Re: comparing with libssh

Daniel Stenberg writes: > On Thu, 11 Jun 2009, Daniel Stenberg wrote: > >> This said, I even failed to build this on Linux... I was tempted to do a >> speed comparison but I'll put that on hold for now. > > libssh2 test: > > $ ./scp_nonblock 127.0.0.1 [user] [password] [big file] >/dev/null >

Re: get off sourceforge?

Daniel Stenberg writes: > Hello > > Peter, you were playing with trac install for libssh2 to take care of bug > reports, how's the status there? Something we should/could switch to? > > We were also discussing moving the mailing list(s) to something that doesn't > pollute each outgoing mail wit

Re: libssh2_knownhost_*

Daniel Stenberg writes: > Hello friends, > > I've just pushed my initial 7 man pages for the known hosts API I've been > working on recently. I'm very interested in comments and feedback on its > functionality. Something is missing? ssh2_exec.c:146: error: ‘LIBSSH2_KNOWNHOST_TYPE_DEFAULT’ und

Re: libssh2 with git

Daniel Stenberg writes: > Hey > > I've switched to git for my libssh2 development, kindly setup and handled by > Peter Stuge for our pleasure. > > So you checkout from here => git://git.stuge.se/libssh2.git > And browse with gitweb here => http://git.stuge.se/ Will it be possible for others t

Re: known_hosts [early PATCH]

Peter Stuge writes: > Simon Josefsson wrote: >> What about a function to add entries to the known_host file? > > Please don't do that. > > Make a createfile or something similar. > > On Win32 there are no files. A helper function doesn't hurt, does it? The

Re: known_hosts [early PATCH]

Daniel Stenberg writes: >> OpenSSH offer a mechanism to configure it to not fail for incorrect keys for >> a particular hosts, though. Maybe libssh2 could support that? I think it >> should still warn, though. > > Well, we have a few challanges related this: > > A) known_hosts is a OpenSSH fi

Re: known_hosts [early PATCH]

Daniel Stenberg writes: > On Thu, 7 May 2009, Simon Josefsson wrote: > >> What about a function to add entries to the known_host file? Maybe it is >> already there, and I didn't read carefully. > > No, that's one of the still missing pieces. I definatel

Re: known_hosts [early PATCH]

Daniel Stenberg writes: > Fire away! I read it and it looks fine to me, but I didn't test it. It seems possible to use even for non-OpenSSH style files, if the application wants to go through the trouble of supporting it. What about a function to add entries to the known_host file? Maybe it i

example/simple/ssh2_exec.c

This examples doesn't cross-compile to mingw: ssh2_exec.c:15:24: error: netinet/in.h: No such file or directory ssh2_exec.c:16:24: error: sys/socket.h: No such file or directory ssh2_exec.c:17:23: error: arpa/inet.h: No such file or directory ssh2_exec.c: In function ‘waitsocket’: ssh2_exec.c:34:

Re: host + key pairs for known_hosts

Tor Arntsen writes: > On Wed, May 6, 2009 at 11:54, Daniel Stenberg wrote: > >> Now, this seems to be a description of the known_hosts file currently in use >> by openssh: http://nms.lcs.mit.edu/projects/ssh/README.hashed-hosts >> >> IOW, they no longer store the host name in the plain, but only

Re: known_hosts support (first take)

Daniel Stenberg writes: > On Mon, 4 May 2009, J.T. Conklin wrote: > >>> A read the known_hosts file(s) and keep a set of hosts/key pair in memory >>> B compare a host/key with the known set to see if it exists and if so if >>> it >>> matches the key >>> C add/replace a host key/key in

Re: RELEASE: libssh2 version 1.1

Daniel Stenberg writes: > Hi friends, > > I'm happy to announce that I've just now packaged and uploaded libssh2 1.1: Great! I tried to build binaries for Windows using my makefile, and fortunately this release built fine as well. Here are the links: http://josefsson.org/gnutls4win/libssh2-1.

