pyopenssl-list — General discussion list for users and developers of pyOpenSSL

[pyOpenSSL] memory leak?

[Zoltan F. <zol...@us...>]

hello list,

i have a problem with pyOpenSSL, as follows:
if i create a network client, i must explicitly delete the context of it 
and the client itself to free the memory it is using. ie. a class with 
the ctx = SSL.Context, conn = SSL.Connection attributes is not removed 
from memory, if i do not delete the conn.ctx and the conn attributes. i 
was testing it with a simple loop that just created a thousand 
connection objects without connecting them to anywhere, and it used up 
about 20 megabytes of memory. besides that, there is another problem i 
could not work around, which arises at the server side of an SSL 
connection. using the simple server in the examples directory of the 
package, a thousand clients connecting, sending some data and then 
disconnecting also made the server to use significantly more memory than 
the amount it was using at the startup. is it a problem with me doing 
something evil, or is it a problem in pyOpenSSL or even in OpenSSL? in 
the first case what the heck am i doing wrong? the environment is Red 
Hat linux 8.0, OpenSSL 0.9.6h, pyOpenSSL 0.5.1, Python 2.2.2.

thanks, zoltan

ps: sorry for the poor english
ps2: if i was not clear/understandable, or the subject needs more 
testing/whatever, i have the time and willingness...

[pyOpenSSL] PyOpenSSL bug

[Kevin L. <kli...@mk...>]

Hello, I've been using pyopenssl with python's BaseHTTPServer and have
stumbled on to racy bug.

The server sometimes blocks indefinetly during the SSL handshake during
a read(). It appears todo this when there are reverse dns issues with
the client ip or dns server. Maybe if the dns server is slow to respond...
Basically everything is well and speedy when I added my client's IP to 
/etc/hosts on the server. (linux 2.4.19, debian woody). 

After alot of debugging it seems the following is happening (I'm not very
familiar with SSL or pyOpenSSL internals) so here is the high level list
of events:

1) Browser Connects to server
2) Server does handshake
3) Server blocks on read()
4) Browser waits for response from server indefinetly

This only happens with Mozilla based browsers, IE and Opera timeout after
5 seconds and will resend the connection request, Mozilla doesn't so will
spin forever.

It is a bit racy and only occurs when the server is serving 10 or so images
at once and I hit reload multiple times.

I've tested it without threads and the problem still exists. Its the only
issue I can see with pyOpenSSL, other than that its been excellent.

If anyone has any ideas or would like more info on debugging this thing
let me know.

Kevin-

---------------------------------------------------
Kevin Lindsay
Debian Developer
Fingerprint: 81E 58A3 B49A 580E EE3D 8CF0 519A 55F0 746C 51F4
Key Id:      746C51F4

[pyOpenSSL] WIN a FREE Wheelbarrow screener and make your own topsoil!

[originalwheelbarrowscreener.com <the...@ao...>]

