For the past three years, an open source consulting firm appropriately named the Olliance Group has put on an event called the “Open Source Think Tank” .

Nowhere else can you spend two concentrated days talking to open source execs, thought leaders and CIO’s. It’s good for expanding the Rolodex. More importantly, its good for expanding my perspective on open source.

Last year, I was left with mixed feelings on the “state of open source”. I think timing had a lot to do with this. At the time, we were wrapped around the proverbial axle on licensing issues. As a result, much of the conversation was consumed with speculation around GPLv3, its impact, etc. While licensing is obviously is an essential underlying piece of infrastructure required to support the ecosystem, these discussions can quickly slide into the arcane and are often far removed from our core goal of making life better for end users.

This year, the primary conversation was different. While licensing remains an evolving and nuanced issue for all of us in the open source ecosystem, the topic has (fortunately) drifted into the background. In contrast to last year’s discussion which was internally focused on issues within the open source ecosystem, this year’s conversation focused externally on the end user perspective, needs, wants, likes, etc. Over the last year open source has made a leap forward as a valid approach to developing and distributing enterprise software. SaaS and commercial open source are entering the mainstream.

My summary observation is that over the last year open source as a viable approach to enterprise software has made quantum leaps in terms of entering the mainstream.

The best evidence of this progress is that the debate over whether “commercial open source” is really an oxymoron has all but died among this group. While there is plenty of divergence in terms of approaches to commercial open source, it is clear that free software, community and commercial business are symbiotic. It is now plainly clear that open source is a permanent and pervasive driver in enterprise software.

While we are each contributing at various levels to bringing open source to the mainstream, no doubt the MySQL / Sun transaction has moved the needle most significantly. I was sure to given Marten Mickos a big thank you and clink his glass of Pinot Noir when I saw him at the event this year.

For first movers and purists, there is always some ambivalence to the mainstreamifying of any concept. As a beta-geek myself and having made a career operating in the early phases of new IT paradigms (e.g. client-server, web, SaaS), I can relate to this odd mix of emotions.

For all involved however the progress is unequivocally good news. For us (the provider), the evolution is obviously positive. Mainstream adopters are the lifeblood of a vibrant software business (closed or open source). Users and contributors (community and commercial) clearly gain as well. Only a healthy and growing business can afford to invest in enabling the community and the next realm of innovation — and give most of it away!

Not long ago Redmonk analyst Michael Cote coined the term “Little Four“ to describe a group of open source IT management solution vendors. This up and coming group of software companies includes Zenoss, GroundWork, Hyperic, and Qlusters (openQRM) and they are seen as alternatives to traditional vendors BMC, CA, HP and IBM (the Big Four). In one of his blogs (Origins of The “Little Four”) whurley of BMC stated the distinction may be short lived. whurley wrote “My moving from CTO at Qlusters (one of the Little Four) to BMC is a clear sign that at least one of the Big Four understands the value of open source in systems management.”

I agree with whurley’s first statement that the distinction of “Little Four” will be short lived, but not for the same reasons. Here is what I predict will happen:

• In a short amount of time (next 18 months) names of “Little Four” will be well known the world over. How you say? Zenoss alone was downloaded for more than 250,000 times in the past 9 months. The pace of downloads is accelerating for Zenoss and the rest of “Little Four” while increasing number of large companies are deploying the solutions in enterprise environments.
• Enterprises looking to deploy new IT management solutions already started to include Zenoss and others in their evaluation process. The pace of this inclusion will also accelerate to the point where members of “Little Four” will always be included in every IT management evaluation process.
• Today, “Little Four” successfully win against and/or replace the Big Four in some competitive cases. The win/loss rate for “Little Four” will increasingly tip in their favor until it reaches a balance.

When my above mentioned scenario becomes a reality in the near future, it will no longer be necessary to distinguish Zenoss, GroundWork, Hyperic, and Qlusters as the “Little Four.” They will simply be considered part of the leading group of IT management solution vendors. whurley is right to say that the distinction will be short lived, but not for the same reasons he may be thinking.

What will happen to the Big Four? Will they understand the value of open source as whurley states? Do they understand it now? I think that in most cases, open source to Big Four is nothing more than Linux OS support for their software or hardware products. In some cases, open source means donating some old code that no longer brings any value, but can be a tax right off. In the near future, the meaning of open source to Big Four will change dramatically. I predict that as they feel competitive pressure from “Little Four”, they will slowly adopt open source business practices. Eventually, we may see Big Four with up to 50% of its software portfolio in open source. Big Four and Little Four will meet somewhere in the middle of the road and race to the end. May the best vendors win.

