Hi Matthias, I've canceled the NMU, the time for us to discuss this issue. I want to make it clear that my goal is to improve bash, not go over the head of its maintainer.
On Tue, 2024-11-26 at 12:34 +0100, Matthias Klose wrote: > On 24.11.24 11:15, Baptiste Beauplat wrote: > > I've prepared an NMU for bash (versioned as 5.2.32-1.1) and > > uploaded it to DELAYED/10. Please feel free to tell me if I > > should delay it longer or cancel it. > > I don't think these should be configured by default. The config files > should be minimal, and not adding some "preferred" aliases. Especially > for ll I've seen to many different ll aliases over time. I agree with you that defaults should not add some "preferred" aliases, but I would argue that those 'll' defaults are not meant to be preferred but rather universal defaults. Just to give you an example, I myself use a variant of that, adding the '-h' to have human size enabled. That's my preferred aliases and I would not want to configure it as a default. However, the goal of defining a minimal common alias is to allow our users that run a shell from an untouched configuration to have a basic set of feature that may go beyond what upstream has defined. Take PS1 for instance. The default for bash is currently set to: bash-5.2$ I think we can all agree that's not a desired default. Therefor, we came up with a better default that I would think, is beneficial for all our users. I would argue the 'll' aliases fit this case. I would also argue there is no downside of having those 'll' defaults enabled and a tremendous benefit from having it. Now, as to the matter at hand: How can we objectively know if those default are preferred or universal? (or somewhere along those two). I guess we need data points to mesure how users are interested in this aliases. I'll gather some stats about whether or not linux distributions enabled those by default. I might also setup a poll for DD to get their feedback on whether they would see those aliases enabled by default an improvement or not. Best, -- Baptiste Beauplat
signature.asc
Description: This is a digitally signed message part
