* What does IBM have to gain and lose in Aperi?
What’s it all about, Aperi?
Is it just for the moment we live…?
Last time I asked two questions: When is a “standard” not a standard, and when is “open” not open?
The Aperi mission statement states: “The mission of the Aperi Initiative is to create a vendor neutral, open, standards based storage management framework and to cultivate both an open source community and an ecosystem for complementary products, capabilities, and services …”
Is Aperi really open? In a technical sense, yes. Just about anyone can join in. In a practical sense, perhaps not.
IBM finds itself in an awkward position regarding Aperi. It has, in a moment of extreme generosity, donated a million lines of code to the project in order to get Aperi rolling in the right direction. This is a good thing, but … have you ever tried to work on a million lines of somebody else’s code? Well-documented though it may be, such an effort is a heck of a challenge.
Some readers may suspect that IBM’s gift is something of a Trojan horse, containing code that will show IBM’s own products off in a better light and trusting to the idea that enough of its core ideas will make it through the review process that IBM will derive some goodness far beyond what the entire Aperi community will get. Some of you, less jaundiced perhaps, will just say that IBM has done a good thing and let it go at that.
The industry is likely to benefit form Aperi, but most of the “anti-Aperi” group are likely to worry that IBM will benefit more than most. This may be a reasonable assumption. After all, driving industry standards is an established and effective way for companies to make progress in the marketplace. IBM does it, as do all its competitors – at least those of them that are large enough to make throwing their weight around noticeable.
The whole Aperi movement would have been better served if it had been more open from the get-go. The idea of a group of companies working on something for weeks (months?) and then extending an invitation for others – key players at that – to join them only hours prior to the announcement hardly speaks in favor of openness. EMC has publicly claimed this is what happened. After all, IBM and the other Aperi founding members would certainly not expect to make corporate commitments on such short notice. It should then be no surprise that EMC, HDS, HP, Sun and Veritas have expressed zero enthusiasm for the idea of playing on IBM’s turf.
It would be nice to take IBM’s contribution of all that code at face value – certainly, end users are likely to benefit. But experience shows that it rarely makes sense to put much reliance on corporate altruism, no matter how much you like or respect the source. And in a world where perception shapes reality, how things appear is a key issue.
Thus IBM, perhaps something of a victim of its own success, may be damned if it does and damned if it doesn’t. It has been generous, perhaps overly so, but surely not selfless. We should thank IBM, but should also forgive its competitors if, in their own self-interest, they choose to beware of geeks bearing gifts.
Finally, is Aperi going to be a standard? Not likely.
Work on Aperi will be standards-based, but clearly will not itself be a standard. Why? Because standards need either to be approved by an accepted standards organization such as ANSI (a “de jure” standard”) or need to take the marketplace by storm, and become a standard by force of their overwhelming presence in the market (a “de facto” standard).
The non-participation of EMC, HP and the others, which collectively represent about 50% of the world’s enterprise storage investment means that the appropriate standards organizations – all of which have representation by both groups – would be hard-pressed to put their seal of approval on the code that comes out of Aperi. Whether or not the result takes the market by storm is of course another issue.
SNIA has given us specifications that the industry has turned into standards. Why then is Aperi doing development outside SNIA? Next time, we’ll finish up by looking at that.
