#No Imagine Blues2
24 messages in this thread
To ramble on further….
The deeper I've gotten into Imagine, the more I am realizing the problems
of documenting such a program. There is simply so much that can be done
with this program. I suspect that you could write an encyclopedia of books
simply about working with Imagine. I think about the manuals and articles
that I've read, totaling what must be several hundred pages, and there are
still so many parts of Imagine that I still don't understand.
Look at it another way: Has anyone tried to learn AutoCad? You
know….that insanely expensive program with documentation that loses you
in the first few pages. It's tough to learn because it is so incredibly
powerful. Go up to Walden Books or Dalton Books and count the third party
books on AutoCad. They don't write so many different books on this simply
because its such a simple program to use (though I try not to think of the
fact that some are catering to the pirated software market…..but that's
another topic of digression…).
Anyway, a program such as Imagine, because of its complexity and its sheer
power, is going to have a steep learning curve. I'm sure Autodesk has
better documentation for its "3d Studio" software, but it also costs
several thousand bucks more than Imagine…..and, frankly, I have a
difficult time believing that it could be any beter than Imagine.
Steve,
I agree with you completely. My mind reels every time I think about what
must be involved in developing and writing the code for a program like
Imagine.
And I agree that Mike's manual is not as bad as some people are saying. On
the other hand, it's a tutorial, not a reference; so when you're in the
middle of a project and need to know how to do something, you will probably
have a heck of a time finding the relevant material. And what you do find
may not be of any help.
Have you seen the new book by Steve Worley, UNDERSTANDING IMAGINE 2.0 yet?
It's a pretty exhaustive reference on Imagine, well indexed and carefully
cross-referenced: well worth its $30 price if you really want to learn the
program.
Pete
No I haven't seen it. I hope to buy it shortly
I've just got Steven Worley's Understanding Imagine 2.0 and it is possible
to write decent documentation about Imagine. He throws in some wonderful
tips on memory conservations, multi-tasking with Imagine and textures.
Worth double the price for the information you get.
Absolutely. I'm very impressed with the book (I recently found out I have
the very first copy ever distributed!). I wish Steve great success with it
— he clearly put in a mammoth amount of effort, and aside from a minor
glitch or two (which I'm sure will get fixed quickly) it is an absolutely
spectacular reference.
Pete
One thing he recommended that I couldn't get to work is to change Imagine's
task priority so I can multi-task it nicely. He said he uses Xoper which I
have around here somewhere but can't find right now. Or – taskpri -1 then
run Imagine from the CLI. Well maybe somebody can give me the correct
syntax to do this from the CLI, I've just been running it from the Icon.
When I launch Imagine from the CLI it runs for a second or two and then
returns to the CLI and is obviously not running. Que pasa?
Can't tell you about task priority, but to run Imagine from the command
line you need to be in the Imagine directory (because it looks in the
current directory for various configuration files). If you are in that
directory and it's still aborting — I don't know. Haven't had that
problem here… Maybe someone else can help.
Pete
Just a note to thank you for posting info about Worley's book. I ordered
it and agree it's a must-have, well worth the price.
I just installed 2.0 and started through the Impulse book. At first, I
thought it a frustrating agglomeration of ambiguities so badly written and
designed it makes IRS instructions look inviting by comparison. Then I
realized it must have been intended as packing material, to protect the
disks during shipping.
Anybody who thinks this is excusable because software is hard to document
ought to compare the middle 2 graphs in Impulse's manual pg 36 with the
same paragraphs on page 67 of Worley's. Same subject, even the same
example. Worley shows empathy for the reader, Impulse shows contempt.
I don't know that "contempt" is the right word. I just think that
programming and writing are often mutually exclusive talents. The problem
is not so much that Impulse has contempt for its user community as that
they try to save some extra money by writing the documentation themselves.
I agree, it's a bad mistake. OTOH, considering the difference in cost
between their program and something like Caligari (whose documentation I
have NOT seen), even when you add in the cost of Steve's manual, you come
out ahead.
My overall experience with complex programs (I'm thinking mostly on the PC
side, with things like Microsoft Word, WordPerfect, Lotus, Harvard
Graphics, etc.) is that the manual supplied with the software is third-rate
and you have to go out and purchase a book to learn to use the product
properly. I agree that Imagine is an extreme example, though.
Pete
While it's common for programmers to attempt to write, and for their
attempt to turn out badly, this can be extended to all walks of life. The
more general rule is, non-writers aren't very good writers. 🙂 With
practice, anyone can improve their writing. It takes time to become a good
manual writer. I haven't seen the Imagine 2.0 manual. Do you know who
wrote it? I don't think it was the programmer.
