January 19, 2012 1 Comment
Right from the start I know this won’t be a pretty post. :) Sorry about that, about the rant… its just the way I see it.
Let me say honestly that my contribution to MS connect has been non existent… shame on me, no excuses, that probably somehow undermines some thoughts & perspectives that follow… But honestly, I’m irrelevant here, that’s not the point.
My friend & team mate Rui Romano was pointing me to a recent issue on MS connect (see Subscription and Alerts Issues with Analysis Services in SQL Server 2012 by Teo).
Reading (and voting btw) on the issue submitted by Teo led me to browse & search for lots of related connect issues on analysis services (mainly dimensional)…
And again it bothers me, as it appears that (to me at least) , regarding ssas dimensional, either there is no team anymore or that team is actually doing zip-zap-nothing as since the 2005 release I would say that the effort on improving the product (dimensional) is extremely lacking….
Almost all the suggestions regarding ssas dimensional, since the 2005 release, are still active or worst, closed as won’t fix.
And the few features that made to 2008 are not even rather memorable… I prefer bids helper for aggregations, the designer for attribute relations is imo worst and confusing. (linear restore/backup time is one of the few I actually find useful, on par with block mode optimizations)
It seems that ssas dimensional model boils down to some lonely guy, intern, working part time, doing….what?…documentation?
I really do find extremely valuable and needed that MS has now a very strong self service BI platform, the promising powerview and –finally- data alerts! (although I can’t believe that this amazing feature is actually “locked” on reporting services reports….. ). But aside from that, honestly, letting go of a great (and I mean great!) product like analysis services dimensional… what the hell?
By now MS should really be clear on this, where are we going, with what tools & models… and assuring that we’ll have a better platform, not just another-different- one.
And don’t even start with BI “Semantic” Model! I don’t see any clue of that for now (remind you its RC0). There are bism models, which are tabular, and there are udm which are dimensional. I’m yet to see the bi semantic model that will properly bind and unify these two worlds.
The famous post from Chris Webb comes to my mind frequently. Chris wrote at the time:
“The point I was trying to make in my original post was that the announcements made at PASS, as I and everyone I spoke to there interpreted them, made me very concerned (to say the least) for the future of the UDM and the multidimensional model. First of all there was the news that Microsoft was putting all of its development efforts into Vertipaq and BISM, while the UDM was (for yet another release) getting very few obvious improvements. Then there was the news that Project Crescent was only going to support BISM as a data source and not the UDM, which made it seem like the UDM was a second class citizen in this regard. And finally there was a lack of clarity in the roadmap which meant I wasn’t sure whether BISM was meant to replace the UDM or not, or whether BISM would ever be able to do the same things that the UDM can do today.”
The concern seems actually pretty up to date to me. Actually a little reinforced as we are now on rc0.
Its really unfortunate that so many obvious, extremely requested & frequently questioned by customers, issues remain on connect without any kind of follow up (makes sense if there’s no one on the team anyway).
Some of my favorites… still waiting…:
(ok, this one is priceless :) I forgive the team for this one)
So I would end with a public connect request: Can the ssas team please go public the ssas udm server code base?… so that at least anyone can take it further where it could be today, seven years passed the great & amazing 2005 udm release? :)
Not a pretty post as I said :), hoping that I’m dead wrong too!
Ending with a more positive tone. Comes to mind that MS should start working on a proper Data Server, the ultimate real semantic model server that will join together olap with tabular, odata driven & enabled, with real data alerting, incorporating also the kind of scenarios that data explorer is targeting today. Running on the cloud or on-premise. That would by Data Intelligence on steroids! :)
Sorry about this one!