The title is a little obscure and I leave it as an exercise to the reader to discover the full meaning. However, that being said, the short hand is that decisions matter, and we all only have once chance at each decision. Now onward and upward.... 

Today is day 6 of my visit to ERI, the Beijing Engineering Research Institute.  This is a fascinating country and it has some excellent engineers. We are lucky to have 15 or so working on our team in Beijing.

My thougths are tending to the quixotic this week. I find a strange sense of beauty and harmony in the Beijing traffic. The traffic, to the unpracticed eye, seems chaotic and dangerous. I see it as a quintessential negotiation. The driver is negotiating with pedestrians, big cars, small cars, bikes and stop lights. To the western eye the traffic is maddening with frequent lane changes and equally as frequent stops and goes. However, it really is a harmony of small concessions and little aggressions. I have yet to see anyone get angry or frustrated with traffic. I am sure there are moments, but by even the most sanguine stretch of the imagination, most US drivers would have suffered an aneurism ten minutes into Beijing traffic.  

With that set up, I have my small thought of the day. It is a bright faced chubby little boy kind of thought. It is a wearing the tin pot on the head and banging it with a spoon kind of thought. It is a rumbly tumbly tummy pooh bear kind of thought, but it is my thought and I like it.

Beijing traffic works for two reason. First, drivers are transparent. They telegraph their movement with horn beeps, signals, and edging intentions (how's that for the turn of a phrase?).  Certainly there are accidents, but not like there would be if you drove this way in the US. There apparently is a pretty clear set of rules to the road and expectations around driver behavior. The first rule is, drive as though you mean it. The next rule is to watch what other people are doing and communicate your intentions.  Finally, perhaps most importantly, don't do anything unexpected. If you stop when you should go, an accident will happen. If you are polite when people are expecting aggression, an accident will happen. If you are too aggressive, an accident will happen.

In some sense open source is a bit like Beijing traffic. You are guaranteed transparency. You can't get the unexpected with open source (well, you can, but at least you can see why you got into your accident, if you will pardon the over stretching of a metaphor). The counter play of negotiation between driver and stack is evident. In the event of overly aggressive drivers or limited driver stacks, corrective action becomes possible by the many and not just the elite technocratti.  

Whew. All of that said and what I really want to say is this:

To my colleagues at Qlogic, thank you for taking the plunge and being the first to open source the HBA driver in the leadville stack. Way to telegraph your intentions to support the leadville stack and lead the charge as a Sun partner. Well done. There is a lot to be said for being first.  

In life we constantly meet momentum. Early in my career I discovered two essential truths. First, rarely does meeting force with force succeed. Success that is predicated on winning or losing, at least in the management sphere, generally results in a tax. There are times when you have to pay that tax, but, as a general rule, avoiding that tax is the best course of action for all involved. This line of reasoning helped me formulate a concept I call "managerial judo".

Judo is an art and a science. It involves anticipation, positioning, momentum and the judicious use of force to accomplish an objective. (It is a bit like hockey but no sticks and less blood). Managerial judo is all of that but also the ability to take negative energy and turn it into positive momentum. Managerial judo seeks to influence and impact things not through force on force negotiations, but by positioning and creating energy by combining forces and accelerating them to achieve more interesting results.

An example you say? Sure. Here is one: At Sun we have a flexible work place policy. Driving to work and spending two hours at the begining or ending of your day is a tax on your employees. If they are commuting, they aren't working. Conversely, if they are commuting they are expending energy and time to accomplish nothing for your company.

Managerial judo says: Take the expenditure of energy and time that is not currently benefiting me and try and turn that into a positive result. The value prop? Instead of driving for some number of hours to spend all day in you office, why don't you work from home a couple of days (or more) a week and use that extra time to be productive (either for your self or the company)? The risk is that you gain no benefit in actual work but the positives are enormous: Intensely loyal employee base, more productive employees (on average), decreased facilities costs (if you don't have space allocated for your home assigned employees you don't have to pay for it). There are additional benefits as well. People hear about it and really really want a chance to do great work in bunny slippers and bathrobes. Some of the visuals are horrible I assume, but managerial judo does not require that I visualize the result. I take great solace in that.

This is only one example. Taking an employee that isn't doing well in a job that they really don't want and helping them to work on something else in your organization that they REALLY want to is managerial judo. Figuring out how to work the process to produce excellent results is managerial judo. Getting people to do things because they want to do things you want done is managerial judo. Changing your organization just a little bit to gain enormous results is managerial judo. Earlier I talked about force multipliers. Managerial judo is the manifestation of a force multiplier.

