Nimble Code: Category Net Culture

Stop Kicking The Enterprise

harrisj@nimblecode.com (Jacob Harris) — Fri, 06 Jan 2006 13:13:00 -0800

I have to agree with Obie Fernandez when he complains in his posting Enterprise Is Not An Insult about some of the recent derision directed at the Enterprise lately from the avant garde in the Ruby On Rails community lately. The problem is most of the mockery comes over a misunderstanding that “enterprise” means “insanely large”, which Obie points out is completely different from any definition used by people who actually work in enterprise software:

Plenty of us say “enterprise” when referring to internal-facing systems written by corporate IT departments and it doesn’t carry a stigma for being too large or too small, too simple or too complex. Enterprise projects might be large, take years, straddle business units and have global effects or they might be small, departmental efforts with only a handful of end users. So many factors influence the success of enterprise software in a given organization, only few of which have anything to do with technology.

Yes, there are large products that call themselves “enterprise-class software”, but that doesn’t mean that all enterprise software has to be big, slow, and hard to develop. But unfortunately, mocking the state of enterprise software has become an oft repeated dumb joke lately. This is like when a comedian throws out a laugh line – like slamming Microsoft or calling France “surrender monkeys” – to get a quick laugh; it’s funny at first but soon loses its novelty.

Giving a definition of my own, “enterprise software” means “business software.” All businesses want to increase their profits, and they do that by reducing inefficiency (ie, “we can track all of our operations within Excel”) or expanding into new niches (ie, opening overseas offices or offereing new business). Enterprise or B2B software – whether shrink-wrapped or custom – is what helps them get there. The best enterprise software becomes as essential as the CEO or infrastructure to a company’s success. And smart companies look to develop such software as quickly as possible, both to increase their profits quicker and gain advantage over their competitors. Yes, there is slow and clunky enterprise software, but there is also lightweight and rapidly developed enterprise software too. Speaking from personal experience, at Alacra, we work with a lot of vendors both large and small, and I can assure you that agility and profitability are far more important than slavish adherence to some awkward and enormous toolkit. Consider that the agile methodologies of Extreme Programming and Scrum were first heavily used in such large enterprise environments as Chrysler and Honeywell respectively. And I would argue that the success of such radically-new-at-the-time technology like the Web, XML, or Java was entirely due to the B2B community recognizing their value early and embracing them fully.

Which is why I think the jokes have to stop; does Rails want to continue to present itself as a niche tool for small B2C development or does it want to be a B2B player? Why have we gone from Rails does too scale! to Uptime is overrated? Yeah, 99.9999% uptime is overkill for almost all software, but if you’re going to sell to business you shouldn’t mock people if they want 99% uptime. They’re not whinging; they want to make your product an essential tool for their business. And so I think Rails needs to embrace enterprise needs. I think David is right to resist such code additions to the core, but I think it has a place in third-party contributed plugins and code snippets. To be an enterprise player, people will want to plug Rails into working legacy database schemas, handling composite keys, potentially even interfacing with back-end service-oriented architectures; maybe even hacking lightweight support for such horrors as SOAP or CORBA. And I think such work is what Rails needs to really succeed. Maybe you think it’s selling out, but I’d rather see Ruby in the enterprise than C#. There is a middle ground between obscurity and (code) obesity; let’s not mock people for trying to find it.

Web 2.37

harrisj@nimblecode.com (Jacob Harris) — Fri, 26 Aug 2005 00:40:00 -0700

Mark my words, Hoodwink.d is one of the silliest and yet coolest new things to hit the web in a while. I won’t tell you more, but I hope to see you there soon.

Update: I stand corrected. Hoodwink.d is really a vile plot to destroy the Internet as we know it and rob us of all we love and hold dear. The resulting collapse of faith in our electronic friends will of course trigger despair, disillusionment, doubt, and more consumption of rich chocolately ice cream (it’s not all dystopic). It’s dire, but not inevitable. This can be averted now before it’s too late. Do your part. Keep using Internet Explorer. Try not to tinker. And definitely steer clear of a man called Json Parser. Thank you.

O'Reilly Emerging Technology Conference (March 6-9, 2006)

harrisj — Wed, 03 Aug 2005 16:47:00 -0700

Check out the Call for Papers of the O’Reilly Emerging Technology Conference This looks like a pretty amazing subject and I am curious what the presentations selected will be. It looks like an amazing forum for really thinking about how the make the new web.

When The Web Was New

harrisj — Wed, 06 Apr 2005 22:05:00 -0700

Here's a flash from the past, courtesy of Technology Review. It's a reprinting of an article about the growing pains of the Web back in 1995. And what's especially funny to me is the screenshot of Mit's Homepage back in 1995. If you squint really closely, you can see the address of MIT's official server back then was web.mit.edu, because the domain www.mit.edu was actually the location of an earlier server set up by the esteemed Student Information Processing Board (SIPB). Originally running on Plexus, a PERL web server implementation, it was one of the first 100 web servers in the world.

Why do I mention this history? Well, back then I was a student at MIT and a prospective member of SIPB, and someone suggested I might want to get involved in being a webmaster. And so this has brought on a bit of a wave of nostalgia for those heady really early days when HTML tags were strictly semantic, a URL on a billboard seemed an insane novelty, and we edited pages on the server by hand in ed. Good times.

For more information, check out original 1996-era homepage or the exciting history of the original MIT web server.