I played the Secret of Hutton and Trilby's Notes and I'm convinced that text parser's are definitively NOT user friendly compared to point and click but, they do offer a bit more "flexibility" and thought-process for the player than instead just clicking a bunch or stuff to see what happens. text parser's are funner IMO, you get to mess around and write stuff like "look in ass" or "use gun on self" and then laugh at the strange comments the avatar says or end up killing yourself for kicks. In point and click, you can only do what's actually intended for the game.
Hutton crashed when I said "take cupboard" in the janitors room, so I'm guessing that using a text parser opens up a Pandora-box of errors and difficulty trying to anticipate what the user might type in. Like I said, users type in some really stupid stuff just to see what happens.
What I think sucks though is the guess the noun or verb for of text parser's while in point and click you get stuck less often since all you need to do in such situations is click every single pixel on the screen until something happens.
I am thinking though of using a text parser since I intend to make a static adventure game, like princess tomato in salad kingdom ( one of my favourite NES adventure games, that and De javu ) just without the user interface for the NES, I think a UI can really obstruct the visuals and force them into a tiny box.
I want to make a visual novel styled game with static images and text parser technology, no animation, no walking around,no collision detection or physics or whatever, no fancy User interface with a bunch of look,take,examine buttons, just text and static-pictures and a command prompt for the player to input commands.
sort of like Steins;Gate: Hen'i Kuukan no Octet visual novel...
I tried making an engine like this from scratch with LOVE2d in Lua but coding the text parser was murder so I gave up (-_-).