[9fans] python and mercurial

[9fans] python and mercurial

[ron minnich]

here is the status of building mercurial

cpu% python setup.py install
running install
HERE we are in init_plan9!
running build
running build_py
running build_ext
building 'mercurial.mpatch' extension
arg 0 is rc
arg 1 is -c
arg 2 is /bin/date
Fri Apr 13 22:04:03 PDT 2007

CMD ['8c', '-p', ' -T', '-I/sys/lib/python2.4/include', '-c', '-o',
'build/temp.unknown-2.4/mpatch.8', 'mercurial/mpatch.c']

cpp: mercurial/mpatch.c:45 Could not find include file <arpa/inet.h>
cpp: mercurial/mpatch.c:46 Could not find include file <inttypes.h>
arg 0 is rc
arg 1 is -c
arg 2 is 8c -p  -T -I/sys/lib/python2.4/include -c -o
build/temp.unknown-2.4/mpatch.8 mercurial/mpatch.c
/usr/bootes/src/mercurial-0.9.3/mercurial/mpatch.c:236[mercurial/mpatch.c:8936]
syntax error, last name: uint32_t
/usr/bootes/src/mercurial-0.9.3/mercurial/mpatch.c:237[mercurial/mpatch.c:8937]
syntax error, last name: uint32_t
/usr/bootes/src/mercurial-0.9.3/mercurial/mpatch.c:238[mercurial/mpatch.c:8938]
syntax error, last name: uint32_t
/usr/bootes/src/mercurial-0.9.3/mercurial/mpatch.c:380[mercurial/mpatch.c:9080]
syntax error, last name: uint32_t
/usr/bootes/src/mercurial-0.9.3/mercurial/mpatch.c:381[mercurial/mpatch.c:9081]
syntax error, last name: uint32_t
/usr/bootes/src/mercurial-0.9.3/mercurial/mpatch.c:382[mercurial/mpatch.c:9082]
syntax error, last name: uint32_t

arg 0 is rc
arg 1 is -c
arg 2 is /bin/date
Fri Apr 13 22:04:05 PDT 2007

CMD ['8l', '-u', '-o', 'build/lib.unknown-2.4/mercurial/mpatch.d',
'build/temp.unknown-2.4/mpatch.8']

arg 0 is rc
arg 1 is -c
arg 2 is 8l -u -o build/lib.unknown-2.4/mercurial/mpatch.d
build/temp.unknown-2.4/mpatch.8
??none??: cannot open file: build/temp.unknown-2.4/mpatch.8

building 'mercurial.bdiff' extension
arg 0 is rc
arg 1 is -c
arg 2 is /bin/date
Fri Apr 13 22:04:06 PDT 2007

CMD ['8c', '-p', ' -T', '-I/sys/lib/python2.4/include', '-c', '-o',
'build/temp.unknown-2.4/bdiff.8', 'mercurial/bdiff.c']

cpp: mercurial/bdiff.c:36 Could not find include file <arpa/inet.h>
cpp: mercurial/bdiff.c:37 Could not find include file <inttypes.h>
arg 0 is rc
arg 1 is -c
arg 2 is 8c -p  -T -I/sys/lib/python2.4/include -c -o
build/temp.unknown-2.4/bdiff.8 mercurial/bdiff.c
/usr/bootes/src/mercurial-0.9.3/mercurial/bdiff.c:55[mercurial/bdiff.c:8769]
not a function
/usr/bootes/src/mercurial-0.9.3/mercurial/bdiff.c:55[mercurial/bdiff.c:8769]
syntax error, last name: rol32

arg 0 is rc
arg 1 is -c
arg 2 is /bin/date
Fri Apr 13 22:04:06 PDT 2007

CMD ['8l', '-u', '-o', 'build/lib.unknown-2.4/mercurial/bdiff.d',
'build/temp.unknown-2.4/bdiff.8']

arg 0 is rc
arg 1 is -c
arg 2 is 8l -u -o build/lib.unknown-2.4/mercurial/bdiff.d
build/temp.unknown-2.4/bdiff.8
??none??: cannot open file: build/temp.unknown-2.4/bdiff.8

building 'mercurial.base85' extension
arg 0 is rc
arg 1 is -c
arg 2 is /bin/date
Fri Apr 13 22:04:06 PDT 2007

