Thanks Tal and Hyperion for the welcoming. It's nice to be talking to you both after so many years reading your posts
Also, nice to see that my concerns are shared by so many great users of this forum and that those who don't share were not pissed off with my post.
I've read a few people saying things like "Building tools like this one earlier rather than later means you can get more out of it", "I get the feeling not everyone understands the significance of the work that's been done", or "[those tool-building efforts] are going to future-rpoof the game". I would be very scared if Josh either shares such thoughts and/or is still in such stage of development after more than five years. Also, my experience is that the idea "You gotta build tools to build a game" is just plain wrong, but I won't get into that.
In terms of "tools for the development of the game", I think that the most important thing a programmer has to *always* keep in mind is the trade-off calculation (a.k.a. there is no free lunch for programmers). It's obvious that *any* new tools have benefits. Even more, any well implemented gamedev tools boos productivity. Since that is true by definition, it should by itself raise the flag that the first crucial question is "how much" benefits is yielded given the cost of building/buying/learning the tool. The second crucial question is "how much time have I already spent and how much time do I have left". Supposed you have a deadline of, say, 4 months to implement a program and with the tool at your disposal you estimate it takes you 5 months to finish the task. Now, suppose that you have an idea for a new tool that would let you do the job in just 3 months, but building the tool costs you 2.5 months. Well, with the new shiny tool you would not save time. Someone could then claim that Josh is a stellar programmer that builds amazing tools in 1 to 4 weeks, which can help improve productivity for many months to come. All that is true, if and only if at some point Josh *stops* building tools and start *using* them.
However, I think that the problem is not merely of Josh never stopping to build tools for him or his team to use. I think that even more important is the issue of not realizing that as much as many players would love to have a deeply moddable game (myself included), he is not developing a game engine for others to build games. The praised deep moddability of LT already costed at least half of LT's development time and there is little to no progress game-wise. What is better, to decrease the depth of a great released game's moddability or to have an amazing unprecedented level moddability to an imaginary game that never happens to be implemented? In my humble opinion, Josh's main mistake has been to confuse, for years now, the nicety of giving us a deeply moddable game with his passion for implementing gamedev tools for others to use.
All that said, I deeply disagree with the following:
Grumblesaur wrote: ↑Sat Mar 10, 2018 10:54 am
Internal development should stay internal; whether LT is built with vim, LTEditor, Notepad++, or a carefully-wielded electron emitter pointed directly at a RAM chip is beyond the scope of the community's interests. Hungry, slavering barbecue guests probably aren't as interested in the technical advancement of the host's grill as much they are in the meat sizzling on its grate.
I fully agree with Damocles above. I love seeing the technical details. I bet that the possibility of seeing the internal development of LT is actually what brought many of the initial supporters of the game on-board over the years. The problem is not seeing the tool-building, the problem is not seeing game-building progress.