Re: no users destructors if stack 0 deleted

[Joel E. Denny]

On Tue, 6 Dec 2005, Paul Eggert wrote:

> "Joel E. Denny" <address@hidden> writes:
>
> > The combined diff for the two patches appears below.
>
> That patch was corrupted, unfortunately.  I guess your Pine client
> munged it?  There were some leading spaces in it.

This is my fault not pine's. Personally, I find tabs maddening and am in 
the habit of not using them, but I've been trying to adhere to the 
conventions in "glr.c" anyway.  I should've used s/// to find these 
invisible inconsistencies. Sorry about that.

> It's helpful to bundle all the related items in to the same patch,
> along with a ChangeLog entry for it.  Something like the following,
> which I just now installed.  (It reformats the comments a bit, to use
> the usual GNU style.)

I'll do this in the future.  Thanks for installing it.

> Thanks again for the fixes, and for persevering to get them installed.
>
> 2005-12-06  "Joel E. Denny" <address@hidden>
>
>         * data/glr.c (yyprocessOneStack): Synchronize the shift for all
>         stacks, and iterate another stack in order to call user
>         destructors.

There are significant changes in yyparse() in addition to 
yyprocessOneStack().  Is this important?

Joel