Here is another example: Recently we added a customer advocate to my team. She is a blogger of rare talent and has an unusual talent for breaking every piece of software I ever wrote but also for making my software better and building intensely loyal user communities. She is helping to manage my hub. Look for her blog soon. Deirdré is all about managerial judo and I look for some awesome throws!  She has a knack for turning things just a bit and making the end result much better. Welcome Deirdré.

I have a new job at Sun, prompting a new blog category. (Total ripoff of Hal Stern's opening line in his Change in Longitude blog). I am in the Solaris organization and will be working closely with many old friends (and some new ones) to help manage the SAN, iSCSI, AVS, Multipathing and what ever else my boss decides to throw at me.

I normally don't get too excited about things, preferring to keep things pretty low key. However, looking around, I see things at a tipping point in Solaris.

AVS is open. Really. This very cool application is file system agnostic, volume agnostic, application agnostic and is foundational to a solid storage strategy. And it is open. This host based data service rides on top of Solaris 10 and frees your from boundaries. Check it out. Compile the source if you are so inclined. Very cool.

S10 has huge adoption.

Nevada Development is rocking right along. Check it out here.

 OpenGrok  rocks. I use it a frequently.

So I am not normally a "cheer leader type", but every once in a while things coalesce and start to build internal momentum. I think we are there with open Solaris. Very cool. There are lots of other things going on that are pretty exciting but not yet public. More to come.

 Dan

 

 

 

I have been working from home these last couple days. I do that not so much out of choice but rather because I have two little ticking biological time bombs and they both devested me on Saturday. I love my kids but they bring home some nasty goop from their colleagues at school. They are relatively immune, I, on the other hand, I suffer for days if not weeks.

Anyway, since I have had to spend lots of time on the phone, I have discovered two very very cool things. First, I use skype for all of my conference calls. Since I am dialing 800 numbers, there is no charge and it keeps my home and my cel phone free for really important calls.

Secondly, if you need to talk to someone and you don't know their current schedule, call them five minutes before a significant change in time (at the :25 of the :55 mark.) Given our busy schedules and our compulsive need to schedule on the hour or the half hour, you have a better chance of having your phone call answered. They are either freeing up from a meeting or getting ready to go to another meeting. People try and get caught up before their next meeting....

 

For many years I have talked about the levels of engineering. As a practitioner of the craft of engineering I went through the stages as well. Early in my career I spent time learning about the tools of the craft. Intimate familiarity with the compiler, source code control system, bug tracking system, installation tools and the libraries we use. This is frequently fodder for religious wars about the "best" compiler, OS, widget, editor (VI of course), etc...

After this stage, one learns about crafting sweet code. Now the definition of sweet code is different for everyone. My definition of sweet code is simple, elegant, very readable and efficient. And this brings my to my point.

Excellent engineers are like artists. The very best engineers are artists. You can be very good at what you do and still be a craftsman. In other words, as a craftsman you write functional, solid code. Most engineers don't attain artist state. Most engineers are craftsman. I spent years writing code, and although at times I approached beautiful code, I never considered my self an artist. I wrote solid, well structured, somewhat utilitarian code. Don't get me wrong, I am quite proud of my code and what I did. However, I wasn't an artist. It is probably why I became a leader of artists. My natural talents lie in finding the very best folks I can find and stuffing them with pizza and beer and harvesting their code. (Perhaps I am sort of like a bee keeper). Maybe sometime I will share the optimal pizza/beer to work ratio. 

Artists write code that leaps ahead of functional. It isn't obscure or overly clever, it is beautiful in both  functionality and the structure. A staff engineer that works on my team and is the lead of my HBA engineering organization said it better than I do. Code is like poetry.

Code that sings when you read it is rare, but man when you see it, wow. This probably classifies me as a true freak, but the aesthetics of what we do is very important to me. For instance, I take great pride in my slide show transitions and "harmonizing the synergies of the talent pool". (Retch)

Seriously, at Sun I am surrounded by people that sculpt amazing code, compose sonnets that make the FC stack sing, write symphonies of file systems and every so often create a haiku that wakes the soul and makes you weep and rejoice at the same time.
 