Re: the web site

Daniel Stenberg writes: > Hey > > I've modified my web pages about libssh2 and host them here: > > http://libssh2.haxx.se/ > > Consider it my suggested remake of the main web site. Nice work! I kind of like the flexibility of a wiki homepage, though, since it (in theory) allows more peopl

Re: memory (leak) debugging

Daniel Stenberg writes: > On Mon, 30 Mar 2009, Simon Josefsson wrote: > >> I gave up on dmalloc a long time ago, in my experience valgrind leads to >> better results and doesn't require changes to the build. >> >> In most of my other projects, I always se

Re: memory (leak) debugging

Daniel Stenberg writes: > On Sun, 29 Mar 2009, Dan Fandrich wrote: > >>> Does anyone have any positive experience with something existing to share? >>> Like with http://dmalloc.com or similar? That looks like a fine lib to work >>> with for our use case... >> >> I've played with glibc's mtrace

Re: RE : RE : RE : For the interested ones :libssh2_userauth_publickey_fromfileandlibssh2_userauth_hostbased_fromfile with no need of the publickeyfile

Alexander Lamaison writes: > 2009/3/27 Simon Josefsson : >> "Jean-Louis CHARTON" writes: >> >>> maybe my suggestion to make the publickeyfile parameter optional is >>> not interesting at all. >> >> I haven't understood exactly when t

Re: For the interested ones: libssh2_userauth_publickey_fromfileandlibssh2_userauth_hostbased_fro mfile with no need of the publickeyfile

Peter Stuge writes: >> I could look into implementing this for libgcrypt if you just separate >> the OpenSSL specific stuff into openssl.h/openssl.c. It could also be >> acceptable that the libgcrypt-port does not have this feature until >> someone implements it. > > I think the patch will linge

Re: RE : RE : RE : For the interested ones :libssh2_userauth_publickey_fromfileandlibssh2_userauth_hostbased_fromfile with no need of the publickeyfile

"Jean-Louis CHARTON" writes: > maybe my suggestion to make the publickeyfile parameter optional is > not interesting at all. I haven't understood exactly when this would be useful? I could look into implementing this for libgcrypt if you just separate the OpenSSL specific stuff into openssl.h/o

Re: RE : For the interested ones : libssh2_userauth_publickey_fromfile andlibssh2_userauth_hostbased_fromfile with no need of the publickey file

"Jean-Louis CHARTON" writes: > Hi Mikhail > > The requested diff. Hi! Thanks for the patch. You can't use OpenSSL specific functions or data types outside of openssl.c/openssl.h. Instead, you need to define a new meta-function in openssl.h and gcrypt.h and call that in userauth.c. The OpenSS

Re: blocking redone committed

Simon Josefsson writes: > /Simon > > -- > ___ > libssh2-devel mailing list > libssh2-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/l

Re: blocking redone committed

Daniel Stenberg writes: > On Fri, 27 Mar 2009, Daniel Stenberg wrote: > >> Uh, I can and yes it's most probably related. I'll dig into it! Thanks. > > Fixed now, turned out I was just stupid as usual! ;-P Works great now. Anything holding up the v1.1 release now? People tend to test things mor

Re: blocking redone committed

Simon Josefsson writes: > Daniel Stenberg writes: > >> My rather biggish overhaul of how to deal with the blocking API has now been >> committed. Everyone please try out CVS HEAD and see how it runs for you! > > Running 'make check' just hangs for me, can you

Re: blocking redone committed

Daniel Stenberg writes: > My rather biggish overhaul of how to deal with the blocking API has now been > committed. Everyone please try out CVS HEAD and see how it runs for you! Running 'make check' just hangs for me, can you reproduce that and do you think it is related? /Simon -

Re: libshh2 & minGW

Patrick Masotta writes: > Hi > I'm just arrived to libssh2 > I'm struggling trying to compile the library on Vista with minGW. > I understand libssh2 depends on libgcrypt-1.4.4 & libgpg-error-1.7 making > things a bit more complicated. > > I'd appreciate if some list member using minGW can send

Re: [ libssh2-Bugs-2528572 ] libssh2 on Solaris

Peter Stuge writes: > Daniel Stenberg wrote: >> > I don't think the best fix is to redirect the emails, I would >> > prefer to replace the origin. >> >> Is the savannah version any better? > > I haven't used it myself. > > >> Or is there another good replacement? > > I like trac a lot. One gotch

Re: more about the mailing list

Daniel Stenberg writes: > Hey > > I'd just like to mention that I find sourceforge's hideous ads they add in > every single darned mail we send/get on this list totally outrageously big > and > annoying. In many mails the ads are larger than the actual content. > > For this reason alone I thin

