ftp.xmission.com/2014.06.ftp.xmission.com.tar/ftp.xmission.com/pub/lists/fractdev/archive/fractdev.200112

Internet Message Format | 2001-12-17 | 19KB

From: [e-mail]
Subject: (fractdev) Xfractint assembler code
Date: 03 Dec 2001 21:06:16 -0600 (CST)

Folks,

I now have successfully assembled (with nasm) and linked the pentium
mandelbrot code with Xfractint.  The outside options still need to be fixed.
Should be ready by this weekend.

Jonathan

Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"


-------------------------------------------------------------------------------

From: Tim Wegner <[e-mail]>
Subject: (fractdev) Possible change in fractdev list
Date: 05 Dec 2001 16:35:23 -0600

I have been having some problems with majordomo in administering the
fractint and fractdev lists. The kind xmission folks who host this
list are suggesting that I move to different (newer) software -
mailman. If this happens, I may be able to just move the list members

to the new software, or I may just ask members to sign on to the new
list when it comes on line.

This is just a heads up. I am not sure what the timing will be. I may

consider other hosts also if I am making a change, though I must say
that the xmission relationship has worked very well for us, with only

a few problems, so staying with them is fine with me as long as they
will have us.

If anyone has any thoughts on majordomo vs mailman or other list
software or hosts, I suggest we declare this "on topic" for this list

for a short while.

Tim Wegner
fractint/fractdev list administrator


Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"


-------------------------------------------------------------------------------

From: [e-mail]
Subject: (fractdev) Julia iterations
Date: 06 Dec 2001 20:25:40 -0600 (CST)

Tim,

Take a look at the default julia fractal, then look at marksjulia at
(0.3,0.6).  I would think that these two images should be the same.  It
turns out that the mandel code is not decrementing the counter after doing
the first iteration, before entering the loop.  Do you have any recollection
of why?  That's a rhetorical question, of course.  8-))  How do you think we
should fix this?

I found this while I was trying to get the Xfractint C-based mandel code in
sync with the DOS version.

There was also a problem with the outside=real and outside=imag options.  It
turns out that the compiler evaluates these two lines differently (duh!)

coloriter += new.x + 7;
coloriter += (long)new.x + 7;

The first is what was in Xfractint and the second is what Fractint does.

Jonathan

Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"


-------------------------------------------------------------------------------

From: Tim Wegner <[e-mail]>
Subject: Re: (fractdev) Julia iterations
Date: 07 Dec 2001 20:32:33 -0600

Jonathan wrote:

> Take a look at the default julia fractal, then look at marksjulia
> at
> (0.3,0.6).  I would think that these two images should be the same.  It
> turns out that the mandel code is not decrementing the counter after doing
> the first iteration, before entering the loop.  Do you have any recollection
> of why?

I think this was done because of the mandelbrot. Bert Tyler made z(0)

= C rather than Z(0) = 0 for the first Mandelbrot iteration to save a

cycle, and because he liked a blue rather than a black background. So

this is a quirk of Fractint's history.

> That's a rhetorical question, of course.  8-))  How do you think we
> should fix this?

In the past, we would be very loath to change something like this
because our artists would go nuts wanting to protect their legacy
images. But since we don't hear from them anymore, I'm not sure it
matters. But while fractint is no longer a dominant program, it still

does have a following.

Let me pose a question to this list. Is Fractint backward
compatability still an issue for you? Do you mind if we eliminate
quirks and make the program more uniform, or are you concerned that
legacy images render the same.

Hmmm - upon reflection, since we have things like Jim's "fractal of
the day", there still are accumulating libraries of PAR files. Maybe
it is important that we keep backward compatibility so as much as
possible fractint renders old PAR files the way it used to.

> coloriter += new.x + 7;
> coloriter += (long)new.x + 7;
>
> The first is what was in Xfractint and the second is what Fractint does.

Those two interpretations don't seem very different to me. Coloriter
is long, right? When would these two approaches give different
results?

Tim


Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"


-------------------------------------------------------------------------------

From: [e-mail]
Subject: RE: (fractdev) Julia iterations
Date: 07 Dec 2001 23:02:00 -0500

Tim,

I have many, many legacy pars, and I am still working on some new formulas.
I would not like to see the mandel code changed.

Ron Barnett

-----Original Message-----
[mailto:[e-mail]]On Behalf Of Tim Wegner
Sent: Friday, December 07, 2001 9:33 PM


Jonathan wrote:

> Take a look at the default julia fractal, then look at marksjulia
> at
> (0.3,0.6).  I would think that these two images should be the same.  It
> turns out that the mandel code is not decrementing the counter after doing
> the first iteration, before entering the loop.  Do you have any
recollection
> of why?

