Dave Jenkins

Most IT execs try to boil down their decisions to simple dichotomies: build vs. buy, distributed vs. centralized, minimum ante vs technology leadership, good vs. oracle, freedom vs. microsoft. This pattern repeats amongst the developer crewmates: visual studio vs. rational rose, DOM vs. script, cron vs. UP, and the most ancient of wars: vim vs. emacs.

Now that I am on the business side, and not held directly responsible for the tactical stuff, I can see why the business suits always simply stare at IT people in agogged wonder– these tight decision trees have nothing to do with the real world. Developers and engineers live their whole day in an artificial construct of reference hash tables, primary keys, routing diagrammes, and copies of 1s and 0s that need to be shephered from here to there and back safely. The engineer’s job consists of either a) building more of that artificial construct, b) cleaning up someone’s poor interpretation of that same construct, or c) defending their version against someone’s proposed revision.

Business people think in analog– sales are way up, slightly up, break even, almost to goal, a little off, under plan, or ‘in need of budget revision’. Notice the complete lack of any diacritical statements next time you talk to marketing– it’s all shades of orange (the new gray), very little black and white. Notice how the IT guys will pepper endless questions trying to make some logical tree out of statements like “make it cool”.

So what? Well, I don’t know yet. I’ve spent several years trying to come up with the magick formula, the correct set of questions to ask, the right analogy to frame things for the business owners. I hve learned the following points (in no logical order):

• never start with a stark choice– it scares the business types, and makes them inherently defensive
• when picking your analogy, try something close to the listeners’ heart: cars seem to be popular here in the midwest, while history worked well on the east coast. Japanese like to use organic metaphors: seeds, vines, roots (nemawashi), etc. Be careful with chess– your listened either has no clue about the game, or is a grandmaster– either way, they’ll start asking questions about your analogy for which you don’t have answers (you didn’t think that many moves ahead– oh the irony!)
• When laying the groundwork for your eventual “this vs. that” question commital point, make sure you attribute all the incoming data to someone else: bonus points for attibuting the background research/material to the same person you are about to ask for a decision– they feel smarter already.
• multi-variate choices are much better in terms of quick understanding, but they usually require a whiteboard to lay out the different factors involved (i.e. cost and complexity, time and ROI). At this point, the best you’re gonna get is to have your decider play pin-the-tail-on-the-donkey somewhere on the whiteboard. This is a false answer– it is still analog-y and relative to their opinion. You still don’t have a hard decision (maybe that’s enough?). Warning: do not attempt multi-variate using only verbal communication. Double Warning: don’t try this with a metaphor, as your listener will forget the question and only answer their opinion about dogs/cars/football teams/naval battles.
• Whenever possible, practice your Fractal Management.

If you’ve read any of my previous posts (all 12 of you, according to my stats), you know that I’m hip to the Open Source. you also know that i work inside a large company where the only open source is a recently installed MediaWiki, but everything else comes from Redmond.

In that context, I’ve had at least 6 people come up to me at work asking for my take on the Microsoft bid for Yahoo. In short: meh. Do I think Open Source is doomed at Yahoo? No. Do I think that Ballmer will make sweeping changes at Yahoo to have them tear out their BSD machines for IIS? No. Will Yahoo see talent drain out the backdoor? No more than is currently happening.

If anything, I see Yahoo infecting Microsoft much more than Microsoft dictating terms to Yahoo, much like Hong Kong is affecting the PRC after Beijing retook the colony. Just because MS is the one forking over the moolah, that doesn’t mean they want to dictate terms, it means they want to get something for their money: ad revenue, dynamic online channels, and perhaps (ZOMGWTFBBQ) some street cred with the online and open source market in that they might be able to show that even MS can own some datacenters with BSD and perl and php and apache and not muck things up. I’m willing to give them a shot.

Let’s remember what we all have been praying for: the end of the Windows Hegemony. It may actually be closer than we anticipated, and it may be a lot more peaceful than we thought (or ‘’hoped'’ in our bloodlust). At this point, MS is a big fat company that realizes they cannot develop the way they had for the last 15 years. Vista is the last hurrah. MS execs have said as much, and told us where they want to go. For once, I actually believe them. MS is a big fat wallet that needs to get the best ROI for investors– that’s what the stock market is all about. Yahoo may bring some new juice to the bar, and it will be up to MS to either do something good with it, leave it alone, or fuck it up. Either way, the knowledge transfer will be going from San Mateo to Redmond, not the other way.

In short, lighten up, Francis.