Re: [ libssh2-Bugs-2528572 ] libssh2 on Solaris

Daniel Stenberg writes: > On Thu, 19 Mar 2009, Simon Josefsson wrote: > >>> In file included from libssh2_priv.h:120, >>> from channel.c:40: >>> libgcrypt.h:164: error: syntax error before '*' token >>> libgcrypt.h:169:

Re: [ libssh2-Bugs-2528572 ] libssh2 on Solaris

"SourceForge.net" writes: > In file included from libssh2_priv.h:120, > from channel.c:40: > libgcrypt.h:164: error: syntax error before '*' token > libgcrypt.h:169: error: syntax error before '*' token It seems gcrypt.h didn't define 'struct gcry_sexp' for you -- can you check

Re: mailing list administrativa

Daniel Stenberg writes: > On Wed, 18 Mar 2009, Peter Stuge wrote: > >>> * I changed Reply-To: to point back to this list >> >> Not as cool. >> http://woozle.org/~neale/papers/reply-to-still-harmful.html > > Yeah, I know this is religous and for each of the points in that document > there are reb

Re: make check failure on cygwin

Yang Tse writes: > Daniel and Simon, > > sshd -ddd clearly shows that the problem is on my system $HOME user's > permissions: > > debug3: secure_filename: checking '/home/yang/libssh2/tests/etc' > debug3: secure_filename: checking '/home/yang/libssh2/tests' > debug3: secure_filename: checking '/h

Re: make check failure on cygwin

Yang Tse writes: > Hi, just in case this should be addressed... I think it should. > If libssh2/tests/ssh2.sh is modified to add -o 'UsePrivilegeSeparation > no' linke in: > > $SSHD -f /dev/null -h $srcdir/etc/host \ > -o 'Port 4711' -o 'Protocol 2' \ > -o 'UsePrivilegeSeparation no' \

Re: [libssh2] Wrap up a release!

Daniel Stenberg writes: > Hey > > This is just a note that I plan to put together a libssh2 1.1 release in > exactly week. I aim at March 24th for release day, but if something pops up I > can move it a bit. > > I also won't commit any feature changes or anything before the release, but > I'll

Re: [libssh2] libssh2_userauth_publickey_fromfile won't work with passphrase

ll...@163.com writes: > Hi Heiner and Alex, > Thanks for your reply. > > Heiner, I am pretty sure that the libssh2 you are using was compiled with > openssl. I took a look at the source codes and I did find out that the > passphrase was implemented in openssl module but NOT in libgcrypt module

Re: [libssh2] [ Bugs #2476770 ] Win32, static lib (using MinGW in MSYS shell) (fwd)

Daniel Stenberg writes: > Hi friends, > > Can anyone who actually uses mingw to build libssh2 help me out with some > feedback on this. I know we've poked a bit on these files lately, so I don't > like to just blindly commit this contributed patch and I just don't have any > windows setup near

Re: [libssh2] suggestion: libssh2_version()

Daniel Stenberg writes: > On Mon, 23 Feb 2009, Simon Josefsson wrote: > >> const char *libssh2_check_version (int req_version_num); >> >> Which could be used by applications like this: >> >> if (!libssh2_check_version (LIBSSH2_VERSION_NUM)) { >> fprintf

Re: [libssh2] suggestion: libssh2_version()

Daniel Stenberg writes: > Hey > > I'd like to suggest that we add a libssh2_version() function that returns > information about the lib. I just noticed how debian unstable got a fresh > libssh2 1.0, but curl is still thinking it is 0.18 since it got the version > number at build-time instead o

Re: [libssh2] Exported private symbols

Dan Fandrich writes: > On Mon, Feb 16, 2009 at 10:33:40PM +0100, Simon Josefsson wrote: >> I believe so. I propose to use the following patch which makes libtool >> hide the other symbols. We should also bump the revision field of the >> ABI version before we make the n

Re: [libssh2] Exported private symbols

Mikhail Gusarov writes: > Hello, > > I was packaging 1.0 for Debian and found out that there are private > symbols exported by the library: > > + _libssh2_cipher_crypt > + _libssh2_cipher_init > + _libssh2_dsa_new > + _libssh2_dsa_new_private > + _libssh2_dsa_sha1_sign > + _libssh2_dsa_sha1_verif

Re: [libssh2] wishlist?

