From: Lars Lundback
Subject: Genera pointing at the Way To Go?
Date: 
Message-ID: <367513E5.83A55B51@eralslk.ericsson.se>
Hi,

I have not dared to raise any LispOS/LispVM topics. As you know, all
arguments, and all views possible, can be found in the mailings list.

But quite a few of the postings to this group carry undertones that
something is missing in the current CL environments. (ie. freely
available, users of the commercial systems should make their comments).
We see this expressed in terms of

- Portability
- Ease of use when developing (ie: IDE:s)
- Efficiency vs performance
- Conventional OS trade-offs
- etc

CLIM, Garnet and other kits are frequently recommended. It would seem
that widgetry provides the road to interactive applications, however,
the main difficulty lies in marrying a problem/abstract model to a
viewing model, be it CAD or Lisp Classes.

I find to my delight that Rainer Joswig has put up introductory
information pages on Genera at his Internet site, which is well worth a
visit. Kelly Murray, Rainer, maybe others, have also made suggestions
for the "Apps in Lisp" request to this newsgroup. Finally, we have the
ever-present topic of GUI toolkits; CLIM, Garnet, etc, come to mind.

Take Genera. I quote from Rainer's Genera introduction: 

  - "Self-revealing system: Information about the internal workings
of the system is always available." 

  - "Definition editing: Programmers often work on big programs,
involving thousands of routines defined in hundreds of files.
The logistics of that are staggering on conventional computers.
When you need to modify a function, you have to find it first;
you end up being a bookkeeper as well as a programmer." 

  - "Mouse sensitivity: is an important part of Genera's user
interface."

  - "Help and documentation always present, and context-sensitive."

A basic developer's environment should have at least those four
characteristics, and it can, to some degree, be built on top
of a conventional CL. It requires that an X-kind of Editor and an Y-kind
of Inspector is constructed, both having built-in mouse sensitivity, and
a few graphic drawing primitives.

I can testify that bringing about even the simplest environment of this
nature is immensely satisfying. It also helps in understanding how to
construct an interactive application, and in deciding when a "foreign"
subsystem, eg OpenGL, should be brought into the application.

So, instead of getting stuck in OS trivia, much can be accomplished
through observing eg. Genera functionality, and bringing some of it on
top of a CL implementation, tied to a conventional file system. Then, I
would say yes to the topics list that Rainer had in a recent posting.

Something should be said about sharing experiences and views. A forum
for that can perhaps be provided through ALU, possibly like the Message
Board that Corman Common Lisp has started (
http://www.corman.net/CormanLisp.html and
http://www.dejanews.com/~cormanlisp/)

Perhaps we could then view Genera the way Rolls Royce talks about
horsepower:

  "There is enough."

And an emerging LispOS could make use of the environment I have
indicated. For that environment we could say with the VW Beetle:

  "Acceleration 0-60 mph?   Yes."

/Lars Lundback

From: Axel Schairer
Subject: [Noise] Re: Genera pointing at the Way To Go?
Date: 
Message-ID: <ydj1zm0hxq8.fsf@ws-423.ags.uni-sb.de>
Lars Lundback <·······@eralslk.ericsson.se> writes:
> indicated. For that environment we could say with the VW Beetle:
> 
>   "Acceleration 0-60 mph?   Yes."

That's the old beetle, isn't it?

Axel
From: David B. Lamkins
Subject: Re: [Noise] Re: Genera pointing at the Way To Go?
Date: 
Message-ID: <fwZd2.3954$Oh4.2207188@news.teleport.com>
In article <···············@ws-423.ags.uni-sb.de> , Axel Schairer
<········@dfki.de>  wrote:

>Lars Lundback <·······@eralslk.ericsson.se> writes:
>> indicated. For that environment we could say with the VW Beetle:
>> 
>>   "Acceleration 0-60 mph?   Yes."
>
>That's the old beetle, isn't it?

Downhill, maybe :)

---
David B. Lamkins <http://www.teleport.com/~dlamkins/>
From: Lars Lundback
Subject: Re: Genera pointing at the Way To Go?
Date: 
Message-ID: <36793DFD.39525C68@eralslk.ericsson.se>
Hi guys,

Well, I heard that the new Beetle is promoted along those lines, and I
liked the phrasing. Anyway, one of the companies bought the other, and
it wasn't RR that put up the money ...

Rollses are marvellous, no question about that. But I'd rather go for
vacation in a Beetle, nicely and with company, than stay alone in town,
yearning for a Rolls that I don't really need in the first place. I bet
it's raining too, and the bus isn't coming anyway. Brrr.

