Why CMIS might just work

The voting for CMIS v1.0 is underway, and bar some unforeseen hitch is likely to become established at the end of this month.  Three cheers for the new industry standard, which will make life easier not only for third party vendors creating tools for the conforming platforms, but also for all those companies who use more than one platform internally – and let’s face it, are there any out there that don’t?  Which means life is set to become easier all round.  Or is it?

Before we cheer too loudly, let’s be silent for a moment and think of all those past standards that have bitten the dust, or failed to become universally accepted and used, at least for content management.  WebDAV, where are you now?  Well you are still around, but you didn’t catch on for the CMS industry.  A few companies put a lot into this, but as an EMC source involved in their CMIS implementation said to me at the AIIM conference yesterday, it never reached critical mass and so the work was effectively wasted – leading at least this person to feel “jaded” about any new initiatives including CMIS. JSR170 and 283 took over 7 years to get finalized.  In the pharmaceutical industry, those like us who lived through and participated in the birth of the eCTD standards know how much work was required by many individuals, companies and regulatory organizations in order to come up with a stable, usable agreed standard – and can see how easily such standards fall into difficulties such as the European PIM standard for product labeling.

So what makes anyone believe in CMIS?  A standard that just OASIS was calling for participants for just 18 months ago?

Well, there are many reasons in my view.

1. There has been an immensely impressive speed of development, agreement and involvement among industry vendors on this standard.  To have arrived at a v1.0 in a short time, and many of the vendors have already built and issued their CMIS layer, notably Alfresco, Nuxeo, IBM and EMC.  Microsoft, after keeping everyone guessing for a while, has just announced their SharePoint support for CMIS in June this year.
2. The involved vendors are actively reaching out to third party vendors to work with them on CMIS.  In March we released our CARA3 product, the world’s first full functionality web application that can connect to multiple repositories using CMIS + repository-specific extensions, and we were almost immediately requested by several of vendors, including Nuxeo, IBM, GateIn and HP, to extend our testing to their platforms from our initial Documentum and Alfresco releases.  This indication of willingness to collaborate is very positive.
3. Those involved in the definition are all in the same line of business; unlike some standards such as those for the pharmaceutical industry that required vendors, customers and regulatory authorities, who all had very different requirements and end goals, to collaborate, the interest group for CMIS is more focused and can therefore advance in a more unified manner.
4. The standards have started with a reasonable scope – v1.0 is of course the initial “lowest common denominator”, and lacking many of the more advanced services and features that most repositories offer, but it is enough to perform the common basic CMS functions, without trying to do too much too soon.

So will it succeed?  I like to think so.  Sure, there is a way to go in getting to a v2.0 or v3.0 that contains a richer set of functions – and at least one vendor expressed hesitation about getting there, because at that point there would be very little in terms of barriers to entry to the market or customers switching from one repository to another, so the industry leaders involved might just be building in the demise of their own dominance; and of course, there are a good number of vendors still out there (who I spoke to at AIIM) who look blank when you mention CMIS, and therefore are clearly a long way from implementing it.  But overall, with the momentum it has behind it, the underlying technology being mature enough to handle it, and the stability of the initial standard, as well as the obvious desire for those involved to see it succeed, I do believe that this time we might just have something that will work.

Advertisement