Path: ...!eternal-september.org!feeder3.eternal-september.org!news.eternal-september.org!eternal-september.org!.POSTED!not-for-mail
From: =?UTF-8?Q?Arne_Vajh=C3=B8j?= <arne@vajhoej.dk>
Newsgroups: comp.os.vms
Subject: Re: Local Versus Global Command Options
Date: Fri, 21 Feb 2025 19:45:05 -0500
Organization: A noiseless patient Spider
Lines: 35
Message-ID: <vpb6mh$3krmq$1@dont-email.me>
References: <vonh1j$3fukn$1@dont-email.me> <vonr8r$3ghb6$1@dont-email.me>
 <voo40m$3jej5$1@dont-email.me> <vookpo$3mk0q$1@dont-email.me>
 <67afe79c$0$719$14726298@news.sunsite.dk> <voqpjj$5mpo$1@dont-email.me>
 <votq85$ptnb$1@dont-email.me> <vp015a$19ajm$1@dont-email.me>
 <vp0qua$1dbh3$1@dont-email.me> <67b54873$0$709$14726298@news.sunsite.dk>
 <vp3hlk$n28$1@reader2.panix.com> <vp3mel$1m224$1@dont-email.me>
 <67b647c3$0$712$14726298@news.sunsite.dk> <vp5jsd$j2u$1@reader2.panix.com>
 <vpas10$3irvn$1@dont-email.me> <vpau3j$3j7d2$1@dont-email.me>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Injection-Date: Sat, 22 Feb 2025 01:45:05 +0100 (CET)
Injection-Info: dont-email.me; posting-host="a8b2652d2544001fa06865b957e7aadc";
	logging-data="3829466"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX18P9Wg3ynTkeHCWg/Q48prA5BFvmE86yHU="
User-Agent: Mozilla Thunderbird
Cancel-Lock: sha1:d2pySFJbMSYZMKDQKAeuchQHt9Q=
In-Reply-To: <vpau3j$3j7d2$1@dont-email.me>
Content-Language: en-US
Bytes: 2843

On 2/21/2025 5:18 PM, Lawrence D'Oliveiro wrote:
> On Fri, 21 Feb 2025 16:42:56 -0500, Stephen Hoffman wrote:
>> Is it feasible for DCL?  Sure. Though it'll likely involve passing some
>> of the syntax to lib$table_parse or ilk for parsing, as do a few DCL
>> commands I've encountered (or have written) over the years.
> 
> What happens when you bypass DCL? So one program directly spawns another
> and gives it a command to execute. Can the receiving program behave the
> same in that case as if it were invoked from a shell?

This is where the difference in *nix and VMS process model
get into the picture.

A "normal" process on VMS got the CLI (DCL) in P1 and
the running image in P0 and/or P2.

Most ways to activate another image (lib$spawn, C system,
C exec* etc.) end up creating a process with CLI mapped.

It is possible to start a process without CLI (sys$creprc
or run process - with an image that is not loginout.exe).

In that case I don't think there are a concept of arguments.
I don't see a way to pass arguments. Arguments are something
CLI provide to image, so no CLI means no arguments.

Such processes exist, but I would consider them relative
rare. There are some limitations on such processes - some
LIB$ functions will return LIB$_NOCLI.

Arne