<DIV class=Section1>
<P class=MsoNormal><B style="mso-bidi-font-weight: normal"><I style="mso-bidi-font-style: normal"><SPAN style="FONT-SIZE: 14pt; COLOR: red">WIN a FREE Wheelbarrow screener and make your own topsoil!<?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" /><o:p></o:p></SPAN></I></B></P>
<P class=MsoNormal><B style="mso-bidi-font-weight: normal"><I style="mso-bidi-font-style: normal"><SPAN style="FONT-SIZE: 14pt; COLOR: red">It’s easy!<SPAN style="mso-spacerun: yes">&nbsp; </SPAN>No purchase necessary, all you have to do is visit<SPAN style="mso-spacerun: yes">&nbsp;&nbsp; </SPAN><A href="http://www.originalwheelbarrowscreener.com/">www.originalwheelbarrowscreener.com</A> and register to win!<o:p></o:p></SPAN></I></B></P>
<P class=MsoNormal><B style="mso-bidi-font-weight: normal"><I style="mso-bidi-font-style: normal"><SPAN style="FONT-SIZE: 14pt; COLOR: red">We will post winners on the site look for your name.<SPAN style="mso-spacerun: yes">&nbsp; </SPAN>This product is perfect for the home gardener!<o:p></o:p></SPAN></I></B></P>
<P class=MsoNormal><SPAN style="FONT-SIZE: 10pt; FONT-FAMILY: Arial">Innovative new gardening tool will help you screen your own topsoil, compost, mulch. This can give you your own unlimited supply of topsoil. Comes with universal clips that fit all wheelbarrows. Made with easy grip handles and a shovel holder. Constructed with a steel screen and plastic outlay. This makes it easy to screen your own topsoil. Place Wheelbarrow Screener on wheelbarrow. Dig the soil you want to use. Dump soil onto Wheelbarrow Screener. Scrape dirt through screener. Remove roots, rocks, and other debris. You are left with clean screened topsoil in your wheelbarrow to place in your garden or into pots. To be removed from this e-mail list please reply to this e-mail and say remove in the subject line.<SPAN style="mso-spacerun: yes">&nbsp; </SPAN>Be sure to visit<SPAN style="mso-spacerun: yes">&nbsp;&nbsp; </SPAN><A href="http://www.originalwheelbarrowscreener.com/">www.originalwheelbarrowscreener.com</A><o:p></o:p></SPAN></P>
<P class=MsoNormal><SPAN style="FONT-SIZE: 10pt; FONT-FAMILY: Arial"><BR clear=all><?xml:namespace prefix = v ns = "urn:schemas-microsoft-com:vml" /><v:shapetype id=_x0000_t75 coordsize = "21600,21600" o:preferrelative = "t" o:spt = "75" filled = "f" stroked = "f" path = " m@4@5 l@4@11@9@11@9@5 xe"><v:stroke joinstyle = "miter"></v:stroke><v:formulas><v:f eqn = "if lineDrawn pixelLineWidth 0 "></v:f><v:f eqn = "sum @0 1 0 "></v:f><v:f eqn = "sum 0 0 @1 "></v:f><v:f eqn = "prod @2 1 2 "></v:f><v:f eqn = "prod @3 21600 pixelWidth "></v:f><v:f eqn = "prod @3 21600 pixelHeight "></v:f><v:f eqn = "sum @0 0 1 "></v:f><v:f eqn = "prod @6 1 2 "></v:f><v:f eqn = "prod @7 21600 pixelWidth "></v:f><v:f eqn = "sum @8 21600 0 "></v:f><v:f eqn = "prod @7 21600 pixelHeight "></v:f><v:f eqn = "sum @10 21600 0 "></v:f></v:formulas><v:path o:extrusionok = "f" gradientshapeok = "t" o:connecttype = "rect"></v:path><o:lock aspectratio="t" v:ext="edit"></o:lock></v:shapetype><v:shape id=_x0000_i1025 style="WIDTH: 252pt; HEIGHT: 189pt" type = "#_x0000_t75" coordsize = "21600,21600"><v:imagedata o:title="girlwithscreener" src = "WIN%20a%20FREE%20Wheelbarrow%20screener%20and%20make%20your%20own%20topsoil_files/image001.png"></v:imagedata></v:shape><o:p></o:p></SPAN></P>
<FORM>
<P class=MsoNormal><SPAN style="FONT-SIZE: 10pt; FONT-FAMILY: Arial"><o:p>&nbsp;</o:p></SPAN></P></FORM>
<P class=MsoNormal><B style="mso-bidi-font-weight: normal"><I style="mso-bidi-font-style: normal"><SPAN style="FONT-SIZE: 14pt; COLOR: red"><o:p>&nbsp;</o:p></SPAN></I></B></P>
<P class=MsoNormal><B style="mso-bidi-font-weight: normal"><I style="mso-bidi-font-style: normal"><SPAN style="FONT-SIZE: 14pt; COLOR: red"><o:p>&nbsp;</o:p></SPAN></I></B></P></DIV>

[pyOpenSSL] Win32

[Dave A. <da...@im...>]

The new SPIKE Proxy for win32 contains an entire Python distro with
binaries of the latest OpenSSL and PyOpenSSL. If you want to quickly
test someone on windows, you can download it (it's 8 megs) and it should
work. :>
-dave

Re: [pyOpenSSL] pyOpenSSL for win32

[Martin <md...@md...>]

tis 2002-10-08 klockan 15.48 skrev Dave Aitel:
> I'm trying to create a standalone release of SPIKE Proxy, which relies
> on PyOpenSSL, but currently am failing to compile pyOpenSSL with mingw -
> is there an environment variable I can setup that sets the compiler?

I have no idea what distutils's inner black magic does, but I saw there
was a thread on python-dev about mingw:

http://mail.python.org/pipermail/python-dev/2002-October/029178.html

(I don't have any experience with mingw so I don't know if it's helpful
or not)


/Martin

[pyOpenSSL] pyOpenSSL for win32

[Dave A. <da...@im...>]

I'm trying to create a standalone release of SPIKE Proxy, which relies
on PyOpenSSL, but currently am failing to compile pyOpenSSL with mingw -
is there an environment variable I can setup that sets the compiler?

