[comp.sys.amiga] CLtd TimeSaver

mp1u+@andrew.cmu.edu (Michael Portuesi) (10/22/87)

blgardne@esunix.UUCP (Blaine Gardner) asked for a review of TimeSaver
sometime back.  Mine just arrived in the mail, so here's a review
based upon my first impressions with the product, and some perusing of
the instruction manual:

> Amiga World's review of Cltd's Timesaver clock-calendar/macro gizmo
> was quite favorable, but I'd like to get a second (third...) opinion on
> it. 
> 
> The method of setting the date (send a ctrl-D, followed by DATE, and
> "Execute startup-sequence" when the keyboard is reset) seems like kludge,
> and I'm wondering how it works with things like VD0: that insist on
> being the first thing in the startup-sequence.

Timesaver sends a ^D when the keyboard is reset.  It then waits for n
seconds (n is factory-set to 11, you can change it to anything from 0
to 31.5 seconds in steps of .5 seconds), then sends the DATE command.
The ^D is intercepted by AmigaDOS at first opportunity and aborts the
Startup-Sequence.  After sending the DATE string, it then sends
"execute s/startup-sequence" (note that there is a slash in
"s/startup-sequence" and NOT a colon).  I must note that the
factory-set value for the delay is right on target, and probably will
need only minor tuning for your system. 

TimeSaver has an option to automatically execute a keyboard macro on
bootup in addition to setting the date, but it executes the macro
*after* executing the DATE command.  Of course, the auto-date set and
the auto-macro features can be disabled.

Basically, the date-set feature is a kludge, and it doesn't really
"respect" things like vd0: which must be first.  but the good news is
that TimeSaver's execution of the DATE command doesn't really throw
vd0: out of whack.  I've seen instances of using vd0: where I have
rebooted under a different Workbench, run some programs, rebooted back
under my original Workbench and found vd0: STILL there!!! 

One caveat is that since the AmigaDOS DATE command does not accept an
argument for seconds, the resolution of the time set by TimeSaver is
accurate only to the closest minute.  This shouldn't be a problem for
most applications, but those of you who spend your time listening the
the Bureau of Standards Atomic Clock may not like this "feature."  I
do not know what the time resolution is for any competing units
(A-Time, TIC, MouseTime, etc. etc.)

> 
> Besides VD0: how does it work with Conman? Is Conman completely
> redundant with Timesaver installed? Are there any conflicts with ANY
> programs? What if DATE and EXECUTE don't happen to be on a disk (a copy
> protected game for example), does Timesaver cause the s-s to die?

With Timesaver, the keystrokes you type are saved in its internal
buffer (about 1K).  When you type HELP-uparrow, Timesaver enters edit
mode.  It erases the contents of the current line, then re-types it
with a semicolon (the AmigaDOS comment character) at the front.  You
can edit the line with cursor-left and right, DELETE, and BACKSPACE.
SHIFT combined with cursor-left and right moves 15 characters in the
specified direction.  When you are done editing, you press the RETURN
key from any position on the line, and Timesaver re-types your line
without the semicolon to the CLI, which then executes it.  The reason
why it prints the line with a comment delimiter on front is that
TimeSaver sends the appropriate codes to the console window to
perform visual editing, which are not part of the command you are
typing.  Hence, the current line cannot be executed by AmigaDOS.  If
you have not typed anything on the current line, it enters a
command-history mode where you can step through the last 1K worth of
commands (which are typed on the current line with a comment
character in front).  You can then edit and execute them.

Since keyboard repeat is handled by the Amiga proper and not the
keyboard, and since keyboard repeat rate is set by Preferences,
keystrokes generated due to key repeat are *not* saved by TimeSaver.
This also means there is no key repeat when in the TimeSaver editing
mode.  Of course, the 15-character move option does alleviate the
problem somewhat for cursor movement.  In addition, TimeSaver is not
able to process command lines that contain unprintable ASCII
characters, nor are macro functions available in edit mode.

TimeSaver doesn't make ConMan totally redundant -- just two of
ConMan's most productive features.  You don't get things like
automatic window iconify/expansion, screen movement, or options to
control window parameters from the command line.  I prefer ConMan's
editing functions to TimeSaver's, but TimeSaver is acceptable and
requires no addtional memory. 

I should point out one fundamental difference between command history
under ConMan and TimeSaver.  Since TimeSaver grabs keys from the
keyboard to save for command history, there is *one* history for all
windows.  For example, the characters I am typing into VT100 right now
will show up in my CLI as a command when I enter TimeSaver edit mode.
This can be useful when juggling multiple CLI windows, but is not
useful when using the CLI with an application that involves heavy
keyboard interaction (like a terminal program or an editor).

The only incompatibility between Timesaver and other software is if
the other software uses a key sequence that TimeSaver has
pre-programmed or that you have programmed yourself as a macro.  For
example, TimeSaver uses HELP + F5 to type the DATE dd-mmm-yy hh:mm
string.  If your software depends on that key sequence, it won't see
it.  If you define ALT-E to run Emacs, then software depending on
ALT-E will be in trouble.  The TimeSaver system functions and
pre-programmed macros all use the HELP key as a prefix modifier; that
shouldn't interfere with any of the software I know of.  You can turn
off all macro processing with HELP + RIGHT-SHIFT + 2, or all TimeSaver
functions altogether with HELP + RIGHT-SHIFT + 0.  With all TimeSaver
functions off, only one key is still eaten: HELP + 0, to turn
TimeSaver back on.  If your software can't live with that, you should
probably find new software.  Selected TimeSaver functions can be
enabled or disabled at will. 

If DATE and EXECUTE are not present on a disk, then Startup-Sequence
will not be executed.  If DATE is not present, but EXECUTE is,
Startup-Sequence will get executed, but the time will not be set.  You
get the idea.

If you have a copy-protected program that does not have DATE or EXECUTE,
your best bet is to disable the TimeSaver auto-date feature before
booting that disk, and enabling it afterwards.  Of course, that means
you can't set the time with that program, but I don't think any of the
other clock/calendar thingies on the market can do anything about that
situation either (since you have to modify the Startup-Sequence to
make them work).

I should point out that TimeSaver currently uses only 4K of its 8K ROM
capacity, and that CLtd is planning to offer upgrades to the ROM for a
nominal fee (between $5 and $15).

TimeSaver is powered by a small lithium battery, which the manual
estimates should last a year or so.  TimeSaver contains a power
arbitration chip that draws power from the Amiga when it is powered
up, and from the battery when the machine is off.  So when your Amiga
is in use, TimeSaver draws no power from its battery.  This is a big
win if, like me, you run your Amiga 24 hours a day.

The manual is quite big, and addresses all aspects of TimeSaver
operation, plus some other things too.  For example, it tells you that
you can cut a hole in the end of the cardboard tube TimeSaver is
packed in to turn it into a piggy bank.  There is even a reference to
a piggy bank in the Index.  In another section, it tells you that one
part Downy fabric softener to seven parts water in a spray bottle
makes an excellent anti-static agent.  This little gem carries a
reference in the index as well.  It does contain examples for nearly
everything you would want to do with TimeSaver, a section giving
technical details and theory of operation, and hints for some creative
ways of using TimeSaver.

Disclaimer: I have no affiliation with CLtd except for being a
satisfied (so far) customer.


Michael Portuesi / Carnegie-Mellon University
ARPA/UUCP: mp1u+@andrew.cmu.edu
BITNET: rainwalker@drycas (a uVax-1 run by CMU Computer Club)

"Boys living next door are never what they seem"
		--Bananarama, "Robert DeNiro's Waiting"