BONUS: The Economist thinks the deal is a steal for investors, and they use the Ballmer dance-monkey-dance video as their evidence.

Someone I respect greatly once said that ‘Linux will be the last OS.’ Perhaps he meant it literally, perhaps that no matter what comes along, someone somewhere will try to adapt it into the Linux kernel or driver set. It occurs to me that the same applies to rock music. No matter what new trend comes along, somone puts a pounding 4/4 time underneath it, breaks down the guitar licks, and bingo– folk rock, rock opera, rap rock, and even “That stupid punk rock“.

This isn’t such a big revelation, but just something to think about next time people write off Open Source Software– it’ll be around for a while. Sure, the skins will change, and the emphasis will move around, but the kernels will always be there: Linux, Postgresql, mysql, perl, and yes– vi. People continue to riff off of these core pieces, just like musicians remix the masters

People remix the core bits of open source, and invariably the great riffs make it back into The Canon, only to be recycled and spun off in some new direction. I realize that many of the readers here know this already, but I offer up the analogy to those who have a hard time ‘getting’ open source: it’s rock-n-roll, man. It’s cool, and adapts to whatever the kids are into these days.

Huzzah for backcountry.com, on taking a big step toward full open source citizenship. My former employer announced last week that they would release some postgresql code back to the community. “Finally stop mooching” is how their usual informal press release put it, and props for that. The code allows for master< ->master replication across postgresql databases. It’s called ‘bucardo’ after an extinct goat that lived in the Pyrennes mountains that might be cloned back into existance (replicated, get it?)

Backcountry.com has a webfarm at a hosting center, and then a rootdb back at the warehouse– pretty standard setup. For the sake of speed, however, the company wanted to hold as much information (i.e. dynamic content) as close to the web servers as possible, and therefore wanted to replicate a good chunk of rootdb out at the ISP. If they were using Oracle, this would be a relatively straightforward configuration. However, because the system was built around Postgresql, we needed to configure somrething out. Enter the geniuses as End Point Communications — an entire system of flags sit around data tables, and these flags are synced back and forth, to determine which data then needs to be pushed to the other side. The trick here is to have the 2-way asynchronous flows of information while maintaining basic ground rules of data integrity.

The system has had several months of getting the shit beat out of it, literally. Early on, syncs would hang on a missing flag, or an unsuually large amount of data, and things would back up, only to then try and flow through like a traffic jammed freeway once the flags fell back into place. Well, redundancies have since been built into the messaging system, and the logic has been wrung to optimised levels where flags that don’t need to be in place are simply bypassed afte x minutes. It’s all rather elegant.

Props to all those that were involved with this, and special props to those at End Point who suggested the path to enlightenment lie in giving back to the community.

The devil is sending his minions to collect. A patent troll company is suing our good friends at Red Hat and Novell for technology to power multiple desktops (i.e. multiple screens of data through one monitor, with some hotkey switch between them).The funnytragic part here is that the suit is pretty traceable back to Redmond. A couple of MS lawyer-types joined this troll company not more than 4 months ago. When Novell signed a deal with the devil, I doubt they figured he would come collecting this soon. Novell will bear the brunt of this much more than Red Hat, in my opinion, for a number of reasons:

1. Novell is already perceived as ‘tainted’ by their previous licencing agreements, whereas Red Hat follows a much quieter path
2. Red Hat has better pipeline into the technical arguments that may be needed to defend this (think old-skool UNIX interfaces)
3. Red Hat has better cashflow

In the end, I wouldn’t be surprised if Microsoft ends up ‘coming to the rescue’ to side with Novell (and grudgingly Red Hat). Afterall, one of the major (the only one I’ve seen?) eye-candy bits to Vista is their multiple desktop window 3-d thingamajig (which is crude and stupid ompared to the beauty of Beryl).

My shiny new laptop that runs WindowsXP (company policy: no Ubuntu for now) brought me a funny little icon this morning. The Java icon was bright orange. Curious, I clicked, and was invited to install OpenOffice.org, right there, right now. Well! It looks like Sun has finally grown a pair, and is going to attempt to “infect” as many machines as possible with their app-killer open source freeware. How might Microsoft react? Shut off Java? Not possible. Limit items in the icon tray? Again, too draconian for all of their own bloatware. I would like to see how this campaign plays out in the next few weeks.

Overall, I’ve been happy with OpenOffice.org, but I can certainly see how the power users of MS Excel are very very wedded to their pivot tables and other mojo packed into the newer versions. Mircosoft’s best bet may be to flaunt this extended feature set– Excel is much more powerful than OpenOffice.org’s spreadsheet (there, I said it) for those who know what they’re doing. Granted, 95% of the people will never be at this level, but that other 5% are usually the thought-leaders on this kind of stuff in an office environment.

