With regret I’ve decided to finally admit to myself that CFTracker is about alive as a Monty Python parrot. It was never the easiest project to wrestle with, especially having to play Sherlock Holmes to work out what Adobe ColdFusion’s internal bits and bobs were for (thank the flying spaghetti monster that Railo was open source!).
Last orders at the bar!
If there’s someone out there who wants to take over or team-up (CFAvengers Assemble!), feel free to get in contact. I’m more than happy to share any knowledge and still contribute when I can, just can’t manage it on my tod
What was the plan?
My aim was to release a rewrite (version 3) with most of the effort going into how the code is organised. The core code that dabbles with Java wouldn’t have changed that much. I did make some progress with the core API, but never did find the time to get any of the frontend together. In additional there was going to be a few new features that would have made it much more useful:
- Full RESTish API using Taffy.
- Optional database support for additional features (alerts, rules, logs, recent statistics, users)
- AngularJS application, talking to the Taffy powered API.
- Frontend to use nothing serverside except for API calls, making it easily portable / standalone.
- Allows multiple servers to be monitored (just talk to multiple CFTracker API’s).
What is the plan?
The source will still be available on github. To where I’ll also be moving a cut down version of the website (stating that it’s discontinued) to use github pages for hosting. If I do ever get any pull requests, I’ll review them and pull them in for a release still but I won’t be doing any active development from the looks of things.
Anyone with an existing installation, I’m still available to help out if you need it, but within reason
For anyone curious, here are my main reasons for putting the brakes on:
- Shockingly, I don’t get to use it. I only use Adobe ColdFusion in my workplace but they don’t use CFTracker.
- Wrestling with Adobe ColdFusion internals is stone cold hard. In theory I could just support Railo / OpenBD but…
- I’m not experienced enough with Java, so working it all out is slow going.
- Every engine is different, every engine version has it’s quirks (testing involved CF8, CF9, CF10, Railo3, Railo4 and sometimes OpenBD). Dabbling with internals is not like writing normal CFML that should work on most engines
- Lack of feedback: Since I don’t get to use CFTracker, it’s hard to make decisions on what was needed or directions to take.
- Competition: Adobe ColdFusion Server Monitor (Enterprise only), FusionReactor / FusionAnalytics, SeeFusion, Railo Memory monitor plugin…
- Support: For the price (free) I still provided a lot of help to those who contacted me, a small minority did take advantage a little.
- Security: I stumbled onto the odd issue with Adobe and Railo. The Railo team were amazingly agile and quick in discussion, solution and release. Adobe disappointed me in communication, turnaround and views of certain issues. Without going into too much detail, it did delay me waiting for opinions/decisions.
…so that’s it. Comments are welcome, suggestions too. Sorry if this depresses you in any way