(or a precompiled win32 binary I can shove somewhere?)

-dave
=20

Re: [pyOpenSSL] Hmm...

[Martin <md...@md...>]

m=C3=A5n 2002-10-07 klockan 23.58 skrev Dave Aitel:
> has anyone compiled this on OS X?

Yes, I have. But I had openssl-dev installed from fink (that's why
setup.py explicitly adds '/sw/include' and '/sw/lib' when on darwin). I
don't remember if there's a OSX machine in the sourceforge compile farm
(but it seems to be down at the moment anyway).

I don't have a mac myself, so it's kinda hard to give an estimation on
when I'll be able to test it...


/Martin

[pyOpenSSL] Hmm...

[Dave A. <da...@im...>]

has anyone compiled this on OS X?

I got this from a user:


funny thing, maybe it is just me -but I am having quite a time to get
the PySSL stuff to compile on Mac 10.2.1...
Have you heard of any success stories?

Dennis


-dave

Re: [pyOpenSSL] PyEval_RestoreThread: NULL tstate

[Dave A. <da...@im...>]

It's no biggie for me. I'm not having any problems with
OpenSSL.SSL.Connection, despite being multi-threaded. I can wait as long
as it takes.
-dave


On Mon, 2002-10-07 at 13:30, Martin Sj=F6gren wrote:
> m=E5n 2002-10-07 klockan 12.28 skrev Dave Aitel:
> > can we just fix tsafe for .2? I'll force all my users to upgrade. :>
> > -dave
>=20
> Yeah I suppose so. Are you in a hurry? I'm swamped with work at the
> university (trying to study full time AND teach).
>=20
>=20
> /Martin

Re: [pyOpenSSL] PyEval_RestoreThread: NULL tstate

[Martin <md...@md...>]

m=C3=A5n 2002-10-07 klockan 12.28 skrev Dave Aitel:
> can we just fix tsafe for .2? I'll force all my users to upgrade. :>
> -dave

Yeah I suppose so. Are you in a hurry? I'm swamped with work at the
university (trying to study full time AND teach).


/Martin

Re: [pyOpenSSL] PyEval_RestoreThread: NULL tstate

[Dave A. <da...@im...>]

can we just fix tsafe for .2? I'll force all my users to upgrade. :>
-dave


On Sun, 2002-10-06 at 15:49, Martin Sj=F6gren wrote:
> s=F6n 2002-10-06 klockan 21.29 skrev Dave Aitel:
> > Does a tsafe connection not support this?
>=20
> The reason for this is that the tsafe.Connection hasn't been updated
> when the SSL.Connection has been. :/ I blame my bad memory. :-) If you
> check the code in tsafe.py you'll see it's easy to fix ;) but I think it
> might be a good idea to use a getattr descriptor instead of this ugly
> hack. If anybody have suggestions I'll gladly hear them.
>=20
>=20
> /Martin

Re: [pyOpenSSL] PyEval_RestoreThread: NULL tstate

[Martin <md...@md...>]

s=C3=B6n 2002-10-06 klockan 21.29 skrev Dave Aitel:
> Does a tsafe connection not support this?

The reason for this is that the tsafe.Connection hasn't been updated
when the SSL.Connection has been. :/ I blame my bad memory. :-) If you
check the code in tsafe.py you'll see it's easy to fix ;) but I think it
might be a good idea to use a getattr descriptor instead of this ugly
hack. If anybody have suggestions I'll gladly hear them.


/Martin

Re: [pyOpenSSL] PyEval_RestoreThread: NULL tstate

[Dave A. <da...@im...>]

Does a tsafe connection not support this?
-dave

Exception in thread Thread-1:
Traceback (most recent call last):
  File "/usr/lib/python2.2/threading.py", line 408, in __bootstrap
    self.run()
  File "./spkproxy.py", line 700, in run
    self.connection.startSSLserver()
  File "./spkproxy.py", line 98, in startSSLserver
    self.mysocket.set_accept_state()
AttributeError: Connection instance has no attribute 'set_accept_state'

