home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
BBS Toolkit
/
BBS Toolkit.iso
/
doors_1
/
mta_vf10.zip
/
Q&A.MTA
< prev
next >
Wrap
Text File
|
1992-05-03
|
5KB
|
110 lines
Some frequently asked questions (important) and their answer:
Q: Why is it so difficult to install MTA, why isn't there an install
program;
A: MTA is very transparent. It is possible to manipulate archives and
related material in as many different ways as there are users. The
result is a program that (could) need a lot of special options in
the configuration file. If an installation program was included,
you still needed the documentation and the installation program
would only be an intelligent editor;
Q: Why can't the configuration-file be compiled so MTA can read it
very quickly.
A: Almost all options in MTA.CTL have to do with the 'outside world'.
In general I dislike programs that report things like the following
(to commonly) line:
'Runtime error 003 at 0F11.AA76'
Because of this, MTA will test all access to this 'outside world'.
Tests extend to the presence of directories, files, memory and so
on. These tests take most of the time. A benchmark on MTA.CTL of
60K (a LOT of options) WITHOUT these tests resulted in a wast of
around 2 seconds (386/16 without disk-cache). The tests result in
a slower thruput but (in general, bugs not included) it will result
in a safe operation;
Q: I am loosing paths inside my archives. The source archives did
contain them but after conversion the target does not contain
them anymore. Why ?
A: You did not include the SubstDrive options in your MTA.CTL. Read
the entry for this option carefully and try to implement this
option (when having problems, contact me). This is only a part of
the answer, the second part will deal about the reason.
In the early days the archive programs did only store path infor-
mation in a relative way. If TEMP\WHY.EXE was included in the
archive, it would unpack as .\TEMP\WHY.EXE (so as a directory
UNDER the current one). At this moment there are two popular
archive programs that (can) use ABSOLUTE directories. An entry
like \TEMP\WHY.EXE would unpack as \TEMP\WHY.EXE, so WHY.EXE
would be in the directory TEMP under the ROOT directory. Even
this could be solved in a normal way but what if someone included
\DOS\COMMAND.COM (or any other file) and under your OWN root you
already have a \DOS (and a COMMAND.COM). This would result in
(dangerous) overwritten files and worse, MTA would create a new
archive with ALL files from \DOS (also those that were already
there) and the called archive-program would delete all these
files also (resulting in a non-bootable machine).
Why not allow for a virtual drive (my RAM-disk) that is really
empty ? Simple, if the source archive contains absolute directories
AND some nested archives (archives in archives) which in turn would
also include absolute directories, the new archive would not look
the same as the original archive, in fact, given a product that
contains:
BLURB.CTL
\BIN\BLURB.EXE
BLURB286.ZIP ---> contains \BIN\BLURB.EXE
BLURB386.ZIP ---> contains \BIN\BLURB.EXE
It would result in:
BLURB.CTL
\BIN\BLURB.EXE (the last of the three)
And the two other BLURB.EXE files (the XT and 286 files) were gone.
To overcome all these problems, MTA will map a substituted drive
over the temporary directory (the temporary directory will become
the root of this new drive) and will extract the files as normal.
With nested-archives, MTA will swap the substituted drive to the
next temporary directory under the previous and again will assign
this a the new root. In this way extraction can not overwrite your
own files AND the target archive will be the same as before.
If you do NOT include the SUBSTDrive support, MTA will not be able
to carry the directories over into the new archive. There are some
related options (SubstOn, SUbstOff) that will even make it possible
to use this mechanisme on networks that don't allow the usage of
SUBST.EXE. In this case you can use the network-version of such a
program (like MAP, PAGEDR and so on).
Q: I get 'Extended test failed' regulary. Why ?
A: Because you don't use SUBSTDrive options in MTA.CTL;
Q: I don't WANT to use SUBSTDrive options, should I switch this option
(ExtendedTesting) off ?
A: NO. The error is correct. Given the archive that contains:
TEST.TST
TEST.NEW
TEMP\TEST.NEW
And given the fact that you don't use SUBSTDrive (so the archivers
are told NOT to create directories), the result would be a TEST.TST,
and a TEST.NEW. TEST.NEW in TEMP\ would overwrite the original file
TEST.NEW and the total of uncompressed bytes on disk would not match
the calculated value which is taken from the archive internally. Also
the first TEST.NEW would be lost and the target archive would now
contain 2 and not 3 files. All in all, you have corrupted the archive
in this way. You should use ExtendedTesting unless VERY specific
problems occur;