Tuesday, 24 March 2009

"The Magic of BizSpark"

Today I attended a one-day event, hosted by Microsoft, at their London Victoria offices, entitled "The Magic of BizSpark", intended to promote Microsoft's new BizSpark offering which is aimed at helping new start-ups who want to get started using the Microsoft technology stack, but would ordinarily find the cost prohibitive to doing so.

It's a no-brainer!

I should say, right from the get-go, that if you're familiar with the Microsoft technology stack and favour it over other open source alternatives, the BizSpark initiative is a complete no-brainer. It's an extremely generous offer from Microsoft, and one that will undoubtedly help them break into new markets. It's a bold and brave move. If I had a serious software product I wanted to launch, and had formed a new company to do so (the initiative is limited to NEW start-ups), I'd be all over this initiative like a rash! You get 3 years of free software (and much more complete software than I get with my £2000/year MSDN subscription) and free 'Production usage' licenses on top. That's a very sweet deal!

Notes on the Initiative

DreamSpark is a similar initiative to BizSpark, albeit one aimed at the student community. I believe it's possible to migrate from DreamSpark to BizSpark if that makes more sense (although I may be getting confused by my notes - it's possible that DreamSpark was only mentioned as being an initiative that helped motivate the launch of BizSpark).

BizSpark is an accelerator program that was launched globally in October 2007, involving 161 companies worldwide. To date Microsoft have invested and sponsored more than USD100 million in software licenses and over 2,100 hours in workshops and training (when I say 'over' these figures are apparently several months old, so likely to be significantly higher by now).

When you sign up to BizSpark you sign up to a 3 year programme with no initial cost, and just a USD100 'exit fee' at the end of the 3 years. If your start-up goes bust, or disappears in the first three years then that's it; otherwise you start to pay for what you use in your fourth year.

The web sites MicrosoftStartupZone.com and BizSpark have more information.

Production licensing is through hosting partners (or you can self-host). Currently there are four network partners who sponsor startups, which include (I think - my handwriting is poor and I may have misread the company names!) NTT, Attend, 7Global and Rackspace. There has been some criticism of the cost of these hosting partners but Microsoft are actively looking for other, cheaper hosting companies and expect to announce more partners shortly.

You get two support incidents/year included.

Membership of BizSpark now automatically gives you entry to Metro, a pre-release software programme that includes not just pre-release software, but also support for that pre-release software and training too. The Metro portal allows you to enter your project details and the pre-release technologies you're interested in, and will give details of (mostly) free training available in Europe.

The program also offers free one-to-one 'Fast Track reviews' where you can discuss your architecture and use of different technologies to get advice from Microsoft on possible alternatives or additions.

The slide presentations from the day's talks can be found here. General queries about the initiative can be sent to ukdev@microsoft.comn.

Like I said, it's a sweet deal!

Why I came Away from Today's Event Disappointed

With little to go on other than the title 'The Magic of BizSpark' I had expected more on the initiative itself (we got a 20 minute introductory talk on this, the main points of which I've summarised above, and it may be this was all that was needed, but I'd expected more discussion around the criteria for entry)

The bulk of the day was really very much a high-level set of pitches on the beta/CTP Microsoft technology stack. I'm often called a 'Microsoft bigot' but even I found the talks at times misleading and, I would argue, potentially dangerous for those who might be making decisions on the future of their business based on what they were told about these technologies today.

For example, we got what I felt was an over-emphasis on use of the UpdatePanel in AJAX, in a 60 minute talk that had to cover every single technology in the RIA space. OK, so maybe I'm biased because a greenfield project I worked on got canned when we discovered, naive fools that we were, much too late in the day that the UpdatePanel doesn't scale or perform with anything other than a few simple controls embedded in it; or maybe I'm biased because I once worked for a client where the CTO advised his staff that if he ever found anyone using an UpdatePanel control they'd be fired, but I've since heard other Microsoft staff give quite clear warnings on avoiding use of the UpdatePanel. The reality is people NEED to be warned that if they drop a datagrid or a hierarchical tree control into an UpdatePanel it won't perform or scale for more than a handful of concurrent users (at least if my experience is anything to go by).