CMD ['8c', '-p', ' -T', '-I/sys/lib/python2.4/include', '-c', '-o',
'build/temp.unknown-2.4/base85.8', 'mercurial/base85.c']

arg 0 is rc
arg 1 is -c
arg 2 is 8c -p  -T -I/sys/lib/python2.4/include -c -o
build/temp.unknown-2.4/base85.8 mercurial/base85.c

arg 0 is rc
arg 1 is -c
arg 2 is /bin/date
Fri Apr 13 22:04:07 PDT 2007

CMD ['8l', '-u', '-o', 'build/lib.unknown-2.4/mercurial/base85.d',
'build/temp.unknown-2.4/base85.8']

arg 0 is rc
arg 1 is -c
arg 2 is 8l -u -o build/lib.unknown-2.4/mercurial/base85.d
build/temp.unknown-2.4/base85.8

running build_scripts
creating build/scripts-2.4
copying and adjusting hg -> build/scripts-2.4
Traceback (most recent call last):
  File "setup.py", line 102, in ?
    welcome='contrib/macosx/Welcome.html')),
  File "/sys/lib/python2.4/distutils/core.py", line 149, in setup
    dist.run_commands()
  File "/sys/lib/python2.4/distutils/dist.py", line 946, in run_commands
    self.run_command(cmd)
  File "/sys/lib/python2.4/distutils/dist.py", line 966, in run_command
    cmd_obj.run()
  File "/sys/lib/python2.4/distutils/command/install.py", line 513, in run
    self.run_command('build')
  File "/sys/lib/python2.4/distutils/cmd.py", line 333, in run_command
    self.distribution.run_command(command)
  File "/sys/lib/python2.4/distutils/dist.py", line 966, in run_command
    cmd_obj.run()
  File "/sys/lib/python2.4/distutils/command/build.py", line 112, in run
    self.run_command(cmd_name)
  File "/sys/lib/python2.4/distutils/cmd.py", line 333, in run_command
    self.distribution.run_command(command)
  File "/sys/lib/python2.4/distutils/dist.py", line 966, in run_command
    cmd_obj.run()
  File "/sys/lib/python2.4/distutils/command/build_scripts.py", line 53, in run
    self.copy_scripts()
  File "/sys/lib/python2.4/distutils/command/build_scripts.py", line
107, in copy_scripts
    "python" + sysconfig.get_config_var("EXE")),
TypeError: cannot concatenate 'str' and 'NoneType' objects
'/env/PYTHONINSPECT' file does not exist

Those files trying to include the arpa/inet.h and inttype.h are doing this:
#else
/* not windows */
# include <sys/types.h>
# include <arpa/inet.h>
# include <inttypes.h>
#endif

i.e. you ain't windows, you must be something that has this stuff.

Silliness ...

This is mercurial-0.9.3.

I will beat on this for a few minutes, would be interesting if it
actually worked ...

ron
p.s. but we sure could use ssh2, hint hint.

Re: [9fans] python and mercurial

[Federico Benavento]

I've always wondered why python's port doesn't use ape...

-- 
Federico G. Benavento

Re: [9fans] python and mercurial

[Russ Cox]

> I've always wondered why python's port doesn't use ape...

It doesn't need to use APE.  Python doesn't require Unix.
(It runs just fine on Win32, for example.)

Russ

Re: [9fans] python and mercurial

[ron minnich]

> > I've always wondered why python's port doesn't use ape...


Why not Ape? Because we need to get non-Plan 9 project thinking about
the Plan 9 universe. It's nice to start getting these native ports
working and, most important, planning to get those ports back into the
mainstream. We need to get people outside the Plan 9 world to start
thinking, e.g.,  about a world where everything is not an int (i.e.
signals). I can imagine a python expert finally getting the niceness
of notes, and then wondering why Linux doesn't do it that way. It gets
interesting when we start providing extremely powerful Python modules
that use capabilities only available in Plan 9. You want people to
start getting frustrated with old, primitive OSes :-)

thanks

ron

Re: [9fans] python and mercurial

[W B Hacker]

Russ Cox wrote:
>> I've always wondered why python's port doesn't use ape...
> 
> It doesn't need to use APE.  Python doesn't require Unix.
> (It runs just fine on Win32, for example.)
> 
> Russ
> 
> 
.. and several other non-*n*x OS'en...

(though one could, from time-to-time, argue the '..just fine..' point)

;-)

Bill

Re: [9fans] python and mercurial

[Federico Benavento]

> It doesn't need to use APE.  Python doesn't require Unix.
> (It runs just fine on Win32, for example.)
>

yes, but what happens when you need modules that depend on
stuff that hasn't been ported to Plan 9 natively?
I have had problems with apps that needed the select module
and also required that the socket one to support ssl (which needs
openssl).

I guess we should start writing emulation wrappers or whatever
they're called for the stuff we don't have

-- 
Federico G. Benavento

Re: [9fans] python and mercurial

[ron minnich]

On 4/14/07, Federico Benavento <benavento@gmail.com> wrote:

> I guess we should start writing emulation wrappers or whatever
> they're called for the stuff we don't have

only if you are writing them with the intent of jamming them back into
Python mainstream. Fix Python. Plan 9 isn't broken, after all ...

Python has all kinds of guck in the modules, to try to figure out what
it has and what it hasn't. So for select support, you would want to
fix that for the plan 9 case, in the usual manner (whatever that might
be ...)

It's ugly, I know, but there are not many good choices here, and if we
get Python, we get a lot. That's not saying I like Python, just that
we need it. Backward compatibility is hell, but we have it easy. Just
talk to an Intel architect sometime about backward compatibility.

ron