Tuesday, December 1, 2009

The editorial strikes back

FROM THE EDITOR-IN-CHIEF

By David Gewirtz

Last week, we ran an editorial "Sloppy analysis at the core of another Domino vs. SharePoint report", where I picked on the sloppy analysis that was at the core of yet another Domino vs. SharePoint report.

As you might imagine, the author of that analysis was less than thrilled (I get that a lot) and he sent me a a sternly-written follow-up email (I get those a lot, too). I've posted his letter below, and then I'll follow it up with some of my comments (which will likely result in yet another sternly-written letter...sigh).

Email from Adriaan Bloem, Analyst

Hi David,

First off: I'm not very charmed by the continuous suggestions that CMS Watch, or myself, would have any agenda to promote SharePoint over Notes/Domino. It feels like an easy vilification of anyone that dares criticize the Notes/Domino platform.

Of course, I shouldn't have suggested that the platform is going away. We've apologized profusely over that though.

But what surprises me is the assumptions in "The next part that frustrates me is when someone makes technical criticism based on something they know nothing about." Really. I do know. I've been there. And anyone who has done anything meaningful in Notes/Domino has been there. So I'm sure you're well aware you can't do @dblookups in a view column.

So it seems a bit disingenuous to pretend the problem doesn't exist.

Sure, you could say the way I put it was simplistic. It's much like saying you can't multiply in 8-bit Z80 assembly. (Of course you can; you just have to write the routine yourself.) Likewise, I remember quite a few of the workarounds. (My most ambitious one, if I recall correctly, involved something like running a nightly agent that would load the two views in separate arrays, loop through them, store the results in a third array, and then writing it to another view). And yes, now you can do that relatively easy in Xpages, and since 8.5.1, you can even do that for Notes (not just the web). But "relatively easy" is just that, because writing Javascript versions of @formula language to run as server-side webscripts (...and then using that in Notes?) is not, well, the most elegant of solutions. And yes, you could use views on a database instead, but that's sort of overkill for most of the simple uses you'd want to do those lookups for.

In short, it was just an illustration to show Notes/Domino isn't meant to be a relational DB. It really does treat documents as documents: that's also it's strength. That's probably a more interesting thing to focus on. And I would have preferred to hear more rational explanations of Notes/Domino's specific strengths. That's why I ended the blog post with "I'd welcome any additions or corrections."

Cheers,
Adriaan M. Bloem
Analyst, CMS Watch
Get the Real Story: Evaluations, Education, Advisory Services

David replies

Short answer: I think Adriaan played fast and loose with his analysis to the detriment of a group of dedicated IT people and an exceptional enterprise IT platform. The problem has been that this same fast and loose analysis has been repeated over-and-over by other analysts, creating a false sense of doom and gloom for an excellent product line and fostering a sort of conventional wisdom that's neither accurate nor wise.

Lotus is not phasing out Notes. Domino is a fine product. And SharePoint is not the be-all and end-all of collaboration systems.

Some of Adriaan's concerns about the capabilities of native views are relatively valid. However, that's not what he wrote in his analysis. Analysts, like journalists, have a level of responsibility that goes beyond that of the typically opinionated blogger. When you set yourself out as an analyst, the assumption is you did some analysis. When you represent yourself as an expert in a field, you are expected to be able to actually demonstrate expertise in that field and not just make stuff up.

When, as an analyst, you suggest a platform is going away, that's a very dangerous statement to make without serious corroborating evidence. Adriaan claims to offer "the real story" and then he pretty much invented a story, one that's hurtful to a lot of people, and represented it as real, considered analysis. Apologies after the fact just don't cut it.

Perhaps, in the future, when Adriaan posts something that he calls "analysis," he'll actually support his assertions, display the nuance that one expects from carefully considered analysis, and consider carefully before blindly supporting one vendor at the expense of the ecosystem of another vendor.