Peter Stuge writes: > Daniel Stenberg wrote: >> Apart from better performance (higher transfer rates), what are the >> primary things you miss in libssh2? > > I'd like a Windows library binary download, maybe static MinGW? > > SSH is fantastic technology and should be made as available as > possi

Re: [libssh2] Memory leak on MinGW

double writes: > Hi, > > The following line produces a memory leak, compiled > on Win32 MinGW (dmalloc) - no problems on Linux: > > libgcrypt-1.4.1/cipher/pubkey.c:2197 > > I will try to get the call stack. Please do, or at least get a valgrind-style stacktrace. Did you use libgcrypt on the non

Re: [libssh2] release thoughts

On Sun, 2008-12-21 at 21:59 +0100, Daniel Stenberg wrote: > Hey > > I'm pondering on a 0.19 release in a week or so. Is there anything in > particular someone is working on or feel is necessary to cram in before we > can > release anything? > > Furher, and I know I've mentioned this before, I'

Re: [libssh2] libssh2 install fails on AIX 5.3

I'd say the make implementation is broken... Others seems to have had similar problems with other projects: http://gcc.gnu.org/ml/gcc-help/2004-08/msg00173.html http://marc.info/?l=xalan-c-users&m=103611476029916&w=2 According to http://www.realvnc.com/pipermail/vnc-list/2001-June/023527.html

Re: [libssh2] libssh2 install fails on AIX 5.3

"David M. Funk" writes: > Makefile: > +690 > +691 .PHONY: ChangeLog > +692 ChangeLog: > +693 if test -f .cvsusers; then \ > +694cvs2cl --utc --fsf --FSF --usermap .cvsusers -I ChangeLog > -I .cvs; \ > +695 fi > +696 > +697 dist-hook: ... >> "Makefi

Re: [libssh2] libssh2 install fails on AIX 5.3

"David M. Funk" writes: > I am getting an error when I try to install libssh2-0.18 on my aix server. > Anybody run into this? Any help will be much appreciated. > > I run ./configure and that works ok. But when I run "make" I get this > error: > > r...@aix1/usr/opt/perl5/mods/libssh2-0.18--> ..

[libssh2] Daily build on i586, x86_64, macosx, and mingw

All, Just to let you know that I've set up daily builds on some common platforms: http://autobuild.josefsson.org/libssh2/ Right now mingw and x86 works, and powerpc and x86_64 fails. When I have time, I'll see if I can debug it... /Simon --

Re: [libssh2] docs/libssh2_session_block_directions.3 missing?

Daniel Stenberg <[EMAIL PROTECTED]> writes: > On Thu, 27 Nov 2008, Simon Josefsson wrote: > >> make[1]: *** No rule to make target `libssh2_session_block_directions.3', >> needed by `distdir'. Stop. > > Oops, I

Re: [libssh2] Fix -I/usr/include aka Mingw patches

I'm re-adding the list to the discussion.. "Yang Tse" <[EMAIL PROTECTED]> writes: > Hi Simon, > > 2008/11/27, Simon Josefsson wrote: >> Hi! This is a definitely a bug -- could you send me a config.log and >> output from running ./configure, from a run w

[libssh2] docs/libssh2_session_block_directions.3 missing?

make[1]: *** No rule to make target `libssh2_session_block_directions.3', needed by `distdir'. Stop. /Simon - This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based appli

Re: [libssh2] Code coverage?

Peter Stuge <[EMAIL PROTECTED]> writes: > (Whatever way you are using GMANE it seems to mess with the email > headers a lot.) I'm posting via NNTP now, let's see if this works better. > Simon Josefsson wrote: >> There is a nit with OpenSSH run as non-root: it di

Re: [libssh2] Code coverage?

Dan Fandrich <[EMAIL PROTECTED]> writes: > On Fri, Nov 21, 2008 at 04:03:58PM +0100, Simon Josefsson wrote: >> There is a nit with OpenSSH run as non-root: it dies trying to update >> the last-log on the system, which is understandable because it does not >> have appr

Re: [libssh2] Fix -I/usr/include aka Mingw patches

"Yang Tse" <[EMAIL PROTECTED]> writes: > 2008/11/18, Simon Josefsson wrote: >> Hi! This patch fixes the detection of openssl/libz to use more normal >> header+library detection code. It should work fine on common platforms, >> but it may change detection

Re: [libssh2] Code coverage?

