I've built Python, but can't figure out how to package it for windows
Python 2.6.4 is built, and I found a bdist_wininst project and wininst-8 project. How do I manage to build the msi for this thing? -- http://mail.python.org/mailman/listinfo/python-list
Re: I've built Python, but can't figure out how to package it for windows
That was so simple, thanks. I scanned all the folders for inst, install, setup, but since msi was the expected output extension, I didn't see that! On Feb 9, 6:14 pm, "Gabriel Genellina" wrote: > En Tue, 09 Feb 2010 19:55:30 -0300, Mark Jones > escribió: > > > Python 2.6.4 is built, and I found a bdist_wininst project and > > wininst-8 project. > > > How do I manage to build the msi for this thing? > > See the Tools\msi directory; and look for some posts last year from Tim > Golden regarding some issues with the directory layout and other details. > > -- > Gabriel Genellina -- http://mail.python.org/mailman/listinfo/python-list
Re: I've built Python, but can't figure out how to package it for windows
Turns out there is an tools/msi directory and in there is python code
to help build the MSI from the tree you built. Only problem is you
can't use it without having python and PythonWin installed. So I
grabbed 2.6.4 python and pythonwin and installed them.
It uses COM objects and the CabSDK from MS to build the MSI file. And
then it has a couple of "Issues" that I had to resolve. First you need
a VS2008 shell so you can
nmake -f msisupport.mak
then you need to grab a copy of TIX (I didn't have to build it, just
have it in place fore the license.terms file (probably could have just
removed that list member for the same effect, but I was worried about
something else being needed down below)
("Tcl", "tcl8*", "license.terms"), ("Tk", "tk8*", "license.terms"),
("Tix", "Tix-*", "license.terms")):
had to be changed to:
("Tcl", "tcl-8*", "license.terms"), ("Tk", "tk-8*", "license.terms"),
("Tix", "Tix*", "license.terms")):
because the package names have evidently changed in the not to distant
past?
After that, I ran c:\python26\python msi.py and then it griped about
the python264.chm being missing, so instead of trying to build it, I
grabbed the one from the copy of python I had to install in order to
build python and dumped it in the expected location.
Oh yea, I also had to go to the PC directory and
nmake -f icons.mak
This gave me a runnable msi file to install python (which was already
installed, so that I could build the msi file to install my own
version). Oh well, at least it is built now. Whew!
--
http://mail.python.org/mailman/listinfo/python-list
os.stat and strange st_ctime
I've read the part about these being variable
Note
The exact meaning and resolution of the st_atime, st_mtime, andst_ctime members
depends on the operating system and the file system. For example, on Windows
systems using the FAT or FAT32 file systems, st_mtimehas 2-second
resolution, and st_atime has only 1-day resolution. See your operating
system documentation for details.
However, given the fact that NTFS supports 100ns resolution, the results
here just look wrong. st_ctime doesn't show a different time until there is
a ~15 second gap in the creation times (then however is shows the full time
gap) Notice
Linux doesn't do this, the times look normal. Seems strange to break cross
platform compatibility like this...
Here is the test code:
import os
import time
def runtest(sleepSecs):
print "Sleeping: " + str(sleepSecs)
fA = open("A", "w")
fA.close();
if sleepSecs:
time.sleep(sleepSecs)
fB = open("B", "w")
fB.close();
info = os.stat("A")
print "A:",info.st_ctime, info.st_mtime, info.st_atime
info = os.stat("B")
print "B:",info.st_ctime, info.st_mtime, info.st_atime
os.remove("A")
os.remove("B")
When run under Win7 on NTFS
for i in range(1,20):
runtest(i)
>>> for i in range(0,20):
... stattest.runtest(i)
...
Sleeping: 0
A: 1286770829.44 1286770829.44 1286770829.44
B: 1286770829.44 1286770829.44 1286770829.44
Sleeping: 1
A: 1286770829.44 1286770829.44 1286770829.44
B: 1286770829.44 1286770830.44 1286770830.44
Sleeping: 2
A: 1286770829.44 1286770830.45 1286770830.45
B: 1286770829.44 1286770832.45 1286770832.45
Sleeping: 3
A: 1286770829.44 1286770832.46 1286770832.46
B: 1286770829.44 1286770835.46 1286770835.46
Sleeping: 4
A: 1286770829.44 1286770835.47 1286770835.47
B: 1286770829.44 1286770839.47 1286770839.47
Sleeping: 5
A: 1286770829.44 1286770839.48 1286770839.48
B: 1286770829.44 1286770844.48 1286770844.48
Sleeping: 6
A: 1286770829.44 1286770844.49 1286770844.49
B: 1286770829.44 1286770850.49 1286770850.49
Sleeping: 7
A: 1286770829.44 1286770850.5 1286770850.5
B: 1286770829.44 1286770857.5 1286770857.5
Sleeping: 8
A: 1286770829.44 1286770857.51 1286770857.51
B: 1286770829.44 1286770865.51 1286770865.51
Sleeping: 9
A: 1286770829.44 1286770865.52 1286770865.52
B: 1286770829.44 1286770874.52 1286770874.52
Sleeping: 10
A: 1286770829.44 1286770874.52 1286770874.52
B: 1286770829.44 1286770884.53 1286770884.53
Sleeping: 11
A: 1286770829.44 1286770884.56 1286770884.56
B: 1286770829.44 1286770895.56 1286770895.56
Sleeping: 12
A: 1286770829.44 1286770895.57 1286770895.57
B: 1286770829.44 1286770907.57 1286770907.57
Sleeping: 13
A: 1286770829.44 1286770907.58 1286770907.58
B: 1286770829.44 1286770920.58 1286770920.58
Sleeping: 14
A: 1286770829.44 1286770920.59 1286770920.59
B: 1286770829.44 1286770934.59 1286770934.59
*Sleeping: 15*
*A: 1286770829.44 1286770934.6 1286770934.6*
*B: 1286770949.6 1286770949.6 1286770949.6*
Sleeping: 16
A: 1286770829.44 1286770949.61 1286770949.61
B: 1286770965.61 1286770965.61 1286770965.61
Sleeping: 17
A: 1286770829.44 1286770965.62 1286770965.62
B: 1286770982.62 1286770982.62 1286770982.62
Sleeping: 18
A: 1286770829.44 1286770982.63 1286770982.63
B: 1286771000.63 1286771000.63 1286771000.63
Sleeping: 19
A: 1286770829.44 1286771000.63 1286771000.63
B: 1286771019.63 1286771019.63 1286771019.63
--
http://mail.python.org/mailman/listinfo/python-list
