Andy's World: Microsoft

Your XBox360 can blog - the beginning of the end?

Wed, 06 Feb 2008 07:08:08 GMT

The text is obviously templated (like some of those bad online astrology forecasts), but the concept is pretty funny - a blog for your XBOX 360.

Check out what consoles around the world are saying about their owners...

I may have to let my box indulge... ;-)

Living the digital media lifestyle

Fri, 12 Oct 2007 08:36:01 GMT

I'm a big believer in the power and value of continuously improving technology. It shouldn't be a surprise to anyone who knows me that my home is a veritable bounty of laptops, desktops, DVRs, HD video, media players, game consoles, digital cameras, and wireless networking.

I love it when stuff just works. Apple gets praised for this all the time, and I give them props for much of what they do, but just because the chains are gold, doesn't make them any less restrictive.

Getting an end-to-end experience to work and hold together is challenging, but ultimately very rewarding for the customer. Some recent personal examples - my DirecTV TIVO DVR. XBox Live Arcade + 360. Halo 3.

Getting an experience to work with an ecosystem is just downright hard and not for the faint of heart. It so frequently fails (e.g. Windows), that customers are surprised when something actually works. Recent example - I have a Creative Zen Microphoto that was a pain in the ass to get working with my Dell running WinXP. Lots of software to download, install, etc. While I was on my recent trip to Toronto, the battery was about dead, so I decided to try plugging it into my Toshiba running Vista, even though I hadn't installed any software or drivers. Lo and behold, in under 10 seconds, Vista recognized the device, configured it, and offered to sync.

Nice!

What really sucks though are when things you expect to work, don't.

Example 1 - HD DVR from DirecTV. My family was so happy with our Tivo DVR from DirecTV that we held off on getting HD until this summer. It looked like they had worked through most of the bugs and the box was ready for rollout - it was on sale, things looked good. Unfortunately, the whole experience was a complete disaster. The installer was late, took forever to complete, and left the old dish lying on the ground outside our house. The unit itself kept locking up and had to be restarted multiple times a day. Shows wouldn't record, or it would look like they were recorded, but they would disappear after a restart. After a couple more software updates, things got a little better and the thing has somewhat stabilized, but it's really soured my family's relationship with DirecTV. What a bummer.

Example 2 - Music subscription from Urge / MTV. I refuse to buy in to the Apple golden-chain model. We've got a number of Creative Zens in the house, so I wanted a service that would support them. I also really liked the new media library and searching I got from WMP 11 and Urge. We have a pretty extensive CD collection, but we're always on the lookout for new music, so a subscription seemed like a great idea. At first it was. But when I started trying to actually use the download and sync with media device, I could never get it to work. After multiple calls to Urge support, I gave up trying to get subscription music onto my Zen. To add insult to injury, the whole Urge / Rhapsody merger now means I have to ditch WMP 11 and start using RealPlayer (ugh). What a piece of garbage.

So - experiences matter. Now I just have to get them working...

Software development at Microsoft and the pernicious spec

Wed, 07 Feb 2007 08:34:10 GMT

Having worked for a number of product development teams at Microsoft over the last nine years, I can tell you that there's one constant about the Program Management job - every team tries to be rigorous about their feature design process, and every team ends up doing it differently. (the more things change...)

Enter the pernicious spec.

In concept, specs are intended to capture the detailed requirements and design for one or more functionality aspects of a software product. In reality, Specs are both evil incarnate and the best thing since sliced bread.

Huh?

Their duality is completely reflective - they are what you make them. And therein lies the key to walking the fine metaphoric line between heaven and hell.

I've been on a number of teams that treated the specification process very loosely - if it got done, great, if not, as long as development felt they knew what had to be built, that was OK too. The eventual problems with these situations is pretty obvious - how do you know when something isn't working the way it's intended? Without a written statement (e.g. a spec), the options are a) guess, b) use your best judgement or c) get the team together and decide. To be fair, spec or not, this still happens a lot later in the development cycle when you start finding 'bugs' that no one expected - that's what team bug triage is for. BUT - if you take that route from the very beginning, you're asking for trouble - you're either going to get a failed feature or a very small set of people who actually know what the feature is supposed to do (try not to think about what happens to your product development when those people move to a new team or company).

Scrum seems to be the latest favorite excuse for some teams to stop writing things down - "We don't need to write specs - everything we need to document is in the code!".

Yeah, right.

OK - so not writing specs is bad. What's wrong with writing them?

