Wikileaks (below) and Wikipedia are constantly seeking funding.  A better system for accummulating and spending public funds would help ensure these guys get their funding.

SS has not delivered yet.  It is partly my fault because my attention has been upon design rather than development.

I have stopped using HTML mockups to model the application layout but I am continuing to use them to experiment with style:

Now I model the layout in a wiki document.  Wiki documents on a server can be accessed from any computer and, since wiki software records all document versions, I can edit the documents without accidentally losing valuable ideas.

To ensure LH finished the job as required, I created user cases and agreed to pay a $100 bonus.

The next job is to enable topic members to choose their topic Administrator. I have accepted SS bid of $100.  He is very enthusiastic about the application  (he has read the entire blog).

I learnt, again, that UK and USA rates are 3 times more expensive for, apparently, equivalent skill:

• LH to whom I ended up paying $300, is based in USA
• SZ, based in the UK, quoted $300 for the job
• GS, based in Lithuania, did not bid but has done the same amount of work for $100

LH has identified a weakness in my method; I cannot quality check code (see below).  I do not know how to fix this problem yet.

As I go through the code, I notice how badly written it is. The code is sloppy, inefficient, and under regular traffic, it will certainly bog your server down. It is slowing down the revising process drastically as I have to spend a lot of time figuring out how the code works since it has no real structure. I highly recommend you rewrite the code after this project.

I am having trouble topping up my Scriptlace credit.  It is capped at $200 per month for most payment methods.  I have attempted a Bank Wire to get around this but the payment seems to have been lost along the way.  The manager of Scriptlance, Rene Trescases, gets a bad rap online but, in this instance, it is not his fault.

[Update: 14th January 2009]

Like the phrase ‘frequent difficulty of describing what their invention was for’:

But the founder of the Academy of Applied Science, and the patent lawyer, regularly recognised himself in the inventors, old and young, who came to see him.  The streak of craziness; the thrill of the quest; the frequent difficulty of describing what their invention was for; their vulnerability to ridicule, because what they had done or thought was new; and the need to protect that thought, as something interesting and precious in itself.

Economist magazine, “Obituary of Robert Rines, scientist and Nessie-hunter”, November 21st 2009

In principle, there are three ways of limiting human environmental impacts: through population policy, technology and governance.

Economist magazine, “Falling Fertility”, October (?) 31st 2009

The Economist said that world population growth will drop below its replacement rate before 2050 and there is not much that need or can be done to speed it up. This means that, if there is a panacea to be had, it will be of technology or governance.

If it is of governance then it will be either a great leader or a great system of governance. A great system of governance finds and empowers the world’s best leaders, wherever and whoever they may be. Is this a mission statement?

Whilst thinking on these lines, I happened across this excerpt on Wikipedia:

The power of Mass-collaboration is evidenced by the scope and accuracy of the Wikipedia project. On his website Mass-Collaboration.net, Kevin St.Onge discusses the opportunities that mass collaboration presents to the world in achieving a global democracy. He argues that governments as we know them will be dissolved and replaced by the collaborative efforts of the entire world population.

Found article in The Times newspaper which considers how differences between African and British behaviours impact feelings such as companionship and lonliness.  Excerpts:

In Britain we shut ourselves off from other people and leave the lonely to themselves, especially at Christmas. Loneliness and depression are serious afflictions, created by the way we live.

A student friend from Ghana tells me that the first time she felt lonely was when she came to London.

While Westerners tend to shed family members, Africans greedily gather and hoard them.

Outside South Africa, very few Africans have lost contact with the village they come from. So even in modern towns, village ways persist.

In the past, the worst punishment in many African societies was expulsion. To be excluded was worse than death.

This communalism ensures that no one is left alone, but it has negative side-effects. For example, distant family members can call on you for money. They will turn up unannounced and expect to receive hospitality. You cannot refuse. When rich men die, their fortune is pulled to pieces and squandered by the many people who can claim a gift from the departing relative. And in most families there is a delinquent who has broken the rules or is disliked. They — and their offspring — are excluded or tolerated, but exploited. These days, when labour is becoming more expensive, the traditional practice of taking the child of a poor relative into one’s family to help them has led to exploitation. Where the child is a girl it has even ended in a relationship of slavery and rape.

Communalism can also make societies deeply conservative. Where maintaining the community is the ultimate goal, important but divisive truths cannot be discussed for fear of creating a rift, so decisions are left untaken. And the African family ensures there is no such thing as a self-made man: the classic rootless entrepreneur of 19th-century Europe or America who tears up the rule book and builds a new world.

