Thus Spaketh Idd Salim

4 BabySteps

Banx suggested that I stick to being a tech blogger and a gadgets reviewer for the next 3 months until watu watulie. Moving on swiftly.

Well, @smutinda and @mbuguanjihia represented Kenya in the just-concluded DEMO launchpad event – www.demo.com. I was privileged to get an exclusive on what happened and who won what and why. It was a success story. Letting everyone out there know that there are real coders and thinkers in Africa, contrary to common belief.

This was an awakening.

Shimba ranked highly in their category and was behind a solution from Stanford with their MedAfrica product. The Stanford team, with funding from Eric Schmidt and the best tools (more about this later), had made a gadget that you attach to your back and it gives you a slight electric vibrations when you sit in a bad posture. It helps you sit better and protect your backbone. [Read More here]

MedAfrica was something different altogether. Given the sad fact that there are ONLY 7, 000 doctors tasked with serving 40, 000, 000 Kenyans, MedAfrica uses the fact that there already 10M mobile users in Kenya to create eDoctors. MedAfrica would cut the Doctor:Patient ratio from1:5714 to 1:4.  Here is the presentation.

The 4 lessons.

Lesson 1: Relevance Matters

Ofcourse, In being from a First-world country, this was a difficult sale. With all the online services like my MyPhysician and the rest, it was difficult to explain how so man people could not have access to basic human right like medicare. So, MedAfrica tanked in this initial round. Luckily, not winning the DEMO accolades did not amount to total failure.

Lesson 2: Funding Matters

I can’t talk enough about this. Linet always frowns at me when I talk about this. We always seem to have a “Develop your solution, generate interest and the funding will come” VS “Give me the funding and I will give you the BEST solution” debate. I never seem to be able to explain the later, a mantra I subscribe to.

Seed capital is all developers ask for. Not Millions and Millions. Google would never have been afloat without the US$100,000 funding from Andy Bechtolsheim [Read More Here]. Facebook would not have kicked off without the first investment of US$500,000 in June 2004 from PayPal co-founder Peter Thiel [Read More Here]. I could go on the WHOLE day.

But then again, we are in Africa and funding/seed capital is a pipe-dream.

So cometh the challenge to the Universities and the Government.

Lesson 3: Government Matters

In the US, all Universities offering Computer Science have funding available for the MOST promising students. Government subsidizes and offers security for loans to IT initiatives from Government Banks. Need KSHS 5M to do a project, no problem. Your University and the Government just needs 20% shares of your company, and you will get the loan tomorrow. Try that in Kenya.

Our old-money-old-ideas mental setup, colonial-thinking and defensive business models make this impossible.

So, the innovators and thinkers are still REQUIRED and EXPECTED to fend for themselves. Think about your product, then think about the money to make the product a revenue-generating solution. Once you start making money, KRA and KAC will be sent over to see ‘how’ you make all this money. Then the ‘investors’ will come. When you don’t need them.

People who invested millions in FB and Google, ventures that had NO revenue potential or defined streams, right now are smiling all the way to the Keyboard to refresh their profits screens. There are no risk-takers here. Yet.

Lesson 4: Name Matters

Think of the power in a Name. UoN is know for Riots. Harvard is known for Excellence. Most people would still rather partner/invest in a C student from Harvard, than an A student from UoN. [If the GradeNazis would allow my use of Kenyan Grades for Harvard]. Until maturity comes to our students to elevate our name in such a manner that the very mention of UoN, Strath, JKUAT or CUEA would put a smile on investors and partners, the prejudice will remain.

Also, what we learn or are perceived to learn matters. It makes me sad that we are still teaching PHP, CSS and Javascript in Campus, 4th year. I will not belabour this point.

Kids need to leverage the Fibre and use the 4MB Ngwati links to learn things that blow the socks off investors. Talk QR Codes, NFC, e/m Payments, Cloud, Anything. Si Css.