On Sun, 2002-10-06 at 14:22, Martin Sj=F6gren wrote:
> s=F6n 2002-10-06 klockan 19.28 skrev Dave Aitel:
> > On Sun, 2002-10-06 at 06:46, Martin Sj=F6gren wrote:
> >=20
> > > If you're running multithreaded, you'd do well to use
> > > OpenSSL.tsafe.Connection instead of OpenSSL.SSL.Connection (tsafe sta=
nds
> > > for thread safe :))
> >=20
> > hmm. How does one do this exactly?
> > -dave
> >=20
> >=20
> >         self.mysocket =3D OpenSSL.tsafe.Connection(ctx, self.mysocket)
> >     self.mysocket =3D OpenSSL.tsafe.Connection(ctx, self.mysocket)
> > AttributeError: 'module' object has no attribute 'tsafe'
>=20
> D'oh! I must have forgotten to import tsafe from the __init__.py file.
> If you do import OpenSSL.tsafe, or from OpenSSL import tsafe, it works.
>=20
>=20
> /Martin

Re: [pyOpenSSL] PyEval_RestoreThread: NULL tstate

[Martin <md...@md...>]

s=C3=B6n 2002-10-06 klockan 19.28 skrev Dave Aitel:
> On Sun, 2002-10-06 at 06:46, Martin Sj=C3=B6gren wrote:
>=20
> > If you're running multithreaded, you'd do well to use
> > OpenSSL.tsafe.Connection instead of OpenSSL.SSL.Connection (tsafe stand=
s
> > for thread safe :))
>=20
> hmm. How does one do this exactly?
> -dave
>=20
>=20
>         self.mysocket =3D OpenSSL.tsafe.Connection(ctx, self.mysocket)
>     self.mysocket =3D OpenSSL.tsafe.Connection(ctx, self.mysocket)
> AttributeError: 'module' object has no attribute 'tsafe'

D'oh! I must have forgotten to import tsafe from the __init__.py file.
If you do import OpenSSL.tsafe, or from OpenSSL import tsafe, it works.


/Martin

Re: [pyOpenSSL] PyEval_RestoreThread: NULL tstate

[Dave A. <da...@im...>]

On Sun, 2002-10-06 at 06:46, Martin Sj=F6gren wrote:

> If you're running multithreaded, you'd do well to use
> OpenSSL.tsafe.Connection instead of OpenSSL.SSL.Connection (tsafe stands
> for thread safe :))
>=20
>=20
> Regards,
> Martin

hmm. How does one do this exactly?
-dave


        self.mysocket =3D OpenSSL.tsafe.Connection(ctx, self.mysocket)


Connected to by ('127.0.0.1', 36388)
Exception in thread Thread-1:
Traceback (most recent call last):
  File "/usr/lib/python2.2/threading.py", line 408, in __bootstrap
    self.run()
  File "./spkproxy.py", line 699, in run
    self.connection.startSSLserver()
  File "./spkproxy.py", line 94, in startSSLserver
    self.mysocket =3D OpenSSL.tsafe.Connection(ctx, self.mysocket)
AttributeError: 'module' object has no attribute 'tsafe'

Re: [pyOpenSSL] PyEval_RestoreThread: NULL tstate

[Martin <md...@md...>]

s=C3=B6n 2002-10-06 klockan 12.20 skrev Zoltan Felleg:
> hello list,
>=20
> i have a problem with pyOpenSSL, namely as follows:
> i have created a client/server application, where the clients=20
> communicate with the servers on a secure channel. When a client tries to=20
> log in to a server (ie. after the SSL handshake it tries to send a=20
> message), the client dies with the subject, and the server gets a=20
> "connection reset by peer" exception. i have two questions about this:
> a.) has anyone seen this before?
> b.) is there an "official" way of handling the WantXYZError exceptions=20
> besides ignoring them?

The WantXYZ exceptions are tricky, but that's because it's tricky in
OpenSSL! If you get WantReadError that means OpenSSL wants to read from
the socket but couldn't, so after you've made sure that the socket is
readable, you should call *the same method* again, with *the same
arguments*. It's symmetrical for WantWriteError.

So, if for example you get a WantReadError when you do ssl.write('foo')
you have to wait (using e.g. select) until the socket corresponding to
`ssl' is readable, and then call ssl.write('foo') again. Yes, you can
get WantReadErrors on writing, and WantWriteErrors on reading, since
OpenSSL does handshakes transparently.