Sphere: Related Content

]]> http://blog.zenoss.com/2007/08/29/hey-big-four-meet-the-little-four/feed/ http://blog.zenoss.com/2007/03/26/all-i-want-is-five-pieces-of-feedback/ http://blog.zenoss.com/2007/03/26/all-i-want-is-five-pieces-of-feedback/#comments Tue, 27 Mar 2007 00:42:01 +0000 Mark Hinkle http://blog.zenoss.com/2007/03/26/all-i-want-is-five-pieces-of-feedback/

Today was a big day here at Zenoss. We were the recipient of the Slashdot effect. Which is great because it allowed a lot of people find out about Zenoss which we think is one of the coolest things to happen to systems management in a long time. However, that’s not good enough for me, what I really want to hear about is how people are using Zenoss and how we could make Zenoss better. Nothing’s cooler to me to hear how people all over the world are using something written by a bunch of goofy guys in Annapolis,MD .
So as I am grateful for the 10,000 or downloads we got today I would be just as happy with Five Pieces of Feedback. In the time that I started righting this post, Buanzo dropped me a line to tell me that Zenoss runs on Gentoo without any problems. Awesome! I have been a Gentoo fan since way back in the day when Daniel Robbins started the project. What other Linux distributions or platforms are you guys running Zenoss on?

That’s one, now I need four more before I hit the sack tonight. You see I can tell you why I think why Zenoss is great, I can even tell you some places I would like to see us improve. However, it means a lot more coming from people using our stuff every day, so post a comment to this blog entry or send me an email at mrhinkle(at)zenoss.com and let me know what you think.

Not once — ever — have I been tempted to get a PDA. I’ve used friends’ devices and they make me nuts. Too small, too time-consuming to sync, etc. I’m generally a tech-head, but PDAs don’t do it for me.

However! Thanks to dried on #twisted (IRC) providing a link today (admittedly old, 2004), I may have a way of incorporating this functionality into my life, and in a manner that suites me — despite the Stone Age technology. Okay, maybe not stone age… it’s metal and paper, after all.

Here is the article that introduces the Hipster PDA.

And here are some great links to help you get started with your new hardware:

• Templates - print these puppies on 3×5 cards to help keep your junk organzied
• More templates - ditto
• Direct-to-3×5 printing instructions - how to print the templates on a mac
• 3×5 printer - get a dedicated printer for your cards (I’ll probably get this one)
• Stylus - A pen you can attach to the binder clip
• Leather cases and another - for the chíc

As mentioned earlier, there was a surprise Twisted contingent at PyCon this year, and I had the opportunity to hang out with both those of the gang that I’d already met and done some hanging with, as well as those that I had not. Brian Warner was one that I hadn’t had a chance to chat with much until PyCon, and it was amazing. I plan on interviewing him in the future regarding Foolscap, his thoughts on E and python, as well as the innovative business he’s working for right now.

But what the hell, you ask, does this have to do with “Bouncing Balls”? Well, in addition to going (or not going) to talks, attending BoFs, engaging in extensive business networking, and eating/drinking copious quantities food/drink stuffs, we also had time for good old-fashioned geek fun: watching movies. I do use the term loosely, though, indicating essentially anything *.mov, *.avi, *.wmv, etc. Brian shared with a whole gaggle of us the joys of The IT Crowd (all six episodes), Stalking Santa (preview), George Lucas in Love (which I had never seen), and the apparently famous (and rightfully so) Bravia commercials. The paint “fireworks” advert filmed in Scotland was just brilliant and hilarious (be sure to watch the making of), but I left my heart in a superball, bouncing down the streets of San Francisco. See if the same doesn’t happen to you when you watch the Bouncing Balls.

That video simply mesmerizes me. I’ve watched it tens of times and it still makes me feel like a joyful little kid. Just amazing. Side by side with the insanely awesome cinematography and the astounding colors, the music makes this a stunning classic. The song has been haunting me since PyCon and last night I finally had a chance to sit down and search for the artist. I didn’t have to search long, though, as the Bravia and last.fm pages share all. Turns out the song is “Heartbeats” from José González’s album Veneer, a cover from eccentric Swedish band The Knife. After reading more about Veneer, I bought it immediately. I highly recommend it.