Why not define your goals first. If you discover that you really don't
have any, why then worry about transportation?

My own reasons are pragmatic, even naive:

- I don't anticipate that Genera (or suchlikes) will be available on low
budget machines,
- I'm not going to sell problem solutions, or CL-based software,
- I can dedicate a PC for CL experiments,
- Creating a CL environment and tools from bottom up appeals to me,
- device drivers aren't the most pressing problem now,
- I have no particular hurry,
- there are free CL:s for existing OS'es anyway,
- basic multi-processing mechanisms (possibly threading too) may needed
later, I don't know.
- I don't want to hazzle with pico/micro/macro/super kernels in the
first place. Others do it better,
- dada dada,

The two-way approach I have selected is; Use a free CL for testing
above-CL stuff, while developing bottom-up. No buts and ifs. I didn't
find it very difficult to make my decision.

Regards, Lars Lundback
From: Rainer Joswig
Subject: More screenshots
Date: 
Message-ID: <joswig-0301990112130001@194.163.195.67>
In article <·················@eralslk.ericsson.se>, Lars Lundback
<·······@eralslk.ericsson.se> wrote:

> I find to my delight that Rainer Joswig has put up introductory
> information pages on Genera at his Internet site,


More Symbolics Genera screenshots:

 The Concordia Authoring Environment
 http://www.lavielle.com/~joswig/genera2/concordia.gif

 Previewing documentation records
 http://www.lavielle.com/~joswig/genera2/concordia-previewer.gif

 An application for drawing graphics
 http://www.lavielle.com/~joswig/genera2/graphics-editor.gif

 Editing bitmaps
 http://www.lavielle.com/~joswig/genera2/bitmap-editor.gif

 The Font Editor from Symbolics Genera
 http://www.lavielle.com/~joswig/genera2/font-editor.gif


I also located a screenshot for the Xerox Lisp machine.
See http://www.parc.xerox.com/istl/groups/nltt/medley/ .



Happy New Year,

Rainer Joswig

-- 
http://www.lavielle.com/~joswig
From: David Lichteblau
Subject: Re: More screenshots
Date: 
Message-ID: <slrn7927ki.10i.david.lichteblau@lambda.dummy.de>
In article <·······················@194.163.195.67>, Rainer Joswig wrote:
>More Symbolics Genera screenshots:
[...]

Thanks for all the nice screenshots, they give a good impression of what
the Lisp Machines looked like.  The introduction you made available
sounds impressive, too.  For someone (like me) who has never seen, let
alone worked with, a Lisp Machine, it's hard to understand what's behind
these windows and promises, though.

To explain how Lisp Machines worked, would it be possible to make more
complete documentation publicly available, e.g. in the web?  Are there
other ways to find material about the subject?  Of course I know the
Symbolics Museum and the issue has also been discussed on the Lisp OS
list, but as far as I have read the archives, there doesn't seem to have
been a conclusive answer to the problem.

I'm sure there are lots of other curious persons like me who would very
much like to discover this world they have only heard about if they
could, so please tell me that there is documentation available
_somewhere_.

Hoping,
David
From: Rainer Joswig
Subject: Re: More screenshots
Date: 
Message-ID: <joswig-0401992145270001@pbg3.lavielle.com>
In article <·······························@lambda.dummy.de>, ················@snafu.de (David Lichteblau) wrote:

> To explain how Lisp Machines worked, would it be possible to make more
> complete documentation publicly available, e.g. in the web?

In principle it should be possible to serve the (almost 10000 pages -
I have 24 book - from a couple of pages to over 600 pages in my Genera 8.3
documentation collection) documentation to the web. Nobody has
written a converter yet (but this should be possible using something
like the scribe converter of Concordia as an example). Few have time.

>  Are there
> other ways to find material about the subject?  Of course I know the
> Symbolics Museum and the issue has also been discussed on the Lisp OS
> list, but as far as I have read the archives, there doesn't seem to have
> been a conclusive answer to the problem.

There is a book "Lisp Lore, Programming the Lisp machine".

> I'm sure there are lots of other curious persons like me who would very
> much like to discover this world they have only heard about if they
> could, so please tell me that there is documentation available
> _somewhere_.

Yes - I think it would be good if Symbolics at one point
just moves the doc to the web. I don't see it soon.
I wanted to show some pictures to make sure people know
that life could be so much better. If you happen to visit Hamburg,
I could give you a demo, though. ;-)

-- 
http://www.lavielle.com/~joswig