Mailing List CGatePro@mail.stalker.com Message #99642
From: Karl <kgretton@mac.com>
Subject: Re: The future of CommuniGate (was: Clarification and discussion on roadmaps)
Date: Thu, 04 Mar 2010 07:14:51 -0500
To: CommuniGate Pro Discussions <CGatePro@mail.stalker.com>
X-Mailer: Apple Mail (2.1078)
Comments below.

Karl

On 2010-03-03, at 8:37 PM, Andrew Hotlab wrote:

>
> ----------------------------------------
>> To: CGatePro@mail.stalker.com
>> From: kgretton@mac.com
>> Date: Wed, 3 Mar 2010 10:08:23 -0500
>> Subject: Re: Clarification and discussion on roadmaps
>>
>>
>> Why are we even bothering with MAPI any more? Microsoft are dropping support for it in the next version of Outlook and also in Exchange.
>>
>> Rather than waste resources on supporting the various hacks that exist in a protocol created in the 1980’s, CGS should be focusing on the future. Exchange Web Services has won the Enterprise market - support that instead.
>>
>> Communigate remains the most robust solution out there but it seriously needs modernizing in the Outlook and Mac connectivity areas. Zimbra and a host of other have implement EWS with relative ease. The folks at Apple had little trouble implementing the client side according to my sources. EWS is the future.
>>
>
> Sorry, but I did not find anything about Zimbra support for EWS. Since I'm really interested in the adoption of EWS by 3rd parties, I ask you to give any link to documents or news about Zimbra.

They originally hacked together a solution but I am informed (I didn’t realize that this wasn’t in the current release) that they switched to web services.

>
> EWS inclusion in CommuniGate was one of the first ideas that came to my mind when I realized that poor support for native clients would have slow down the adoption of CGP in enterprise environments.  I have not yet evaluated economic implications (I guess that MSFT would request some type of royalty for inclusion of its technology in 3rd party's servers) and, from a technical point of view, the freely redistributable managed API SDK allow to develop only client applications that use EWS... but I wish to share with you some thoughts about this matter.

Well the situation for MAPI is anything but clear.  MAPI is not a published, open standard.  Its Microsoft proprietary.  EWS is published and, as part of the settlement with the EU, anyone is free to implement it.  I cannot find any specific licensing fees associated with it.  If you have any links, let me know.

CGS pay royalties to a number of different licensing organizations.  Historically, MAPI was a separate licensed component for a significant premium over the standard feature set.  Any license obligations can be easily accommodated.

>
> It's evident that the rich user experience we need to deliver nowadays requires something new and different from the protocols created in the past decades.  Today we have to support not only e-mail and groupware services, but IM, presence, voice and secure generic data sharing flowing in and out our corporate perimeter, for a ton of new kind of portable/mobile devices.  We are asked to do all this while providing comprehensive yet friendly interfaces to our users.  Oh, and obviously, we are expected to keep the TCO of this solution as lowest as possible to maintain our enterprise competitive in the market.

I think we need to differentiate between _baseline_ features and _value_added_ features.  Jon is right: CGP has an amazingly broad feature set.  You are right: CGP supports a whole bunch of things that require separate software.

The baseline, though, is that, whether on Windows, Mac OS X or Linux, we do a few basic things: email, contacts, calendars, shared folder.  CGP works reasonably well with MAPI / Outlook but pretty poorly with anything else and this needs to change.

For the value added stuff, there can be multiple solutions.  Outlook with EWS + Counterpath plug-in or CGP Pronto plug-in?

>
> I'm not a developer, thus I may be a bit too optimistic with my assertions, but I think that Stalker has done a really remarkable work by creating the XIMSS API, long before Microsoft created EWS.  Moreover, XIMSS enables users to access a far more complete set of communication services than EWS currently does. Pronto! web interface proves that value pretty well by delivering a real (even if still improvable) UC user experience.

Actually, EWS was around in 2005 but was first shipped as part of Exchange in the 2007 edition.

XIMMS is impressive.  You can’t knock the CGS development team - they have some brilliant people.  Its more complete.  However, it doesn’t replace the baseline.  Outlook will still want to use EWS by default as will Mac OS X 10.6.

Pronto is not a native experience and never will be.  Mac OS X’s UI will change in the near future.  Even if Pronto adapts to the UI that exists today, its going to be behind the curve on the next iteration.  

Jon and his team made the right decision when it came to the iPhone: they went with ActiveStink because its an M$ standard and Apple licensed it.  Not the best protocol in the world, but why beat your head against the wall developing an alternative for the _baseline_ functionality?  Pronto (in Obj C) for iPhone would be a nice addition.