Daniel Stenberg writes: > On Fri, 21 Nov 2008, Simon Josefsson wrote: > >> I'm thinking that the self-tests could start 'sshd' with appropriate >> parameters that would work for ordinary users, with a dummy host key, and >> then invoke some libssh2 sel

Re: [libssh2] Code coverage?

A patch against libssh2 that makes 'make coverage' generate a code coverage report in docs/coverage/ is as below. You'll need to install gcov+lcov. I'll wait with committing this until we have more self-tests in place, for now I just wanted to illustrate how easy it is to generate the reports. /

[libssh2] Code coverage?

It would be easy to create code coverage reports such as this one for libssh2: http://www.gnu.org/software/libidn/coverage/ However, right now the self-tests are pretty minimal, so it may look rather embarrassing. I'm thinking that the self-tests could start 'sshd' with appropriate parameters th

Re: [libssh2] Fix -I/usr/include aka Mingw patches

Peter Stuge <[EMAIL PROTECTED]> writes: > Simon Josefsson wrote: >> What do you think? Ok to commit? > > Yes! Done. I'll see If I can arrange daily builds of mingw packages too, to help catch regressions of the now good mingw status... FWIW, on a debian x86 bo

[libssh2] Fix -I/usr/include aka Mingw patches

Hi! This patch fixes the detection of openssl/libz to use more normal header+library detection code. It should work fine on common platforms, but it may change detection on weirder systems, but nothing that cannot be fixed by appropriate use of --with-foo-prefix. It also sets up -R parameters if

Re: [libssh2] Fix -I/usr/include aka Mingw patches

I forgot a patch for README. /Simon --- README 12 Apr 2007 23:40:22 +0200 1.78 +++ README 18 Nov 2008 17:58:15 +0100 @@ -48,49 +48,43 @@ the older more reliable method. * --with-libgcrypt + * --without-libgcrypt * --with-libgcrypt-prefi

Re: [libssh2] [PATCH 0/5] libssh2: Enable cross-compilation from Linux to Windows using MinGW

Daniel Stenberg writes: > checking for libz... Found in /usr/include /usr/lib I guess this is the problem: the libz autoconf test seems broken, it shouldn't search through the system but instead try to link to the library and use it if it works. The openssl test might be equally wrong. I'm in

Re: [libssh2] help me clean up the tracker!

Peter Stuge writes: > Daniel Stenberg wrote: >> https://sourceforge.net/tracker/?func=browse&group_id=125852&atid=703942 > > The SF tracker and the emails it sends make my head hurt! :\ Ditto! Browsing the web pages suggests there really are some issues that should be taken care of though... Ma

Re: [libssh2] wee neat

Guenter Knauf <[EMAIL PROTECTED]> writes: > Hi, >> Has anyone mingwed libssh2 with success yet? > yes, there are mingw makefiles in the ./win32 folder which work with plain > MingW32 - not yet tested from MSYS though; but I do use same makefiles for a > non-configure build on Linux with MingW-

Re: [libssh2] wee neat

Guenter Knauf <[EMAIL PROTECTED]> writes: > Hi Simon, > >> difficult to fix it. Seems to be the usual winsock-related problems. > nope, the code compiles with MingW32; > configure needs to be fixed; I do already non-configure builds with > mingw32-cross on Linux, and that works fine, and also

Re: [libssh2] wee neat

Peter Stuge <[EMAIL PROTECTED]> writes: > Hey all! > > Just found the project (via c-ares via Purple Labs via THC GSM > hacking via hackaday/CCCamp wiki) and wanted to give a few thumbs up. > > As a side note I also hang around OpenSSH-portable development and > know a little about that code and

[libssh2] Updated CVS statistics

Hi! For some reason, the CVS statistics had stopped working and I didn't notice. Now, for us who like this kind of thing, I made it work again. It should run daily from now on. http://statcvs.josefsson.org/libssh2/ /Simon ---

Re: [libssh2] indent mini script

Daniel Stenberg <[EMAIL PROTECTED]> writes: > On Tue, 7 Aug 2007, Simon Josefsson wrote: > >> Can't we use a make target for this, rather than a script? I'm >> thinking 'make indent'. > > Yeah, sure. But I don't think people will use it un

