Zsh Mailing List Archive
Messages sorted by:
Reverse Date,
Date,
Thread,
Author
Re: optioninfo.so
- X-seq: zsh-workers 10723
- From: Tanaka Akira <akr@xxxxxxxx>
- To: "Andrej Borsenkow" <Andrej.Borsenkow@xxxxxxxxxxxxxx>
- Subject: Re: optioninfo.so
- Date: 13 Apr 2000 16:32:58 +0900
- Cc: <zsh-workers@xxxxxxxxxxxxxx>
- In-reply-to: <000601bfa516$701681d0$21c9ca95@xxxxxxxxxxxxxx> (Andrej Borsenkow's message of "Thu, 13 Apr 2000 11:03:57 +0400")
- Mailing-list: contact zsh-workers-help@xxxxxxxxxxxxxx; run by ezmlm
- References: <hvo66tmh4s6.fsf@xxxxxxxxxxxxxxx> <000601bfa516$701681d0$21c9ca95@xxxxxxxxxxxxxx>
In article <000601bfa516$701681d0$21c9ca95@xxxxxxxxxxxxxx>,
"Andrej Borsenkow" <Andrej.Borsenkow@xxxxxxxxxxxxxx> writes:
> Yes, it is not very portable. Do you suggest to use it outside of Zsh?
> Or in Zsh completion? If it can be made Zsh module - why not. But
> please, no PRELOAD fiddling.
Zsh completion system should work without LD_PRELOAD. But is there a
reason that the system automatically detect LD_PRELOAD and use it only
when it is working? Of couse, if there is no way to detect that, it
should be disabled by default.
Writing a completion function which supports various versions of a
command is very difficult. (See _patch, _telnet, _finger, _make,
etc.) So I think it is useful even if it is not work on every
environment.
Also, it may be useful for writing template of completion functions if
a completion function writer can use LD_PRELOAD. :-)
--
Tanaka Akira
Messages sorted by:
Reverse Date,
Date,
Thread,
Author