Back to code

Wazi.

InMobi : CPC is flawed

I used public transport today and it did not take me long to realize WHY I hate this mode so much. Apart from the usual risk of pick-pockets and catching the smell of cheap-cologne, there is the annoying trend of Matatu Drivers and sad Kenyans to listen to Classic105 FM.

The last thing I wanted in a Monday Morning especially after NIC Bank had frozen my funds last night was to listen to the same old regurgitated jokes from Mwalimu the CopyPaster and Minor KaGayNi talking about relationships and love for females. Ohh, the Irony. It is like me talking about how sweet pork is. Na sipendi pork. I guess once you are on TV, you become an expert on everything.

InMobi Please

So, on to inMobi. As a techie and businessman, I was at MoMo last week at the iHub and one of the talks was about inMobi. One of the phrases that had everyone wet was the fact that inMobi churned out 15, 000 Ads per Second. Now, this is good stuff. Ignoring latency and process-lags, this is still a SUPER performance by the inMobi Servers.

After the meeting, I met Wesonga and he wanted advise on how/if and when to Integrate inMobi into his mobile App. I responded the way I always respond to issues. With facts, bila bias ama matusi.

Remember, these are just my views and might be changed later as I get more information and material. As a developer, one needs to decide on the CVB [Cost Vs Benefit] of any feature you add in your app. Especially a mobile app.

The Concern:

Here is my problem, and hence a honest question and concern. InMobi charges the advertiser (and thus can ONLY pay the developers) for Clicks. Not Impressions. So your App gets the COST of getting the ads from inMobi servers (Phone process and network costs + data cost to the app user) and shows the Ad on your App (space cost on the small screen), BUT you will only get paid when a user interacts with the AD.

To me, this is a flawed business logic for the Mobile space. Mobile is about eyeballs. Not fingers. In the web-pages, interation matters and IS possible. But how does one interact with the AD on a mobile phone. How else, other than VISUALLY. InMobi should charge for impressions. Or at least the developers should be allowed to charge inMobi for Impressions. Not clicks. Not on mobile.

Consider this fictitious scenario: PSI have a ‘Mangika na Trust promotion‘ where Trusts are on sale for 2 bob. This is ALL I need to read on the mobile app to get the message. I don’t need to click or ‘engage’ the Ad. Visual engagement is enough. I think inMobi is using a Web model on a Mobile space. Wrong Move!

And this is EXACTLY why I will not use or advise anyone to use an CPC model advertising system on their app. There is NO CBV. Your app will become bloated with Ads and you will increase someone’s sales, but NEVER get paid for Anything. Or utalipwa 200 bob per month.

Then again, maybe I am wrong. Si vita. I would be glad if someone from inMobi would contact me and clarify anything I might have misreported.

Back to code.

Wazi!

THINK for us, please

I have just finished the sad story of Spotify and it nearly made me spitify on my keyboard.

In the US, a lazy bum can just register a patent and sit there for years waiting for someone to ACTUALLY invent something related to the patent and the SUE them and live happily and rich ever-after.

Just a few weeks after Spotify entered the US Market, it hit 700, 000 users in the US and KUJA HAPA, they got sued for Copyright Infringement.

Pathetic. Sad. Really.

And so, you want to know, “What is this brain drain that you speak of, Salim?”. Well, read on. I will not leave you hanging.

Absence of Patent/Copyright

To my own knowledge (and I will graciously accept correction) IP law in Kenya is as nonfunctional as VB code on a Mac. Wine or no wine. Let us say I invent a way to make coders understand what all the fuss is about Justin Bieberre (See, I can’t even spell it.), It will take me 100 times longer, 100 times more hustle to get a patent for that in Kenya than in the US.

The bonus is that if I were (God forbid) a US Citizen, then it would be smooth sailing. I can come to Africa, Listen to an Idea and go patent it in the US. Then wait for these bloody Africans to go big and enter the US market with ‘My product’… and… Kaching!!

