On 02/22/2015 08:03 PM, Kurtis Rader wrote:
Ray, You're missing the point. Every feature has costs as well as benefits. The costs are not limited to writing the original patch to the source code. Every new feature makes future changes more difficult. Every new feature requires additional verbiage to the documentation and thus a potential source of confusion. Every new feature makes it more likely that someone will be surprised by the behavior of the program. Every new feature increases the cost of supporting the program. Etcetera.
I couldn't agree more. There is always a 'price'--several prices, some of them up-front and obvious, but always some lurking in the dark waiting to ambush you.
The value of the proposed feature has to be considerably larger than those costs.
Of course. I argue for it, but the devs will decide if the 'price' is worth the minimal utility. I'd suspect that the price would be very low, but it's not my call. I mostly like it as a bash compatibility thing--maybe this is important to bashers, who knows. One for sure. It's not worth any trouble.
I love zsh. I also hate a lot of the "features" that have been added which benefit a very tiny percentage of its users at the expense of everyone else. Including the people trying to help new zsh users and keep the project alive and the preferable choice from the alternatives. There are a million features that could be added to zsh which would make someone happy. Implementing all those features would kill zsh.
Again, I couldn't agree more. We see Bart and Peter and others wrestling with this kind of stuff every day, it's an intractable mess. Nope, even if it were my call, I'd only do this BG color thing if it were dead simple, absolutely clear and predictable and reliable, and of real worth to bashers.