> ps: both the client and server are multithreaded, the SSL connection is=20
> nonblocking, and the pyOpenSSL version is 0.5.1

If you're running multithreaded, you'd do well to use
OpenSSL.tsafe.Connection instead of OpenSSL.SSL.Connection (tsafe stands
for thread safe :))


Regards,
Martin

[pyOpenSSL] PyEval_RestoreThread: NULL tstate

[Zoltan F. <zol...@us...>]

hello list,

i have a problem with pyOpenSSL, namely as follows:
i have created a client/server application, where the clients 
communicate with the servers on a secure channel. When a client tries to 
log in to a server (ie. after the SSL handshake it tries to send a 
message), the client dies with the subject, and the server gets a 
"connection reset by peer" exception. i have two questions about this:
a.) has anyone seen this before?
b.) is there an "official" way of handling the WantXYZError exceptions 
besides ignoring them?
thanks a lot, and sorry for the poor english.

zoltan

ps: both the client and server are multithreaded, the SSL connection is 
nonblocking, and the pyOpenSSL version is 0.5.1

[pyOpenSSL] pyOpenSSL 0.5.1 released

[Martin <md...@md...>]

The bugfix-release, 0.5.1, of pyOpenSSL is out. This version fixes the
python 1.5 and 2.1 breakage ;) and also introduces some basic PKCS12
support.

The release is on SourceForge:
http://sf.net/project/showfiles.php?group_id=3D31249&release_id=3D111204


Regards,
Martin

[pyOpenSSL] 0.5.1 RC 2

[Martin <md...@md...>]

Since there has been quite a few fixes since 0.5.1rc1, I thought it best
to err on the safe side, so here's RC2 :)

Again, it would be very nice if you guys could have a look at it and
tell me what's broken so I can fix it.

Get it from CVS (:pserver:ano...@cv...:/u/cvs/pyopenssl) or
the tarball here:
http://www.strakt.com/~martin/pyOpenSSL-0.5.1rc2.tar.gz

Some of the changes since RC1:

* Explicit type checks in crypto.dump_* and many other places (thanks
Iko!)
* Fixed includes for AIX (I hope? I have no AIX to test on :)
* Added SecureXMLRPCServer example (thanks Michal!)
* Fix sendall() method. It segfaulted because it was too generous about
giving away the GIL.
* Fixed build requirements for rpms.


Thanks,
Martin

[pyOpenSSL] 0.5.1 RC 1

[Martin <md...@md...>]

Taking the advice of people smarter than me, I've decided to do an RC
before releasing 0.5.1 ;)

I would really appreciate it if people had a look at it, poked it, et.c.
so we can find any stupid mistakes before I make the actual release.

You can check the CVS or you can get a tarball here:
http://www.strakt.com/~martin/pyOpenSSL-0.5.1rc1.tar.gz.

Thanks!


Regards,
Martin

Re: [pyOpenSSL] CVS Access

[Mihai I. <mi...@re...>]

On 17 Aug 2002, Martin Sj=F6gren wrote:

> fre 2002-08-16 klockan 22.43 skrev Mihai Ibanescu:
> > On 16 Aug 2002, Martin Sj=C3=B6gren wrote:
> >=20
> > > fre 2002-08-16 klockan 20.09 skrev Mihai Ibanescu:
> > > > It looks like pyOpenSSL does not run on 1.5.2 anymore. From a qui=
ck check:
> > > >=20
> > > > PySequence_Fast (can be worked around with PySequence_Tuple) - in=
=20
> > > > src/crypto/x509req.c
> > > > PyObject_AsFileDescriptor (not in 1.5.2) - in src/ssl/connection.=
c
> > > >=20
> > > > I know, 1.5.2 is old :-)
> > >=20
> > > I *meant* to support 1.5. And I noticed this too, but too late. It'=
s
> > > fixed in current CVS, and will be part of the 0.5.1 "heh, oops" rel=
ease
> > > ;)
> > >=20
> > > This is a case of bad testing, and is really a wake-up call that I
> > > should do an RC before a real release next time ;)
> >=20
> > I'd be happy to QA the CVS version. Probably next week.
> > Any idea about the 0.5.1 timeframe? :-) Sort of too early to ask.
>=20
> Thanks for the offer!
>=20
> If you can QA it next week, "after that" is the best estimate I can giv=
e
> right now ;) I've got a patch for PKCS12 support from Mark Welch that
> didn't make it into 0.5 that I want to put in 0.5.1 too, so I will try
> to do that asap. Hmm, if I can release it next weekend, that'd be good,
> it's kinda sloppy to have a broken version out for too long. :/
>=20
> Seeing as the freshmen come to the university on Monday, my next two
> weeks will probably be spent in a happy beer-haze *cough* but I will tr=
y
> to get *some* work done :)

Yesterday's CVS snapshot works fine on 1.5.2, AFAICT.