Tech-Challenges

We all know African Tech Solutions. Ushahidi, Mpesa come to mind. Apps made by Africans for Africa. Because African solutions are made on an empty stomach, they address a REAL need. A real problem. Not AngryBirds. HungryCoders. No one will give you USD 1M to start an experiment and ‘see how the market responds’ in Africa. So, Tech-challenges leave a lot of coders flat-nosed.

Tech-Challenges present another problem. We see them every now and then and I am big Fan of them, because of the investment opportunity they give local developers. But what about Idea Protection. Are we in a position to protect out ideas.

I am not going to be all nasty and disrespect IPO48, Pivo25, AppCircus or any of the local developer challenge initiatives. These challenges offer a NOBLE and REAL opportunity for Devs to get their app to the next level. But what happens after the events? What happens to the 17 who miss out after the top 3 slots have been taken.

What stops vultures from taking their ideas, shipping them to China or India and using the resources the developers don’t have (time and money) and killing another Kenyan Dream?

So, what now?

Are we fucked? It all depends. You need to decide what you want as a developer. Obviously, there is NEVER room for HelloWorld Apps in developer contests. But what about the Mutindas, the Wesongas, the HildaSams and the Mwais of the local space. How do we protect these people. How do we make sure that not winning in App contests does not spell the end?

What models can we adopt to make sure the Investor’s money is just a by-the-way. The only person I know who believes and invests actively in Local Techies financially is JM. The rest are just happy to invest old-money on old-models. 100 bob making 120 bob, instead of 100 bob making 600 bob.

I don’t yet think Mbetsa is a Millionaire as he should be. He invented a Kenya’s-First and possibly Africa’s-First. But what next after the invention?

Someone once suggested that we start a Kenyan SharkTank. But who will be the hosts?

What do you think can be done to salvage the sitiuation? Tell me. Discuss [20 Mks]

Back to code…

Wazi.

They all come in different shapes, sizes and color. Anything really. Cars, pens, niniis and kereas.

The PHP frameworks cannot be left behind. I have tried all with some various degrees of success.

I nearly gave up searching and learning and decided to design my own Framework doing all the common things I do in a pre-packaged manner. That is what a framework should do. Then cameth the enlightenment.

Kohana, YII (former PRADO), DOOPHP, Zend, CakePHP, Symphony all aim to address the purpose of a framework. To ease software development. But if I have to spend a month or 2 studying a framework, a new templating dialect, I am better off doing Erlang or Lua with my time. Seriously. By the time you start to master version 1.1, version 1.2 is released. With a completely NEW model.

Rasmus Lerdof suggests that NO ONE should use a framework. He suggests we all write a set of functions to do our common tasks and make our own ‘NoFramework’. I totally agree with this. Also, an analysis of all frameworks and models is here. CI wins.

The common failings are:

• Bloat-ware : Most systems come with alot of bloat. To just say HelloWorld, you need to load 100KB of files. Whatever happened to a simplicity?
• Tight-Coupling : Each library depends on 7 others to perform its basic tasks. One missing file, despite being totally unneeded, will render your app unusable.
• Cryptic OO – Most Framework developers are there to show off and even the most mundane PHP code is OO. Java-like classes to do simple things like DB disconnects, all for showoff. Programming/Coding is hard enough. Why add another layer of complication?
• Code that Makes coffee – Most development tasks are routine and if specialized code is needed, then plugable libraries are easier to develop/plug-in that trying to understand bundled one-solution-fits-all ‘FrameWorks’

So, Why Code-Igniter?

CodeIgniter is everything a PHP coder tired of doing the same code over and over would wish for to Santa.

I won’t type/copy-paste everything here and pretend to have come up with it so as to get traffic or an iPad, so here goes: Click and learn why.

I will start posting sites I will have done using CI here, soon.

Watch this space.

Back to code…

Wazi.