"Exchange is dead" ?

At Lotusphere, various of us discussed the viability of Exchange in relation to Domino. After three years of market confusion, we felt that the "domino is dead" meme had finally been put to rest. And so we thought of Exchange.

This week I had the pleasure of attending a technical meeting, ostensibly to discuss with an organisation which platform they should choose. Unfortunately, before I got there, they had decided (at a senior level) to go for Exchange for whatever reason. My input in this meeting was therefore somewhat diminished. (Its important to note that this discussion is not about this customer, but on the exchange/domino technical discussion)

However, listening to the discussion was fascinating.

  • Since Exchange is tightly bound to the underlying Active Directory infrastructure, this company had to basically sort that out first, migrating three separate environmments into one. Which was going to push their project back several months.
    Compare and contrast: Domino has no AD dependancy. It *can* use AD for single sign-on, but does not rely on this.
  • Exchange *can* be ran on a "several forest" AD infrastructure, but exchange users on different forests cannot "see" the lists of other users on other forests. I guess this is analogous to separate Notes domains, where we'd just replicate the "other" domains' directory and use directory assistance to offer it to the users. Talk was of a third party tool which would copy across the "other" forests exchange users on a scheduled basis. This seemed like a huge hole to me.
    Compare and contrast. Separate notes domains can share directories and use them "out of the box"
  • Clustering. Long discussion on clustering their exchange servers, either using MS technology straight out of the box, or third party tools. No-one around the table seemed to think it was at all reliable. Some testing had been performed, simulating an exchange server failure. Result: fatal database corruption.
    Another fairly fundamental point - exchange clusters seemed to be grouped into "Active/Passive" pairs. Meaning that only one out of the two Exchange servers were actually being used, and the other was there for just failover. In a Domino cluster, all servers are "active".
    This is scary stuff. I mean, if you cant trust the clustering, what can you trust ?
    Compare and contrast. Domino clustering had been publically available since the release of v4.5, and is supported on between 2 and 6 servers (all active at the same time if you choose). (I've seen clustering on more than six - its just not supported). You can cluster between differernt versions, different platforms, etc. And because each domino application is held in a separate database, you can (as a function of database deployment) choose which way the users will fail-over in the event of hardware failure.
  • Database infrastructure. Exchange 2003 is still a "Jet-based" (pre MS-Access v2) database infrastructure - MS killed off the "Kodiak" database enhancement. Reliability, anyone ?
    Compare and constrast. Domino ".nsf" files have evolved with each major release, and are backward compatible with all previous releasess. I can still create a domino v4 style database with my domino v7 client, by using the database extension ".ns4" instead of ".nsf". Domino databases have increased in speed and stability in each release, and of course in v7, you have the option of using DB2 as a database store also.
  • Notes Applications. This customer had a number of notes servers with 30+ critical applications, which they have to support for some time. So they need to web-enable these applications and support at least a single domino server until they migrate to a new application infrastructure. Note that the "new" application infrastructure is likely to be MS-SQL and VB.
    Compare and contrast - Notes applications are upwardly compatible from version 2 to this years hot new number - version 7. So you can take a v2 app, put it on a v7 server and it works. No "install this DLL" VB-style application distrbution, etc. Notes applications can be  web, fat client, mobile, or all three.
  • Futures. The latest exchange announcement - "E12" - talks of not a lot of enhancement in the latest version of exchange, and since its likely to be based on Longhorn, is in the dim and distant future. We were wondering why the announcement was called "E12". Well, one possibility is that since Exchange first came out (in beta) as v4.5 - in order to mirror the Notes versions  - that this version of exchange was going to come out around about the same time as Domino/Notes v12 - sometime around 2018... (Okay, we're kidding on that one..).
    The MS collaborative application infrastructure seemes to have swung back to sharepoint, and to building more capabilty into Word and Excel. But till we see them released, no one knows.
    Compare and contrast. Domino v7 - due this summer - is on public Beta. Domino 8 - Next autumn. Domino 9 - year after that, etc. And since there's no "rip and replace" between versions, you can choose what you wish to do. Another point to note is that as part of a migration exercise, its possible (and fairly typical) to run mixed-versions in a single environment. (For example, in another customer site, we have OS/2 v4.5 servers, NT based v4.6 servers, win2k based v5 servers and are migrating to v6.5 based win2k servers. No problem).
    One point to note. According to Garner, Domino has 48% of the market, compared to exchange on 42%. Garner obviously is an independent analysis firm. Whilst Mr Gates may talk of "Exchange has 75% market share", I have not yet seen any validation of this figure.
    Secondly. Consider the IBM Workplace services express offering. A very sexy, "built on portal" sharepoint killer. Who is it aimed at ? My guess is the 50% of exchange sites still on v5.5 (now out of free support) of Exchange.

The meeting obviously contained a lot more discussion than these few points.

I guess the important thing to remember is that Domino is (certainly in this case) technically a far more capable application and mail infrastructure than exchange, and has been out there in a stable, supported manner for far longer. We as Domino consultants should keep hammering away at our customers, showing them at a business level why the two products are *not* the same. That our offering is far more capable, scalable, and business-centric.

Its going to be an interesting year.