In the late 1990s my martial arts coach asked me to help keep track of the gym's members as his current way of doing it with a pen and paper was getting complicated as his business grew.
As I had been programming since I was 13 or 14, and had just left school and had begun working as a trainee programmer, I started developing a system to do it.
Version 1
It was first designed for Windows 3.1, written in VB5 and stored the data in a text file, one row for each person and all their details.
The original name for it was Visual Gym Manager (partly because it was written in Visual Basic).
Version 2
Towards the end of the 90s it had progressed. It was now a VB6 WinForms app and I had migrated the backend to Microsoft Access, with a relational table structure.
It was around this time too that I renamed the product to be Arnold - not directly because of the famous body-builder actor, but rather because that was one of the server names at the company I used to work at, and I thought it sounded cool as a name for software too.
Then, in 2002 came the big launch of .Net!
Version 3
Given that I had never gone to university, and was almost entirely self-taught … I went out and spent $2000 on Visual Basic .Net (remember those days? buying a box of CDs) .. I also bought a “teach yourself VB.Net” book and got to work re-writing the entire front end in VB.Net.
It remained relatively stable in this format for a number of years. Regardless of where I was living (Sydney or Rio de Janeiro), I would use my spare time away from work (which was mainly bouncing at this time) and jiu-jitsu training developing the software.
Version 3.5
When I returned to Sydney full-time in 2006, and resumed working 40 hours a week as an IT professional, I gained more exposure to SQL Server (which I had first touched whilst working in Rio). So it was in circa-2012 that I migrated the backend from Access to SQL Server. The front end still remained a WinForms VB.Net app but the backend was now using SQL Server Express.
This was obviously a step up but I still had the problem of having to install it on location at Boxing Works’ reception PC and at my own gym’s premises, and also at a 3rd location that I had picked up.
And everytime I made a change, I had to roll out the executables and go physically to each gym and do SQL.
Version 3.6
The next logical step then was to move the database to an online server, so I did that. And this made doing back end support a lot simpler … but I still had the problem of installing the front end on each PC which became a nightmare, especially as I had now picked up clients in New Zealand and Darwin!
Anyone know the pain of DLL Hell ???
Different versions of Windows, and time zones to support … not to mention I was working 40 hours a week programming, and teaching my own jiu-jitsu classes … I didn’t have the time to do the support.
Version 4
There was only one logical step to do if I wanted to keep my software alive - make it a web app. But I had never written a web application before. Up until now I had only written millions of lines of Visual Basic for windows, and become very good at SQL Server.
So I googled how to write a web application in Visual Basic, and found some tutorials on ASP WebForms. Keeping it in VB allowed me to use most of the .Net Classes and libraries I’d already coded and so I just had to redo the screens, which was still a lot to learn.
But, after a few months of late nights I got it done and working. It looked really ugly … but it worked!
Version 4.5
It was now 2020 and I had become a Microsoft Certified Expert in SQL Server (I somehow found time to study too) and the software was working … but I wanted to integrate with a 3rd party who could do the direct debits in a PCI compliant way. I spent hours going through their documentation but couldn’t figure it out. I remember speaking to their support guy and I said something like “I’m a SQL Server developer and I don’t know how to do web programming and your documentation sucks” … his reply was “well mate, you’re trying to do web development so go and learn it!” … it was the blunt truth.
So I phoned an old colleague I used to work with and paid him some cash to help me with the integration. He got that working, and I mentioned how the thing looked ugly, and he said “oh that’s easy, just find a bootstrap template and use that”. He was right. I’d never heard of bootstrap until then … so another $20 for an online course which I watched on a trip to the city and back (I was working as a DBA at PwC at Barangaroo by this stage) and I got the hang of it … and yeah, it now looked presentable!
Version 4.6
Just after I got it looking better, the webhost where I had it hosted suffered a massive DDOS attack which took the application offline for almost 2 days. And as much I like supporting small businesses (I still use the same company for my sales websites), I lost 2 clients as a result of this downtime and so it was time to scale.
Today both the web application and the sql database are hosted on Microsoft Azure.
Version 5
The shiny “new” version 4 did really well. But … there is a reason that ASP WebForms applications are almost extinct. The whole postback thing where the entire page gets reloaded almost every time gets annoying, especially if the user navigates via the browser buttons which then can cause multiple actions of the same thing happening …
I knew what I had to do. I’d been putting it off long enough … rewrite it as an MVC App.
It’s taken me the best part of 6 months worth of late nights and weekends, but I’ve finally finished converting 90% of the code from Visual Basic Webforms to C# and Javascript in a new .Net Core 6 web application. I’ve had to learn so much more for this too (authentication and authorisation for example) that it has been a long road, but very rewarding. In fact 3 years ago I’d never written a line of javascript in my life, and now I have this application which is probably 25% js.
There is still some VB code there, the nightly jobs and things like that which don’t involve user interaction will stay for a while. But the user experience is so much nicer and faster.
The main thing I am super proud of is the underlying database structure. It’s remained almost unchanged from the original design I did in Access. With one click of the button, the business user can enrol the member into the class, and it verifies their payments are up to date and calculates all other stats like classes for grading and sale expiry etc.
It’s proof that if you build the engine of the system correctly, then everything else is just window dressing. It’s like the chassis of a car … do a google search for "cars with the same chassis" and you'd be amazed at how many different cars actually use the same basic blocks!
I don’t know how to do a lot of things … but one thing I know how to do is build a database. One that’s fast and works.
Coming Up Next
It’s now time to go back to market and focus on getting some new subscribers, and in the meantime my development time will be returning the Apps … yes, I also wrote an Android and iOS app for mobile phones too which allows members of the gyms to check in and receive push notifications of alerts. But it’s clunky and also needs a re-write. Will need to find a new course to take and start the process again.
It’s been 25 years of development, and I reckon I’ll do another 25.
I’m super proud of this product and I honestly couldn’t run my gym without it.
If you know someone who runs a martial arts, dance or yoga studio .. or even a small PT style gym, and needs some simple software to help run their business, please do me a favour and share the link with them. My software is free for businesses with less than 25 members and I know it’ll make a difference.
Thank you.
Rodney Ellis
ps. Link to trial can be found here: https://gymdatabase.co
No comments:
Post a Comment