#3D Studio & PowerPC
40 messages in this thread
Gary, that may be true right now this minute, but don't hold your
breath for long. The 601 is about as fast as the 75MHz Pentium, and
the 603 is about as fast as the 99MHz, and both are available now.
The 604 will be faster still, and the 620 will be faster than an
Alpha AXP. Because of the very architecture of the PowerPC vs. the
Pentium (RISC vs. CISC), the PPC will be able to increase in
performance both more rapidly and to a higher ultimate level than the
Pentium's ultimate potential, not to mention far cheaper.
Keep in mind that much of the Pentium's vaunted speed is due to
its superscalar nature, which is practically cheating: they basically
have the equivalent of two CPUs in one chip, running in parallel.
That's one reason the Pentiums are so big (in die size) compared to
the PowerPC. Big means more power draw, and more heat generated, and
fewer chips per wafer for lower yields, and thus higher per-chip
costs and even higher system costs (chassis have to be
thermally-engineered, costlier power supplies, etc.).
The PowerPC 615 from IBM (not available from Apple) will be able to
run 486 code at Pentium speeds for a fraction of the cost of a
Pentium. And it (the 486 code) won't even have to be recompiled to
achieve that speed (the Pentium requires recompilation for full
speed, remember?)! It does this by having 486-compatible microcode
on-chip, so that it can execute 486 code without emulation, at the
full native speed of the PowerPC.
3DS's performance and features are hot stuff now, but the interface is
really showing its age. I mean, come ON!! MONOSPACED TEXT in a GUI!?!? Even
the LISA was more advanced than THAT and that was over a DECADE ago!! Even the
late and not-so-lamented GEM was more advanced than THAT!
I recommend that to keep up with the future, R4 should be rewritten
in a more portable form, using one of the user-interface libraries
that enable porting to Mac, Motif, Open Windows, X-Windows, NT, OS/2,
Workplace OS, Taligent Pink, etc. etc. etc. with simple
recompilation, with the UI also translated to the appropriate form
(radio buttons look like MAC radio buttons on a Mac, like Windows
radio buttons on Windows or NT, etc.). That would make porting to
whatever new CPUs may come out much easier. 3DS should not be so
tied to a particular hardware platform, IMO.
>> The PowerPC 615 from IBM (not available from Apple) will be able to run
486 >> code at Pentium speeds for a fraction of the cost of a Pentium. And it
(the 486 >> code) won't even have to be recompiled to achieve that speed (the
Pentium >> requires recompilation for full speed, remember?)! It does this by
having
>> 486-compatible microcode on-chip, so that it can execute 486 code without >>
>> emulation, at the full native speed of the PowerPC.
If it doesn't require recompilation, what's the question?
>> 3DS's performance and features are hot stuff now, but the interface is
really >> showing its age. I mean, come ON!! MONOSPACED TEXT in a GUI!?!?
Even >> the LISA was more advanced than THAT and that was over a DECADE ago!!
>> >> Even the late and not-so-lamented GEM was more advanced than THAT!
More advanced, or more needless garbage to look at? It certainly
seems like the prettier something is on a PC, the slower it is. Even
with native PowerOpen applications, PowerPC Macs will still be
running a GUI-based operating system, which will be very fast with
things like Photoshop and what have you, but as a 3D rendering number
cruncher it would be shooting itself in the foot. Granted, CISC
processors' days are numbered, and indeed should never have been used
in the first place, but the Pentium will still be a faster DOS-based
number cruncher than PowerPC with a GUI.
Intel will eventually run into a brick wall with CISC, and I expect
they'll realize when it happens and move on to RISC, but until then
the real-life processing power will be roughly comparable.
Now, however, it is at best premature to say that 3D Studio would be
faster on a PowerPC.
On the other hand, though, why aren't Alias Poweranimator or Wavefront
being ported from SGI to Mac or PC? Sometimes things work better on
the platforms they were built for.
Re: the PPC 615, I was simply pointing out that it will probably be the
fastest CPU for 3DS rendering if 3DS is not recompiled for optimized speed
under the Pentium or the other RISC chips. The Pentium is only twice as fast
as a 486 at the same clock speed for 3DS as it is now, but with the
recompilation it would be four or more times as fast. The PPC 615 would bring
Pentium-class performance for unrecompiled 486 code!
