Linux PPC build?

[richmond62]

OK: I just tried that (sorry, slight hiatus to cook supper) and got

EXACTLY the same error messages as before.

[richmond62]

Obviously "everything" is against me, LOL!

bonkers.png

[richmond62]

There was no Linux PPC engine for any later version of Metacard than 2.4.

[richmond62]

Strictly for nutters: http://www.filewatcher.com/b/ftp/ftp.ge ... ard-0.html

[mwieder]

(catching up here)

Richmond- the warning you got from "make clean" is proper - trying to remove a directory that hasn't yet been built. No worries.

Fraser, Richmond- I also admit to having no idea what the score.txt files are for. Or where they came from.

Richmond- what do you see as an output from "which gcc"?

Fraser, Richmond- Lines 48-50 in file livecode/thirdparty/headers/linux/include/lcms.h look suspicious:

Code: Select all

// Uncomment this one if you are using big endian machines (only meaningful
// when NON_WINDOWS is used)
// #define USE_BIG_ENDIAN   1

I think that should be switching on #ifdef __ppc__ rather than forcing an uncomment, but at any rate you *do* want USE_BIG_ENDIAN defined here for a ppc build.

[richmond62]

As far as I can tell I need, somehow, to switch OFF little endian.

[mwieder]

...try removing the "//" before that #define line. That should help.

[richmond62]

Thanks to the good offices of the Wayback Machine I have been having a "Back to The Future" Session:

http://web.archive.org/web/200202061425 ... /mcmirror/

backToTheFuture.png

Metacard 2.4 running on Ubuntu 12.04 with LXDE.

And, "So What?" I here you all cry.

Well, failing all else, one can at least author some sort of content delivery stacks for educational purposes, and build standalones (of a sort) from Livecode stacks saved in the ultra-legacy
format.

[richmond62]

This got badly stuck:

humpf.png

[richmond62]

This is what RunRev 1.1 Linux should look like, as here it is running on UbuntuStudio 13.10 Intel right now.

RR111_INTEL.png

[richmond62]

I'm a lucky bloke: Monday, 3rd of March is Bulgarian National Day, so here I am, at home, continuing in my Hank Wangfordesque search for "The Lonesome Yodel"

eckie_thump.jpg

[or, as an ex-Amish friend of mine once put it; "The Lonesome Yoder"].

However, Mark, after a fantastic night dreaming of a functioning 6.5.2 on Linux PPC, I'm afraid your edit was insufficient:

noJoy.png

Althoug, rummaging through thirdparty/headers/linux/include/lcms.h I did wonder about this:

wonder1.png

[richmond62]

Does "if macintosh" refer to a build on a PPC machine under Mac OS, or a build on a Macintosh PPC machine running Linux PPC, or both?

[richmond62]

Well, as nobody exactly leapt into the breach to help me, I decided it was (err . . . perhaps) time to learn C++.

Now, if you stop and think about that, that is a slightly ridiculous decision for a 52 year-old bloke who has a full-time job running an English as a Foreign Language
school to take . . .

. . . not that my life (and those of very many other people) has not been littered by daft decisions . . .

Anyway; I went here: http://ocw.mit.edu/courses/electrical-e ... ure-notes/

and downloaded the set of 10 pdf documents constituting a set of lectures on C++.

Having had an initial look, in the light of my "murky past" (FORTRAN, PASCAL, BASIC, ZILOG), it doesn't look, perhaps, as opaque as a lot of people
[mainly that horrible species; "computer experts", that lurk in funny places in technical institutions here in Bulgaria who won't throw you a crust of bread even if you are on your last legs; so intent are they on maintaining the
mystique of their craft] would have me believe.

Well; I cannot, however, see me becoming God's gift to C++ anytime soon; but I am sure that with a few cups of coffee and a pair of horse-blinkers (to aid focus), I can get a nodding knowledge of the thing;
and possibly just enough to get a Linux PPC build up and running.

Why am I posting this, slightly OT bit here?

Because I am a great believer that people who hang around and wait very rarely get anything beyond a slightly sour taste in
their mouth.

[richmond62]

This is part of the code section of the document livecode/thirdparty/headers/linux/include/lcms.h:

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

extract 1

// Uncomment this one if you are using big endian machines (only meaningful
// when NON_WINDOWS is used)
// #define USE_BIG_ENDIAN 1

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

extract 2

// Uncomment this one if your compiler/machine does support the
// "long long" type This will speedup fixed point math. (USE_C only)
#define USE_INT64 1

// Some machines does not have a reliable 'swab' function. Usually
// leave commented unless the testbed diagnoses the contrary.
// #define USE_CUSTOM_SWAB 1

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

Presumably I should uncomment extract 1 as a MacMini PPC is a Big-Endian machine.

Now I do not know whether a MacMini PPC supports the "long long" type (whatever that actually is?)

Nor do I know anything re a 'swab' function and a MacMini PPC.

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

I would be grateful if someone could either answer my questions or tell me how
to find ouht directly on my machine.

[LCfraser]

GCC supports long long on all platforms so go ahead and uncomment that part.

"swab" is (I think) a function for swapping byte orders, i.e little <-> big endian. Given that PowerPC has machine instructions for doing such things, I think it's safe to say you don't need to uncomment that part.

On a tangent - what version of UbuntuPPC did you install on your mini? I've tried a couple of different images but not had any luck getting a bootable system on my G5. The only other big-endian machine I have lying around is an UltraSPARC - maybe I should give up on PPC and fix the big-endian Linux build issues on that instead