I put a new nl-xml.c in http://newlisp.org/downloads/WinCE/ which might solve the xml-parse problem on the PPC (and Sparc). The newlisp.exe is also redone with the new nl-xml.c
Lutz
pocket newLISP?
I have asked Luis to test the new EXE.
I have started to do some research for PPC-hardware.
Since the new VGA-based PPC are on the horizon, it
maybe worth to wait the time they get into the markets.
Luis and Alexander gave some advice about their hardware
(Dell X30 and imate Pocket).
As always get as much speed and memory as possible! ;-)
Dream-PPC: VGA/624MHZ/128MB RAM
Pocket Windows 2003 SecondEdition
I have started to do some research for PPC-hardware.
Since the new VGA-based PPC are on the horizon, it
maybe worth to wait the time they get into the markets.
Luis and Alexander gave some advice about their hardware
(Dell X30 and imate Pocket).
As always get as much speed and memory as possible! ;-)
Dream-PPC: VGA/624MHZ/128MB RAM
Pocket Windows 2003 SecondEdition
Hans-Peter
So xml-parse is working and all other problems have their explanation.From: "Luis Carvalho (Kozure)" <kozure@c...>
Date: Tue Aug 24, 2004 4:44 pm
Subject: Re: newLISP compile possible?
Hallo,
Sorry again for the offtopic...
> Lutz get his compiler to work and after I had also problems
> with xml-parse on Solaris Sparc, he thinks that he found the problem.
> He compiles the new EXE and post all changed sources here:
>
> http://newlisp.org/downloads/WinCE/
>
> Can you test the new EXE with the qa file and give info about results.
> Lutz also registered your and Alexander's email to get always through
> his mail-filter.
Everything works right now. I'll tried to write to you and Lutz again,
with no success. Please tell him that qa has only minor problems:
- test-exec and test-process, simply because windows mobile does not
contain native _ls_ and _cat_.
- test-pipe, probably because pipe is not properly ported yet. Check the
compilation flag I put in nl-sock.c regarding this.
xml-parse works now.
Cheers,
K, I think spam is a big problem in Deutschland... :)
--
Life is the childhood of our immortality.
-- Goethe
--
Luis Carvalho (Kozure)
kozure at cox dot net
PGP key: A616822D
So no more experimental?
;-) Still no PPC-Hardware.
No idea with the mail-problem.
I have no personel mail-filter. (Maybe t-online does some)
Hans-Peter
Thanks again to Luis. I am glad 'xml-parse' now works on both, Sparc and PPC, probably it was a byte-order problem.
I am currently working on a platform independent QA file, which also gives more diagnostics on failure and keeps on testing although a function has failed. I have got it more or less working for Win32 and Linux, but still have problems on BSD. Its mainly test routines where 'process' or 'exec' are involved, where the timing and handshaking is hard to get right.
The next qa by default will not test pipe, process, exec and udp functions but all other net-working functions (they are not tested currently).
Supplying an extra option it will also test pipe, process, exec and udp. I can test udp functions only when I have 'process' to start a second newlisp session.
I will mark the PPC version as non-experimental in the next version and document what should not be expected to work. It will also get it,s OS Id as 7 in 'sys-info'.
Lutz
I am currently working on a platform independent QA file, which also gives more diagnostics on failure and keeps on testing although a function has failed. I have got it more or less working for Win32 and Linux, but still have problems on BSD. Its mainly test routines where 'process' or 'exec' are involved, where the timing and handshaking is hard to get right.
The next qa by default will not test pipe, process, exec and udp functions but all other net-working functions (they are not tested currently).
Supplying an extra option it will also test pipe, process, exec and udp. I can test udp functions only when I have 'process' to start a second newlisp session.
I will mark the PPC version as non-experimental in the next version and document what should not be expected to work. It will also get it,s OS Id as 7 in 'sys-info'.
Lutz