 

I last wrote about a book called Moneyball that I was just starting to read. It is by Michael Lewis and I really like this book. If you recall, I said the book was ostensibly a book about baseball. It is not. It is about looking at things, systems and components and how you derive the best results given certain constraints. 

 Rarely have I come across a book or blog that so accurately reflects what I believe and value. The essential elements of the book sing to my engineering soul (if you will pardon the allusion).

Short summary:

* Measure

* Measure the right things

* Objectivity is good. Subjectivity is bad

* Ruthlessly decide based on data

I like this book. 

 

 

After 25 years of using the keyboard as an interface to my data, we may be approaching a convergence where this tool is not only obsolete, but also the least efficient tool to get to my data. I am amazed that the younger generation can use the predictive text on phones to thumb hundreds of messages a day. Now Jeff Han uses multi-point touch screens to access data. Each key stroke on a traditional user interface is limited to a single character. Take the Windows UI and you get increased efficiency with respect to the  UI (alt F X means make the application go away). With Dr. Han's advances in multi point touch screens the possibility for efficiency gains for each given stroke (finger stroke efficiency ratios?) becomes nearly self evident. If you can't see the flash below, click this link:

http://www.youtube.com/v/LByAFrcOwJk 

 

 Interesting when you talk about collaborative computing as well. Since there are multiple touch points, you don't have input contention issues. Two keyboards anyone?

I think a lot about management styles and approaches and I have heard arguments about how to lead effectively that make sense. One style I have never particularly cared for is the "screamer" or the butt kicker.

As a commander in the Field Artillery I never raised my voice or verbally abused anyone. I agree that it is a style that can work, but it isn't me and so consequently I could never wear that hat effectively. One of my heroes is Orde Wingate. I don't like everything he stood for nor everything he did. I promise never to stand in front of my troops clad only in a shower cap and dispense orders.

However, he was quietly competant. He wasn't the General MacArthur type (loud, brash, wearing an ivory pistol coolly talking during an artillery barrage) but rather the quiet "lead by example" type.  Again, not everything he did was something I admire (advancing a personal agenda to the detriment of the organization is something I despise) but he took on the Italian army and by many accounts, won.

Anyway, one of the things I admire is leadership through enabling the team to do what they do very well. A colleague of mine recently forward me this link about the "screamer" as a leadership style. I thought it was worth reading and I agree completely with the rebuttal. I think the "enable the team to lead" approach is not compatible with the autocratic, kick butt style of leadership. If we hold as truth that most people most of the time are looking for ways to be successful in their job, we are compelled to enable them to be successful. Sometimes this means showing folks how to do things, sometimes it means letting them try and fail and sometimes it means helping them choose a career that is better suited for them. For me, it never means shouting, kicking butt or denigrating them.

Thanks for the link Michael!

I have been fascinated by China for years now. The combination of growing up reading James Clavelle (Noble House) and having my folks live and work in China and the emergence of China as a dominating marketplace has caused my interest to grow over time.

Years ago I wrote the business plan that started the Network Storage office in Beijing (Sitting right next to the Solaris guys made this an easy sell), and so I remain interested in how our China office does.

I stumbled across Sin-Yaw Wang's blog and it is fascinating. He writes about some of the same kinds of things I write about, but he includes history and an eclectic mix of culture and science and strategy. I love his tips.

Anyway, read his blog.

http://blogs.sun.com/syw/

Regards,

 

Dan

I was interviewing a candidate for a manager slot for one of my colleagues the other day. I asked him one of my favorite questions as part of the interview process.

"Tell me about the last couple of leadership books you have read." I like this question for a lot of reasons. The first order approximation gets me an the answer for this question "Do they read leadership books." Next, the way they discuss the books tells me about how or whether the candidate thinks about leadership. Finally, the really sophisticated candidate knows that pulling a book off the shelf with "leadership" in the title isn't what I am talking about. Almost any book can be a lesson in leadership. I was trying to explain this to a candidate that didn't understand why I was asking about the kinds of books they are reading.

I read "The Hunt for Red October" by Tom Clancy for about the 10th time the other day. It is a fictional novel about a brilliant sailor and a CIA analyst and, well, leadership. The leadership of a member of the Soviet military and the hard choices involved in doing what is right. The leadership of a brilliant man (Jack Ryan) and how even in the face of great adversity he was willing or able to accomplish the mission. The suborning of self for the cause at hand (saving lives on a ship) and and and...

