Corporate Sponsorship

It's unfortunately true, the market does need one or more big corporations to give their blessing to something in order for it to start appearing in the list of needs and ways to fix them.

Remember LDAP? Remember LDAP before Windows 2000 came out? The LDAP market grew hugely when Microsoft started talking about directories even if they never got around to implementing the actual protocol, their blessing on the concept was a good thing for the market.

Just to give a second example, would most organizations that are now thinking about Big Data would have entered the bandwagon if Oracle hadn't started talking about it?

I came back at thinking about this because of some talks I've been having about Fujitsu's role and potential role in today's market. It seems there are several people that don't think a company can be determinant as an Hardware seller only.

I find that idea as consistent as someone saying that no electricity company can generate money, the value is higher up the stack like, … Household appliances or that an oil company is sure to be broke soon, the value lies elsewhere. But, I'm willing to concede that the IT industry has to deal a lot with perspective and, well, most of IT analysts that came out on the news I read are basically ignorant fools (remember Nicholas Carr?)

So, while I don't totally subscribe that a full stack is needed for a company to succeed, is that really a show stopper?

When I started working in a media company, living it's day by day, I came across a totally different reality from my time in pure IT companies. The fact that our IT was managed by a real visionary where our job was to give managers, reporters and all the people in the company superb flexibility and excellent reaction times, combined with a team smaller than what could be considered decent, really pushed us to go for the throat attitude.

So, when selecting our tickets platform, we had in the table all the big players, Remedy, HP, CA, but, amazingly, the single best ticket platform in the market today is RT. Even if you don't consider the time it takes to implement an RT solution (measured in hours or days) against a service center deploy (weeks or months), RT is simply superior in every single way.

When looking at content management solutions, hell, we actually implemented half of them, only to throw them out. True the OSS market doesn't have the panacea of content management where one tool is the best for every purpose but, the reality is that vendors only say they have it.

Notice that no where here I mentioned the price, I wasn't paying for it anyway (the contractor already had enterprise agreements with most of the vendors) but, I did had my neck on the line on several of this choices but, it came to a point where I couldn't, in conscience, recommend commercial software that was clearly outclassed, had less features, a less resilient architecture and it would give us more trouble to implement, customize and support.

I leave to the folks at the Opensource Zealots institute the flag carrying of free speech and whatever. All I really wanted was stuff that worked, performed, scaled, had good resilience features and had a good support. Most of the stuff we got from vendors failed in ALL of this points.

And, it was with that it mind that I thought about Fujitsu's way to take the market by storm. A small fraction of Fujitsu's Professional services could, in a very short time, define, implement and productize a complete stack of Opensource Software to fill it's customer needs. In a near free way, it could became the place to go in terms of Opensource Software and, if Fujitsu could learn with the mistakes made by RedHat and Suse, it could really put itself in the center of corporate innovation.

Just some food for thought

My imaginary monologue with Masami Yamamoto

Along my career, I was fortunate enough to get to meet several people that clearly helped shape the world we live in. Of course, most of them wouldn't recognize me if we crossed in the street but, their views on the world, on IT and on what we, as a class, should be doing in this world left a lasting impression with me

But, if I could ask for one person to listen for 5 minutes what I had to say, it would be to Mr. Yamamoto.

Masami Yamamoto is the President of Fujitsu, a man that, when combines his age with what he already accomplished in Life, would swat me to the side but, yes, given a chance, I would go to him and tell him:

Mr. Yamamoto, we need more from you.

When I first started working with some of your machines, buying Fujitsu was like cursing someone. The machines were bad, would fail more than my home PC, Fujitsu's support was non existent and, buying something more complex than a simple machine from Fujitsu was a task worthy of all the 12 tasks of Hercules.

Then, … things got worst and we had all the Fujitsu's to handle. A time of competing proposals from several Fujitsu's, of high prices, of “we'll give you the proposal next week, ok?” Not such good times, specially when no one actually understood in what exactly was Fujitsu Services different from Fujitsu-Siemens (I'm sure the differences were plenty but, no one was interested in getting a PHD in Fujitsu's board).

Recently, I had some encounters with Fujitsu technology and, every time, I jumped back thinking, hell, no way but, after having my hands tied, I really had to take a professional look at what Fujitsu was proposing.

Started with the Primergy line, right, who cares about X86 hardware, right? Well, at least until I opened it. Yes, it's still a crappy Intel processor but, holy that actually seems like a real machine, … It has remote console, good Electrical Isolation, pretty sound engineering in terms of reliability and it performed amazingly well in our benchmarks, where I expected it to be the same crap as the other machine we had there, it rocked and blew other X86 vendors.

Then, I saw your work in Sparc. I left the Sun reseller world more or less at the same time as the APL came out so, I became suspicious of the I/O capabilities of Sun (now Fujitsu's tech) high end machines. Hell, I was so wrong it actually hurts. The M series (I tested the M5000) seriously kicks some bad behinds. In most workloads I tested, it seriously kicked IBM's Power line and, when you start from a clean state (dedicated Storage, Dedicated backups, Q&A, etc) I could build a complete solution for less than half of IBM's offer (list prices but, then again, I'm not disclosing the details of the deal and the tests so, I see this info is not all that relevant to anyone but myself).