Thanks,
Misa

Re: [pyOpenSSL] CVS Access

[Martin <md...@md...>]

fre 2002-08-16 klockan 22.43 skrev Mihai Ibanescu:
> On 16 Aug 2002, Martin Sj=C3=B6gren wrote:
>=20
> > fre 2002-08-16 klockan 20.09 skrev Mihai Ibanescu:
> > > It looks like pyOpenSSL does not run on 1.5.2 anymore. From a quick c=
heck:
> > >=20
> > > PySequence_Fast (can be worked around with PySequence_Tuple) - in=20
> > > src/crypto/x509req.c
> > > PyObject_AsFileDescriptor (not in 1.5.2) - in src/ssl/connection.c
> > >=20
> > > I know, 1.5.2 is old :-)
> >=20
> > I *meant* to support 1.5. And I noticed this too, but too late. It's
> > fixed in current CVS, and will be part of the 0.5.1 "heh, oops" release
> > ;)
> >=20
> > This is a case of bad testing, and is really a wake-up call that I
> > should do an RC before a real release next time ;)
>=20
> I'd be happy to QA the CVS version. Probably next week.
> Any idea about the 0.5.1 timeframe? :-) Sort of too early to ask.

Thanks for the offer!

If you can QA it next week, "after that" is the best estimate I can give
right now ;) I've got a patch for PKCS12 support from Mark Welch that
didn't make it into 0.5 that I want to put in 0.5.1 too, so I will try
to do that asap. Hmm, if I can release it next weekend, that'd be good,
it's kinda sloppy to have a broken version out for too long. :/

Seeing as the freshmen come to the university on Monday, my next two
weeks will probably be spent in a happy beer-haze *cough* but I will try
to get *some* work done :)


Regards,
Martin

Re: [pyOpenSSL] CVS Access

[Mihai I. <mi...@re...>]

On 16 Aug 2002, Martin Sj=F6gren wrote:

> fre 2002-08-16 klockan 20.09 skrev Mihai Ibanescu:
> > It looks like pyOpenSSL does not run on 1.5.2 anymore. From a quick c=
heck:
> >=20
> > PySequence_Fast (can be worked around with PySequence_Tuple) - in=20
> > src/crypto/x509req.c
> > PyObject_AsFileDescriptor (not in 1.5.2) - in src/ssl/connection.c
> >=20
> > I know, 1.5.2 is old :-)
>=20
> I *meant* to support 1.5. And I noticed this too, but too late. It's
> fixed in current CVS, and will be part of the 0.5.1 "heh, oops" release
> ;)
>=20
> This is a case of bad testing, and is really a wake-up call that I
> should do an RC before a real release next time ;)

I'd be happy to QA the CVS version. Probably next week.
Any idea about the 0.5.1 timeframe? :-) Sort of too early to ask.

Misa

Re: [pyOpenSSL] CVS Access

[Martin <md...@md...>]

fre 2002-08-16 klockan 20.09 skrev Mihai Ibanescu:
> It looks like pyOpenSSL does not run on 1.5.2 anymore. From a quick check:
> 
> PySequence_Fast (can be worked around with PySequence_Tuple) - in 
> src/crypto/x509req.c
> PyObject_AsFileDescriptor (not in 1.5.2) - in src/ssl/connection.c
> 
> I know, 1.5.2 is old :-)

I *meant* to support 1.5. And I noticed this too, but too late. It's
fixed in current CVS, and will be part of the 0.5.1 "heh, oops" release
;)

This is a case of bad testing, and is really a wake-up call that I
should do an RC before a real release next time ;)


Regards,
Martin

Re: [pyOpenSSL] CVS Access

[Mihai I. <mi...@re...>]

On 15 Aug 2002, Martin Sj=F6gren wrote:

> I forgot to mention this in the announcement mail. We've set up a
> pserver for anonymous (readonly) access to the pyOpenSSL CVS, generousl=
y
> hosted by Strakt.
>=20
> The CVSROOT is :pserver:ano...@cv...:/u/cvs/pyopenssl, the
> module is pyOpenSSL.
>=20
> Happy hacking, and send lots of patches ;)

It looks like pyOpenSSL does not run on 1.5.2 anymore. From a quick check=
:

PySequence_Fast (can be worked around with PySequence_Tuple) - in=20
src/crypto/x509req.c
PyObject_AsFileDescriptor (not in 1.5.2) - in src/ssl/connection.c

I know, 1.5.2 is old :-)

Misa