3DS is a GUI anyway. As such, it is slower in operation than a
text based program would be. I know from personal programming
experience that displaying monospaced text on a pixelmapped graphic
screen isn't that much faster than displaying proportional text
unless the monospaced text is forced to byte-aligned pixels (which it
would be anyway at 256 or more colors, since ALL pixels are
byte-aligned at those resolutions!). But proportional text is MUCH
easier to read, and, perhaps more important, much more space-efficient
on the screen. Buttons could be smaller and still have legible
labels. A text-based interface is only drastically faster than a
graphics one when you're talking about TRUE text-based — as in
running in text MODE. As in Video BIOS Mode 2 or 3. As in 80×25 or
maybe 80×50 text. As in NO GRAPHICS WHATSOEVER. This does NOT
describe 3DS.
But there are more problems with 3DS's interface. It just isn't
that well layed out. For instance, in the Materials Editor, View
Image and File Info are both buttons with drag-and-drop ability
(NICE!). Their purposes are closely related: either view info on a
bitmapped image file, or view the image itself, to help decide which
image file to use. However, in the other modules, File Info is
located in the File drop-down menu, while View Image is located in
the command column under "Render"! This makes NO SENSE and is
confusing to new users. The only reason it EVER becomes easy to use
(not learn) over time is by brute force of habit.
3DS's interface needs a COMPLETE OVERHAUL, in my opinion, to keep
it competitive with the flood of new 3D and multimedia software
coming from Microsoft and other software giants who could buy and
sell Autodesk before breakfast. Sure, 3DS has a TECHNICAL head
start, but for new users, the ease of learning that a
well-thought-out and well-implemented and, yes, ATTRACTIVE
user-interface engenders is of more immediate importance than whether
the rendering engine does 64-bit rendering with gamma correction.
As for Intel's future dominance in the face of the PowerPC and
RISCs, remember, Intel was king of the CP/M hill with the 8080, and
one day there came into existence this little upstart company called
Zilog which put out a little upstart chip called the Z-80, and
suddenly Intel was an also-ran! It happened before, and it can
happen again. In fact, I dare say that if IBM had not gotten into
the PC market (and it was touch-and-go there for awhile — IBM's top
execs were halfway convinced to stay out of it after the dismal
failure of their 5100 — remember that?), or if they had but had gone
with some other 16-bit CPU (say, the Zilog Z-8000, or even the 68000)
then there would be no such thing as Intel today.
Marion:
>>>But proportional text is MUCH easier to read<<<
Do I understand you correctly? In trying these typical commands: LOAD
PROJECT or SELECT/OBJECT/BY NAME if you have naming conventions that
use any form of character grouping (ie. EXT666, INT111, etc.) are you
saying that it would be easier to read if the columns of
characters_didn't_line_up? Hope I am misunderstanding you because
I'd prefer a sharp stick in the eye to proportional text in selection
lists.
Of COURSE not for selection lists! That's why Mac has Monaco, and
Windows has FixedSys, and both have Courier. Monospace does indeed
have its place. But that place isn't "anywhere and everywhere there
is text."
As far as the GUI goes, yes, 3DS has a GUI, but it is not constantly
working at the same time as a render is going on, as both System 7
does and PowerOpen will. 3DS runs directly from DOS and does not
rely upon another constantly running program (unless you count DOS)
for either graphical instructions or memory management, there is no
way it couldn't be faster. 3DS has a GUI, but it doesn't run UNDER a
GUI, and that is what makes the difference.
I feel no particular kinship with Intel, and as a company I much
prefer Motorola, but until something better that is not encumbered by
an overlaying GUI comes along, 3D Studio is where it belongs.
Personally, I wouldn't mind an Alpha version of 3DS, and a
Plain-vanilla R4400 UNIX version wouldn;t hurt either, for that
matter.
If Intel ever dies, we would probably have to equate it to progress
more than anything else (but of course Intel doesn't, and probably
cannot, make a deal with a system manufacturer to build a machine
around the chip like Motorola did; that would be like trying to make
a deal with a whole orchard of Apples!).
