On 08 May 2014, at 09:21, jay binks <jaybinks@gmail.com> wrote:

can you explain what the differences are when you say "avp stored in the transaction" and process variables ?
do you literally mean variables in my module here, or are these some kind of structure attached to the sip_msg ?
Read the docs on AVPs, shared variables (shv) process variables (var). They are all stored in different memory and related to different things.

Thanks, Ill look into these.
Great! Reading docs usually doesn't hurt people. A few animals where killed in the process though ;-)

 
effectively what im looking to do is store some custom data against sip_msg* that I can use later if I need to.
When is later? In another request or in the response to the same request? If there's a stateful transaction,
AVPs are reachable in response processing.

Sorry... by later, I mean, later while processing the same message.
Not later on in the call, in that case I know it must be persisted in a dialog.
Then you want to use the process memory - $var.

my reason for doing a module is that I had a few things I wanted to do in a module ( looking up large data structures for performance reasons ), and im implementing a few other bits of business logic in this module.
Still doesn't explain why you need to code it in c and make it a modules. You can do a lot in the existing
logic.

Fair call,  I know I can do a lot in kamailio, its quite surprising actually what you can do.
Ill re-evaluate my reasons for writing a module.
Please do. During many years of using Kamailio, I've been able to do almost everything I want to do with the existing set of modules. Developing our own module put a lot of burden on the company - someone will have to maintain it for as long as Kamailio is used.

We also have the lua, perl, python and java modules that you can use to extend our script language even more.

 
im happy to discuss further if you want to gtalk me ( on this email address )
I would suggest that you read up on the available docs and then find us in the IRC channel where we try to be available and answer when it works for us. If you really need fast help and responses, many of us are available as consultants.

Ill try IRC again, but yea... there were not many people around last I was on there :)
I understand maybe time zones didnt work, or people were too busy. 
It's not as active as a few other projects. But it heats up from time to time. ;-)


 
In general, you have to be a bit more exact in your questions to get good answers on -dev, a bit less on -users. Your original e-mail was a bit too inexact so it wasn't easy to answer, which lead to everyone thinking that someone else may understand and may answer...  Before reposting to a mailing list, try to figure out why no one answered and add information to make it easier to answer.

Fair enough, sorry for being vague. 
Don't apologize, just learn from the experience. Open Source is a different world and everyone has to learn how to interact with the community to get answers and provide knowledge back.

/O