[vtls] vtls background

[Tim Ruehsen]

On Monday 09 February 2015 16:24:40 Daniel Stenberg wrote:
> On Mon, 9 Feb 2015, Tim Ruehsen wrote:
> >> Making libcurl dlopen() crypto libraries sounds like a really bad idea to
> >> me.
> > 
> > It's an idea, but not a 'bad' one.
> 
> I'd also maintain it is a bad idea. A road full of pain and tears.

I have closed-source pluggable SSL library code running at customers sites, 
but I never experienced 'tears' or 'pain'. Some challenges, yes. Challenges 
are one of the reasons why I like my job ;-)
But again, I am not fixed on dlopen etc.

> Sure, you _can_ use it but you have to do a lot of the work manually instead
> of having the linker and tools do it for you.

+1 against runtime loading

> And the portability problems are not even mentioned yet...

(I already mentioned it, no real reactions so far.)

We need a list of systems to support (for a start, can be extended later).
I guess, you have the broadest experiences regarding what systems to support. 
What do you think which systems should we  support ? And what systems can we 
support with the available resources we have right now (dev/testing) ?

What I know we have here:
Darshit Shah (Arch Linux, OpenSSL)
Dagobert Michelsen (Solaris 9,10,11 - he even has a build farm and maybe could 
automate building / error reporting)
Tim Rühsen (Debian testing/unstable, Cygwin 64bit, GnuTLS)
We can automate tests via Travis-CI (Ubuntu 12.04, OSX ?, does it already 
support Windows ?)

I added 'Supported Platforms' in the wiki: https://github.com/darnir/vtls/wiki

Tim
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: This is a digitally signed message part.
URL: <http://lists.mayfirst.org/pipermail/vtls/attachments/20150209/59324828/attachment.sig>