Not quite. 3DS is constantly updating that little red status bar
and the rest of said display while it's rendering, and that's GUI
stuff. Not much, mind you, but it IS there. And a GUI operating
system is not a function of the CPU itself (GUIs exist even for the
1MHz 6502-based C-64 — GEOS was originally for the C-64 and 128 and
later the Apple II before it ever appeared as GEOworks on the PC).
Modern UNIXs almost all have GUI front-ends, but they are just that:
front ends, just like Windows is to DOS. They are NOT the OS. UNIX
works just fine on a plain dumb ASCII terminal connected to a
mainframe, with absolutely no graphics ability whatsoever. Macs
don't have a text mode, but that's not because of the 68000, but
rather it's because the Mac video hardware has no text mode (text
modes are actually more complex to implement, since the data sent to
the screen has to go through TWO memory locations to get there for
any given pixel: the ASCII character code stored in the character
screen map, and the shape of that character stored in the
character-generator ROM [or optionally loaded into RAM in the case of
the EGA, VGA, C-64, C-128, Apple ///, the old Hercules InColor and
Hercules Monochrome Plus, etc.], whereas for graphics modes only one
memory location [or group] has to be accessed for each pixel.).
68000-based computers such as the TRS-80 Model 16 and the Tandy 6000
did have a text mode, and indeed it is the native mode of the
operating system (a UNIX clone).
There is no reason that a 3DS rendering engine couldn't be made
that runs in text mode and renders to a non-display buffer, with
text-based status displays. Such a renderer would be far more likely
to be able to run in the background under Windows (which can't
multitask DOS graphics programs in the background, but can multitask
DOS text-based programs in the background). Granted, the rendering
would be slower, but rendering could go on while you're doing
something else with the same computer: say, getting Compuserve
messages with WinCIM or Navigator.
Hopefully, 3DS rendering modules, text-only or otherwise, will be
made for the various non-Intel chips out there, that would work with
Network Rendering.
Yes, the little red bar IS constantly being updated, but that's hardly
on the order of the constant polling of Windows and System 7.
I really would like to see 3DS running on a PowerPC under PowerOpen or
what have you…it would be interesting to see, but I honestly doubt
that it would render as fast, much less faster, than a current Intel
chip. Even with all their shortcomings with newer and better
features being built on top of outdated technology, Intel chips have
the same kind of integer performance that some of the fastest RISC
chips (R4400 and DEC Alpha, for instance) are blessed with.
The idea is that a system like 3D Studio, with its enormous need for
processing power, should not also be slowed down further by being
asked to calculate the space between pretty font characters just so
that we'll all lose our eyesight that much sooner.
Bottom line for me is: what platform that we can all afford, other
than Intel, do you think 3D Studio would actually perform better on?
>3DS has a GUI, but it doesn't run UNDER a GUI, and that is what makes
the difference.<
Hmm. If you are thinking about a System 7 type environment/operating
system, the thing that affects the speed going to 3DS is that 3DS is
not treated as "God" of the machine, merely part of the whole. There
are other things going on besides that one program. There are all
kinds of things added which enable a multitasking environment (among
other things), which take away from the raw performance of the
individual program being "speed checked". The main issue is not
whether the program has a graphical user interface, but whether it
running in a large, sophisticated OS or a small, simple one.
(Both 3DS and Mac/Windows/etc. type programs have GUI's. It's just
that the Mac type program has more "stuff" to deal with.)
<stepping up to the soapbox.. <g> >
3DS is definately honed to suck as much out of the system as
possible, but does this by sacrificing on how people work
(multitasking). This is where I believe a Win 32 solution comes in.
If 3DS were developed as a Win 32 program, it would run on many
different types of processors, including the many PowerPC's without
modification. Windows NT is already running R4400 systems, Alpha AXP
systems, MULTIPROCESSOR Pentium systems, and more. It allows
portability, scalability, and allows the user to take control of the
system. One could render in the background and work on other
things. Right now, the best solution is running it (a tweaked DOS
program) in OS/2. It works, and increases flexability, but is
ineffecient.
Imagine a workgroup of, say, a R4400 machine, a couple 486/66's, a 2
processor PPC 620 machine, and a Pentium 75 machine. They are all
running Windows, either Win 4 (on the Intel's) and NT on the rest.
