I am culprit too. Bohoohoo!!
I know. Chora blog.
There are about 18 blog posts there about ‘software projects that Salim has not finished.’ Some are true. Some are just written by shoe-shiners who want to lecture us about how to code. Some are just plain vindictive.
But from each, I learn a lesson. You are just as good as your LAST complete project. And this makes me better.
Being in the front-row of the battle gives me the facts, authority and balls to stand infront of men and women and lecture them about software development. I am a consultant. Not an employee. So I can lecture you about #biashara here all day from the angle of a true code-hustler. Not arm-chair lectures from someone behind a desk.
So, the big question is : “How does one hire Kenyan coders for a software project and see it COMPLETED?”
I will address this issue in a 2-phased point form. First of all addressing the problem, then recommending solutions I have known to work.
The big-chinese-peenus coder
I once wrote a blog-post about the Chinese peenus that, as always, annoyed alot of people. This is described as the bike owner who claims to have a Hummer 3. This type of coders always are in LUCK of getting projects with beautiful proposals and candy-talk. 1 year experience in Java and 24 hours with a “Teach yourself PHP in 24 hours” and they start lecturing the seasoned masters about code.
These ‘yes-men’ and ‘yes-galz’ will promise heaven to the client and deliver Liddos. All due to having pure willingness and no skills. You cannot blame the coders for over-estimating their capabilities or under-estimating the code/time demands of a project. Nor can you blame a zealous client for hiring mediocres.
As a client, you should do due-diligence on the coders assigned to do your project. Ask the company to send you CVs of the actual people who will be working on your system. Then do peer reviews and background checks.
The disgruntled master coder
I have always said this. Nothing is as useless to an organization as a disgruntled master coder. It is better to hire a wannabe helloworld guru from Campus and pay them well, than hire a seasoned master and pay them poorly or late… or worse, not pay them. More work will be done by the campuser.
It is not always about money. Sometimes, the coder would be disgruntled about project types. “I came here to do a serious system. Not a payroll. This is a waste of my talent and vast skills.”, I once heard one say. Na si ati anaringa.
Never hire a master-chef, to wash your floor. The floor will never be washed.
The shifting client
“I want the logo blue.” That is week one.
“I think pink will accentuate my inner female.” That is week 2.
“Lilac makes the logo more professional.” Yes? Week 3.
“These stupid designers can never finish a simple job. It is now 4 weeks and my logo is not done.” Week 4.
“Fuck you!!”, Week 5. Designer cancels the job. Client writes a blog about how unprofessional Kenyan designers are.
The ‘must be ready by tomorrow’ hailstorms
My most abbhored client types. Client sleeps, wakes up, goes to shower, and as s/he gently does their shower thingie, comes up with an Idea. Calls Salim. “Sasa, add this feature. I want the system to be able to do H, I, J and W. Can this be done by tomorrow 3PM? Can I see a demo??”
“Well, if by ‘a demo’, you mean my left elbow, then yes. You can see ‘a demo’ even today.”, is always my immediate response. Although what goes out to the client is something to the lines of : “Let me get back to you, sir!!”.
My solution to these clients is a CLEAR contract before any engagements detailing:
- How new feature requests will be made. Officially. In writing.
- The fact that new features will attract MORE pay.
- The fact that new features will be done AFTER the already-scoped features are done
The writing pen; first
A lesson from Mpesa. The most basic system known to Kenya. A system whose workings is as clear to the techie at Safaricom as it is to a cobbler on K-street. But moves gazillions of data and money per second.
I know of many people who have a list of the MANY features mpesa should/could have. Some realistic. Some hormonal. But I was once told: “Make the pen write first. Then improve on how it writes. How it feels to the hand. The color of the ink. The color of the tip. But the PEN, must WRITE first.”
Focus on the CORE. Add the bells and whistles later.
Ni hayo tu.
Back to code.