Contributor stevec

=================================================
Priority: Medium
Realization: (Maybe) Easy
Improvements in terms of usability: Very High
=================================================
Let's say that Bob, an ubuntu user wants to try another ubuntu flavor, he goes to Synaptic Software and selects Install Kubuntu Desktop. Waits, Done, he's got a brand new Kubuntu DE. After one week he realizes that he doesn't want it anymore for many reasons, so he decides to uninstall it. Again, goes to Synaptic and chooses uninstall Kubuntu Desktop.
Whats his surprise? Only the meta package is removed so he has to uninstall MANUALLY EVERY SENGLE KDE APP.

Please make this work in both directions, I'd be nice to uninstall Kubuntu desktop and everything else. With just one click.

Many people in my family use the computer.
My mom, dad, sister and brothers.

Everyone have their own user account, and want to download and save stuff, but the hard disk drive have limited capacity.

Add so you can assign "disk quotas" for the user accounts, so that I can limit the accounts to say 1 gigabyte each.

I frequently watch the system monitor on my top panel, e.g. is my cpu busy, what about my memory, that sort of stuff. Then to check which programs are running I use top or htop in a terminal.
Would it not be nice to combine the two? When you'd hover your mouse over the cpu part of the system monitor, a pop-up would tell how much cpu is used by the different processes (obviously with a threshold or a limit of say 5 listed programs). Same for memory/swap.

Splash Screen doesn't hide bootup messages such as FAIL and PASS. Users have the option to see it like in Fedora if they click the arrow still with the Splash Screen showing. Not all Ubuntu users are newbs, and they want to be able to use Ubuntu as well.

i would like to subscribe to some threads/ideas and to receive some email notifications:

SUBSCRIBE/UNSUBSCRIBE

- on a specific idea that has been discussed on
- on categories, maybe i would like to receive emails on Gaming ideas (Graphics, Hardware, Server, System...)
- on "My ideas", i would like to be notified where someone has a response for my ideas that i've submitted
- on ideas that i promoted or demoted
- on ideas that are "Most popular today", "Most popular this week", "Most popular this month", and receive daily, weekly, monthly notifications
- on ideas that are "Ideas being worked upon" or "Implemented ideas"

so for short i would like in my profile and also on specific pages some links that allow me to subscribe/unsubscribe from email notifications.

When I play chess in real-life with a real chess board, then after we kill a chess piece, we move it away.

Like, if I kill my opponents chess piece, I move it to the left.
If he kill my chess piece, he move it to the right of the board.

So the dead chess piece stands outside the chess board, but on the table.

So I can see and get an quick overview of which pieces I have lost, and which he have lost.
So we can keep track of it.
And we can compare who has taken out most chess pieces, and who has taken out the most valuable chess pieces.

In glChess, after you kill your opponents piece, it just disappears.
I propose you make it so you can see the dead pieces, outside the board.

Image to illustrate the concept:
* http://launchpadlibrarian.net/12657821/chess.png

This way, new users who get Ubuntu, and open Pidgin, see #ubuntu, and can click on it, to chat about Ubuntu and get help if they wonder something.

This could also make it easier to get people involved in the community.
To make it easy for them to get help and to help others.
To make them feel that they are not only using Ubuntu, but that they are actually a part of it.

And hopefully also get some contributors. :)

So if they get involved in the community, and feel they're part of it, they will be like "Wow, I'm in this cool Ubuntu club!" and be proud and go brag about it. :D

The System Log Viewer is too simple.
It needs more functionality.

It needs so you can sort it by event type, date, category, source package, etc.

It needs icons or colors.

And have "severity" so that it displays differently informations, warnings and errors, example different icon or color.

http://tinyurl.com/2tjxl8

I don´t think this is possible rigth now, would be great to display the size of the folders in the tab "size in Nautilus", currently it is just for files...
Maybe cache the results, and only compute the size of a folder if there are changes, so it won´t have to do it everytime ubuntu restarts...
I don´t say to make it default, no, that would probably slow things too much, just add that option, so we can turn it on...

Nautilus is curently very slow especially when opening folders containing many files and sub-folders.

It takes 10 seconds for example to display /usr/bin directory with Athlon XP 2500+ CPU and 512 MB RAM machine. Thunar does the action with no lag on the same configuration.

Pressing CTRL+Z on Desktop and Nautilus undo the last action.
for example when moving a file, CTRL+Z move back that file.

It would be nice if the standard dictionary included with Ubuntu could be spruced up a bit. Here are some ideas:

* Keep a copy of the dictionary database on the computer, so that you can access it offline. As of right now, I can't look up a word if I'm out on the lawn typing. It could be updated when the computer has an internet connection. (Manually/Automatically?)
* Show results as you type. For example, if I typed "tre" in, I would get everything from 'treacherous' to 'trey' displayed in a list from which I could select the word I wanted. This would update as I continued typing.
* Have automatic hyperlinking from all the words in the definitions. If I look up "oxygen", and I don't know what a 'silicate' is, I should be able to double-click on that word and be taken to its definition (back and forward buttons would be useful here)
* Show the thesaurus by default.
* Have a more standard dictionary database. I'm not sure about licensing, but if we could access reference.com 's database, that would be great. Also, let the user select the databases they want to use.
* Tie in all other dictionaries (OpenOffice, Firefox) to one central database for spellchecking. Spellcheck should be a global feature, with every program using one database which would be editable through a GUI or the program you are using.
* Make an overall prettier GUI, but clean and simple. All I want is a search bar, a definition area, and maybe a menu or two where I can customize the above ideas.