They are all networked, and have a background, distributed rendering
program running on idle in the background. If this workgroup is like
most, there is much time wasted waiting for the user to make up
his/her mind on what to do next. The background renderer sits there
for a while, and notices that the machine has been idle for x amount
of time. (the user must have gone to get more cofee) It sends out a
note to the other computers and tells the others that the computer
isn't doing anything. Now, at the same time, there is another
computer that is doing a rendering with some cool new IPAS, and could
use some help. The computer (still rendering) zips over some
instructions on what the idle computer can do, and cuts the time
needed to do the job. Therefore, some of the megabucks spent on that
caffine addict's computer could be re-absorbed.
Other examples include taking away XX% of John Doe's computer power
and giving it over to a distributed rendering pool, available to
anyone who needs it. Or switching back and forth between that really
complex Corel Draw design, last minute Photoshop texture map, and
3DS- all while only using one machine.
What do you all think? I think a Windows 4+/NT (not a windows 3.x)
program would be the ideal platform for R4. (Dos is already dead,
that's evident with the need for the Phar-Lap extention.) Windows 4
(if it's done as they say.. ) will be a full OS, and be 32 bit.
Windows NT is already 32bit, and they will share the ability to run a
common base of programs natively. I think this is the best
solution. NT can be used by those needing non-Intel processors, and
4 will demand the least amount of resources for those who don't
(along with adding cool new system features).
I know many don't like "Windoze", and would rather eat nails than use
it, but what else is there for 3DS? Unix? nope. Mac? nope. OS/2?
nope. I doubt there will be another Dos… it's time to choose.
>>>What do you all think? I think a Windows 4+/NT (not a windows
3.x) program would be the ideal platform for R4.<<<
If R4 can't be run on a DEC ALPHA AXP then I won't be running R4. :-[
You going to make that clean of a break, and dump all your old
equipment? Seems a little early to me. R4 will be out in a year, and
I can't see Alphas or PPC's making that big a splash is so short a
time.
Don't have any equipment to dump, actually. We have 50 DEC
workstations that we lease and will all be going through an upgrade
process to ALPHA's over the next 2+ years. We are preparing an order
for 30 more workstations (all DEC) to be purchased this year. The
CADD package we run, GDS, has been ported to NT running on the ALPHA
so we will be swapping over to Windows NT. The only PC that will run
3D studio hasn't yet been ordered (I will order it by weeks end). I
am buying the DEC XL566 Pentium which allows me to upgrade to an ALPHA
chip via a daughtercard swap. If R4 doesn't run under NT on the
ALPHA then I'll be forced to go to Wavefront (which I have to upgrade
from an old copy I've got running on an old Tektronix XD88) or
possibly ALIAS, whichever will run on the DEC ALPHA platform. My
take on 3DS is that running under windows NT will allow it to
seriously crush competition from above and the sides. Running on the
PowerPC won't make or break 3DS nor will running on the SGI platform
although it would improve the market share. 3DS, IMHO, will go where
AutoCAD goes and that appears to be Windows NT. Please feel free to
jump all over me if I am wrong, use nasty language, too as it will
help me understand how vehement you wish to make your point. :-] I
have a lot riding on my decisions so any objective input can only
help me plot a viable future for my firm.
I share completely your vision of the Alpha processor. Besides, it is
cheaper thann anything there (from the price/performance side). It
cost the same to get a DEC pentium than a DEC Alpha. And for sure
will cost less to the the AXP 150MHz than the Pentium 100MHz. I see
the AXP as an ideal platform at a great price, that not olnly will
provide you with high level performance but with a bridge between the
comic world of DOS to a serious operating system. When I see this NT
drivers for the MGA2 waiting in the disk case, I go mad. Everything I
have here is ready to go to the AXP world, except 3DS. As you see, a
powerful reason to stay with DOS <g>. Once 3DS gets to NT I will say
bye to Intel.
Jeez, I hate it when people agree with me. :^) Let me dream for a
second:
There I am with 8000 frames to render, its 2:00 pm, I need the
animation for a board meeting of the XYZ corporation in the morning,
and I need it on tape, and in enough quantity to pass out a copy to
each of the 12 board members. No sweat, the frames render in 1.5
mins on the Alphas instead of 5 mins on the Pentium. I have 80
workstations available to network render so the completed animation
will be done in 2.5 hours which puts me at 4:30 pm. I need to make
12 copies of my final product so I have all of the frames rendered
directly to the PAR drive which I then record great quality first
generation tapes for each board member. The animation is just short
of 5 mins so the recording is done just before quitting time at 5:30
pm. I go off to the beach in time to catch a couple of 2-man beach
volleyball games as the sun sets on another easy ALPHA kinda day.