I think this was done because of the mandelbrot. Bert Tyler made z(0)

= C rather than Z(0) = 0 for the first Mandelbrot iteration to save a

cycle, and because he liked a blue rather than a black background. So

this is a quirk of Fractint's history.

> That's a rhetorical question, of course.  8-))  How do you think we
> should fix this?

In the past, we would be very loath to change something like this
because our artists would go nuts wanting to protect their legacy
images. But since we don't hear from them anymore, I'm not sure it
matters. But while fractint is no longer a dominant program, it still

does have a following.

Let me pose a question to this list. Is Fractint backward
compatability still an issue for you? Do you mind if we eliminate
quirks and make the program more uniform, or are you concerned that
legacy images render the same.

Hmmm - upon reflection, since we have things like Jim's "fractal of
the day", there still are accumulating libraries of PAR files. Maybe
it is important that we keep backward compatibility so as much as
possible fractint renders old PAR files the way it used to.

> coloriter += new.x + 7;
> coloriter += (long)new.x + 7;
>
> The first is what was in Xfractint and the second is what Fractint does.

Those two interpretations don't seem very different to me. Coloriter
is long, right? When would these two approaches give different
results?

Tim


Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"


Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"


-------------------------------------------------------------------------------

From: Tim Wegner <[e-mail]>
Subject: RE: (fractdev) Julia iterations
Date: 07 Dec 2001 23:03:14 -0600

Ron wrote:

> I have many, many legacy pars, and I am still working on some new formulas.
> I would not like to see the mandel code changed.

Not to worry Ron. By now you have seen my replies. I find this all
very amusing that sleepy email could result in such a furor :-)

The only one doing active development now is Jonathan. Jonathan, how
many hours of your life have you spent working on backward
compatability in Fractint? :-)

(For those not in the know, Jonathan's real name is "Jonathan
Backward Compatability Osuch").

Tim


Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"


-------------------------------------------------------------------------------

From: "Jonathan Osuch" <[e-mail]>
Subject: Re: (fractdev) Julia iterations
Date: 08 Dec 2001 07:01:30 -0600

Tim,

>> coloriter += new.x + 7;
>> coloriter += (long)new.x + 7;
> >
>> The first is what was in Xfractint and the second is what Fractint does.

> Those two interpretations don't seem very different to me. Coloriter
> is long, right? When would these two approaches give different
> results?

Yes, coloriter is long.  I thought I knew the answer to the question of why
the two are different until I started to reply.  I do know that the two
results *are* different.  One would think that the fractional part being
truncated would be the same in both instances.  And, the integer parts being
added together should also be the same.  It may have to do with whether the
compiler decides to truncate or round the result.

Jonathan



Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"


-------------------------------------------------------------------------------

From: [e-mail]
Subject: Re: (Fwd) Re: (fractdev) Julia iterations
Date: 09 Dec 2001 20:20:59 -0600 (CST)

Tim,

> I'm forwarding this in case it doesn't reach the list. I am having
> some mysterious email problems. I'm not sure you have gotten all
> my emails, not that I have sent you a lot. Did you get an email from
> me in response to your telling me that you had ported mandelbrot
> MASM to NASM? I asked you if that meant also porting the
> medium memory model to flat memory. Though if I think about it,
> porting float code this way is not so bad.

I don't seem to have any recollection of an email such as you describe.  I
did get the three you sent to the springnet account, although I didn't check
it from Iowa.

NASM uses the flat memory model.  It is actually pretty easy to do the
conversion.  I have contemplated trying to combining both versions of the
assembler code into one file.  Unfortunately, there are some major
differences and some variables are dword in NASM and word in MASM.  The
floating point registers are st(1) in MASM and st1 in NASM.  The use of
square brackets [ & ] is different with the two assemblers.  They represent
the value of the variable in NASM and the address of the variable in MASM.
I think, anyway.  It's been a while since I used MASM.

The reason I haven't released the patch yet is because I can't get the C
code to produce the same image as the assembly code.  That's a bit
frustrating!

The other thing I'm having to do is require fpu=387 in sstools.ini, since we
don't seem to have a way of telling what we are running on in Xfractint.
Hopefully, it is just a temporary solution.

Jonathan


Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"


-------------------------------------------------------------------------------

From: "Humberto R. Baptista" <[e-mail]>
Subject: RE: (fractdev) Julia iterations
Date: 10 Dec 2001 13:41:19 -0300



> >> coloriter += new.x + 7;
> >> coloriter += (long)new.x + 7;
> > >
> >> The first is what was in Xfractint and the second is what
> Fractint does.

    This may be _completely_ off the mark, but: some C compilers do their
integer arithmetic in normal int, and ONLY when needed convert to long, bu
this would only cause problema in large calcularions and/or divisions. And
the line above doens not seem to have either.

