The Magma system has been in development by 4 programmers at IJW on-and-off over the past 3 years. Back in February we began a pilot/beta site install and since then have had a few problems gathering the information we need to get it working. Last week I think we got enough to get the ball rolling again. I had one of my tenants come in to the office for a couple days to do data entry. Tonight I went back into the office and further cleaned up a few things before barcoding tomorrow morning. Right now I feel kind of excited about barcoding because it's one of the first major steps. We still have a bit more data entry and cleanup to do though.
Mike found out that by switching the web based reporting system's template engine from yapter to smarty, we get an over 900% increase in performance! One example report returned 15800 records in 6.5 minutes with yapter, and just 20 seconds with smarty! This reporting system has been built to work with different IJW systems and it's being used with Crescendo, Kiosk Reports, and Magma right now. Any database can be described to it so it can be used for 3rd party systems too. At one meeting with a different client, the IT staff asked why we didn't use Crystal Reports because it is the industry standard for reports and is very powerful. I have worked extensively with Crystal Reports 7 and 8 and know it's power. The trouble is, most of the people using our systems don't understand database tables, left and right joins, writing scripts to display words depending on a constant value, or anything about our database at all. Not all people running reports have IT skills. These people want to just click click click and get the same kind of powerful reports. In our reporting system you go through a 7 step wizard that gives you the same power as a person typing in an SQL query by hand, but it's super simple to use. In the first step you pick a report type which tells the system which preconfigured joins to use, and which fields are available. For each report type it knows the constants and data types and presents the user with simple to use and understand screens. It can also do grouping and summarizing. I really love it. This is something we've talked about doing for years but never did until just recently.
Hopefully on July 26th (my birthday!) we'll be training a few of the staff to try the system with live data.
Working from his home office in Toronto,
Ryan de Laplante can be found developing software in
Java by day, and obsessing with technology by night.
Ryan has been designing and writing software for
IJW since 1998 and is very passionate about his work.