I was amazed with this “new” Fujitsu. And, so I started talking with people, reading papers, and studying Fujitsu's offer, from Grids to Big machines, from disk to cables and, in a word, … WOW.

But, I must say all this was dwarfed by actually working with Fujitsu.

In a project, things went south, really bad. It was one of those case study scenarios where everything that could go wrong, did go wrong right after production launch and, it was Fujitsu that showed up and offered to help us.

Having spent most of my working life in a reseller, I do know that what happened to us was part luck but, experience shows that luck usually happens to the hard working and better prepared. Fujitsu's behavior, as a benchmark of the company structure in terms of support, presales and willing to commit with the customer was nothing short of brilliant.

So, I have to ask, why isn't Fujitsu head and shoulders above the competition as an IT supplier?

IBM moved from “no one was ever fired for buying IBM” to “buy IBM and risk your job moving into IGS”. They are still selling but every reasonable customers with an actual IT staff should be looking for ways to get out of IBM's rule.

HP's offer ranged from okey-ish X86 hardware to a pretty awful offer based on the crappiest processor the industry put up for sale since Cyrix. Every proposal from HP has to be double (and tripled) checked for errors

Oracle is recognized as the hardest company to make business with, they are lost with their offer (specially in terms off hardware) and their bet on what they call integrated solutions is just diverting their sales force from technology sales into marketing driven sales with crappy technology.

I guess Dell is an OK bet but, you know what you're getting into. They're the entrance level of the server market and have no engineering to enable them to move into the high end.

The Storage will start inverting tendencies soon enough. People who didn't move into real clouds will soon learn again that disk is the ultimate bottleneck of their systems and, Eternus has the possibility to amaze (extra bonus, it's the best Storage one could dream for a Solaris zones consolidation environment).

And, it's in this scenario we have Fujitsu. A company that has the technology, has the services infrastructure, has the know how and the manpower to start building what everybody else should be doing if they weren't so focused on their own belly. A company that understands the market and is willing to partner with customers in order to make things happen but, from some reason, isn't moving, at least not as fast as one could hope and would expect.

Mr. Yamamoto, you are the one in the best position to validate an entire generation worth of work in IT. The principles of IT acting as a business enabler, as a value added provider, as a problem solver can be regained by what we could do with your help. Please, just give the order to go, grow.We, the world need more from you and I'm sure you want more from us also.

Give me Liberty or give me, ... SAML?

Management is growing strange this days, asking for employees to come up with solutions for problems that cost less money but, for the other hand, not wanting to actually change anything.

My new Boss (new as my boss but, we worked together in the first time I worked in PT) just made me a recommendation in Linkedin. Like all my recommendations (and all that I've made), I never wrote myself the text to someone else to publish. The way I see it, when you're speaking about someone, what you say also says something about yourself so, if you highly recommend a brain dead or poorly recommend someone brilliant, you're also presenting who you are to the world.

I must say that his recommendation was pretty good, “one of the most gifted”, “ideas well built” and “focused on solving problems” are some off the expressions he uses but, he also called me rigid. Now, that's a down side, isn't it? So, the most obvious question is how can I overcome this particular side of my, … err, … personality.

I'll give here an example of a real situation where I was "rigid" because I spoke against the trendy decision and didn't look for ways to accommodate what I thought it was a foolish decision (note: I had no vested interest in either way so, the only thing I was about to gain was the pleasure of seeing others do something good).  

The most recent point of contention was about an identity project called PT-ID. That's a project who's main goal is to unify the logins of customers in the PT group of companies.

Throughout it's history, PT has been buying and selling, creating and closing companies as dictated by the management trend of the year to consolidate to better focus on their core business or diversify. That results in a good number of obvious problems that need to be addressed by it's IT. Identity Management is, of course, one of them.

So, being IdM a dear subject to me, I tried to know more about this PT-ID project. When I found out that, they were taking this from the backend side, trying to unify the logins of the customers, I felt like on a trip to the past, a wrong, without value and one that I had hope it was dead past.

As soon as the PT management thinks about buying a new company, this project (assuming it's already completed) will break due to the need of creating an PT-ID v2 to unify this new company with everything else or, if the company is small enough not to warrant this kind of expense, we will end up with 2 ID repositories, the unified one and “the other”.

If the management decides to sell one of the companies, then it will be even worst. We either sell the company without it's customers database (and, if you like this deal, I have an almost new car that you might be interested in) or we let the new owners to take away a database with all the customers of all the companies in the group.

We crossed this bridge in the 90's. We realized this was a lose – lose situation where the IT was drastically reducing the value of a business opportunity (selling a company) and was opening a company into a good number of lawsuits and bad publicity because of the way it handles it's customers data.

It was with this problems in mind that we (the industry) started to create the Federation Standards as a way to develop and delivery crossed services without exchange of personal customer data. A way for IT to act as a business enabler instead of an constrain in what the business side wants to do to better one's organization.

PT needs to start to build cross services between different companies inside the group (and, why not expand it to the outside) but, it needs to do this in a way that's scalable, both from the IT point of view (more users, more transactions per second) and business (other companies, more services in the federation network). It needs to start moving to channel independence where all customers are treated equally, either from a computer, a cell phone, a tablet or the device of tomorrow.

In a world of Big data, this small boy mentality, from a small town that uses small words can't keep up. And this isn't a step in the right direction.