I felt that the demonstration of Silverlight showed none of its strengths, and just made it appear overly complex and confusing. Wow your audience with a real demo from the silverlight.net site, not five coloured buttons in a stack panel or a simple data binding form where it's really not clear what advantages the technology offers over the much more established and seemingly easier-to-understand ASP.NET. Nor did the talk really mention any of Silverlight's current limitations that might force one to choose ASP.NET/AJAX instead (lack of printing/webcam/microphone support until version 4 which is a long way off with version 3 having only just entered the 'preview' stage; the difficulties of debugging 'code' that is essentially just declarative angle brackets, etc).

Several times an assertion was made that when compared to the open source alternatives the MS technology stack was integrated and much easier to get to grips with. When we're talking CTPs, betas and Out-of-band releases (which we were!) I would argue the Microsoft technology stack is typically just as problematic as any set of open source products (especially when so much of it is playing first release 'catch up' with free 'competitive' products that are far more mature, having had several years head start: Unity vs Castle Windsor, ADO.NET Entity Framework vs NHibernate anyone?)

A talk on the new 'Velocity' caching system as part of a 60 minute talk that had to cover ALL of Microsoft's data products got very confusing very quickly, and it was only after some quite heated debate with strong audience involvement that the real advantage scenarios for its use (cf just letting SQL Server take care of caching) became evident. Admittedly I may have stirred the pot by asking why Microsoft's caching advice appeared to have changed (the advice always used to be that SQL Server was optimised for caching and should just be used asis instead of trying to come up with home-grown caching approaches that inevitably caused more problems than they solved) but the marketing advantages and benefits of Velocity (over straight SQL Server caching) should have been understood and related at the start of the pitch, not at the end of it after a rather painful debate! For the record, Velocity looks like it will be of most use where you're using multiple databases or have an MSN 'friends online' -type scenario (where the data can be large and considered throwaway). The advice on Velocity was to hang on for a couple of weeks when CTP3 (the last CTP before RTM) should be released since CTP2 is a few months old now. Velocity is scheduled for RTM release in Summer this year and will be a free Out-of-bound release.

I guess the problems I perceived with the day's main pitches partly arose because Microsoft themselves were unsure exactly who the audience would be. There were 13 of us, comprising two people with PHP/MySQL experience (although both had used MS technology in its pre-.NET days) and the rest was then evenly split between those already on the program who regarded themselves as 'very technically savvy' in the Microsoft stack, and those who weren't on the program, who also identified themselves as being very technically savvy. In all honesty, if I'd known that 'The Magic of BizSpark' actually meant 'MSDN Day Lite' I wouldn't have taken time out to attend!

Am I being too 'mean' ('mean' seems to be Microsoft's favourite new word after 'So...')? I honestly don't think so. One attendee, totally unprompted by me I hasten to add, commented in the lunch break: 'It's typical Microsoft - all thrown together at the last minute, without thinking through what the marketing message should be. They don't seem to realise that start-ups just want clear guidance on what they should use, and when'. We now live in a world where even the likes of Scott Guthrie can be happy using words like 'NHibernate' in his MIX keynote, or Microsoft UK's data evangelist Eric Nelson can be open and honest about the very real weaknesses of the first release of the ADO.NET Entity Framework, so it's hard to understand why start-ups aren't being given the real story about the very real pitfalls they will face with certain parts of the technology stack, effectively being told everything's wonderful and much more integrated and easy to use than anything else out there. If we're told about the weaknesses we can work around them (and wait for the fixes), but if a product is completely oversold and the very real weaknesses covered up such that products are developed with a real risk of failure at a time when it's too late to fix things people will quickly become disillusioned and angry, and that benefits nobody.

All that being said, if you're working with the Microsoft technology stack BizSpark is a complete no-brainer. Use Google to find the 'real' story about the technologies you're thinking of using before leaping in and you should have a very happy experience. The software alone (Not just full-blown Team System Suite edition but production licenses too! For free!) is worth considering. Add in access to Metro (a pre-release programme offering not just pre-release software, but free support and training too) and it's not hard to see why lots of folks are excited about the initiative. I just hope those folks have the common sense to look at what the cost of all that 'free' software will be once their three years are up! The final demo of the day, extolling the virtues of all the options in Team Foundation Server, was impressive but I couldn't help wondering how many people would still be interested once they realised how much this stuff will cost them when they get to Year 4. There are much cheaper (and, I would argue, more mature) products out there for much of this functionality).

No comments: