Archive for September 2008
As I’ve mentioned, my recent thoughts on client tools (see here and here) have prompted a lot of interest around the Analysis Services community. One result is that Panorama have asked me to host a webinar for them where I get to sound off about the state of the client tool market before they show you their latest stuff. Yes, I’ll be paid for it but I’m not going to be promoting their products directly (I feel like I need to justify myself!), just repeating my standard line that if you want to do anything serious with Analysis Services you should at least check out the range of third-party client tools available rather than stick blindly with what MS gives you – and given that Panorama are the largest vendor of third-party client tools for Analysis Services, they deserve to be on the list of tools to check out. Here’s the link to sign up for the webinar:
So the SQL marketing people have put together a website to
plug SQL2008 share the SQL Server 2008 experience:
Actually, there are some quite interesting (in a Channel 9 sort of way) videos up there in the ‘For Developers’ section, including a few interviews with members of the AS dev team. Worth a look.
While I was at SQLBits I had the pleasure of meeting Tomislav Piasevoli (who has come all the way from Croatia especially), someone who has been very active on the Analysis Services MSDN Forum recently and who knows a lot about MDX. His company, Softpro, sells an Analysis Services client tool called Cubeplayer and he very kindly gave up his lunchtime to gave me a detailed demo. As I said recently the general feeling of frustration surrounding Microsoft’s client tool strategy has made me look again at the third-party client tool market and decide to review some of these tools here, and this look at Cubeplayer is the first in the series. Remember, if you’ve got a client tool you’d like me to look at, please drop me a line…
The first thing to say about Cubeplayer is that it’s a tool for power users and consultants, not the average user who might want to browse a cube. As such it’s going to appeal to the fans of the old Proclarity desktop client, which it vaguely resembles in that it’s a fat client with a lot of very advanced query and analysis functionality. It’s not part of a suite – there’s no web client etc – but it includes dashboarding functionality that’s only available through the tool itself, and also has the ability to publish queries up to Reporting Services.
What can it do? Well, the web site has a good section showing video demos of the main functionality, but here are some main points:
- It can certainly do all the obvious stuff like drag/drop hierarchies to build your query, as well as more advanced selection operations such as the ability to isolate individual members in a query, drill up and down on individual members or all members displayed. It also has a number of innovative features like the ability to click on a cell and drill down on it, which means that you drill down on every member on every axis associated with this cell.
- It also supports all the more advanced types of filtering and topcounts that you’d expect from a tool like this; in fact, it seems to do pretty much anything you can do in MDX. This does give you an immense amount of power and flexibility, but sometimes at the expense of ease-of-use. Take the old nested topcount problem, solved in MDX using the Generate function: any advanced client tool will have to handle this scenario and Cubeplayer certainly does in its Generate functionality, but would you even expect a power user to be able to understand what’s going on here and remember they have to click some extra buttons to get this to happen?
- There’s a nice feature where you can choose to display the data in your grid either as actual values, ranks, percentages or other types of calculation. This makes it really easy to make sense of large tables of data.
- It has a whole load of built-in guided analyses such as ‘How Many?’, ‘Show Me’, ABC analysis (for segmentation) and Range analysis. This for me is a real selling point – I’ve been asked several times, for example, about doing ABC analysis with Analysis Services and I’m not sure I’ve seen another tool that does it.
- Another cool thing I’ve not seen before is the ability to put two queries side-by-side and, if they have selections on the same hierarchy, do operations like unions, intersects and differences on the selections.
- There’s an MDX editor (with intellisense and other useful stuff) where you can write your own queries. Again, not something that even a power-user might want to do, but if you’re a consultant who knows MDX it’s a useful feature for those times when you know you can write the query but you can’t get the query builder to do exactly what you want.
- You can generate Reporting Services reports from a query view. You probably already know my opinions on the native support for Analysis Services within Reporting Services, and this certainly does make it much easier to create cube-based reports.
Overall, definitely worth checking out if you’re in the market for this type of tool. There are a few criticisms to be made: as I said, I’m not sure it’s as easy to use as it could be although this is partly the price you pay for the richness of the functionality; there are some strange lapses in UI design such as the way all dialogs have ‘Accept’ and ‘Cancel’ buttons with icons on, instead of the usual plain ‘OK’ and ‘Cancel'; and charting is competent but not up to the standards of the best visualisation tools (I think many vendors would do well to look at their tools and ask themselves "What would Stephen Few think?" – it might not be very complimentary). In my opinion, though, it’s a very strong and mature tool and the positives far outweigh the negatives.
One final point: Tomislav mentioned he was looking for reseller partners outside Croatia. If you’re interested in this I can put you in touch with him.
As you may have already seen, Microsoft has just announced it has closed its acquisition of DATAllegro:
More details will be forthcoming at the BI Conference, CTPs will be available in the next year or so and the product is slated for release in the first half of 2010 (for SQL2010 then?). It’ll be interesting to see what form it actually takes exactly. As I’ve said before I suspect it would be easier to get Analysis Services working with this technology than many people think, and I also still wonder whether MS might also be thinking about buying or building a column-oriented database too.
As a counter to this euphoria, here are some interesting links… First, take a look in the discussion in the comments here:
Plenty of people with an axe to grind, I’m sure, but some points worth considering. Also worth checking out for a dissenting point of view is Kevin Closson’s Oracle blog, which discusses DATAllegro a few times, for example:
It’s always good to take the blog-equivalent of a cold shower before getting too worked up about a new feature or technology. I’m sure that the guys at MS have done their due diligence on DATAllegro, and that come 2010 we’ll have a solution that is way, way more scalable than what we have right now, but let’s also set our expectations appropriately – it is going to be a version 1.0, and the competition isn’t going to stay still in the next two years either.
Thinking more on the subject of Chrome and web-based BI, my lunchtime browsing today brought me to the topic of site-specific browsers, something I’d not come across before although it’s been around for a while. Here’s what Wikipedia has to say about them:
There are quite a few site-specific browsers out there but the Bubbles web site does a good job of showing off the kind of thing that’s possible:
For someone like me without a financial background, the default behaviour of AS when displaying a hierarchy seems natural: when you ask for a set like MyHierarchy.Members, you’d get all the parent members appearing in the list before their child members. But for financial reports you want the opposite, you want to see all child members displayed before the parent members. The other day I was asked how to do this in turn and asked Andrew Wiles (who has a lot more financial OLAP experience than me), and he pointed me to the simple answer – the HIERARCHIZE function.
Here’s an example from Adventure Works:
SELECT [Measures].[Amount] ON 0,
FROM [Adventure Works]
Run the query and you’ll see the following returned, with the parent member Statistical Accounts displayed above its children:
However, if you want to display it in the financial style, all you need to do is wrap your set with HIERARCHIZE( <<set>>, POST) as follows:
SELECT [Measures].[Amount] ON 0,
FROM [Adventure Works]
I have to admit, I’d always wondered what the point of Hierarchize() was – this is the first time I’ve actually needed to use it!
Phew, another SQLBits done and dusted. And I think it was a good one, especially looking at the comments so far:
Thanks to everyone that came and especially to everyone that helped out! And if you weren’t there, the best thing is that Microsoft brought along a camera crew and filmed the sessions (I think all of them, certainly all of the sessions in the room I was monitor for) which should be made available online somewhere soon. Apologies to everyone who stopped me to say hello – I was running around like the proverbial blue-a*sed fly all day so I couldn’t stop to chat with anyone for more than a few minutes, I hope I didn’t seem too rude…