I think the candidate got what I was trying to say, and I hope he recognizes that leadership isn't something you put on or take off. Leadership isn't always the flamboyant "take charge and cross the Rubicon" kind of stuff either. Sometimes leadership is letting others take charge of their destiny. Sometimes it is letting people try and fail (not too badly) as failure can be a great teacher.

One should study leadership and look for ways to improve how and why we do what we do. To that end, "What am I reading?"

Moneyball, Michael Lewis. This book is about the manager of the Oakland A's and how he changes baseball by looking at things differently and using metrics. At least I think that is what the book is about. I hate baseball. It is like hockey for out of shape people that can't skate. But, my sister-in-law suggested the book and I like it so far. Plus, she paid for it, so why not?

Empire of unreason, Gregory Keyes. The third book in the age of unreason series. It is about women and their place in a fantastic (not as in the good fantastic though but rather the unbelievable) society, the role of science and religion, power, restraint, obligation and magic. Kind of fun.

Getting to Yes, negotiating agreement without giving in by Roger Fisher and Willima Ury. It is a book about, well, negotiations.

I am also re-reading Thomas Paine's classic treatise on the "Rights of Man". "That men should take up arms, and spend their lives and fortunes, not to maintain their rights, but to maintain they have not rights, ..." makes one pause a bit doesn't it?

What are you reading?

I vividly recall each time I was promoted in the military. The first real promotion was the day I pinned on my 2LT butter bar. That was a day mostly of pride at my achievements. I felt that I was successful and that I had personally overcome great difficulties and this was my reward. My grandfather and mother pinned my bars on and I recall a mixture of relief and pride at having "made it". I thought about leadership then, but it was mostly abstract. In short, this promotion was only about me. Looking back it was the last promotion about me, as it should be.

My next promotion to 1LT was in front of the troops of my command (I was the XO) and my wife and my commander pinned me. This promotion was significant in that I felt that I had really earned it. 1LT isn't a rare promotion. Most 2LTs will make it. But for me it was the manifestation of what I had learned and that I was a "real leader". I looked forward to my promotion as Captain. To me Captains symbolized the true professionals of leadership. CPTs are commanders and a select few of them get to lead combat batteries (troops, companies). The ones that understand the awesome obligation of leadership are fit to wear the bars. It symbolizes that one has been tested and found equal to the task. It isn't automatic and it isn't about the privilege of rank. It is about a heavier load. It is about (sometimes) the power of life and death. I was mostly disappointed at my promotion to Captain because it wasn't handled well. It was casual, not symbolic in any way and reflected my achievement but didn't articulate the seriousness of leadership. Promotion of the individual generally is for the troops to see how the military rewards duty and obligation. It is also to symbolize the success of the organization. At the CPT level it is not, nor should it be, about the success of the individual. My commander didn't see it that way. Promotions of this magnitude need to be marked. Both for the individual and the organization. It is a rite of passage and a failure to articulate that cheapens the whole experience.

That being said, I view promotion to director at significant technology companies as a big deal. It isn't automatic and few will actually reach this responsibility. Recently a friend of mine was promoted. I wrote this for him:

Know ye by all those present that one man stands before us that has been tested and found acceptable to those that have gone before and those that will come after. We gather to celebrate his promotion on this 23rd day of October, 2006.

 This man stands between duty and courage and those that would have neither. He represents the skills and lesson learned in the iron forge of selfless duty and obligation to those of his charge. He has embraced responsibility and leadership and emerged from the craft of engineering as a leader of men and women.

Therefore, it is with great respect that we honor his achievement and provide these symbolic beverages to celebrate his promotion and remind him of his obligations.

First, we give him Warrior Port. This legendary drink is to symbolize that leadership is about providing safe harbor for the men and women of your command. It is to never forget that leadership is a moral obligation. Be strong, be fair, but take care of those that are in your charge. Additionally it represents that occasionally you come across someone that has the spirit and fire of a true warrior. We  would go to war with those rare individuals. We have found him to be one of those people.

Next we give him a keg of beer to symbolize that leadership is about getting work done and achievement. Results matter. The choices of leadership are like the crafting of a fine brew. A delicate blend of ingredients and time result in a product fit for consumption. As the leader, the symbolic brew master, it is your responsibility to drive excellence and achievement for those that follow. Be right and we all drink. Be wrong and the brew is tainted and the organization sickens. This is not a light responsibility.

