Why didn't linux just add some buttons to upgrade OS?

[enhu]

Most Linux guru recommend not to upgrade by simply executing an upgrade command in the terminal because it will somehow be faulty in the end. So why didn't these developers fix this issue so that users can just click buttons to upgrade their system. Has anyone ever upgrade via issuing upgrade commend in the terminal and find no fault to the result?

[Kerms]

Most linux gurus recommends not to upgrade via console because the inexperienced might break something. The statement is misleading though. If you know (and are careful) enough to upgrade via console, go ahead.

Standard software upgrading, while staying within the OS version should never be a problem--via gui or cli--if your package manager's package source list doesn't have conflicting sources.

Distribution version upgrading is really different, on the other hand.

For example, in OS version 1, function Baz is handled by package Foo, but in version 2, package Bar replaces package Foo to do the same function Baz. However a good amount of packages depends on package Foo...

And the only clean way to do this is by installing packages one by one that don't conflict each other (which is usually done by an algorithm). However, in the user's point-of-view, accomplishing this is no harder than issuing a standard upgrade command.

What breaks the system in the end, is that since entirely new software is introduced different methods are implemented deprecated methods are removed, and various other changes that makes the perfectly working config may not work anymore.
And a simple point and click will never be able to fix this.

This is a thing that differs Linux from Windows. Windows, yes, a point and click fixes everything, but you're trading configurability for simplicity. At its core, each Windows install is more or less indifferent than any other windows install.

And to answer your last question, I never had a problem doing a standard upgrade. I only had problems where I tried to install stuff not maintained by the OS's software maintainer(s) (ie. self-compilations, stuff from another but related distro..).

[ehansen]

Most Linux guru recommend not to upgrade by simply executing an upgrade command in the terminal because it will somehow be faulty in the end. So why didn't these developers fix this issue so that users can just click buttons to upgrade their system. Has anyone ever upgrade via issuing upgrade commend in the terminal and find no fault to the result?

To go on with what Kerms said, the rule of thumb is to actually not do upgrades that you don't know about.

For example, if you're upgrading AbiWord to the newest version, and it needs a new library to do feature X, there's no guarantee that the installed library won't conflict with another one. This is a major fault in rolling releases such as Arch Linux. They push out updates the best they can but you sacrifice updates for safety, essentially.

Then you have OSes like Ubuntu and Debian where they have a longer release cycle, programs go through QA until they know it will work or not, and then act accordingly. You sacrifice not getting a patched PHP version for a month or two for the sanity of safety. You'll know that on a default install with no custom software installs, the patched PHP will work just fine.

There's two different models of thought when it comes to software upgrades in Linux. The closest to clicking a button to upgrade a system you can get is what Ubuntu (and similars) have implemented.

I've upgraded from the terminal many times and have had no issues. But, take production servers for example: rule of thumb on those is to not upgrade unless absolutely necessary.

[Rob]

Yep yep.. on production servers, I'll usually migrate to the newer OS version on another box/vm if it's a full release behind..