[Info-vax] DCL, was: Re: Microkernel

[John E. Malmberg]

On 8/17/2012 6:47 AM, VAXman- @SendSpamHere.ORG wrote:
> In article <k0jvkr$r1q$1 at dont-email.me>,
 > Stephen Hoffman <seaohveh at hoffmanlabs.invalid> writes:
>>
>> I'm not particularly interested in extending DCL and enhancing the DCL
>> buffer, myself.  I'll leave that to other folks.  To me, a different
>> shell would be more interesting.
>
> I'd toyed with the idea of taking the GNV bash and making it a proper
> CLI but I've been busy with other projects that provide slightly more
> income than freeware. ;)

You do not want to consider the GNV Bash 1.14.8 for such a project.

1. It is out of date.  GNU Bash is at 4.2.37.

2. The GNV Bash 1.14.7 port, even with my latest fixes still has serious 
problems.  Some of which I discovered while working on 4.2.37.

3. There is a very gnarly hack in it to deal with it not having a fork().

4. Many of the existing bugs in the GNV bash 1.14.8 were inflicted by 
the inexperience or confusion on the part of the people making the port.
I recently did a compare of it with the official GNU Bash 1.14.7 (There 
was no GNU Bash 1.14.7 release).  It looks like at least 10% of the VMS 
specific changes were made to code sections that where never compiled in 
on VMS, including several modules that were never used in the build script.

I have not had time recently to finish the port of Bash 4.2.37 to VMS, 
but I do have it building and what I have done is recorded at the GNV 
site in the gnv-develop archives, and can be pulled from CVS.

https://sourceforge.net/mailarchive/forum.php?forum_name=gnv-develop

There are two major blocking bugs in the port.

"echo foo > file" is broken and always produces an empty file.

"echo foo | /bin/less" is broken and goes into a continuous I/O loop in 
the parent process where it parent is reading "foo" from the output pipe 
instead of the child process, and then immediately writing "foo" to back 
into the same pipe for the child to read.

These are a bit related to the gnarly hacks needed to get around not 
having a fork(), and having to implement fake child logic.  Somewhere I 
missed something.

Asside from those bugs, I think I have most of the other VMS specific 
changes in it.

Regards,
-John
wb8tyw at qsl.network
Personal Opinion Only