Some of the new things I’ve been working on include:

• Email data into the system
• An API
• Natural language parsing for adding tasks
• Keyboard shortcuts

Now is also a good time to review how I stuck to my original timeline. I’ve never been a big fan of gantt charts (which is why they’re not one of the features of my project). I did however make an effort to stick to the one I came up with in the first week. Instead of using it to mark strict deadlines, I used it to keep an idea of where I should be in relation to everything else. Here is a summary of how it went:

• I didn’t spend 3 weeks researching database systems as expected. Although I did spend a few days researching nested sets within SQL (and opted to use MPTT) I opted to use MyISAM as I didn’t real feel the benefits of InnoDB outweighed the cost in tersm of runtime.
• I did spend about 2 weeks on the database schema as it was originally very complex. It was later mostly scrapped in favour of a much simpler, more streamlined schema based on the ever changing requirements. Had I realised earlier I should descope, I could have saved some time here
• I spent a little more than 2 weeks on the UI and application flow. Although I got the basic site structure up very quickly, I spent much longer thinking about optimal flows
• Project management – I didn’t spend as long as predicted on the project management part of the project. I spent about 2 weeks getting it to allow nested and linked projects, which was later all rewritten in favour of a much simpler approach… keeping projects in isolation and using tickets to separate parts of projects.
• Task management was implemented along wide with the time tracking part as they were very similar and had dependencies between them. This means I spent a less time than expected across the 2 parts as they were combined
• Milestones where also simplified so the time spent was also reduced as result
• The messaging system was also dropped completely – it had only been started at the point of my restructuring
• Tickets took roughly as long as expected
• The scope of files changed to include some new features but the time spent on it stayed the same
• The public beta (alpha) launched on time

One thing I found to be a repeating trend was that I overestimated the time things would take, which is a good thing, as I didn’t allow any time for the more interesting features

• People don’t need nested/meshed projects – One of the things that I was considering was having a project mesh, which allows similar projects to share certain data between them. When I asked people if they thought this was a good idea and would use it, almost everyone said yes. However people don’t actually seem to need this sort of functionality. They just use tickets to separate the project out. As a result, the scope of tickets has changed to allow for this instead of being a change control mechanism.

• People don’t need fine grained permissions – This is another thing people said they needed, but having worked in groups of varying sizes and on big and small projects, people don’t nee fine grained permissions so that has been removed

I was told to focus less on the more traditional and cumbersome parts of the application, such the following sections have been removed:

• The ability to categorise messages, tasks, files etc. No more tags or folders for files – Although useful, this functionality is nothing new and adds little value.
• Search – Although a very useful part of any application, especially one which stores large amounts of data, this also has to go. The time and effort which would need to be invested for an efficient search (which would probably just consist of a 3^rd party library with a few modifications such as Zend Lucene) or a simple LIKE claused lookup provides too little value.

Other things may be removed over time where there is justification.

Instead, I will be focusing more on the innovative features. My supervisor and moderator where particularly interested in the idea of linked resources, so that will get some more attention.

The main things I need to do are conduct surveys and some interviews. The surverys will be sent out in bulk (but not spammed) to people in the correct target market, and I will conduct interviews with several project managers.

I also need to carry on anaylsing existing PMSs, and rate them using some common method. Will produce a table at the end with a score for each site for each piece of criteria.

Aim to get a week by week plan to my supervisor in the next week or 2.

We just discussed what our projects where about and any ideas we had. We also talked about th structure of the project and what needs to be covered. The remainder of the meeting was discussing the alternatives for 2 of the students who hadn’t done any programming before, and how they could tackle the implementation portion of the project.

The next meeting is scheduled for Monday the 19th of October, 1pm