In short - nothing is wrong with specs themselves - they're just documents after all. What *is* wrong is treating specs like they are the code that you are shipping. They're not. Until someone (probably Microsoft Research) invents the magical spec compiler (followed soon by the mystical v2 Executive Edition - the Powerpoint compiler), specs don't ship. Customers don't install specs. Partners don't sell specs. Retailers don't stock specs on their store shelfs. You get the idea. Some teams don't.

Specs are tools used to track and capture the creative and synergistic design work that your software development team does to figure out how they are going to implement a feature. Specs are artifacts of a much more important effort - figuring out what to build. When specs start to stifle that creative energy rather than foster it, you know that you're due for a change.

Of course, getting teams to go through a culture change is probably best saved for another post...

Until then, have fun designing!!

My refresher course as a small business consultant

Tue, 19 Sep 2006 07:56:23 GMT

My kids just started at a brand new private school (Tall Cedars Academy) this fall and they're having a great time. The headmaster, Bill Alsdurf, is a former teacher of theirs, he's run private schools before, and he does an amazing job with the kids. Over the last few years, Bill has become a friend of the family, and he's on a startup budget, so of course I offered to help him get his IT infrastructure going.

Hmmmm - and I'm paying him *how* much in tuition this year? ;-)

Ah, the fresh smell of small business consulting in the morning!

Actually, I kind of enjoyed doing it. See - I used to work at a smaller IBM business partner in San Diego that sold Point-of-Sale and Inventory Management to speciality retailers and the like (e.g. Williams Sonoma, Harry & David, Claires, Petco). In addition to the bigger guys, we sold a lot of smaller deals to - 2 and 3 story chains that were just getting started. It was fun meeting different people and getting the systems set up for their business. I was mostly on the senior consultant side - managing projects, requirements analysis, training, and the like. Occasionally I had to get involved in a lot of crawling under tables and counters to install POS terminals and help troubleshoot wiring. Lots of dust bunnies, wires, bad connections, and occasionally faulty hardware.

Doing the stuff at Tall Cedars reminded me a lot of my time in San Diego.

It took the better part of the Labor Day weekend, but in the end, we got everything up. In the interim, I ended up consulting with Bill on a wide assortment of things - which computers to buy, what hardware he needed for a touchscreen that students could sign-in with, how to set up his phone system, etc. I also did the dirty work of getting everything installed - drivers, cables, USB hubs, print servers, touchscreen, security patches, etc.

The experience left me pondering a couple of thoughts that I figured I'd share:

It's all about Trust, stupid! - I hammer this point on a regular basis with MSCRM partners that I talk to, but I can't over-emphasize how important it is to build trust with your clients. They're looking to you to help them make smart decisions for their business - take the time to learn about what the customer does, how they do it, and why. You'd be surprised at what you can learn. Pragmatic empathy is a great skill for a consultant.
The coming SaaS battle - shame on Microsoft for not doing a better job with providing small business services. The Windows Live services are fairly invisible to the average small business owner and the next step direct from Microsoft after that is the purchase of Small Business Server ($599 for Standard, $2000 for Premium server s/w and 10 CALs). For the moment, he's just doing a peer network between PCs, but we're talking about doing something more sophisticated. Here was a guy who needed *everything* - HR, accounting, student management, e-mail, network access, internet filtering, business web portal, custom domain, fax services, client PCs, productivity software, educational software, etc. How much did he buy from Microsoft? Windows and Office for each PC. Everything else went to a large number of other vendors. I'm a firm believer in the power of on-premise software, but I'm also positive that Bill shouldn't have to think about the complexities of software management and shouldn't have to pay $200 - 500 / month to outsource management of a server to a small business partner. There's no question that Software-as-a-Service is the future in the small business space. As Om Malik points out (and fortunately for Microsoft), no one else is doing a very good job here either (even Salesforce). The next couple of years are going to be interesting...

So - the customer is happy, the dust washes off, and my kids get a kick out of signing in with the new touchscreen. Not bad...

Now I just have to figure out what to do about Small Business Server for him...

Adjusting to Spaces upgrade

Sun, 06 Aug 2006 08:23:01 GMT

The upgrade to the new Windows Live spaces has been a pain in the ass. The site has been effectively unavailable to me for the last few days, so I haven't been able to post anything. I'm hoping the redirect is working now...I'll be interested to what happens with the social circle stuff...

WWPC 2006 Coming to a Close

Thu, 13 Jul 2006 19:43:18 GMT

Today is the last day of Microsoft's Worldwide Partner Conference in Boston. It's been quite a week. The first few days have been great - lots of interesting discussions with partners, lots of excitement over the announcements around CRM (especially CRM Live), and lots of walking :-) (the Boston Convention Center is pretty big, and I have a pretty nasty blister on my heel to show for it).