One of the things that I think makes OS X such a well-designed operating system is how easy it is to create characters with accents. When I had to take Spanish classes for my degree, I always used a Mac to make my life easier. Windows makes it too complicated. Unfortunately, so does Linux.

The way it's done in Mac OS is that a key combination provides the accent and the next keystroke adds the letter. It's also very intuitive, for instance tildes are almost always going to be over the letter n, so to do a Spanish n-tilde as below you type Command+n, n. To do an accented e you use Command+e, e.

It really makes life sooo much easier for anyone typing these foreign characters that I think it would drastically increase the marketability of Ubuntu to the people affected (as well as students fulfilling foreign language requirements).

- Accented vocals: á é í ó ú
- Spanish ntilde: ñ
- Spanish punctuation symbols: ¡ ¿
- Euro symbols: € ¢

The package management of Debian-based distributions is by far the best available; even better than Mac OS X. It's the best part about using Debian-based systems.

All the documentation still suggests people use apt-get to install things, and Synaptic and Add/Remove Programs, to my knowledge, are based on apt.

The good thing about apt is that whenever there's missing dependencies it's smart enough to download and install them itself. The bad part is that once you've installed something, if you choose to remove it you can still be left with programs or libraries that were installed as dependencies but are no longer required.

The reason aptitude is better is because it remembers what gets installed as a dependency, and if you go to remove a program and it has a dependency that aptitude knows is no longer required by any program installed on the system, it'll remove that dependency program also.

Because of that I think making Synaptic and Add/Remove Programs based on aptitude instead of apt would be an improvement.

There should be *some* type of version controlling or incremental storage of /etc so that if you change a config file (or it's changed by an update or some other program) and it screws up your system, you can easily either check the differences to figure out what caused it or simply revert to the working version.

It doesn't have to be a version control system. There's a forums thread on using rdiff.
http://ubuntuforums.org/showthread.php?p=872142

I'll be happy with anything that helps me fix something I accidentally borked with my tinkering.

Administering multiple systems and often through SSH, I spend a great deal of time at the command line. While I know the current version is "good enough" and stable, I think the default command line could use some overhauling.

I think the command line should be as colorized as is humanly possible. Almost every distribution now comes with ls aliased to ls --color. That's an example of what I'm suggesting but only the tip of the iceberg.

I don't know if it's the default but I've inherited a Gentoo machine where the prompt itself is colorized, nano, less, and diff are all colorized.

Vim and emacs should have syntax highlighting out of the box (I think emacs might already, but I know I had to edit .vimrc to add colorization to Vim).

It may sound prissy but if you're staring at terminals all day it really makes things so much more readable.

Ubuntu should adopt standardized usage of IEC binary prefixes (MiB, KiB, etc.) and SI decimal prefixes (MB, kB, etc.) throughout the most visible parts of the operating system.

When referring to sizes based on powers of two, software in Ubuntu should use IEC binary prefixes: KiB for 2^10 bytes, MiB for 2^20 bytes, GiB for 2^30 bytes, and so forth.

When referring to sizes based on powers of ten, software in Ubuntu should use SI decimal prefixes: kB for 10^3 bytes, MB for 10^6 bytes, GB for 10^9 bytes, and so forth.

Like most of us, I grew up writing "MB" and saying "megabyte" when I meant 2^20 bytes, and I expect I'll continue to use SI prefixes in conversation. In those days, the rate of error was minimal -- around 2.5% for kB vs KiB, and around 5% for MB vs MiB. As storage capacity continues to grow, and users deal with larger files, the error rate increases -- around 7% for GB vs GiB and a hefty 10% in the TB/TiB range.

Standardizing Ubuntu's description of data sizes using the IEC and SI prefixes will help to avoid ambiguity, and will also help to prevent the increasingly large errors that result from conflating the two prefix systems.

Now i don't see any informations when my archive has ben extracted.

From #5352 (merged):

When you are extracting an archive in File Roller you get a progress bar with a little box that moves to the left and to the write. The problem is that this gives no indication of how long the file will take to uncompress. I know that in Windows there is a progress bar the moves depending on how much of the file has been uncompressed. You can also check this manually by comparing the size of the uncompressing file to the total size that it should be once completed.

When you click "Save as" menu item in gnome application a dialog windows appears and you naturally start typing new name but that field doesn't have focus and you have to click it. It would be perfect to give focus to the name field automatically.

I've had this in mind for a while and have wondered why it hasn't yet been implemented . . .

What prevents the distribution install, either during the installation, or on first boot, from detecting the specific hardware of the machine on which it's been installed, and producing a custom installation of the OS? Effectively, 'lspci', "Oh, you have that particular piece of hardware, no need for /that/ module, but *this* one is exactly what you need."

Advantages:
* unneeded kernel modules definitely not in memory, because ...
* unneeded kernel modules no longer kept on disk
* if compiled, can be compiled /exactly/ for user hardware
* faster startup as only needed modules are available/scanned

On a related note, what keeps the installation from compiling (perhaps optionally) a custom kernel, with gcc optimizations for the individual machine, processor, ram, etc?