Matthias Andree <ma+ov...@dt.e-technik.uni-dortmund.de> said:

> On Sat, 14 Sep 2002, James Yonan wrote:
> 
> > (1) Forking server support
> > (2) Automatic Secure MTU discovery
> > (3) IPv6 endpoints or IPv6 over tun device
> > (4) Windows port
> > 
> > While none of these (with perhaps the exception of the last :) is
> > rocket science, all require some work, and given that OpenVPN has
> > reached a nice stability plateau, I'd like to hear your opinions on
> > future directions in the development effort.
> 
> ad 1) what's this good for? Sharing one port? Won't really work with
>       UDP. Convenience? Go get daemontools, or let's figure if openvpn
>       can be run from xinetd. Not something that should be done unless
>       you're really bored or have some compelling reason (a malicious
>       person holding a gun against your head might be one). Don't waste
>       your time.

Well actually "Forking server support" is really a misnomer.  It would be 
better titled "Server support for arbitrary number of connecting clients 
without requiring a separate config file and a pre-instantiated daemon for 
every client, or just "scalability support".  xinetd is an interesting idea.  
Anyone using xinetd with OpenVPN?

> ad 2) I'd like to see that somewhen in the future, but I'd give IPv6 the
>       preference.
> 
> ad 3) IPv6 is certainly something that will have to be added.
> 
> ad 4) If they want secure networking, they shouldn't be running an
>       operating system of which the current version forces you to
>       register with the vendor, sending data you don't control.
>
>       It may be a moot point, but IMHO Windows which is made and sold by
>       a company that's always looking at its own profit, acting to its
>       own advantage, no matter how ruthless, and that aims to harm
>       OpenSource just to fortify its monopoly, to make more money, and
>       to control ever more markets to further maximize its huge
>       revenues, should not be supported by OpenSource developers.

While I agree that Windows is marketed by an aggressive monopoly, I don't agree 
that just because a company is anti-open-source, we shouldn't port open source 
software to its platform or make *nix implementations of its protocols.  On the 
contrary, I would argue that the more ruthlessly Microsoft behaves, the more 
that Open Source can benefit by porting to Windows, because once we have a 
large body of cross-platform software, then suddenly the OS won't matter as 
much and people will be free to choose based on quality.  And the ultimate 
weapon against a ruthless monopoly is this: Give customers a choice.

Having said that, and having glimpsed the cries of agony on the cipe-win32 list 
from those who have faced the windows network driver model, I expect that a win 
port will not be a panacea.

James

>       If your opinion differs from mine, that's called freedom. :-)

> Matthias Andree


Reply via email to