text
stringlengths
0
99.6k
The board is currently operating at 300, 1200 and 2400 baud.
We will also try to post a current version of ARC on
COMPUSERVE, (GO CBMPRG and check the HIGH LEVEL UTILITY
section with keyword ARC). You can also check the Commodore
SIGs of DELPHI and GEnie which should also have current
copies of ARC. Note: Also on CompuServe's GO CBMART in
Data Library 1 -- Help/Data Library Tools.
If you are patient, you can also just wait until it is
available on a local bulletin board in your area.
When updating by one of the above methods, sufficient
documentation will be available so that you can figure out
how to use ARC. Complete documentation is only available in
printed form to registered users.
If you wish to post ARC on your bulletin board for
public access, then we would prefer it if you would make the
most recent version available and delete previous versions.
We will be coming out with a number of external utilities
for ARC and will be making the assumption that anyone using
them has a current copy of ARC.
ARC VERSION 2.20 PAGE - 6
When updating ARC, we will try to stick to the
following version numbering scheme:
ARC version X.YZ
If Z changes, then a minor change has been made to ARC,
and it is probably not worth the effort of updating. If Y
changes, then a useful change has been made and it is
probably worth the effort to get an update. If X changes,
then a major change has been made, and you will have to
update if your version of ARC has a lower value of X. X is
the number that you see in the column marked "V" when you
list an archive's directory.
For example, when we added the /w switch to DIR, the
version changed from 1.33 to 1.34. When we added ARC/F, the
version changed from 2.00 to 2.10. When we added three new
data compression techniques, the version changed from 1.34
to 2.00.
Note that if your version of ARC has the same value of
X as the latest version, you will be able to extract files
from archives created by the newer version. If X is lower
than the most recent ARC, then you may come accross some
archives which your version of ARC can't handle.
The actual ARC program itself will be distributed as
"ARCXYZ.H", where XYZ is the version number, and "H" will
indicate what machine the program runs on. At present "H"
can be "4" for a standard C-64 and "8" for a C-64 equipped
with an 80 column adaptor (BI-80). Hopefully "H" will take
on values like "4032", "8096", or "128" in the not to
distant future.
ARC VERSION 2.20 PAGE - 7
Bugs in ARC
As we become aware of them, we will try to repair any
bugs which manifest thier ugly heads. There are a few bugs
which managed to slip by us. Some of them are harmless, and
some are not. There are a few that you should be aware of.
ARC 1.00
Would occasionally make an archive entry one block
longer than it should be. This extra block would confuse ARC
when extracting the archive. If an archive created by ARC
1.00 verifies ok, then it is ok.
ARC 1.xx ( xx not 00 )
These versions had a bug in the pack routine. If a
character repeated n times with n>255, then it would unpack
as n+1 bytes. This would usually result in a checksum error,
but under the right circumstances it would not. (in
particular...in picture files with alot of empty space)
ARCs numbered 2.20 or higher are smart enough to detect this
problem and will un-pack such an archive properly.
There was also another problem with the unpack routines
and the checksum calculation. The last byte of the file
would occasionally be included in the checksum more than
once! This problem was in fact harmless, except for the fact
that it fooled is into thinking that the previous problem