Re: [libssh2] indent mini script

Daniel Stenberg writes: > On Tue, 7 Aug 2007, Thomas Harding wrote: > >> just a one-line script to help with indent. > > I would rather prefer the longer version I posted in my commit message as that > is more elaborate on exactly what it does. I also prefer your version, since I assume it is

Re: [libssh2] ANNOUNCE: libssh2 0.16

Daniel Stenberg writes: > Hi friends, > > I just now packaged and uploaded libssh2 0.16 to sourceforge: > > http://sourceforge.net/projects/libssh2/ Great! Is now a good time to run indent on the source code, and immediately release 0.17? Assuming people still (?) think running indent is

Re: [libssh2] anything blocking a 0.16 release?

Daniel Stenberg writes: > Yes, > > I said I wanted a 1.0 release next, but it feels like instead of considering > if we're at that stage yet I thought we better first ship a 0.16[.0] with the > bug fixes, the new header style version number etc done since 0.15. > > I think we're quite prepare

Re: [libssh2] libssh2 0.15 release candidate

Simon Josefsson <[EMAIL PROTECTED]> writes: > As far as I'm concerned, we're ready to go. Except that don't forget to update LIBSSH2_VERSION in include/libssh2.h! /Simon - This SF.net email is

Re: [libssh2] libssh2 0.15 release candidate

Daniel Stenberg writes: > On Thu, 22 Mar 2007, Simon Josefsson wrote: > >> I think we should do a release. > > And yet again we're back on this subject. Great! > Please everybody, go test the latest CVS version and give it a real spin and > tell us what the

Re: [libssh2] Support for GSSAPI/Kerberos authentication

petesea-jG/[EMAIL PROTECTED] writes: > Have there been any requests or thoughts about including GSSAPI/Kerberos > authentication in libssh2? Your query is the first one, as far as I know. > Specifically I'd like to see gssapi-keyex... but would also be interested > in gssapi-with-mic. I imple

Re: [libssh2] [PATCH] Building static lib on x86_64 linux

elated with this thread > > 2007/3/29, Simon Josefsson <[EMAIL PROTECTED]> wote: > >> I believe this is the wrong solution -- you shouldn't be mixing static >> libraries (non-PIC) with shared libraries (PIC). Either use the >> shared PIC library of libssh2, or

Re: [libssh2] Does not build on Cygwin with --disable-shared

"Yang Tse" <[EMAIL PROTECTED]> writes: > Failure building on Cygwin with --disable-shared (CVS version of course) > > ssh2.o:ssh2.c:(.text+0x104): undefined reference to > `__imp__libssh2_session_init_ex' Hi! Ouch, this is bad. I'm not sure what the problem is, building and and linking to libra

Re: [libssh2] libssh2 compile success on AIX - needs patch

Michael Augustin <[EMAIL PROTECTED]> writes: > Hi libssh2-developers, > > I compiled libssh2 successfull after some days of try-and-error. I wasn't > able to compile the examples, but the ssh2_sample.c from version 0.14 was > build and runnable with version from 0.15pre (cvs-snapshot). >

Re: [libssh2] make clean does not remove .deps

"Yang Tse" <[EMAIL PROTECTED]> writes: > make clean does not remove .deps subdirs > > How to reproduce: > > ./buildconf > ./configure && make > make clean That is standard automake-behaviour. There is 'maintainer-clean' and 'distclean' targets that remove .deps/, but I couldn't find any other ru

Re: [libssh2] 2 bugs in libssh2

out of Makefile.am. That will work around the problem, and it doesn't create any serious problems of its own that there aren't other workarounds available for. /Simon Simon Josefsson <[EMAIL PROTECTED]> writes: > Daniel Johnson <[EMAIL PROTECTED]> writes

Re: [libssh2] [PATCH] Building static lib on x86_64 linux

"Yang Tse" <[EMAIL PROTECTED]> writes: > When static libraries are built for AMD64 linux they _must_ be > compiled using PIC, otherwise linkage will fail. > > The attached patch is the same solution I've contributed to libcurl, > and which has proven to be efective. > > The 'trick' is to set to 'y

