Some Startups Should Be Avoided
|"A Series of Tubes"|
Olympus E-3 with Zuiko Digital 12-60mm
1/160s, f/4.5, ISO 100, 60mm, -1 EV
There's a blog posting loose on the Internet titled "Why we don’t hire .NET programmers" by David Barrett, CEO of Expensify. David Barrett's posting is a classically trollish post illustrating several key flaws found in too many startup founders: a toxic mix of exaggerated self-importance and an unreasoning bias against Microsoft technologies.
David Barrett's exaggerated self-importance is quite evident if you look at his company's website. The company, Expensify, is dedicated to producing an expense reporting module for Salesforce and Google apps.
With all the money they've raised along with all the trash-talk about .Net programmers (and Microsoft in general) and the boasting about the staff, you'd think they were working on a truly hard problem such as practical parallelism or new and novel extensions to artificial intelligence, not a new way to do web-based expense reports for Salesforce.com.
Yes, a gilded web-based expense report. If you read the description of what the company's one and only product does, its first key capability is it "imports expenses and receipts straight from your credit card or bank account." While it appears they get it with regards to security (at least they have a web page covered with security totems) when accessing sensitive information, when you eliminate the human-in-the-loop you increase the chances for other classes of mistakes. Just because you can automate a process doesn't make it any more mistake proof than the process you've replaced.
The piece is larded with left-handed compliments towards Microsoft and .Net. On the one hand "it's dandy", but on the other hand "startups, by definition, need to think of things from new angles, and those angles typically don’t involve .NET." A position which he fails to prove either with solid facts or with his "simile", which, by the way, is actually a metaphor.
Startups should innovate on top of existing software tools irregardless of their provenance. Everything in the IT landscape should be grist for a startup's mill. .Net isn't a language, it's a broad umbrella covering Microsoft's collection of technologies that include languages (C# being the most prominent), underlying infrastructure (CLI/CLR), all the supporting libraries (which you can use or discard as needed), system plumbing, cloud connectivity, tools, and applications that run on top of it all. Startups should understand this, deconstruct it, and incorporate those elements that make sense into their creative plans. This incorporation and re-use is equally applicable to other software technologies, such as Java and Java Enterprise, C++ and the multitude of frameworks such as Qt, Ruby and Rails, Python and Django and Zope, etc, etc, etc. Whatever it takes to fulfill the mission.
Startups should tackle tough, seemingly impossible to solve problems, not pursue some trivial opportunistic angle a buzzword-compliant marketing droid could think up. The only bias a startup should have is against closed minds. Startups should be focused on their primary goal but be aware enough to avoid the technological cliff edges that dogmatic blindness can hide and be nimble enough to execute a course change when change is called for. I see none of this in Expensify, either by reading the CEO's posting or by looking at the Expensify website.
This is neither a company I would want to work for nor is it a company I would recommend to others.
Update 28 March
An amazingly eloquent and humorous refutation: "Why I Don’t Hang Out With People Named David". How I so wish I could write like that.
Update #2, 28 March
An articulate and well-written post that explains what the real problem may be with people who select .Net. I don't agree with what he says, but the presentation is professional and far less confrontational than Barrett's. "Corporate stereotypes, and why Microsoft could kill your startup career" is a thought-provoking piece written by Matt Swanson, who understands how to professionally present his arguments. Swanson is CTO of SpeakerText, what looks to be a pretty interesting startup.