Squeak 2.4 on Casio E105 observations

Russell Allen russell.allen at firebirdmedia.com
Fri Aug 6 01:46:09 UTC 1999

>Serg Koren wrote:
>> 1) Hm I'm not familiar with Alan Kay's recognizer.  Any pointers would be
>> appreciated.
>Do Alt-r in a workspace -- the cursor changes to a dot, which you can
>use to draw whatever forms you wish to use to represent letters (it's
>trainable, so you could use the Graffiti forms or any other you would

As far as I recall, the character recogniser built into Squeak maps one
stroke -> one character; unlike Graffiti in which, for example, an 'X' is
drawn with two strokes, and capital letters are indicated with a preceding
upstroke.  However, within those limitations a "Graffiti-like" characterset
is possible.

Some comments by Alan Kay on the recogniser can be found on the swiki at:


What I would really like to see in Squeak is a fullword cursive recogniser
like the one used in the late lamented Newton -  which is now sold as


but writing one of these is considerably beyond my knowledge and abilities :)



>>> Also, has anyone noticed how Dan Ingall's article in the August 1981
>>>byte on
>>> the
>>> Smalltalk VM implementation talked about "small" VM implementations (i.e.
>>> 10's
>>> of K bytes rather than the 100's of K bytes of most Squeak VM's)?  I can't
>>> beleive that C should produce a VM implemenation that's an order of
>>> magnitude
>>> larger than an assembly implementation.  Where's the bloat coming from?
>>>                                    -Dean Swan
>I would guesstimate that the current VM is probably only about a factor
>of two or three larger than a decent assembly language version would be.
>The "bloat" really comes from a lot more primitives being added (all the
>networking and multimedia stuff for example), along with the
>OS-dependent interface code to support it. You can get an idea of what a
>1981 vintage VM would contain by comparing it to the Blue Book VM
>-- Dwight

Russell Allen

russell.allen at firebirdmedia.com


More information about the Squeak-dev mailing list