>
> So, from a strict technical point of view, I see XIMSS as the best technology to build upon the new generation of clients. The only very big drawback is that XIMSS is poorly documented and sponsored (I don't ever know if it can be freely implemented in other server products), thus the only real implementations (that I know) today are Stalker's and ITooLabs' proprietary software.

But XIMSS extends the _baseline_.  Its value added for sure...a differentiator.  But also too unstable / in-flux.  At Apple, as you undoubtedly know, this would still be considered a private API since it changes too much.

>
> I think it's a shame but... it seems too late for XIMSS to be a viable choice to support the 95% of the clients out there today: the just-born EWS from Microsoft, even if it's less technically valuable, has surpassed the 4 years old XIMSS and it has "nominally" took over that market share (Windows + Apple's SL).

EWS and XIMSS started around the same time (from a public perspective).

>
> Well, perhaps it might be worth with to "consolidate" all development resources now working on the MAPI Connector, the iSync Agent and the CalDAV/CardDAV interfaces in developing a server-side support for EWS.  By doing that, all new Windows and Macintosh clients will be totally supported out-of-the-box, without need to install anything on them to connect to CGP basic groupware services...

The point here is to get rid of the connector.  It must be time consuming and expensive for CGS to maintain.  M$ are moving away from MAPI.  Get over it, move on.  EWS is the future of Exchange and its actually documented and there are a bunch of bureaucrats in Brussels waiting to bang M$ with a fine if they fail to conform to their own specification.  Perfect.

Apple, M$ and Linux all support the ‘standard’.  What is not to like about that?

Then, innovate (not required...already done) and add in additional value (CGS Outlook UC plug-in).

>
> That's not a decision to take without a lot of care. Here are the main issues I see:
> - To support EWS does not mean to support a really unified user experience (we'll still need IM and voice at least).

Sure.  Value added plug-ins to whatever the client is.

> - Jon will say that he will lost control over his company's destiny! :)  Well, even today AIRSYNC, MAPI and bad IMAP and SIP clients are making the life of Stalker's developers somewhat difficult.

....Jon does not control his own destiny or CGS.  Standards and defacto standards do that...

> - Royalties required by Microsoft might be too much expensive to get an acceptable ROI from the development effort.

Nope.  Minimal or non-existant.

> - EWS might cause more technical issues than advantages. I left Microsoft because its development vision made their latest products (Exchange 2007/2010, for example) too much complicated and thus unreliable and expensive to deploy and maintain. Anyone can give any feedback about EWS real-world implementations, seen by his own eyes?!

We only deploy Exchange using EWS.  Its been perfect with Mac OS X clients and Outlook.  Setup is a breeze (if you correctly configure the auto discover).

>
> I would be glad if anyone might contribute to this new thread with his/her professional considerations.
> Please forgive me, but I do not have any interest to waste more time trying to make Jon understand that customer's priorities come a bit before his sales goals.

Jon needs to understand that its not a question of Pronto OR alternatives.  They need to do both things.  Vladimir is a brilliant guy...I am sure he still completely controls what goes on a CGS.  But you have to determine your space.  CGS planted themselves firmly in the Enterprise space with the MAPI connector.  They can either abandon that space (drop the connector, Mac, etc) and go to being carrier-only.  Or they can continue to innovate and embrace emerging standards such as EWS.  It might be crap (ActiveStink, anyone, anyone?).  But its better than MAPI.

> Really, I think that we can spend better this time if we go on and use this list for talking seriously about what Stalker and we can do to address the issues that we all are complaining about. Hoping that someone from Stalker will have the wisdom to join us and make these discussions really profitable!
>
> Sincerely.
>
>
> Andrew
>    
> _________________________________________________________________
> Hotmail: Trusted email with Microsoft’s powerful SPAM protection.
> https://signup.live.com/signup.aspx?id=60969
> #############################################################
> This message is sent to you because you are subscribed to
> the mailing list <CGatePro@mail.stalker.com>.
> To unsubscribe, E-mail to: <CGatePro-off@mail.stalker.com>
> To switch to the DIGEST mode, E-mail to <CGatePro-digest@mail.stalker.com>
> To switch to the INDEX mode, E-mail to <CGatePro-index@mail.stalker.com>
> Send administrative queries to  <CGatePro-request@mail.stalker.com>

Subscribe (FEED) Subscribe (DIGEST) Subscribe (INDEX) Unsubscribe Mail to Listmaster