What made Medley a great developer experience?

45 views
Skip to first unread message

Paolo Amoroso

unread,
Jan 30, 2023, 9:18:45 AM1/30/23
to Medley Interlisp Users/Interest
What made Medley a great developer experience?

Advanced tools like Masterscope, structure editors, and the break package are widely known and often credited with the great developer experience. But are there other lesser known tools, Lisp constructs, or system features that contributed to the experience? In what ways did Medley increase productivity and help manage large code bases and complex systems? Did it support coding for production as well as exploratory programming?

I'm exploring Medley now and never used it back in the day. So I was curious about a similar question posted to LinkedIn, but the post was deleted.


Paolo
--

Paolo Amoroso

unread,
Jan 31, 2023, 3:16:22 PM1/31/23
to Medley Interlisp Users/Interest
Any feedback or stories on what made Medley a great developer experience?

Larry Masinter

unread,
Jan 31, 2023, 3:53:15 PM1/31/23
to Paolo Amoroso, Medley Interlisp Users/Interest
I would look at the contemporary reviews
or even Steve Kailer's books as they were written with early Interlisp in mind.

Some of the enjoyment (for Smalltalk as well as Interlisp) is that the users who got one had been used to a teletype interface to a time-sharing system. Or worse (batch-processing Lisp and debugging with hex core dumps)
The idea of a mouse and keyboard and GUI were novel -- we would start demos by pointing out the box on the desk with a wire from it that 
I used to periodically go on sales calls... some customers (managers) reacted to our claim of "making it easier for programmers to change the software" with horror -- they didn't WANT it to be easy.)

The main focus of Interlisp as an IDE was to be the "Programmer's Assistant" -- with several individual features under that heading. At the time, that was a novel idea -- that the computer would spend valuable time doing things to help the programmer.


On Tue, Jan 31, 2023 at 12:16 PM Paolo Amoroso <paolo....@gmail.com> wrote:
Any feedback or stories on what made Medley a great developer experience?

--
https://Interlisp.org for more details
---
You received this message because you are subscribed to the Google Groups "Medley Interlisp Users/Interest" group.
To unsubscribe from this group and stop receiving emails from it, send an email to interlisp+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/interlisp/2c80dbcf-295c-4041-a474-a5e705f93d92n%40googlegroups.com.


--

Michele Denber

unread,
Jan 31, 2023, 5:13:44 PM1/31/23
to inte...@googlegroups.com
On 1/31/2023 3:16 PM, Paolo Amoroso wrote:
> Any feedback or stories on what made Medley a great developer experience?
I think Beau Sheil's "Power Tools for Programmers" article describes it
really well.

For me, it was the ease with which you could do graphics.  You could put
up a new window with just one command (CREATEW) and see the result
instantly, something that took a dozen lines of code and a compile cycle
in Motif.

            - Michele

Tim Daly

unread,
Jan 31, 2023, 7:34:17 PM1/31/23
to Larry Masinter, Paolo Amoroso, Medley Interlisp Users/Interest
What makes systems like Melody "a great developer experience"
really has to do with the kind of people it attracts and are attracted
to it. The very fact that it can be modified is what makes it
"interesting". Lisp systems, in general, attract people who find them
"interesting" rather than "useful because..."

Every time a research project tries to do something useful it fails.

Research requires a certain personality. You have to be interested in
things without the "because". Researchers will "make up reasons" for
people who need them. You know, I'm researching self-modifying,
self-assembling robots ... but there is no real "because...". Sure, I'm
talking about robots on Mars but that's just for show. I'm researching
transcendental numbers ... but there is no real "because...".  Sure, I'm
talking about quantum-secure encryption but that's just for show. I'm
researching dependent types ... "because" I can merge proofs and
computer algebra. If I had to guess O'Neill was interested in a lot of
useless things but he talked about space cylinders "because..."

The thing I loved about IBM Research was that there were a few
people who didn't need a real "because...". I did rewriteable paper
which forced me to learn molecular chemistry and its reactions under
the influence of lasers. I worked in computer algebra which forced me
to learn galois theory and category theory. The fact that it is useful for
a computer algebra system is interesting but the computer algebra
system itself isn't.

I guess that's the definition of Research for me. "Is it interesting?"
"Because..." is for people who aren't researchers. I rarely even
use the word. The researchers I know always struggle to "justify
their work", usually to get funding.

If I ran a research lab (wow, THERE's a stretch) I'd fire the people
who needed a "because...". "Is it interesting" is the only criteria.

Ask any manager. I'm the most useless person they know.
In any other time in history I'd probably have starved.

What makes Medley a great user experience (similarly with Smalltalk)
is that it allows you to do "interesting" things. It is "without form".
I suspect people don't use it "for their real job because,,," is exactly this
ability to make "interesting things" rather than "real-world things".

Your motivations may vary.

Tim

Paolo Amoroso

unread,
Feb 1, 2023, 12:30:33 PM2/1/23
to Medley Interlisp Users/Interest
On Tuesday, January 31, 2023 at 11:13:44 PM UTC+1 Michele Denber wrote:
I think Beau Sheil's "Power Tools for Programmers" article describes it
really well.

I'd love to read the article but the ScienceDirect checkout fee is, well, unfriendly: $31.50 for 48 hours access. I wouldn't mind if it was a one-time purchase fee.

Is anyone in touch with Beau Sheil? Is he allowed to share the article?

Ron Kaplan

unread,
Feb 1, 2023, 2:10:00 PM2/1/23
to Paolo Amoroso, Medley Interlisp Users/Interest
I looked for it in my files, didn’t find it (I only found papers he had written when we were grad students together).

However, I was able to download it from Science Direct using my Stanford credentials.

This probably shouldn't be posted or distributed, but here it is.  Looks like a scan, so a little fuzzy.

Sheil 1983 Power tools.pdf

Larry Masinter

unread,
Feb 1, 2023, 3:42:24 PM2/1/23
to Ron Kaplan, Paolo Amoroso, Medley Interlisp Users/Interest, Evelyn, Interlisp core
The Interlisp Zotero library
has an entry for this article, with a pointer to the portal where you can pay.

I'm not sure if Zotero lets you store private copies.

There is a youtube video (with terrible audio) in which Beau gives a demo of Interlisp features
as a sales pitch

I thought it might be interesting to revive that demo or something similar. 

I think the discussion (on the Interlisp mailing list) of science / research vs. engineering / product development is interesting.

Interlisp was targeting research scientists doing "experimental programming"
Smalltalk had a different focus ("Learning Research Group")
Mesa / Cedar another.

I don't agree with many of Tim's assertions about the nature of research vs development in software but for a different message




--
https://Interlisp.org for more details
---
You received this message because you are subscribed to the Google Groups "Medley Interlisp Users/Interest" group.
To unsubscribe from this group and stop receiving emails from it, send an email to interlisp+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/interlisp/583713DC-6A7B-4F30-8538-A0A10A825F52%40post.harvard.edu.

P.S. The last I interacted with Beau, about 20 years ago, he was in Hawaii.  be...@tropicblue.net

--
https://Interlisp.org for more details
---
You received this message because you are subscribed to the Google Groups "Medley Interlisp Users/Interest" group.
To unsubscribe from this group and stop receiving emails from it, send an email to interlisp+...@googlegroups.com.

--
https://Interlisp.org for more details
---
You received this message because you are subscribed to the Google Groups "Medley Interlisp Users/Interest" group.
To unsubscribe from this group and stop receiving emails from it, send an email to interlisp+...@googlegroups.com.
Reply all
Reply to author
Forward
0 new messages