Sure do know who wrote it — it was Mike Halvorson. So, no — it wasn't
the programmer (which I guess would be Jack Knutson) but same deal —
Mike's definitely not a manual writer.
Pete
Are you *sure* you know that Mike wrote it? That's been the general
consensus around here, but even before 2.0 hit the streets I heard that
someone else was working on that manual: Steve Gillmour. The "style" of
the book is similar to Halvorson's, but when I read the manual, I came away
thinking that Mike didn't write it.
Steve
Well …
Mike himself TOLD me he wrote it. So if someone ghost-wrote it for him
-he's not telling. Besides, IMHO, it SOUNDS like him. No?
Pete
Steve,
Gillmour told me that Halvorson wrote it and that he (Gillmour) edited it.
-sja
Okay, let me rephrase it. The Impulse manual is not written for the
reader. It's written from the perspective of the developer: here's what we
want you to know. Is that kinder and gentler enough for you?
What I find really annoying is the manual's harping: if only you fools
would understand you have to experiment …. Yeah, no argument. And I'll
bet if you're sitting there with the source code and an intimate knowledge
of how the program works, you still have to experiment a lot to get the
results you want.
But I don't have the source code, and this isn't a box of crayons. It's a
complex machine in a closed room where you can't see it. I'm perfectly
happy designing my own Chippendale table and building it by instructing a
hidden machine to "draw cutting tool X across wood block A at location B,"
and taking responsibility for the result, without any help from a manual
about how to build a table. But it becomes a ridiculous exercise if I
don't know whether "cutting tool X" is a chisel, plane, chain saw, knife or
sissors — and if it's a saw, I'd like to know what its teeth are like:
size, shape, orientation, number, etc. This is the kind of information
that should be clearly and concisely laid out in a well-indexed reference
manual.
Then, if they want to lecture me about experimenting, no problem. And if
they want to sell an optional tutorial, or add $100 to the price to cover
the cost of the reference manual, you'll hear no complaint from me.
I have the most awful feeling that half the time they don't know,
themselves, what the shape, size, or number of the teeth are; or if the
tool is a chisel, knife, or a bottle of nitric acid. From talking with
them at some length, I have the idea they would have to experiment
themselves to have the basis for a difinitive manual. It's as if the
program is SO complicated that merely having the source code isn't enough
to tell them how the thing works, and when Mike ran into places where he
would have to do a week's work to get enough of a handle on something to
properly document it, he instead put in a remark about having to
experiment.
That's what it looks like, anyway.
I wouldn't be so forgiving if (a) the program wasn't as absolutely
fantastic as it is, and (b) Steve Worley's manual didn't exist.
Pete
Their tech help isn't much better than the manual at times. I recently
called them with what I thought was a simple question. I need to know the
formulas when creating primitives so can get a specific number of points,
edges, faces so I can have tow objects which have the same numbers in order
to do a morph. Their answer was to play around with it and I'll figure out
the relationship. I told them I have a client and a deadline and couldn't
affor the time to "play around".There answer was that I should send them a
letter There answer was that I should send them a letter and they would
send me the formulas (which must exist otherwise how can the computer do
what it does!!!). I have deadlines and they say send a letter!! That's just
not professional.
It's as I said: I don't think they know.
I've been wondering how they make this fabulous product and only charge a
couple hundred dollars for it. Now I know — it's because it's not built
on the kind of solid foundation that lets them know all the answers they'd
have to in order to market a product to the "professional" marketplace.
They really just don't KNOW how certain things work. To really know,
they'd have to have a larger development staff and some people who know how
to write documentation. And to support all the overhead, the price would
go the way of Caligari's ($3500+).
Pete
One ideal way to do it right is to write the manual first, and the program
second. 🙂
Isn't that known as "programming by design"? Write the manual, then the
software, then throw the manual away and re-write it to account for bugs.
:-'
The Impulse manual reads like a transcription of a phone conversation of
two 6th graders hacking at a C-64 copy protection scheme…
I thought people where kidding when they said you don't need a manual with
light wave.
You Don't. It took six minutes in lightwave what took me 2 1/2 hours in
imagine. Anything beyond model construction in imagine is insane.
Paul
–pt–not seattle– 😉
That's why you need InterChange Plus, due out in a few weeks. With it, you
can move an Imagine object to LightWave format. Or, you can use ordinary
InterChange today, and upgrade in a few weeks.
John,
Can I transfer objects without lousing up the surfaces? Using Pixel 2.0 I
find it better to just start from scratch in lightwave.
A couple of weeks? Really? 🙂
Paul
–pt–not seattle– 😉