PURPOSE OF THE NEWMDSX GROUP

[Tony Coxon]

Several members have asked me about the group's purposes, so here we
go.
1. Current Status: NewMDSX is a successor to the original 1970s MS-DOS
package, MDSX developed at Edinburgh by Tony Coxon /retired &
UEdinburgh, again; Charles Jones/ now UToronto and David Muxworthy/
retired and UEdinburgh) and distributed by the University Program
Library Unit (David Muxworthy is our current survivor) and later by
MIMAS. Then funding and support ceased, and we explored developing
MDSX in a Windows platform with Alan Brier (then Southampton, now
retired), Phil Hawkins (my partner) and David Muxworthy),
subsequently dubbed NewMDSX [ www.newmdsx.com as I expect you know!].
Since then it has been entirely self-funding and non-profit-making
[out of principle, not just necessity!]. All developments and
distribution have been funded by ourselves and sales.
2. NewMDSX is currently under review by the British Journal of Math &
Stat Psychology (due to be published in November), and we are awaiting
this report before making any major decisions about the continuation,
format and content of NewMDSX. Because the "activists" are retired,
they can no longer call upon University resources, and the actual
costs are borne by ourselves. There are a number of changes under
consideration, including:
-- inclusion of new programs (several under consideration)
-- direct collaboration with other cognate packages/programs or
inclusion within NewMDSX (currently operative with PERMAP interactive
MDS [Ron Heady] and text analysis HAMLET [Alan Brier]. Possibly:
Borgatti's ANTHROPAC, Ramsay's MULTISCALE -- neither approached
directly yet.
-- change of control language and expansion of WOMBATS to act as a
general input/output program between NewMDSX and other packages.
Currently it is possible to turn NewMDSX input into PERMAP input, for
instance
-- expansion of graphic capabilities (internal or link to external
programs)
-- inclusion of other related procedures to be either maximally
related to or included in NewMDSX -- Sorting is an obvious example; my
SORT-PAC3 and Joachim Harloff's SORTKIT (see www.methodofscaling.com)
3. Expansion to form a confederation (?) of related soc. sci. packages
(freeware or not-for-profit), to increase collaboration on design,
documentation, distribution etc; to expand services to users,
including discussions like this but also on any other topic felt
appropriate (choice of algorithms, related packages, formation of FAQ
section and direct queries -- the Web 2.0 horizon!). We have no
intention however to collaborate with the biggies like SPSS and do
their work, like providing comprehensible user-friendly documentation
(looked at CATEGORIES recently?).
4. Seeking new markets, so keep afloat and expand : we have never had
much market presence in North America, and yet there are programs like
Smith's MS-DOS package PC-MDS (http://marketing.byu.edu/htmlpages/
pcmds/pcmds.htm)
which IMHO is far inferior to NewMDSX -- costing $495 single copy
and $2000 site -- claiming to be "the world's most widely adopted
specialty package for multidimensional scaling"! Yet is IS widely
adopted in the US. We need publicity and promotion (Frank B?!) there
and in Europe.
5. Build up of a user-communityusing sites like this, where everything
from numerical and algorithmic techniques to proposed changes in
NewMDSX ... and the newbies corner?

Feedback please ....
Tony