So I'm foremost using a text editor for coding. It's a very bare bones editor; provides mostly just syntax highlighting. But on rare occasions I also need to debug something. And that's when I have to resort to an IDE (mostly Netbeans, but got fiddly Eclipse/Aptana working as second fallback).

For general use however IDEs feel not workable to me. It's a visual thing, being used to console UIs etc. And switching back and forth between a text editor and an IDE is slightly cumbersome too. That's why I'm considering extending the editor, not really into a full-fledged IDE - but at the very least integrate a debug feature.

Since I'm working on PHP, it seems not that much effort. The DBGp allows to externalize a debug handler from the editor, so it's just minor integration work and figuring out how to shoehorn a breakpoint feature into the editor (joe btw).
And while I've also got time to do that, I'm wondering if this is really worthwhile.

• In this case it's not a needed development tool. It's just for convenience.
• And the cause for doing it is basically just not liking the existing solution.
• While over time I might extend and adapt this debugger thing, it initially will be as circumstantial as Eclipse. It inevitably starts out as poor development tool.
• Furthermore there is likely not much reuse. (Okay, this is not an important point. Most such software exists sans much of a use case. And also obviously, similar extensions already exist for emacs and vim, so it cannot be completely pointless.)

But what's a general guideline on attempting to conoct custom development tools, particularily if they are not really needed but satisfy personal preferences? (Usability enhancement not certain.)