Well, it looks like my friends made out pretty good.  Yahoo announced that they are buying up Zimbra.  Good for them.  Optimally, Yahoo would use Zimbra for their webmail offering, as well as commit their designers to helping build out the AJAX interface that Zimbra has.  Yahoo has a pretty good library already, and is a proven leader in AJAX development for GUIs– Zimbra’s deployment– while nice– is a bit overweight and could use some clean up.

I want to introduce a new side project that I have been pursuing together with my friend Matt Libo-on: Wikindex.com is an index of wiki sites. We spider through the statistics of sites running mediawiki software, and gather the daily usage, size, and number of users. These statistics should show a searcher which wikis are active, and which ones are dormant. For example, if you search for “linux”, you’ll see three or four entries, with one (Gentoo Linux Wiki) showing a huge number of articles and users. This is the active one, and probably the best source for information.

On a lighter note– who knew that the wiki for Star Trek is bigger than the wikipedia in over half of the European languages? Those trekkers are nothing if not well-versed.

I was at dinner with one of our good friends from Zimbra, and we were talking about random histories of the big players (MS Apple, Google, etc) and how the OS is rapidly becoming a commoditised moot point. As the conversation progressed, my tin-foil hat started to show, and I began to wonder if Microsoft may be playing a Hudsucker Proxy on our hapless friends in Orem:

1. Microsoft has enough business analysts on the payroll to realize that Open Source is unstoppable. Individual companies may falter, but the code and model are pernicious and maleable enough to resist any attack. It’s like sweeping ants in the jungle– there will always be more.
2. Microsoft knows they must come to Open Source eventually– but where? where to pull in OSS without damaging the cash cow? The OS. Noone really pays for the OS anymore, not real money. Sure, we all pay support contracts, but not big cash. Eye candy makes money. For Redmond, the money is the apps. So, where to find an OS partner… Red Hat? too religious. Ubuntu? too foreign. Mandriva? unknown. Debian? communists. Novell? aha– desparate for the bandwagon, jonny-come-latelys, good market share, lotsa cash laying around. In a word: suckers.
3. Microsoft does a deal, and Novell signs over their souls, and the partnership has begun. Here’s the problem– does Mircrosoft ever really partner with anyone for very long?
4. The next day, sure enough, Steve Ballmer is out making fun of his new friends. Novell desperately tries to paint on some lipstick for their new boyfriend because he’s out trash-talking. The community begins to turn, Novell stock drops
5. Now, a few months later, Novell fesses up and starts to admit that MS is the better ROI. That’s right, bitch– make me a sammich and get me a beer, too. The stock will only drop further.
6. Soon, maybe 18 months from now, the stock will be cheap enough that Redmond will get out their checkbook. What’s worse, with that whole SCO nightmare still loose, MS may find itself the owner of some core UNIX intellectual property. If/when that happens, all our lives get washed down the toilet.

Am I crazy? God I hope so.

Someone called my office this morning. he was offering “training” for my crew. i asked what kind of training, and he started to rattle off the usual suspects: MSCE, Novell++, whatever. I told him we are an open source shop, and I could hear him turn the page and say “you mean, like Linux?” Yes, like Linux. He then told me that he has Red Hat training courses for both 7.0 and 7.2

Heh. 1998 called, they want their distro back.

I then told him “well, we have some pretty smart cookies here. Not to sound arrogant, but I bet my engineers are smarter than your trainers.” He responded that yes, I was arrogant. A salesman told me, his potential customer, that I was arrogant.

I went ahead and confirmed his suspicions by hanging up.

Open Source training, unfortunately, is still an afterthought for many of the large training companies out there. The only people I would really trust to teach the RHEL certification is Red Hat. This kinda puts them in the catbird’s seat, and keeps prices high– too bad. The best thing I think Red Hat and Novell could do would be to (forgive me MLK) ‘flood the classrooms’ with RHEL and/or Linux sysadmins. Any sysadmin worth his salt now can administer Linux boxes. But I am talking about all those secondary people: desktop supports, junior sysadmins, floor managers, mail admins, and all their pointy-haired bosses. Never ignore the power of a technically-exposed PHB. Yes, they get in the mix sometimes when you’re trying to make a decision, but generally they are a Force For Good when it comes to budget time, and explaining your hardware/software stack to the rest of the executives.

Red Hat: make training (almost) free, and watch your martketshare skyrocket.