No, choice is poison to a software program and should be avoided unless there are very good reasons to justify an additional option.

You know, I understand wanting to put a little pressure back on the "make everything configurable" crowd; there are a lot of software engineers who go a bit overboard on the "just make it an option" approach to design, and then wind up with big, bloated, buggy software that tries to do everything but fails at most of it and is frustrating to configure.

However, "choice is poison" is a rather extreme over-compensation for this problem. Sometimes, options really do provide value. And many times, if they are designed correctly, they don't substantially increase cost; if the features are orthogonal and implemented independently, the additional complexity is only linear with the number of features, not quadratic or worse as happens if they are too interrelated.

I understand that we want to get away from the "you need to spend hours customizing your environment to make it usable" approach to design that plagues some open source project, but going to the other extreme and saying that "choice is poison" is just as bad.

I would suggest that retaining an option needed by some users with unusual perceptual requirements is a 'very good reason'

(Personally, I do leave my XFCE4 desktop on default settings, but I do alter defaults in some packages)

If it was the only terminal emulator, there would be a good reason.

The person in the link commenting "PUT IT BACK EVERY OTHER TERMINAL EMULATOR HAS IT" is more baffling.

Both wanting choice and options, but at the same time not wanting to have to change when there are other options.

"If it was the only terminal emulator, there would be a good reason."

Gnome Terminal may (effectively) be the only terminal emulator available to you if you do not have rights to install software or to change default software.

If you are in an environment where you have to take what you are given, then whinging to the software developers to add in the functionality your boss won't let you have seems like the wrong move.

Yes, you're right. They should apparently be whining to their boss to use something other than GNOME instead instead of trying to fix their work environment at the source.

I'm assuming the said boss would not have approved the update to GNOME without first checking that all the features they require the workforce to have in order to work, are present.

Then yes, they should.

If the boss gives you lemons, you don't go to the lemon tree and ask if it can make lemons that read email, browse NFS shares and change screensaver settings. You go to your boss and say "just-lemons isn't good enough".

When you cook, do you just give people a bowl of nutrient paste, saying "that's all you need to remain functional"? But we'd like something with texture and flavour. "So what, it's not like those things are needed to be functional. You can get by without them."

This is a baffling opinion to have on HN.

There's plenty of precedent. The plan9 guys were very explicit about not having variable colours in acme. The suckless.org team say that you can configure whatever you like, but only through changing the source code. BeOS aimed to have no configuration and good defaults.

I'd rather have something that was well thought through and is internally consistant than a grab-bag of customisation, none of which coexists well - like most linux desktops.

I'm not baffled. Giving people fewer options seems like a pretty popular paradigm these days. We tried "Give people choice" in the '90s and '00s, and it turned out people were mostly unimpressed. "Opinionated software" is the paradigm du jour.

Choice may be good, but it always comes at a cost. Each choice takes time to design, to code, to test, to document, to support, and most importantly to learn. Software designers should aim to offer the smallest number of options they can get away with.

> Software designers should aim to offer the smallest number of options they can get away with.

That's should always be second to providing solutions. Solve the problems that transparency solved first. If you aren't aware of the problems that transparency in terminals provided, then removing it is purely an act of ignorance.

The thing is transparency didn't solve any problem. It was just a nerdiness thing.

This is a classic example of someone who doesn't use a feature being unable to see the use cases for it.

You are ignorant of the problems it solved.

> How can I use a full screen terminal window and still read documentation in that same screen?

I've asked this question. I've yet to see a reply.

Try one of a dozen solutions which split your screen into two parts? screen and tmux both do this?

If your documentation isn't viewable in terminal, why do you insist of having a full screen all the time - resize your terminal? If you do insist of having a full screen all the time, why are you using gnome-terminal at all?

I already use tmux. It doesn't solve the problem in the least.

> why do you insist of having a full screen all the time

Because I'm actively working in my terminal. Why should I waste space moving something somewhere else, especially away from where I'm looking?

> resize your terminal?

Maybe you like limiting the size of your workspace, but I don't.

> If you do insist of having a full screen all the time,

Because I'm not always in the terminal, but when I'm in the terminal, I don't see a reason to limit the space it uses.

Anyways, not you are being willfully ignorant. Rather than try and understand the problem, and how transparency was being used, you are being dismissive. You aren't trying to help in the least.

So, unless you want to take the time to understand the problem, why waste your time, and others, adding a comment that is essentially worthless?

How do you know?

1) Have you searched the literature on perceptual defects and enablement? If so I'd love to see references.

2) Have you consulted widely on mailing lists that Gnome Terminal users use and checked use cases? If so, can you link to those? I'd like to see them.

> The thing is transparency didn't solve any problem. It was just a nerdiness thing.

Um, do they know their market? I doubt grandma is clammoring for transparency in her terminal windows but she's not the primary user profile for Gnome, you know?

Nerdiness is evolution. Flowers are colorful and fancy to attract bees. Transparent terminals are colourful and fancy to attract hackers (aka users).

In new software, I can see your point. This issue relates to a removal of a previously available option.

Linux is not about choice (http://www.redhat.com/archives/rhl-devel-list/2008-January/m...) was posted to HN a few days ago.

Also relevant: http://www106.pair.com/rhp/free-software-ui.html This text is the unofficial Gnome UI manifesto. If you don't agree with it then you probably will never be happy with Gnome and should use something else.

The personal opinion of Adam Jackson is the ultimate and only truth.

And still, a lot of seemingly smart people are baffled by the fact that Linux has failed to penetrate the mainstream OS market for personal computing...