Africa’s traditional communalism has a lot to teach a world that suffers from loneliness and depression. Africa still possesses the sort of community that we talk about but rarely experience. And best of all, a society that does not leave its members to grow old and die neglected and alone.

From “In Africa they won’t feel lonesome tonight”, The Times newspaper, December 29, 2009.  Author: Richard Dowden [who] is director of the Royal African Society. His book, Africa: Altered States, Ordinary Miracles, is published by Portobello Books

It is the end of 2009.   My goal for 2010 is to complete the current version of the application (Version 5).

The job is to develop Reply feature for threads in a proprietary bulletin board.

Replying to a comment within a thread creates a new topic.

The rationale is that a reply to a comment within a thread is a deviation (i.e. off-topic) and so should be the start of a new topic.

He took about 8 days and charged $60.  Sadly, for me, GS has now got a job and is no longer available to help.

Unsatisfied by previous description of system for contact management and so spent time working on it.  Finally came up with way to describe it effectively (see pictures).

Some comments from edwardtufte.com on managing and promoting bulletin boards:

Another reason to program in pre-approval only is that eventually the moderators of every online forum find other things to do with their lives. The server doesn’t realize this and soldiers on processing postings. Spammers discover a happy home and the database fills up with crud.

Alas, like all security (antiterrorism, insurance, police, copyright), there isn’t a one-sentence answer except to say every site must make, and continuously review, its own complex cost-benefit assessment and maintain a layered defense

There’s nothing like a good argument to inspire interest among members and to attract attention (and even links) from people with similar interests.

I thought of a new term: collaborative search.  I prefer this to search-on-search.

Took 2 weeks holiday and, on return, motivation seems to have gone.

• Continue or stop developing Mootka?
• Develop Mootka as Wave extension or as standalone application?
• Develop as I envisioned regardless of Wave or adapt because of Wave?

I have chosen to use CakePHP (an MVC framework) on the advice of Aaron. His advice:

You should ensure some sort of standardized plugin architecture is built into core of the script which would allow ‘extensions’ to be added without modifying the core. Creating such an infrastructure in code is not the easiest thing to do (lots of overhead like in writing lots of code that does not do anything directly and documentation on how to use it) and needs a fair bit of planning. Without it, after adding an extension it will be hard to change/remove without breaking things, somewhat like grafting on a finger rather than putting in a hubcap. The extensions will be prone to becoming a mishmash when there are multiple developers who are not working together.

Frameworks are great for such projects since they usually enforce a specific architectural pattern which makes it easier to build and add new components by providing built-in access points to data, functional code and for output control.

My new development method is working very well.

For each version I develop a model.  This describes how the application should look and work once the version is built.  I can continue to perfect it without interfering with development.

I develop the model project by project.  Each project adds or improves a feature or, in some cases, more than one feature.

There have been 3 projects this month:

• 001 – Build A Tiny Bulletin Board – $550 – SZ
• 002 – Changes to Small Application – $100 – GS
• 003 – Improve Search – $80 – GS

Typically, I return in the evening and test the whole application including new work.

I record everything:

Developing a flow diagram for Project 001 turned out to be excellent way  to explain unfamiliar functionality:

I am deliberately trying to chop and change the developers I employ.  Benefits include:

• Less reliance on any one developer
• Can optimize developer skills and price by job
• More likely to spot the bad apple early (peer review)
• Clear end to projects

Bidding tips:

• Make it ‘featured’ if it is a big job
• Make a short list of developers (based on price, ability, integrity, ease of communication, accountability)
• Reply to everyone and explain your decisions
• Tell everyone when you will make your decision and stick to it
• Look for emails that are well written (well written emails is more liklely to mean well written code)

Sample proposal:

The job is to make some changes to a small bulletin board application.

The application URL is Mootka.com.

SKILLS REQUIRED:

CakePHP, PHP 5.1, MySQL 4.1, CSS, HTML

OTHER STUFF:

I have described [changes to] the application using HTML templates [and a flow diagram].
I will help test the application (evenings after work).
I am available via email or telephone.
I will need you to update the application online.
Your code must be fully commented.
You must be willing to fix any bugs with your code.
On completion please supply a copy of the code prepared for localhost.
You must be ready to start immediately.

JOB LIST:

1. Add .bold {font-weight: bold;} to style sheet in order to stlye account holder username in main menu.
2. Change layout of main menu and link My Profile to account holder profile.
3. Link all usernames to user profiles.
4. Add <p>Last logged IN: [DTG]</p> to My contacts list items.
5. Add Description field to Profile and corresponding field in Profile Editor.
6. Add Description field to Thread and corresponding field in Topic Editor
7. Add an About page.
8. Ensure matching words in Results are highlighted as per design.
9. Ensure all notifications in right place and that notification meaning, gramma and punctuation is correct.
10. Ensure appearance (i.e. style) matches the design exactly.
11. Make the radio button for topics on the search form selected by default.

