Re: [Openvpn-devel] [PATCH 0/3] Use the garbage collector when retrieving x509 fields

2012-03-30 Thread Alon Bar-Lev
Hm I missed that one. I don't like this patch... the gc_new() and gc_free() should be called once per function at prolog and epilog. This is the whole point of gc, otherwise there is no difference between it and simple malloc. So: 1. in pkcs11.c it should be moved out of the loop. 2. In

Re: [Openvpn-devel] [PATCH] cleanup: flags should not be bool

2012-03-30 Thread David Sommerseth
On 25/03/12 13:27, Alon Bar-Lev wrote: > Discussed at [1][2][3]. > > [1] https://bugs.gentoo.org/show_bug.cgi?id=297854#c2 > [2] https://297854.bugs.gentoo.org/attachment.cgi?id=214304 > [3] >

Re: [Openvpn-devel] [PATCH] Enable pedantic in windows compilation

2012-03-30 Thread David Sommerseth
On 28/03/12 20:43, Alon Bar-Lev wrote: > Apparently -ansi undef WIN32 macro. > > Signed-off-by: Alon Bar-Lev > --- > configure.ac |3 ++- > 1 files changed, 2 insertions(+), 1 deletions(-) > ACK! Applied to master branches on -testing and -stable trees. commit

Re: [Openvpn-devel] [PATCH 0/3] Use the garbage collector when retrieving x509 fields

2012-03-30 Thread David Sommerseth
On 14/02/12 11:11, Adriaan de Jong wrote: > A number of the x509 functions allocated memory directly, instead of > using the default OpenVPN convention: the garbage collector. This is > fixed in this series of patches. > > Adriaan These patches got ACKed by James and me at the developers meeting

Re: [Openvpn-devel] Buildbot build flags

2012-03-30 Thread Alon Bar-Lev
Safest is to go with all combinations... Or at least: --disable-lzo --enable-lzo --enable-pkcs11 --disable-pkcs11 --enable-selinux --disable-selinux --enable-iproute2 --disable-iproute2 --disable-server --enable-small --disable-plugins --disable-management --disable-crypto --disable-ssl

Re: [Openvpn-devel] [DISCUSS] OpenVPN public repositories at github.com

2012-03-30 Thread Alon Bar-Lev
2012/3/30 Samuli Seppänen : > >> Hi, >> >> 2012/3/30 Samuli Seppänen : > [*] As in "We will have the community services on our own servers. > Period." >> BTW: If you like self-hosting, we're experimenting with Gerrit for >> code review (also

[Openvpn-devel] Buildbot build flags

2012-03-30 Thread Samuli Seppänen
Hi all, Before Alon's buildsystem overhaul our buildbot farm tested building with all unique combinations of the following build flags: --disable-lzo --disable-ssl, --disable-crypto, --disable-management In addition, each commit triggerer a build which used no build flags (i.e. just ran

Re: [Openvpn-devel] [DISCUSS] OpenVPN public repositories at github.com

2012-03-30 Thread Samuli Seppänen
> Hi, > > 2012/3/30 Samuli Seppänen : [*] As in "We will have the community services on our own servers. Period." > BTW: If you like self-hosting, we're experimenting with Gerrit for > code review (also provides Git repo hosting as a by-product) and it's > looking quite

Re: [Openvpn-devel] [DISCUSS] OpenVPN public repositories at github.com

2012-03-30 Thread Fabian Knittel
Hi, 2012/3/30 Samuli Seppänen : >>> [*] As in "We will have the community services on our own servers. Period." BTW: If you like self-hosting, we're experimenting with Gerrit for code review (also provides Git repo hosting as a by-product) and it's looking quite promising.

Re: [Openvpn-devel] [DISCUSS] OpenVPN public repositories at github.com

2012-03-30 Thread Samuli Seppänen
> On 03/29/2012 12:00 AM, Samuli Seppänen wrote: >> That said, the code itself is hosted externally on SF.net, so we have >> the _option_ of using GitHub for that part. I'd first look at the big >> picture of how we want to enhance our development processes and then >> move on to the tools. >

Re: [Openvpn-devel] [DISCUSS] OpenVPN public repositories at github.com

2012-03-30 Thread Samuli Seppänen
>> [*] As in "We will have the community services on our own servers. Period." > I don't understand this argument... Did I say I do? :P > Anyway, if you keep services like wiki on your host, please make sure > change notification is possible, as cooperation without notification > is invalid. > >