Re: [libssh2] [PATCH] Building on Cygwin

Daniel Stenberg <[EMAIL PROTECTED]> writes: > On Thu, 29 Mar 2007, Simon Josefsson wrote: > >> Actually, adding -no-undefined should be correct for all platforms, because >> libssh2 doesn't depend on unresolved symbols. For example, I believe AIX >> also n

Re: [libssh2] libssh2 0.15 release candidate

Heiko Jansen <[EMAIL PROTECTED]> writes: > Am Donnerstag, den 29.03.2007, 10:27 +0200 schrieb Simon Josefsson: > >> Hi Heiko! Your post didn't seem to made it to the list (at least I >> didn't get it through the list), so I'm quoting your entire message

Re: [libssh2] Macro define error while executing of libssh2 in AIX Unix 4.3/5.1

"DAO Anh-Viet" <[EMAIL PROTECTED]> writes: > Remark: I have generated the Makefile for AIX Unix version 4.3 and > CC of following version: VisualAge C++ Professional / C for AIX > Compiler, Version 5 Hm, do you use './configure && make'? Could you try it and see if it works better? /Simon

Re: [libssh2] 2 bugs in libssh2

Daniel Johnson <[EMAIL PROTECTED]> writes: > I just tried the latest cvs HEAD and both issues are now fixed. Thanks! Excellent, I think we are approaching release quality. Let me know if there's anything else wrong with the daily snapshots: http://josefsson.org/daily/libssh2/ /Simon -

Re: [libssh2] [PATCH] Building on Cygwin

"Yang Tse" <[EMAIL PROTECTED]> writes: > Hi to all, first message on this list. > > In order to make libssh2 build on Cygwin I had to make a couple of > changes. After applying this patch it built nicely using: > > ./buildconf > ./configure > ./make > ./make test > > Atached patch is against snaps

Re: [libssh2] Illegal instruction (core dumped) of LIBSSH2_ALLOC on AIX Unix

"DAO Anh-Viet" <[EMAIL PROTECTED]> writes: >> I run the testing program using libssh2 at AIX Unix >> >> I have the problem "Illegal instruction (core dumped)" at the line of > LIBSSH2_ALLOC on AIX Unix. >> >> With the define of LIBSSH2_ALLOC: >> >> #define LIBSSH2_ALLOC(session, count) > session->

Re: [libssh2] Illegal instruction (core dumped) of LIBSSH2_ALLOC on AIX Unix

"DAO Anh-Viet" <[EMAIL PROTECTED]> writes: > Hi, > > > > I run the testing program using libssh2 at AIX Unix > > I have the problem "Illegal instruction (core dumped)" at the line of > LIBSSH2_ALLOC on AIX Unix. > > With the define of LIBSSH2_ALLOC: > > #define LIBSSH2_ALLOC(session, count)

Re: [libssh2] 2 bugs in libssh2

Daniel Johnson <[EMAIL PROTECTED]> writes: > There are a couple of bugs that I've found. One is that in > configure.in, the $SHLIB_SUFFIX_NAME variable never gets defined to > anything. ... I installed this too. There are better alternatives for finding openssl/libz than the current code, but th

Re: [libssh2] 2 bugs in libssh2

Daniel Johnson <[EMAIL PROTECTED]> writes: > The script sets up DYLD_LIBRARY_PATH so that Darwin's linker can find > things. This is definitely a Darwinism needed because of Darwin's > wacky linker. This is with libtool 1.5.22. > > Please let me know if I can provide any more information. Thanks.

Re: [libssh2] 2 bugs in libssh2

Daniel Johnson <[EMAIL PROTECTED]> writes: > The other bug is in tests/Makefile.in. The "AM_LDFLAGS = -no-install" > should be omitted, since it tells libtool to not use a wrapper > script. This causes simple to try to link to an already installed > libssh2, not the one just built. So it either us

Re: [libssh2] libssh2 0.15 release candidate

"Heiko Jansen" <[EMAIL PROTECTED]> writes: >>>> Simon Josefsson <[EMAIL PROTECTED]> 22.03.07 14.45 Uhr >>> > >> Please test the daily snapshots from >> <http://josefsson.org/daily/libssh2/> as if it were the 0.15 release, >>

[libssh2] libssh2 0.15 release candidate

Hi Sara, all! I think we should do a release. The debian folks wants one (see ), and I want one to feel finished with the libgcrypt port. So here is a proposal: Please test the daily snapshots from as

