[RFC] SMPP optional TLV patch
Stipe Tolj <[email protected]>
| Newsgroups | gmane.comp.mobile.kannel.devel |
|---|---|
| Organization | tolj.org system architecture |
| Message-ID | <[email protected]> |
Hi list, some comments and thoughts on Alex's patch: Re: [PATCH] SMPP Optional Parameters Date: 09.11.2007 14:54 Msg-Id: <[email protected]> first of all, great work Alex! I agree that we need a config group to "define" the tag names, it's type and corresponding bytecode. The problems I see: a) without configuring via 'smpp-tlv' groups, we don't support any "known TLVs", ie. callback_num = 123 will show in the smpp_pdu_dump() call, but we refuse it to enpack to SMPP PDU since we don't have a configured group for it. So my suggestion is to allow all TLVs that are "known/standard" that come allong the meta-data fragment. In addition optional TLVs that are "unknown", hence vendor specific can be configured via the 'smpp-tlv' groups. b) while doing smpp_pdu_init() we would need to check that no user tries to define a bytecode tag with a "corrupting" type. Obviously we still need to obey the generic PDU field definitions. Not sure if someone could try to use 'smpp-tlv' to define a known bytecode with a different type and hence we run into "nuts" state at packing time. I'm commiting Alex's patch to a dedicated branch in CVS (meta-data), so we have a baseline for the TLV patch before merging into mainline HEAD when all is done. Stipe ------------------------------------------------------------------- Kölner Landstrasse 419 40589 DÌsseldorf, NRW, Germany tolj.org system architecture Kannel Software Foundation (KSF) http://www.tolj.org/ http://www.kannel.org/ mailto:st_{at}_tolj.org mailto:stolj_{at}_kannel.org -------------------------------------------------------------------