Overall people are very excited and agree that FY06 was great but FY07 is going to be an amazing year for Microsoft CRM!

I've got some expo booth time and a couple of additional meetings this afternoon, but I'll have some additional posts later tonight with some thoughts on things I've seen and conversations I've had over the last couple of days...

Money = Horns?

Fri, 30 Jun 2006 08:27:49 GMT

Phil Richardson agrees that Microsoft CRM can become a billion dollar business. He also makes some interesting analogies about MS culture where products treat their revenue as status symbols. I couldn't agree more.

What's even more interesting is how much large amounts of revenue warp your perspective. Besides leading product teams to make uninformed (or even plain misguided) technology decisions, you lose sight of how amazingly unique it is for a business to make hundreds of millions of dollars in profit. For example, I was meeting with a partner a few weeks back who was ecstatic about how he had spun up a new MSCRM business in 3 months, was hiring people as fast as he could, and was still in the black by $10,000. Would that kind of money make a difference if it was my bank account? You bet. How does that compare with the $12 Billion that the Microsoft Business Division made last year? It doesn't, and it's not supposed to. But it's easy to lose sight of that when you are removed from the business-side of things as often happens with large groups at Microsoft.

It's practically heretical given MSFT cash management policies, but I'd love to see MS institute revenue sharing with all employees and give all remaining profit back to shareholders as dividends. If you're making that much money, don't be ashamed to put it to effective use. On the philanthropy side, that's what the Gates foundation is all about. On the business side, it could show as a lot of things. The revenue sharing would need to have a framework around it (e.g. different models based on business / product stage), but it would be great to see the rank-and-file developer get as excited about landing that big deal as the field does.

I'm likely rehashing the long-famous 'weenies, not shrimp' memo, but there's certainly a lesson here that anyone working on a software project has to internalize for themselves. For just about every other business in the world, every dollar and customer matters. If you're making software for them, it pays to remember...

The WinFS saga - over or beginning again?

Wed, 28 Jun 2006 08:51:33 GMT

Before I joined the Dynamics CRM team, I spent some time on the Windows Shell team responsible for thinking through what the storage, sharing, and sync experience would be for Longhorn (now Vista) when using WinFS. Keep in mind that this was before the infamous 'reset' in Aug 2004 when WinFS got the boot.

Some other folks have done a more elaborate post-mortem of what happened with WinFS and Quentin has been catching a lot of flack for his post last week announcing that WinFS was dead as a stand-alone component.

As someone who experienced a lot of the pride and pain of the WinFS work first-hand I have mixed feelings about the whole situation. I have to give props to Quentin for getting the news out - he has always been a pretty stand-up guy, and I think he got skewered unfairly. That said, his timing (announcing right after a bunch of WinFS sessions at TechEd) was questionable at best.

More importantly - what about the WinFS dream? At the time, I remember *many* long meetings trying to work through the semantics that were necessary to navigate a rich object-relational store while maintaining NTFS compat. It certainly seemed like a big challenge, but ultimately achievable. In hindsight, there's no question that WinFS was certainly a classic victim of scope creep. However, it's also important to recognize that the reality of what people are ready for is also played a huge factor - the general populace are so used to working with simple folders and explicit physical location that the concept of complex object graphs and multi-instance data stores (ala sync) may as well be string theory to them. The perfect proof point here was what happened with 'virtual folders' in Vista. Everyone that saw them in Beta 1 completely barfed on the concept because they felt like they didn't know where their files were ("it just says documents - are the on the c: drive?"). Next Vista CTP - virtual folders are gone.

It's great that we're responding to customers, but it just goes to show that customers (and the 600M desktops running some variant of Windows) are still lagging behind the techno-trend setters that are pissed off about WinFS not getting shipped.

In that respect, the decision to push LINQ and Entities into Katmai /ADO.NET is a great decision. SQL is the storage bet for MS. Would it be nice to have a rich relational store in the OS? Sure. Can I deliver an app built on a relational store (or using relational semantics) before then? You bet! Solutions built on SQL (like Sharepoint and Dynamics CRM) are already delivering richer relational semantics and experiences today that are only going to get better. More importantly, these products provide a great platform for developers to build solutions that work better in the role- and process-centric environments of most organizations. (Yes, I know those products don't address consumer scenarios and that I'm shilling for my own team a bit, but only to make the point that WinFS is/was just a means to an end.)

If there's a lesson to learn out of all of this, I think it's that continuous innovation and rapidly evolving your product based on customer feedback is a surefire way to stay relevant and meet your objectives. As a wise man once said - don't try to boil the ocean. :-)