Last month I asked the question do communication protocols need to be Swiss Army Knives or penknives. I also asked the same question in LinkedIn Answers. The (almost) unanimous answer I got was penknives.
The best part of it was actually getting two great acronyms to use for that do-it-all protocol:
- Steve Michelson suggested GPPTDE: General Purpose Protocol That Does Everything
- Cedric Mauvielle suggested SAKF: Swiss Army Knife Protocol
Cedric believes a SAKF (or a GPPTDE) protocol is doomed, much the same way SIP became bloated over time:
“… What happen to SIP would happen to the Swiss army knife protocol (SAKF). Thinkers will love it and promote it. Consulting companies will create a big buzz around it, Telco will want it … better to have specialized protocol, then the game is clearer …”
I am not that sure penknives are the solution. Martin Sauter shares a similar view:
“I guess I am as torn on this subject as you are. Ims is the best example. Versatile but way too complicated. And it is getting worse by the day. One could argue they should have stuck to voice telephony only… Who knows maybe one day they decide it’s too much and they come up with ims light? OK I am dreaming.
Sorry no magic bullet, we have to live with the extremes and let the market decide.”
I’ve spent a lot of time thinking about this issue since I raised the question. Every protocol that is being heavily used will evolve. This evolution will cause it to change and mutate - either by adding “generic” stuff to it, which was built into its design, or by adding “patches” to it. Protocols that don’t evolve fade away.
To me, every penknife is a potential Swiss army knife, given enough time and interest. This is why every several years we need a new protocol to replace an old one.
Paul Jones wrote in his comment on my Swiss Army knife post that AMS (or H.325), might be that knife. I think that AMS is a very interesting protocol - it is a rather new initiative if you look at today’s alternatives, so it is going to be based on earlier “mistakes”. That said, once AMS becomes a reality, people will start using it. Given time, it will become cumbersome as the rest of the Swiss Army knives we have today.
While penknives are better - they will become Swiss Army knives if given the choice (think XMPP and doing VoIP calls).
Subscribe to this blog
Leave a comment
Add your own
1. Eelco | August 4th, 2008 at 3:27 pm
I still do not share your opinion. I believe that client applications or devices will use best-of-breed protocols for the purpose they are most suitable for and attempts to do otherwise will fail. With respects to SIP/XMPP I believe that in the future SIP will primarily be used to setup multimedia sessions like voice and video and XMPP will primarily be used for presence and IM.
2. Tsahi Levent-Levi | August 4th, 2008 at 6:42 pm
Eelco,
I respect your view, and think it is a valid one. I’d take your approach everyday.
But… in the last decade I have been working with communication protocols and found out more than once that the best protocol usually doesn’t “win” - it is the one which is marketed the most or the one that provides a political solution that wins.
In this case, SIMPLE has been selected for IMS. When IMS happens, we will have both SIMPLE and XMPP used for Presence over VoIP networks. Not because SIMPLE is better or XMPP is better, but simply because this is how the standardization bodies and the involved companies steered the process.
3. Eelco | August 4th, 2008 at 8:05 pm
Tsahi,
That is a very valid reason. The choice by 3GPP/IMS for SIMPLE makes it the best candidate for future telephony networks. However, if I was running a business of making voip client applications I would not bet on just one horse. You don’t see IMS (with SIMPLE presence) widely deployed yet and there are still interworking issues that needs to be resolved with SIMPLE. Adding XMPP for presence to a VoIP client seems trivial to me.
Leave a Comment
Trackback this post | Subscribe to the comments via RSS Feed