The development environment predicament has been a on going thing with developers for years. Constant arguments over the smallest things such as programming languages, version control tools, even the great editor wars. I find it quite intriguing how much developers really like to argue over petty things as it can be quite amusing to read many of the baseless arguments. For me personally choosing many of these items has never been difficult except for one the editing environment. That is what this entry is about trying to make sense of it all.

When I develop code I want to be productive. I think this is the case for everyone. Through the years the one thing I noticed is that the IDE or Editor you are using can have a huge effect on productivity. Not in the sense of tasks being difficult but in the sense of not interrupting the stream of thought you are trying to put into code. For me personally one of the worst things ever is to be working on a algorithm and realize you made a mistake 10 lines up and having to go back and fix it then go back and start working again. Each environment out there be it a IDE or Editor has specific features to help combat this for the most part I would think but do not hold me to it.

The IDE is the modern editor of the day. It contains quick ways to refactor large blocks of code, code completion through syntax analysis and parsing, integrates all the tools you need, and the best of all graphical debugging representations of what you are working on. There is more then just this but a solid sampling of features. The key word here is Integrated everything is there and often works with very little configuration. In my experience however the biggest downfall of the IDE is the editor. When you make that critical mistake you need to stop typing, grab the mouse, and then fix your mistake then go back to working again. The other issue is the fact that many of these features often may not have some sort of quick keybinding causing you to have to go through the menu systems with the mouse yet again. Sure the most commonly used features you have keybindings for and I am sure you have them memorized but it is the odd things that are less common that you happen to use more often then others that hurts. One such example could be the selection of text. You usually have 2 options either the mouse to select the block of code or to use shift+arrow key. This is awkward.

The Editor front you have dumb editors and smart editors. Most use old smart editors like Emacs or Vim. These lack many of the IDE fancy features and if they do have a plugin for it odds are it is not as good. The one place they excel however is editing text. When editing text even the novice with very little experience can really reap benefits. For example I have been experimenting with Emacs for a few days now and man I feel productive editing text. Moving around my characters, words, lines, sentences and rapid selection is just awesome. Want to select a line of code 10 lines up from the cursor is easy... C-u 10 C-p, C-Space C-e then DEL or if you want to cut it C-u 10 C-p, C-k. I think one of the most powerful features here is setting the "mark". You can set a mark set with C-Space C-Space then move and make your edit then use C-u C-Space to jump immediately back to where you previously were. I think the overall benefit of these features is to minimize the amount of thought interruption you have when you need to jump and make a edit. No need to grab the mouse and move the cursor.

I am not sure what I appreciate more when I am writing code. Massive Integration with some powerful features or just a great editor environment that minimizes interruption. Could code completion and refactoring really make you more productive enough to sacrifice the power you get from some of these smart text editors. I find myself making lots of small edits in code rather then massive refactors so something like Emacs makes me personally feel really productive. So it comes down to is a sacrifice for a editor worth graphical debugging tools. I have no idea either way with embedded development you are often looking at hex and binary values as well as assembly code all the time so no GUI debugger really makes it look much better.

So my ultimate question is why can't we have a IDE with a amazingly powerful editor? The best of both worlds without it being a hacked plugin that does not really work like the editor it is trying to emulate in the IDE.

Even after writing this out I still do not know what direction to go. Was hoping the post would clear my mind a bit and help me logically put out what I appreciate in a editor. I guess the issue is I appreciate the features both offer and I want both but nothing gives me both. I am not sure I have the time or energy to develop a new IDE from scratch that works how I want it to work. Eclipse is a huge mess and I doubt I can write a new editor component for it to emulate say Emacs. Ultimately all I want is a environment that understands my code and has a really powerful editor to minimize my line of thought breakage and nothing does exactly that.

What is your take on this leave it in the comments I enjoy reading what other people think and what their experiences are like with odd topics like this. Oh and no flame wars :D