Please don't bring me back to Intel reality!
Robert,
One problem with your dream. Animations are like a gas, they fill the space
that is given them. I worked with Wavefront on a loaded SGI Crimson for a
while. I thought I was in heaven. No waiting! So I added a few hundred more
vertices, a ray trace, a few more high res maps, a few more frames, etc. Soon
I was back up to that old 5 minute a frame time. What the hell happened?!? 🙂
I was having a lot of fun, you mean that an SGI Crimson with 100+MBs can be
brought to its knees?
After working with computers for years now, the one thing I have found is that
faster computers never give me free time for the beach. I just get better
quality or more or what ever I am doing. I agree with you on the need for 3DS
to be able to be ported to other higher platforms. We started with 3DS, and
as my business grows, which it is by leaps and bounds, I hope we can stay with
3DS. The news that there will be an SGI rendering engine is good news. The
need for doing rendering on faster machines is evident. Though as a small
business that has to watch its money closely, I can't justify the productivity
increase by doing the modeling and animating on a SGI or DEC. Use the power
where it is needed most and use lower cost machines for the modeling and
animating.
Andrew
Graphic Alchemy
That explanation makes more sense. IT seemed like a big leap.
I completely agree with the differences you stated between the 3DS GUI
and something like System 7 (or whatever), that is also doing a
thousand other things.
I like the NT idea because it's true you CAN put ten or twelve
Pentiums in the same box and if you have the memory even the most
difficult 3DS render would take a few minutes max. If there's a way
a programmer can tell NT to disregard everything else and work solely
on the render, it would be better still.
Network rendering is kinda sludgy in the first place because there a
number of different machines that all have to have a number of
project-specific items local (maps, IPAS processes, etc).
What would be really nice is if there was some kind of rendering
supercomuter that doesn't have the price tag of an SGI Onyx Reality
Engine that could just take a 3DS project from a workstation and
render it as the workstation goes on to do other things. This
machine could be set up to do the work then "forget about it" by
deleting all the maps, processes, etc that went into completing that
specific render. A Pentium multiprocessor would be a nice machine to
do such a thing.
>>A Pentium multiprocessor would be a nice machine to do [serious rendering].
Yes, definately. A multiprocessor PowerPC 620 machine would be even better.
BTW, I've seen 2 Pentium machines running NT out already. I don't think they
can coerced to run 3DS yet (Phar Lap stuff), but they're pretty reasonable in
cost. (and you can cram all that RAM into one big pool, instead of several)
Great times are ahead for PC animators.. I'm glad to get in now.
JNJ
I'd be glad to argue that a bare-bones, GUI-less render engine on the PC makes
more sense than a WinNT version. 1. Speed. 2. You don't need a GUI on a
"render farm" machine. 3. 16 megs of "wasted" RAM on a WinNT machine. 4.
You've already got an investment in Intel hardware, why not use it? Even if a
render farm machine is *less* powerful than your primary machine, wouldn't you
rather render an animation faster *overall* by dividing it among two or three
machines, than only using one? 5. You'd need to port any necessary IPAS
routines to other platforms. The more dependent your users are on plug-ins,
the more intense the need to port them, too.
Now, if they simply made 3DS.EXE run safely as a standalone render engine under
Win32s/WinNT, then you could have a choice. But don't deny 32-bit DOS's speed
if you don't have to.
I totally agree that, today- as a render farm system, there is nothing that
beats the 32bit dos setup. This, however, does not allow others to use the
other computer at the same time. I understand that if you have dedicated
systems bought for rendering (or working at night), this is not a problem, but
when running it on a large network there are alot of CPU cycles wasted everyday
and my examples would (attempt to) use them effieciently.
