Re: Logo

[Scott Raney <raney@metacard.com>]

On Fri, 13 Aug 1999, Malonowa wrote:

(snip)

> Ian, can I ask you how you like to program for Linux? GTK, QT
> etc. Maybe we could get together on a project. We need to develop a
> way for programmers out there to quickly and easily develop software
> with a minimum learning curve. Maybe Squeak is tha answer, maybe
> logo, maybe TCL/TK, maybe just specific libraries in C ontop of
> GTK. I suspect there's a range of solutions - development
> proliferates when there are good, simple libraries and a software
> base to begin with.
> 
>  People have been talking about the technical merits of various
> solutions such as hypercard, metacard etc.  But I don't see much
> about what would actually be done in terms of content here. Isn't it
> better to discuss content and then decide how it could be
> implemented? When it comes to the hypercard stuff, pupils tend to
> only need the most basic of tools anyway, little more than a
> glorified DTP package.

To speak of MetaCard only in terms of "content" I think is to miss the
point.  MetaCard is an application development environment.  It's
every bit as capable as Tcl/Tk, for example, but is faster, much
easier to learn, and a much more productive tool to work with because
it has a fully integrated development environment which includes
layout tools (something not even available for Tcl/Tk).  MetaCard is
also less expensive to use than Tcl/Tk, assuming you get TclPro (and
you're very hobbled by lacking things like a debugger if you don't).

As for Logo, I must say that I have a pretty low opinion of it, an
opinion shared by everyone I've ever communicated with that knows both
an xTalk dialect (e.g., the HyperTalk language used in HyperCard,
SuperTalk in SuperCard, or MetaTalk in MetaCard) and Logo.  Logo was
designed by Lisp programmers a long, long time ago, and decisions made
in its design probably made sense back then but are really out of
place in today's computing environment.  For one thing, xTalk is so
much easier to learn than Logo, a key factor when trying to introduce
kids to programming.  And in addition to drawing tools, xTalk has
fully integrated widget support, which is crucial for any non-trivial
programming task.

There is only one justification that I can see for including Logo in a
Linux-for-K12 project, and that is if 100% compatibility with
HyperStudio is a goal.  But this is probably not an achievable goal,
and indeed is probably not even desirable.  Logo is very well hidden
in HyperStudio, and in my recent research of the matter in the local
schools, I didn't find a single teacher who used HyperStudio that had
anything more than name-recognition for HyperLogo.  Of course, none of
them taught it to their students.  And from following the Logo and
HyperStudio mailing lists for the past several years, my considered
opinion is that the number of teachers using HyperLogo in any district
is vanishingly small.

I do believe that HyperStudio compatibility *is* an important goal,
however.  But only at the user-interface level.  That is, teachers and
students should be able to pick a compatible product and know how to
navigate through the dialogs, how to paint things on the screen, and
how to save resulting project.  This is where MetaCard comes in: since
it has the same basic infrastructure as HyperStudio (card/scripting),
and is a full-fledged development enviroment, building a
HyperStudio-like interface for it would be an exceedingly easy thing
to do.  It would take maybe 20% of the effort of building a similar
package in Tcl/Tk, and probably less than 5% of the effort of building
something like this in Java or C + GTK/Qt/whatever.  This kind of
productivity advantage would almost certainly be the difference
between getting a project like this finished and having it die out
after talking about it for a few weeks.

And if anyone is interested in taking on such a project, we'd be more
than happy to help in any way we can.
  Regards,
    Scott

PS: is there a digest version of this list?  If not, may I strongly
encourage you to create one?  This is very easy to do with majordomo,
and if the list owner has any questions about how to do it, I would be
happy to walk them through the process.

> Roman.
> 
> 

********************************************************
Scott Raney  raney@metacard.com  http://www.metacard.com
MetaCard: You know, there's an easier way to do that...