Date: Wed, 20 Jan 1999 09:50:58 +0200 (WET)
From: Andris Pavenis <pavenis AT ieva05 DOT lanet DOT lv>
To: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
cc: "Mark E." <snowball3 AT usa DOT net>, djgpp-workers AT delorie DOT com
Subject: Re: autoconf 2.13 test version available
In-Reply-To: <Pine.SUN.3.91.990120091019.2569F-100000@is>
Message-ID: <Pine.A41.4.05.9901200943160.91566-100000@ieva05.lanet.lv>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Reply-To: djgpp-workers AT delorie DOT com



On Wed, 20 Jan 1999, Eli Zaretskii wrote:

> 
> On Tue, 19 Jan 1999, Andris Pavenis wrote:
> 
> > > After taking another look at the autoconf Makefile, it appears that *.inf* 
> > > is good enough.The info files are already provided prebuilt and unsplit 
> > > in the source. And if you delete them, they are rebuilt using the --no-
> > > split option. To build the split files for the binary package, I had to run 
> > > makeinfo outside Windows to generate the .i? files.
> > 
> > Perhaps it would be nice to be able to force this behaviour of makeinfo also 
> > with LFN=Y (maybe with some environment variable)
> 
> Which behavior do you want to force?  If you want it to create short
> 8+3 names like *.i? instead of the long *.info-? ones, just set LFN=n
> temporarily.  After all, LFN *is* an environment variable.

If I'll set LFN=N makeinfo will not found *.texi (unless I'll modify
WIndows registry as mentioned in FAQ). I would like both:
	- possibility to use LFN
	- and generate file .inf, .i1, .i2, ...
simultanously.
 
> > in automake works Ok with LFN=Y. There are still some  failures with LFN=N.
> > I uploaded both sources and binaries of automake-1.4 and they are already
> > in v2gnu.
> 
> FWIW, I think it's really bad that a port only works reliably under
> LFN.  Sadly, several ports done lately are showing a consistent trend
> to be not tested and not reliable enough without LFN.  I wish people
> could make some additional effort to make the ports more clean in this
> aspect, before they upload the distribution.
> 
> It is not right to assume that non-LFN environments are insignificant
> and can be ignored.
> 

I have not it tested very seriously with LFN=N. Maybe I have something
else broken with LFN=N as I didn't use NameNumericTail=0 and only
reinstalled autoconf-2.13 from my binary archive with LFN=N under plain
DOS

Andris