    Best to all,

    HB



Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"


-------------------------------------------------------------------------------

From: [e-mail]
Subject: (fractdev) Patch 20.2.02
Date: 10 Dec 2001 21:41:36 -0600 (CST)

Folks,

The diff for patch 2 to version 20.2 is now available.  I've updated the
Xfractint source only at this point.  Tomorrow I'll get the DOS versions and
the what's new files.

Jonathan

Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"


-------------------------------------------------------------------------------

From: [e-mail]
Subject: Re: (Fwd) Re: (fractdev) Julia iterations
Date: 14 Dec 2001 15:50:53 EST


--part1_14a.5c4f99f.294bc02d_boundary
Content-Type: text/plain; charset="US-ASCII"
Content-Transfer-Encoding: 7bit

In a message dated 12/9/2001 6:17:08 PM Pacific Standard Time,
[e-mail] writes:


>
> Tim,
>
> > I'm forwarding this in case it doesn't reach the list. I am having
> > some mysterious email problems. I'm not sure you have gotten all
> > my emails, not that I have sent you a lot. Did you get an email from
> > me in response to your telling me that you had ported mandelbrot
>
I don't know who you are trying to reach, but my name isn't Joe. Wanted to
let you know so you wouldn't thing Joe was ignoring you.
If you have any questions feel free to ask me.
R


--part1_14a.5c4f99f.294bc02d_boundary
Content-Type: text/html; charset="US-ASCII"
Content-Transfer-Encoding: 7bit

<HTML><FONT FACE=arial,helvetica><FONT  SIZE=2>In a message dated 12/9/2001 6:17:08 PM Pacific Standard Time, [e-mail] writes:
<BR>
<BR>
<BR><BLOCKQUOTE TYPE=CITE style="BORDER-LEFT: #0000ff 2px solid; MARGIN-LEFT: 5px; MARGIN-RIGHT: 0px; PADDING-LEFT: 5px">
<BR>Tim,
<BR>
<BR>> I'm forwarding this in case it doesn't reach the list. I am having
<BR>> some mysterious email problems. I'm not sure you have gotten all
<BR>> my emails, not that I have sent you a lot. Did you get an email from
<BR>> me in response to your telling me that you had ported mandelbrot
<BR></BLOCKQUOTE>
<BR></FONT><FONT  COLOR="#ff00ff" SIZE=2 FAMILY="SANSSERIF" FACE="Arial" LANG="0">I don't know who you are trying to reach, but my name isn't Joe. Wanted to let you know so you wouldn't thing Joe was ignoring you.
<BR>If you have any questions feel free to ask me.
<BR>R</FONT><FONT  COLOR="#000000" SIZE=2 FAMILY="SANSSERIF" FACE="Arial" LANG="0">
<BR></FONT></HTML>

--part1_14a.5c4f99f.294bc02d_boundary--

Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"


-------------------------------------------------------------------------------

From: Tim Wegner <[e-mail]>
Subject: (fractdev) Migration of list to mailman
Date: 16 Dec 2001 21:52:53 -0600

I am planning to migrate the fractint and fractdev lists at
xmission.com from majordomo to to mailman.

The new lists are already set up.

See:

   http://mailman.xmission.com

I am planning to migrate everyone tomorrow. However if you want to
try it before that you can subscribe yourself by going to the above
web address.

If any question or comment let me know.

Tim



Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"


-------------------------------------------------------------------------------

From: Tim Wegner <[e-mail]>
Subject: (fractdev) Fractint and fractdev list migration
Date: 17 Dec 2001 18:49:17 -0600

I have completed adding all the fractint and fractdev and
corresponding digest members to the new mailman software at xmission.

Please send all new posts to the new list
([e-mail] and [e-mail]).
The majordomo list will go away very soon. I am investigating options
for what to do with the old archives. I'll make sure they are not
lost.

You should have gotten an email with the password for your list
configuration web page. You can change your password, but don't re-
use an important password since this password provides only mild
security!

One thing that is a little different is that the digests are not
separate lists. Some of you subcribed to both the regular list and
the digest. You can select which you want on your list configuration
web page. I tried to set up previous digest members to get the
digest.

For a short time I hereby declare that questions about the list
software and how to use it are ON topic. However if you notice
particular things that need configuring, I suggest emailing me
directly and not bothering the list.

I have already changed the reply-to from the message poster to the
list. For our list I believe it works best for most replies to go to
the list.

Let me know if you are having any problems.

Tim Wegner


Thanks for using Fractdev, The Fractint Developer's Discussion List
Post Message:   [e-mail]
Get Commands:   [e-mail] "help"
Administrator:  [e-mail]
Unsubscribe:    [e-mail] "unsubscribe fractdev"