help key wanted.
Regards,
ABW
--
Alaric B. Williams (alaric AT abwillms DOT demon DOT co DOT uk)
---<## OpenDOS FAQ ##>---
Plain HTML: http://www.delorie.com/opendos/faq/
- Raw text -