Development tips:

• Do not change the goal posts.  It is unfair and will dishearten your developer.  Keep a list of things you want to change for the next project.
• Build your application bit by bit.  Do not give your developer indegestion.  Lots of small projects is better than a few big ones.
• Develop a list of tests. Use the last test you did as the template for the next test.  This way your tests get more thorough and you can confirm the issues from the previous test were resolved.
• Give the developer only the failed tests.  Developers are busy and this saves them time looking through whole list of tests.
• For a failed test, say what you were Not Expecting and what you were Expecting.  What is obvious to you is not always obvious to the developer.

Spotting Look2Look look-alikes all over the place now. None have cornered the market.  Judging by the increasingly imaginative marketing campaigns, people still believe that the problem is one of critical mass.  For example, one start up called Placeconnect offered to share 50% of its value with its first 1m users.  I switched to Mootka from Look2Look because I think it is about offering more.  Look2Look was like building a classified ads site (such as Gumtree) with only one category available (Missed Connections).  Mootka is an attempt to apply the search-on-search technique in a more general purpose way.

Since the Scriptlance hires had not worked out, I started looking for local programmers. I found a guy on Gumtree who charged £25 per hour. A reasonable rate for London.  I met up with him.  My target budget was £1-£2,000 to build complete application (Version 4).  He estimated 6-8 weeks work which blew my budget.

Since local programmers are too expensive, I re-examined why the Scriptlance contracts were not working out and made some changes. The first change was to make each Scriptlance project smaller.  Development is now feature by fetaure, not version by version.  This makes it easier for programmers to quote for each job, it makes each job more manageable, it means I need spend less time talking through the functionality and it means that I do not have to expose the whole design at the bidding stage.  The second change was to include a flow diagram in the design documents (in addition to the commented HTML templates).  This should also reduce the time I spend talking through functionality.  These changes were substantial.   To celebrate I have assigned the current version, Version 5.

The first job I posted on Scriptlance was to build the core. The core is Version 4 stripped of as many features as possible.

I paid an extra $19 to get my project ‘featured’. I did not provide a guide budget. I provided my private email address.  This proved useful insofar as developers were still able (and did) contact me after the bidding closed.

My project had 15 days to run but the interest started to plateau after 24hrs and so, 48hrs after the project was posted I selected a bid. The bids ranged from from $100 in 5 days to $600 in 8 days.  The locations included Canada, UK, India, China, America, Pakistan, Thailand and others.  I shortlisted 3 from a possible 20 developers (including 10 bids).  I discounted programmers who replied with stock answers, bidded too high, did not appear to be truly interested or had dodgy reviews.

From my shortlist, I selected a $300 bid by a UK developer. UK was a big selling point because of increased accountability (subject to same laws) and easier communication (face to face if needed).

Some guys suggested using development frameworks such as Zend or CakePHP. Initially I was concerned that this would limit future hires by adopted framework.  Latterly, I realised that a framework would help homogenise coding standards throughout the application and provide a ready structure for modular development.

Finally, this month, my laptop broke and I got swamped at work. Now, I have got a new laptop and some leave so we should see some progress in September.

Chose to use INNODB database engine because I can test my schema and queries without needing to learn PHP. The INNODB engine enables you to knit data together using primary and foreign keys.  The alternative, the ISAM engine, requires you to maintain referential integrity programmatically.

Since returning to work post-shoulder op, I have spent a very small amount of time on development.

• XAMPP Lite which gave me:
  • Apache – local server (http://localhost/)
  • PHP – I know it works because my PHP scripts work
  • PHPMyAdmin – interface that makes database stuff a little easier
• PHP & MYSQL Web Development – bought off one-time IT student, now colleague
• W3Schools – excellent start point. Also taught me all about HTML and CSS some years back
• Google – if in doubt, Google it.  Frequently, first result is MySQL.com
• Notepad++ – I examined various development environments but settled for the simplicity and familiarity of Notepad++

Discovered a beautifully designed bulletin board called designateonline.com that, somehow, appeared to be lifting my ideas:

But a quick search on The Way Back Machine revealed that their work pre-dates mine.  Their website looked like this in August 06, well before bounceideas.com was online:

Incidentally, try putting Google into the Way Back Machine for a fascinating insight into their origins.

Hiring non-local programmers (through sites like Scriptlance) to do non-standard jobs, so far, has not proved effective. I will still use Scriptlance but will be more inclined to hire local, and therefore more expensive, programmers – you can knock on their door.