Re: [Netstack] Data extensions in quantum

2011-10-11 Thread Salvatore Orlando
I don't have yet performed a complete review. Anyway, I was wondering whether we can avoid breaking existing plugins by using **kwargs, and not removing existing parameters Example: create_network(self, tenant_id, net_name, **kwargs) I also think Ying had another important point on how the API

[Netstack] No netstack meeting today

2011-10-11 Thread Dan Wendlandt
Hi folks, As we discussed at the summit, we'll skip having a netstack meeting today and resume our normal schedule next Tuesday. If you have not already added your slides/etherpad links to the NetStack Essex wiki page, please do so: http://wiki.openstack.org/NetstackEssexSummit Also, everyone sh

Re: [Netstack] Data extensions in quantum

2011-10-11 Thread Brad Hall
Replies inline.. On Tue, Oct 11, 2011 at 7:54 AM, Salvatore Orlando wrote: > I don't have yet performed a complete review. > Anyway, I was wondering whether we can avoid breaking existing plugins by > using **kwargs,  and not removing existing parameters > > Example: create_network(self, tenant_

Re: [Netstack] Data extensions in quantum

2011-10-11 Thread Edgar Magana (eperdomo)
Hi Brad, After taking a look to both review diffs, indeed adding **kwargs makes a cleaner diff. If this change will be good enough, I think we should stay with it but if you are planning to make more changes I do agree with you that the sooner the better :-) then we will all have more time to m

Re: [Netstack] Data extensions in quantum

2011-10-11 Thread Brad Hall
Sounds good to me.. thanks to both of you for the input. When you get time if ya'll could review it formally on gerrit that would be appreciated. Thanks, Brad On Tue, Oct 11, 2011 at 11:51 AM, Edgar Magana (eperdomo) wrote: > Hi Brad, > > After taking a look to both review diffs, indeed adding

Re: [Netstack] Data extensions in quantum

2011-10-11 Thread Ying Liu (yinliu2)
Hi Brad, I like the idea of adding **kwargs and still keeping existing parameters. Thus, we just need to change plugin method's signature, instead of changing the method body itself. For extended data parameters, we can pass in a flexible data structure. Thus, plugin provider can use it to pas

Re: [Netstack] Data extensions in quantum

2011-10-11 Thread Ying Liu (yinliu2)
Actually, it's surprised to me that Python does not support function overloading. Otherwise, we can keep both create_network(self, tenant_id, net_name) and create_network(self, tenant_id, net_name, **kwargs) Problem will be easier. Ying > -Original Message- > From: Brad Hall [mailto:b..

Re: [Netstack] Data extensions in quantum

2011-10-11 Thread Edgar Magana (eperdomo)
Done it! Regards, Edgar -Original Message- From: Brad Hall [mailto:b...@nicira.com] Sent: Tuesday, October 11, 2011 12:57 PM To: Edgar Magana (eperdomo) Cc: Salvatore Orlando; netstack@lists.launchpad.net Subject: Re: [Netstack] Data extensions in quantum Sounds good to me.. thanks to