Also, I did not mention this to be an NT only solution. I see little reason to
run NT on Intel machines right now, especially in 3D. My example was a Win 32
program, which would run on Win 4 (32 bit) as well. NT would be ideal only for
multiprocessor machines, and PowerPC/Mips 4X00 and Alpha chips. Windows 4 will
probably demand 6-8 MB. Still alot, but less than NT. I have no real idea on
how effiecient this all would run. It's hard to play "what if", but we've got
to look past dos for our future, and I see the Win 32x software as the future
of our PC platform. (Which still seems to be the best bang for the buck
platform)
JNJ
Coming from the Amiga market, where multitasking was the norm and there were
plenty of 3D programs, I can reassure you that you (or anyone else on the
network) will not want to share their processor cycles with a background
rendering task. Even with preemptive multitasking, you're still losing those
cycles. I wish DOS would go away, too. 🙂
John
>I can reassure you that you (or anyone else on the network) will not want to
share their processor cycles with a background rendering task. Even with
preemptive multitasking, you're still losing those cycles.<
But with Pentium and PowerPC type systems being a norm in the near future,
wouldn't many machines have extra cycles being wasted? Just a thought- I've
done some rendering in the background in Win 3.1, and have noticed little power
draw on many applications I use, like this comm program -where there is little
need for massive cpu cycles. I don't need a 486/66 for most of my apps, but
there is a small majority that demand it. (like 3DS)
For smaller businesses, where there is not the budget for having multiple
machines per person, multitasking is essential to productivity. There are many
times where I wished that I could write papers, make bitmaps, etc. when I had a
rendering to do. The rendering was not top priority, but was something on my
list of things I needed to get done. I would have loved to render in the
background while doing other work. Based on my Playmation (windows OR dos)
rendering experience, it's nice to have that flexability. I can render in dos
when speed is essential, and I can render in Windows when multitasking is
essential. 3DS denies me that flexability. I'm still trying to get OS/2 to
work right for me. That should help, at the expense of learning a new OS.
<ugh> Basically- some need it, some don't. I'd like the option. 🙂
JNJ
Don't worry, Microsoft will find a way to eat those cycles and those
spare memory chips. 🙂 As much as I'd hate to quote him, because
he's a goof in so many other ways, but one of Pournelle's Laws is
"one person, at least one CPU." And in this case, a rendering task is
a "person." You'll need extra memory for 3DS and the rendering task
and WinNT's many processes, on top of what you want for the person
who's using the machine. It's a tough fit.
And if WinNT's preemptive multitasking is done the least bit sensibly,
and although I haven't studied the details of it, it appears to be
done well, then there are no "extra" cycles. Many background tasks
are in a "wait" state: waiting for something to happen, an
interrupt, a mouse event, some other event. They don't "busy-wait"
or consume many cycles. The multitasking manager gives control to
each task based on priorities, and among processes of equal weight,
many might be doing nothing, and will surrender their time slice.
This is a simplified view, but it shows that your foreground task
might actually eat more than 1/(N processes) cycles, so there may be
no "spare" cycles that can be eaten without noticeable degradation in
the user's programs. To put it another way, the user gets very tuned
to the time that operations are "supposed" to take. If they take
longer, they'll be visiting your office shortly. 🙂
>…If [the foreground processes] take longer, they'll be visiting your office
shortly. 🙂 <
hehe, I'm sure. It may be worth it though. These are the best ideas I can
come up with (using new technology). I never said it would be invisible to the
user. A bothered user could shut off the availability while he/she's working
and still allow it to run in "screensaver mode"- rendering when he/she's gone
to a meeting, etc. What do you see as better? Just curious.
JNJ
Your concern over the interface of 3DS is right-on target. I have
had the product for a couple of years, but have had difficulty
getting others to use it because THE LEARNING CURVE IS TOO LONG. The
curve is long, not because of the technology or vocabulary which must
be overcome, but because the product is not well constructed from a
new-user point of view. Sure, it is VERY powerful, but it takes too
damn long to figure out how to use it all. I think AutoDesk could
learn something from studying a program like CorelDraw. I quite like
the new CorelDraw interface. Very powerful, but direct. What do you
think?
<< review the GUI for improvement..>>
I'll take that as a R4 wish.
