[Coco] NitrOS9 and 6309 code

Stephen Fischer SFischer1 at Mindspring.com
Sun Jun 9 23:51:56 EDT 2019


As I turned wildcards on and relied that in the scripts I wrote, we are 
in agreement. I was in a world of one using the Delphi / CI$ programs 
and system patches. NitrOS-9 appears to be done by people that did not 
follow what was posted there.

The other options I would need to reread the documentation to say 
anything except, the option to log commands to a file would need to be 
implemented in a different way. But I used that only once but it needed 
to be turned on when starting the shell to be useful. While I used that 
feature in my Super Computer scripts it was too slow to be used on OS-9 
all the time.

There are many other changes beyond the shell that broke my work. Many 
are changes for the good.

I consider many things in Shellplus to be implemented in awkward ways, 
but that was perhaps due to the ipatch method of distributing changes.

Still I still would like to look at the original source code to see if 
there was a simple way to pass parameters to scripts converted using the 
datamod program. That was the biggest shortcoming which has stuck in my 
mind.

I would like someone besides Allen to check the size of the NitrOS-9 
shell in the latest version and to compare with the shellplus 2.2a that 
was also distributed. If I was doing anything useful I would change away 
from the 50% sized shell for sure.

SHF

On 6/9/2019 8:14 PM, Jeff Teunissen wrote:
> On Sun, Jun 9, 2019 at 9:19 PM Stephen Fischer <SFischer1 at mindspring.com> wrote:
>>
>> With a Shell that is ~ the size of Shellplus 2.1A now what does not work
>> may be much less, it would take somebody trying each feature in the
>> documentation to find out what.
>>
>> But the fact IMHO remains that the scripts to configure Shellplus 2.2a
>> have not been supplied as the addresses have changed. But then perhaps
>> that also has been fixed as I have no way of checking as I am busy
>> watching fish in Japan right now.
> 
> It was probably a terrible idea for the Shell to be configurable in some ways.
> 
> For example, wildcards -- they should be on for everyone, or off for
> everyone, otherwise scripts can't be portable. So far, all of the
> programs I have seen assume that wildcards are off by default, and
> require a leading ":" on a given command line to enable them for that
> command. I see, in makefiles, ": del foo cstr.*" for 'clean' targets,
> which wouldn't work if someone had used the wild.on.scr patch, because
> if you enable wildcards by default, a leading ":" then _disables_
> them. A similar problem exists for variable expansion -- if you turn
> it off (it's on by default), then scripts break.
> 
> Other features, like logging by default, are totally innocuous.
> 


More information about the Coco mailing list