built using a very lightly modified GPL version of Jexer. TJ provided
a real-world use case to shake out numerous bugs and limitations of
Jexer.
+
+
+
+Maintainers Wanted
+------------------
+
+Both Jexer and TJIDE are seeking additional maintainers. I am not in
+a position in life to take on significant off-hours programming work,
+and am willing to hand these projects over to one or more persons with
+time and interest.
+
+My personal code design philosophy for TJIDE/Jexer is outlined at
+https://gitlab.com/klamonte/tjide/blob/master/java/docs/code_design.txt
+. I realize that some of the features listed below may require
+deviations from this philosophy, but this is what I have built so far.
+
+Some of the areas that will likely require significant efforts are:
+
+ * Editor improvements. The editor is currently very minimalistic,
+ much closer to MS-DOS edit.com than a real programmer's editor.
+ Users will probably desire many more features: drag-and-drop, real
+ syntax or at least regexp highlighting (not just keywords), paren
+ matching, paragraph/comment reflow, and dozens more. The
+ underlying Document/Line/Word model is not going to be sufficient
+ to meet these features. It would be up to you if you wish to
+ retrofit any improvements you make to the MIT version of Jexer.
+
+ * Better Windows and OSX support. It would be nice to ship a
+ jlink'ed JVM on these platforms with the JRE, JDK, and JPDA
+ modules all together. For Windows, it might be preferable to
+ consider doing any of the following: ship a third-party terminal,
+ use PowerShell, or use the newer ConPTY for TTerminalWindow.
+
+ * Project window awareness for modules. Currently the project
+ window knows only simple file paths (where it assumes paths are
+ subpackages) and jars. It would be nice to also have module
+ awareness here, since modules are the future for Java.
+
+ * Bug fixes. The (G)Jexer codebase is quite large despite my best
+ efforts. Bugs are typically very small to fix, but can take some
+ time to find: a simple NPE or AssertionError can sometimes take
+ 4-8 hours to squash. Fortunately, fixing issues in one place has
+ not often led to breakages elsewhere.
+
+ * Plugin support. TJIDE is the first truly new TUI-based IDE to
+ emerge in quite some time. It was deliberately built to resemble
+ IDEs used to teach computer science courses for decades, and it is
+ hoped that instructors will be able to use it soon. A plugin
+ system to define new kinds of targets, editors, and debuggers
+ would go a long way to using TJIDE for other languages.
+
+ * New (G)Jexer applications. So far as I know, (G)Jexer is the only
+ mouse-supporting full TUI windowing framework with sixel image
+ support in existence. I cannot predict what kinds of applications
+ could be built out of it, and how those needs will push back to
+ the framework.
+
+These are what I can clearly see right now. Obviously users are
+capable of finding many more.
+
+I intend to continue poking on TJIDE and Jexer, and will maintain a
+branch to be "the fastest and simplest Java language IDE available",
+which will deliberately remain small.
+
+I hope that other languages choose to transliterate MIT Jexer to
+provide TUIs to their own platforms. I will be happy to help them
+understand the code to support those efforts.