[Scummvm-devel] RANT

[Torbjörn Andersson]

Max Horn wrote:

> Alas, I was a fool to act that way in the first place. Sure it'll get  
> abused, but who cares -- likely nobody besides me.

I wouldn't assume that nobody cares. More likely people cared enough to
look at the SCUMM game detector, thinking that was how it should be done.

> But since typically my attempts to start discussion on design of
> backend/frontend/glue code leads to no response from the Big Black
> Hole, and in the end I am alone to decide what to do anyway, I
> prematurely cut that discussion out. Which clearly was not a good
> idea. :-/

Speaking only for myself, I haven't yet come across many cases where I
needed special backend/frontend/glue features myself. In those cases, I
tend to simply look at the mail and think "well, this is a bit over my
head but these people usually know what they're doing". That doesn't
mean I don't appreciate the work they - or in this case you - are doing.

> If somebody *does* want to talk about FSNode and how  
> to improve it, that'd be most welcome, but it won't be me starting  
> such a discussion. Because my experience is that nobody on this list  
> is going to participate in the "discussion" (for whatever reasons --  
> because they "do not feel qualified", lack the time, the interest,  
> because they are scarred of me or feel I'll do whatever I please  
> anyway, whatever...).

In my case, that's because the only use I know of for file system nodes
inside a game engine is for game detection. The Broken Sword 2 engine
doesn't seem to require any fancy game detection. There are almost no
hard-coded special cases for different versions of the game. Again, I
figured that "well, these people usually know what they're doing".

> Only fools keep talking to walls, and since I don't like acting a fool
> that much, I am not really inclined to keep this up (my other mail
> regarding String is a last attempt, but I guess it's doomed from the
> start, too).

I don't think so. I would guess that Strings are much more widely used,
and even where they currently aren't, they could be.

> Normally, I'd expect people to speak up when  
> they face problems, to ask for the API to be improved, but that  
> hardly ever happens

If I *know* I'm doing something the wrong way, yes. If I *think* I'm
using something the right way, no. (Of course, there's a comment in
compress_sword2.c which contradicts this. I'm not sure how I'll talk
myself out of that one, so I won't even try...)

> Therefore, I'll grant myself a timeout and will stop monitoring, will  
> stop making cleanups or refactoring or working on the middle end, or  
> anything like that. Most everybody seems content with the current  
> state of things anyway, so my efforts seem misplaced. I feel that all  
> I achieve by trying to be a "leader" with a "vision" is that I turn  
> myself into laughingstock, and cause additional work for myself and  
> everybody else, which is little appreciated.

I haven't noticed anyone laughing, and I hope you don't take my
grumblings about recompiling seriously. (Though my dad did that very
misinterpreting when I helped him move a piano the other day, so who
knows...) Just because I don't have any leadership skills or grand
visions myself doesn't mean I don't appreciate those who do.

But I certainly won't argue that you deserve a timeout if you feel you
need one. I do hope you will be back later, though.

Torbjörn