Xemacs21 problems with ESS 5.9.1 ?? [FWD from ESS-debian]

[Martin Maechler]

>>>>> Dirk Eddelbuettel <edd at debian.org>
>>>>>     on Fri, 4 Jun 2010 10:55:42 -0500 writes:

    > On 4 June 2010 at 17:45, Martin Maechler wrote:
    > | >>>>> Sebastian P Luque <spluque at gmail.com>
    > | >>>>>     on Fri, 04 Jun 2010 10:08:49 -0500 writes:
    > | 
    > |     > On Fri, 4 Jun 2010 12:17:42 +0200,
    > |     > Martin Maechler <maechler at stat.math.ethz.ch> wrote:
    > | 
    > |     >>>>>>> Sebastian P Luque <spluque at gmail.com> on Thu, 03 Jun 2010
    > |     >>>>>>> 23:40:47 -0500 writes:
    > | 
    > |     >>> Hi, I've committed a few changes to ess-rutils.el that Debian sid
    > |     >>> XEmacs can compile with the same inocuous warnings that GNU Emacs
    > |     >>> produces.
    > | 
    > |     >>> Incidentally, I also found that XEmacs doesn't know `delete-dups'
    > |     >>> neither, which I used in ess-help.el to find '\alias{}'ed objects, so
    > |     >>> I replaced that with `ess-uniq-list'.  What a trip into "Emacsen
    > |     >>> interoperability lane" this was!
    > | 
    > |     >> Thanks a lot, Seb!  Yes, it's a painful exercise sometimes.
    > | 
    > |     >> Thanks for having done it!
    > | 
    > |     > And I thank Dirk for motivating me to do it.  He's wondering whether to
    > |     > wait for the next release to produce a new Debian package with these
    > |     > changes, given that 5.10 is already slated in the source.
    > | 
    > | (and I see he has decided already *not* to wait.)
    > | 
    > | Dirk, have you confirmed that the changes do "close the bug"?

    > Yup. Seb was able to replicate the bug, fix it, and validate that it was no
    > longer tickled.  That is certainly good enough for a Debian point release
    > upload 5.9.1-2 which is meant for _bug fixes_.  Release early, release often.

    > | If yes, I would indeed have planned to release  5.10  ASAP;
    > | The problem will also affect non-debian users of old Xemacsen.

    > The Debian builds, to my knowledge, do not involve installations of xemacs so
    > waiting will not get us additional datapoints.  Plus this is a non-binary
    > package so no additional builds will happen.

    > You can wait for Vincent, Michael and Johannes to update the CRAN backport
    > for the deb to trickle into your Ubuntu box though.  Up to you.

No, no!  I'm very glad Seb has done the work 
(of installing and using Xemacs ..  Xemacs almost always manages
 to mess up my ~/.emacs, so I'm happy for each time I do *not*
 have to ...)

But inspite of the good reasons for making 5.9.1-2 (deb),
I still think it would be fine to release 5.10,
.. if only just to revert to the "normal state" of having an  
<n>.<m> version number.

Martin