> Subject:       Re: Hypermedia and data analysis: Bruce Mason's dream system
> From:          [log in to unmask] (Michael D.Fischer)

> Bruce Mason, <[log in to unmask]> wrote of his dream system: (at end)
> 
> Birrell Walsh replied:
> 
> This sounds suspiciously like HyperCard for the Macintosh, or its
> supersets (SuperCard et al.).  They are not Qual Software, but it would
> be very very easy to do most of what the above spec calls for with these
> products.
> 
> 
> However Birrell highlights the real issue. Hypercard is not a hypertext or
> hypermedia application, it is a tool building environment that will indeed
> do most of the things set out in conjunction with other tools. Research
> using computers is better done using a more modular approach where a number
> of different, rather simple programs, can be used to support each other
> rather than trying to define a single program of the one size fits all
> variety.
> 
> This is the approach generally taken with unix and Macintosh OSs. It is
> theoretically possible within the Windows 95/8 and NT OSs, though
> apparently this is not a part of the user culture of these OS.
Granted.  I have programmed in PROLOG and used a variety of OSs and 
of them all Windows is my least favourite environment, however 
beggars can't be choosers.
> 
> Hypercard together with other modular tools can do a lot. However, lately
> we have been taking a more www based approach (beginning to set up projects
> with continuous dissemination. A mixed bag, but look around
> http://lucy.ukc.ac.uk). The www approach has the additional advantage of
> being collaborative accross different platforms, which hypercard cannot
> easily be.
Using the www for dissemination is certainly an option but it doesn't 
provide much help when trying to explore your own data.  This also 
does lead onto another siginificant option: how many specialized 
skills should a social scientist be prepared to learn to work within 
a computer environment?  Secondly how many skills should a "reader" 
of a hypertext-based presentation of research.

Ideally I would like a  hypertextsystem that is not much more complex 
than ATLAS/ti into which a developer can import data and then analyse 
it.  But also which can than "package" the hypertext into a 
distributable form which can be read by others without access to the 
hypertext program.  Again ATLAS/ti does export to HTML but the output 
is pretty bizarre and unreadable unless you understand ATLAS/ti.  One 
could try hacking the output around but then as far as I can tell you 
have to put all the links back in and so on.

The process outlined by Michael below seems to me to be the end 
result of years of research and practice and would appear daunting to 
99% of all potential social science users.  So I suppose finally any 
"dream system" is also relatively easy to use and doesn't have too 
steep a learning curve.  The project I'm involved in is assessing the 
potential use of hypermedia in qualitative research and one of the 
aspects of that concerns the demands placed on a potential user.  The 
more computer literate the user needs to be the narrower the list of 
potential users is.

Granted that Michael's approach allows a user to pick and choose 
tools for the choice rather than being constrained by a 
single program, I suspect that it requires a wide-enough 
range of skills and experience that a vanishingly small 
number of researchers would be able to undertake this kind of 
research.  

---Bruce
> I do substantially what Bruce is talking about on a Mac using the following
> programs (which are also available for windows):
> 
> 
> Visual Page (Symantec) -- html WISIWYG editor with excellent drag and drop
> capabilities (available for windows and mac)
> Movie Star Maker (Intelligence at Large) -- an excellent utility for
> constructing Quicktime movies from a range of sources - makes video
> annotation on a frame to segment basis a snap. (available for windows and
> mac)
> MoviePlayer (Apple) free ware with sufficient power to make clips from
> larger media bases that can then be dragged and dropped into Visual Page
> (available for windows and mac)
> Site Mill (Adobe) for displaying and modifying the entire hyptertext base
> (available for windows and mac)
> Project X (apple) another way to view a hypertext base
> httpd - a www server (available for windows and mac)
> a range of cgi programmes for searching (available for windows and mac)
> a range of freeware and shareware tools.
> 
> There are a mostly equivalent range of things (with even more power)
> I run on a sun ultra 10 workstation, but which are generally
> available for Unix with Xwindows (including Linux (freeware), which has the
> power to transform a slow outdated windows machine (P200 or less, or even a
> 486 or 386) into to a powerful piece of hardware).
> 
> XEmacs - a very powerful editor with integrate www browser and  integrated
> programming language (elisp).
> Framemaker - cross platform editor with good indexing and hypertext
> abilities, images video etc.
> xanim - a program for displaying with great control quicktime, avi, mpeg
> and other video and sound formats,
> xv a power image viewing and manipulation package
> apache - a powerful www server that you can run as a private program.
> hotjava or netscape - good browsers for unix
> 
> + hundreds of relevant standard unix commands (grep, sgrep) and hundreds of
> other tools for working with texts, and comprehenive suites of tools such
> as the Humanities toolkit, the Scholars workbench, Unixstat.
> 
Bruce Mason, SOCAS, UWC.
Research Associate: Hypermedia and Ethnography
Email me <[log in to unmask]> for more information about this project
or check the following url:
http://www.cf.ac.uk/uwcc/socas/research/hyper/


%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%