Third we share with him our Mandarin Vodka to symbolize that we are a global organization. We are Chinese, Indian, American, Russian,  European and many other nationalities as well. No longer shall leadership be about what is best for a particular locale. Rather, the obligation of global leadership is to do what is best for the organization. Your view has expanded, along with your team.

Finally we share with him this bottle of old Scotch for quiet reflection on the metaphorical cold and stormy night. Leadership is a long, lonely road. Choices of leadership are hard, and sometimes the only comfort one has is that of a few close confidants and a sip of rare scotch to reflect on a job well done. It is an acknowledgment that your finest achievements as you progress up the chain of command are not your individual achievements but rather the achievements of your organization. Ultimately we are measured on results and the success of the teams that we lead. The burden of leadership is heavy. May this lighten your load on occasion.

 Congratulations from all of us on a job well done. We salute you!

Robert Scott Tracy, Promoted to Director, Solaris IO, Sun Microsystems, Inc.

23 October 2006

A while back I posted about how I mentor and learn from my staff engineers. I said that some of my thoughts about how I lead and teach came from a document that David Butterfield wrote. I consider it a foundational work. I had it wrong, David didn't write the document. It was written by Mark Kampe and he has been kind enough to submit the URL to his doc here:   http://mark.kampe.googlepages.com/lancelot.txt

Thanks for pointing this out Mark and apologies for misquoting you David....

I was listening to the Collected Stories of Arthur C. Clarke the other day. One of my favorites is a story called "Superiority". This a story about a General who finds himself on the losing side of a battle for control of the universe. They had a vastly superior force at the start of the war and spent so much time making their weapons better that they ended up losing the war.

It is the canonical requirements analysis and risk tolerance that engineering constantly struggles with. "If I only had a few more weeks I could add this feature and it would be perfect...."  If I had a few more resources... There is a funny segment in the story about the fact that the engineers designed the weapon to spec and the support infrastructure used to control the weapon was so heavy weight that they could not field the system with out a massive unarmed space tug. Of course the enemy tubed this ship immediately and made the tool useless.

Several things that struck me about this story:

1. The enemy of the "good enough" is the "better".
2. Fielding a tool before it is ready is bound to be devastating.
3. Simple generally wins out over complex and always does given enough time.

This story became required reading at MIT or so the Internet tells me (perhaps apocryphal but whatever...)

Here is a link to some of the excerpts from the story and a couple of comments from an engineer that aren't too dissimilar from mine.

http://catless.ncl.ac.uk/Risks/12.23.html#subj7

Regards,

Dan

I have been a bit quiet these last couple of weeks. I never learned a damn thing when I was talking. Trying to listen to the heartbeat of the org.  Found out some cool things about our servers, storage connectivity solutions and the sucess of our open source software model. All of these things are being blogged on, so I won't waste your time. It does bring me to my thought for the day:

Listen to people. You will likely find something out.

Not complex, but perhaps profound.

Regards,

Dan

As an officer in the Army I knew that there were going to be times when I need to make a decision without all the information. We few, we happy few, had an obligation to be technically and tactically proficient, as well as exercise judgment and  initiative. As a 2LT we talked about our "velcro butter bar", a wary nod towards the obligation we had to make sure our artillery rounds landed where they were supposed to land. We were told to put velcro under our bar so it was easy for "them" to take off when they took us away to jail. The way we talk describes what we value. Then we valued our bars, our responsibility not to be wrong and our enormous obligation. As a battery commander I rarely had to make "go to jail decisions", (only every time I engaged a target) but I had to enable my young lieutenants to bet their bars. Worse than a "go to jail mistake" was the potential for putting a round on top of friendlies or civilians. I think the Officer Corps kept rolaids in business.

 

It isn't and wasn't as cavalier as the phraseology makes it sound. I slept with my TFT and prided myself on my ability to mentally gauge deflection and quadrant based on a series of grid coordinates. Rarely was I spot on, but I was ALWAYS close. This gave me confidence in my teams and my procedures.

 

Today I had the opportunity to chat with a senior director at a prominent connectivity solutions company. It was gratifying to have him nod his head in complete understanding when I was talking about having to "bet my badge". With nary a pause he responded with "We bet our badges every day".

 

What am I saying? Well, it's simple:

 

1. Results matter.

2. Leaders make choices and are held accountable to those choices.

3. Make choices. Be right.

 

Simple. Elegant. I bet my badge on my choices.