While I'm not married to the current GUI (it's not THAT bad…), I'd hate to
see it be like Corel Draw =-( Why do people keep comparing 3ds to corel draw
lately? Corel Draw is neat, but c'mon…
John Tissavary (La Luna cie)
A custom interface for each person will be the right way to go. Then
everyone can upload theirs and we can let Gary and Co. to say
"something" about them. Oh, we can organize a contest. It is so easy
to design graphic interfaces for programs like 3DS! I bet that the
Yost Grup will include an IPAS for interface design for advanced
animation programs ALL IN 3D! Even John Ellis is willing to see it
ported to Windows because he will have hear that you can edit icon
files for Windows applications. Wow, isn't that great?
Gary, I love that neat 3D buttons.
ana /krazy image
ps. I suppose that you have see the giant inmense gigantic <G>
watermarked in this message, John 🙂
Definitely, designing graphical interfaces for 3DS is a breeze! I
think 3DS ought to not come with one at all, so you have to create
your own right from the start (show all those lazy user's what's
what!). <ggg>
I would love to judge a contest like that. It would certainly be a
lot easier than what I'm currently doing.
Most of the times it is easier to judge than to create the real things
in life. I know that you have a highly developed common sense <g>
and that you will take critics into account… but with the correct
value and weight for each one. As always, keep the good work. That's
what you and the Yost Group know to do better 🙂
ana /krazy images
>>Your concern over the interface of 3DS is right-on target. I have had the
product >>for a couple of years, but have had difficulty getting others to use
it because >>THE LEARNING CURVE IS TOO LONG. The curve is long, not because of
the >>technology or vocabulary which must be overcome, but because the product
is >>not well constructed from a new-user point of view. Sure, it is VERY
powerful, >>but it takes too damn long to figure out how to use it all.
That's certainly true. The major problem I had was going from
drag-and-drop Windows spline-based systems to 3DS. Kinda like the
difference between Macintosh and DOS. Oh well, if everybody could do
it easily, animators wouldn't get paid so well.
I have a friend who has a whole slew of Mac modellers and renderers.
I think the user-interfaces are much better, generally, on the Mac
than on the PC. My friend keeps ranting and raving about how much
better his stuff is than mine! And I agree. Until he renders. Than,
glad to say even with the poor interface, 3DS beats him cold in speed
and quality!
>The only reason it EVER becomes easy to use (not learn) over time is
by brute force of habit.< yes.
>3DS's interface needs a COMPLETE OVERHAUL, in my opinion< mine too,
I suggest under Win 32. My suggestion for the designers is to borrow
a Mac, and a bunch of top 3D/paint/multimedia programs, immerse
themselves with those products and see how these programs deal with
integrating all the functions into a manageable, smart interface.
IMO, design of a gui is as important as the code behind it. I want to steer my
ferrari. <g>
>
Gee… I was proud of our monospaced text until you said that.
And we even put it in those neat little shaded buttons and things.
Awww….
– <G>
I _told_ you we should have used the Victorian Script font!
– J
Ouch.
– G
>>I _told_ you we should have used the Victorian Script font!
>> – J
A nice cyrillic font would have pleasant too, and it'd pretty much completely
preclude anyone using 3D Studio illegally without a manual.
-MF
gee all i did was ask one little question. . .
Yeah, Marion, and make 3DS to cost $1,000 more. Spend a lot of time
rewriting the -for sure advanced- r4 and "rewrite" the cost and the
price too. I can't understand why people is making so much noise with
PPC. Really, when Microsoft talk about Chicago, that's vaporware.
When Apple talks about PowerPC and the Newton, those are benchmarks,
features, and realities <sigh>
Anyway, I don't want to live in promises. I will make changes when I
see things running not when a bunch of computer critics spend their
time looking at their magic balls to tell me how nice is the
information superhighway, how wonderful Taligent Pink is, how
mac-like is Chicago and how fast will be the PPC645 and the Sextium
or Septium chip.
I believe that Gary want to make available (he did it trice, in fact)
a technology that was not available for the PC platform at the right
price. Make the tools available for everybody and not only to a few
mad scientist and production houses. I think that this is the reason
we can run now 3ds in cheap machines, and build cheap networks with
cheap PCs to speed up rendering time of this wonderful tools. For
sure, 3DS will be availble to more platforms, but I think only if
they become popular and cheap enough. If the PPCs live to the
promises to cheap SGI power, maybe we will see R4 running on it. For
now, let's hope the PPC to take off only, ok? 🙂