And, no, I’d never see this ad before; I don’t have TV service. I am of the “pull” school for entertainment, versus “push” (or “always on”).

Anyone who’s been around the tech industry for any length of time knows what these three words, and their acronym, NIH, mean. To summarize, NIH refers to the organizational malady one observes in many project leaders and others in positions of control who frown upon, and frequently seek to crush, all ideas, projects and products not initiated by them. Often, if you want your idea adopted by someone with a severe case of NIH, you have to essentially make it their idea – you basically let the NIHer “steal” the idea. Yeah – you know what I’m talking about, ‘cuz just like me, you’ve done it – played doormat for the greater good, because it was more important to you to get the idea adopted than to get the credit.

As technologists (as undoubtedly you are if you’re reading this blog) let us pose this question: Where does NIH come from? What causes people to become so singularly and doggedly focused on promoting their own pet projects and agendas that they will subjugate the greater good - however measured - to their own selfish interests? I submit for your consideration (and hopefully debate) that NIH stems from, and is part and parcel of, the culture of secrecy that pervades so many technology companies.

Give the Gift of Good Karma
I had the pleasure of participating recently in a conference call discussion on the topic of consumer experience with several leading thinkers from the Open Source, New Media and Massively Multiplayer Online Game (MMORPG) communities, the purpose of which was to begin a conversation around how and why community-driven, or at least community-inclusive, products and projects seem to deliver superior results. The President of MMORPG company eGenesis offered an interesting thought about how communities’ basic ground rules – like secrecy or openness - influence culture and behavior. He described his recent experience at burning man, and specifically its gift economy, where strangers give strangers gifts, and how this fostered a very welcoming and relaxed atmosphere. This experience led him to conduct a thought experiment: what if, rather than a gift economy, this gathering were based instead on a theft economy, where the ground rules were, if you succeed in stealing other people’s stuff without getting caught, you get to keep the loot. This would lead, he postulated (and I agree) to an environment of secrecy, suspicion, hoarding and deceit. OK - so you’d be forgiven if right now you’re asking yourself “what the hell does this have to do with systems management?” bear with me.

Stop and think for a second about the lengths to which most large software companies go to try to prevent their code from being copied and their license terms from being breached. And consider further the sheer amount of collective money and energy that these companies and their customers spend on lawyers and lawsuits, auditors and audits, and compliance officers and compliance. I’d say that today’s proprietary software world is pretty darned close to the the anti-burning man theft economy dystopia envisioned by Andy.

If one were unlucky enough to be living in such a world, it’s logical to expect that NIH would be epidemic and acute. Because when you create a culture of secrecy and distrust, it is only natural for people to look out exclusively for their own narrow interests. Over time, those who steal and hoard well would get richer and more powerful, and most others would be scraping by at best. Think contemporary Rio - Think computer industry of the 1990s.

And when communities spawn a culture where NIH is rampant, and collaboration scarce, they produce inferior results. Think IE. Such communities produce inferior results because, over the short term, many good ideas never see the light of day, and over the medium to long term, because good people self select out of these environments.

Not Invented Here
Now consider the meaning of these three words in the context of Open Source. Perhaps better than any other three words I can think of, they convey the power and culture of Open Source. In this context, Not Invented Here means openness, inclusiveness, collaboration, debate, meritocracy – the best idea wins regardless of its source.

For producers and users (or prosumers to borrow a term from the book Wikinomics) of Open Source, Not Invented Here means faster product innovation through user and solution provider code contributions; Not Invented Here means greater utility for Open Source users because the product is, at least in part, invented by them; and Not Invented Here means lower R&D, sales and marketing costs for the Open Source company, which translate into lower acquisition and ownership costs for their customers. In Open Source, Not Invented Here is a beautiful thing.

To summarize, ground rules determine culture, culture determines behavior and behavior determines outcomes. Openness leads to collaboration and trust, collaboration and trust lead to meritocracy and meritocracy leads to better results. In contrast, secrecy and being closed leads to NIH and hoarding, NIH and hoarding lead to autocracy or oligarchy, and to inferior results.