[libssh2] libgcrypt port finished

Simon Josefsson <[EMAIL PROTECTED]> writes: > What remains is to fix RSA signing, and to run libssh2+curl under > valgrind to find obvious problems. I reckon I'm now 90 % finished. RSA signing now works. It was a silly error -- for some reason the RSA p and q values were

Re: [libssh2] non-blocking status

Daniel Stenberg <[EMAIL PROTECTED]> writes: > On Fri, 26 Jan 2007, Simon Josefsson wrote: > >>> libssh2_sftp_read() might need attention anyway since it is documented to >>> return -1 on error which is hard to do on unsigned variables, and because >>>

Re: [libssh2] non-blocking status

Daniel Stenberg <[EMAIL PROTECTED]> writes: > On Fri, 26 Jan 2007, Daniel Stenberg wrote: > >> I think I like this idea. It would also make it clearly which functions that >> would work non-blocking since we'd only implement those that truly are... > > Ok, let's say I take that approach for funct

[libssh2] CVS statistics

Hi! The following URL should contain daily CVS statistics for libssh2 that may be interesting for those of us that likes statistics: http://statcvs.josefsson.org/libssh2/ While I'm at it, I'll remind you about the autobuild service and the daily snapshots: http://josefsson.org/libssh2/daily/ ht

Re: [libssh2] non-blocking status

On Fri, 2007-01-26 at 12:40 +0100, Daniel Stenberg wrote: > 2 - for LOTS of functions we need to change the prototypes to allow them to > return a "EAGAIN" error code (instead of blocking) until they are done > and > can return ok. A few examples are libssh2_sftp_init(), > libssh2_s

Re: [libssh2] PEM file reader and another libgcrypt status update

Sara Golemon <[EMAIL PROTECTED]> writes: >>> At this point, I think it would be useful for others to review the code and >>> comment on it. I'm sure there are mistakes, but I'm to deep into the code >>> to see them now. >> >> Nice work Simon! >> > Agreed! Excellent work! This doubles the op

Re: [libssh2] PEM file reader and another libgcrypt status update

Daniel Stenberg <[EMAIL PROTECTED]> writes: > On Tue, 23 Jan 2007, Simon Josefsson wrote: > >> At this point, I think it would be useful for others to review the code and >> comment on it. I'm sure there are mistakes, but I'm to deep into the code >> t

Re: [libssh2] PEM file reader and another libgcrypt status update

Simon Josefsson <[EMAIL PROTECTED]> writes: > [EMAIL PROTECTED]:~/src/libssh2/src$ nm -B .libs/libssh2.so|grep -i openssl > U BN_bin2bn@@OPENSSL_0.9.8 > U BN_bn2bin@@OPENSSL_0.9.8 > U BN_clear_free@@OPENSSL_0.9.8 > U BN_CTX_free@@OPENS

[libssh2] PEM file reader and another libgcrypt status update

I just installed a PEM file reader and simple ASN.1 decoder into lib/pem.c. See libgcrypt.c:_libssh2_{d,r}sa_new_private for how to use it. Btw, --with-libgcrypt can now read RSA/DSA private keys and use them to sign in authentication. DSA authentication works reliably, but RSA authentication fa

Re: [libssh2] Libgcrypt port update

"Mononen Jussi" <[EMAIL PROTECTED]> writes: >>> > I thought this was only about the disk-format, does it >>> affect the wire >>> > format also? I'd assume that the protocol clearly specified the >>> > format of public keys on the wire. >>> > >>> The format of the blob which is sent over the wir

Re: [libssh2] Remove OPENSSL_NO_SHA ifdef's?

Sara Golemon <[EMAIL PROTECTED]> writes: >> I have no idea how to test the server mode of libssh2 though. Is that >> even supported? I see there are some RSA signing stuff going on, and >> I'd assume that is for the server side, but right now I'm to deep into >> details to remember how things

Re: [libssh2] Libgcrypt port update

"Mononen Jussi" <[EMAIL PROTECTED]> writes: >>> Is there a standard on what the SECSH filenames are? > > AFAIK there are no restrictions, so far I have seen various version > containing more or less host/user/strength information in the filename. Sigh.. I was hoping for some well-established de-f

  1   2   >