I live with my family in Raleigh, North Carolina (a seemingly random fact, the relevance of which will become apparent in a sec). My wife teaches ballet to children, and so she’s always good for a couple funny kid stories each week to lighten me up. One day, she brought home a particularly funny, and as it turns out, very relevant story. At the start of each class when she takes roll, she likes to ask her kids a question to get them engaged. One day she asked what their parents do for a living, to which one little girl replied “My daddy works at Red Hat – they saved the computer business.” When my wife first told me this, I thought it was an absolute riot, and I also thought it was more than a little presumptuous of that little girl’s daddy. The more time I spend around Open Source, though, the more I think her daddy was right.

As some folks know, I worked with Huw Wilkins on the new redesign of the Twisted site. Huw did the design, CSS etc., and I hacked around on trac to make his vision work within that framework. He made my life very easy, though, as he produced the entire design to fit perfectly within trac.

The new Twisted site has been a success — we’re getting many more visitors and they are staying longer than ever. I believe that the design and the improving info architecture are the most significant contributors to this. And after seeing these changes and the assumed enjoyment that folks are deriving from them, all I could think was “This was so easy to do, there’s no excuse not to have a good-looking trac instance.”

Having been inspired by Huw’s great design for Twisted, I did the same thing for The Zenoss internal trac instances. I had a great time doing this, having explored a lot with the Twisted site redesign. What’s more, with the use of the trac dev branch (forth-coming version 0.11) the Genshi templates really simplify the process. With the process I had established with the Twisted work and the Zenoss work, I was have the courage to update my own massive number of trac projects: 3 domains with each one being a trac “multi-sites” setup.

People don’t realize how quickly you can take a generic trac and have something quite unique. Here’s all you need to do:

• install Genshi and 0.11 Trac from subversion
• design your look and feel and then prepare your images
• edit the trac.css file
• update the site.html and layout.html Genshi template files

Now, things do get a little complicated when you have trac multisites running. You have to write code that will render the sites index the way you want. Then you need to consider how to manage the templates across all the sites.

As an example, I’ve put the trac customizations in subversion here:
http://tinyurl.com/y35qxh

Technorati Tags: design, software, web

Mark Hinkle has been exploring Flock (Mac OS X) for use with the forth-coming Zenoss blogs, and that reminded me of a brief conversation with radix about a year ago where we were discussing blog editors. Thus re-motivated, I did a little research and evaluation of blog editing/posting software. The one that seems to be the most complete and featureful (to me) is ecto. Getting it set up with the new blogger seems a little awkward (the default “access point” doesn’t work), but there’s another version in the works where the kinks will be worked out (or be moot).

Here’s what I had to do to get it (version 2.4.1 Universal + blogger beta support) to work with the new blogger API:

• Hit the “Add Account” button on the Accounts window
• Give my blog URL, as requested
• Set the System as “Blogger”
• Set the API as “Atom”
• Set the Access Point as http://www.blogger.com/feeds/default/blogs
• Be sure that all the RSS feeds are “full” (blog settings in Blogger admin UI)

ecto was then able to get a list of my blogs for various account names as well as recent entries for each one.

One problem I have run into, however, is the multiple accounts thing. I have to quit ecto and restart it in order to post to blogger under a different user name. It seems to be caching login info and not using the user name and password that the particular account is configured with.

Despite that minor inconvenience, I will probably purchase the full version of this software and begin using it extensively. The single driving motivation for this is the multiple accounts: seeing all the accounts and blogs I have within one application acts as a reminder; I will be more likely to post to my other blogs now

For part of this week, I’ve been flown back to Annapolis, MD for a few days, ostensibly for the holiday party, but in addition, to coordinate in-person with Mark Hinkle and Rusty Wilson on the infrastructure and code needed for the next 6 month community push. In an effort to jump-start a whole set of community tools that Zenoss wants to provide, I’ve finished up some pretty sweet (yes, unit tests are sweet and so it automated web application/form processing/monitoring) additions to Zenoss Core and I am now switching gears to integrate community communications and work on the Super Secret Zenoss.net project (lots of Zope3 + Five + Plone hacking). Pretty fun stuff.

It’s been really cool being back with the Zenoss core team this week — lots of changes have happened since August: additional (and pretty sweet) office space, lots more people, lots of awesome code being produced by Erik Dahl, Eric Newton and Chris Blunck that I haven’t had a chance to see until now. There a bunch more stuff that I can’t let out of the bag yet, but is most excitimentful. Well, it is if you’re into monitoring and systems/network management… and the software that helps you do that