Contributor glibik

Add gns3 (Graphical Network Simulator) to repositories.

Many people have trouble installing Ubuntu alone. I will list here some common problems people find and in the end I will propose some solutions.

Attitudes like: "If the user doesn't know how to partition he/she should ask for help from someone else more experienced" are not really the solution to attract new users (mainly people having their first Linux experience or without internet access).

Some problems:
1. One of the first things the partitioner does is to ask what kind of partition should be created. Well, more than 50% of the users would say: What is a partition ? (before even have to ask themselves "Why should I do it?").

2. People don't know they should create a "/" partition (does it sounds any intuitive at first glance ?)

3. People don't know why do they have to do it because "in te system X I don't need to do it ! Everything simply works out of the box". We should not think it as any natural concept.

Off course it has many more problems. Let us stop here and show some possibilities:

- An introductory screen with short and acurate information. One suggestion is:

First screen:
"Now you may choose to have your disk partitioned. This process can make your system more secure, responsive and easy to manage. "

Second Screen:
"Some advantages of partioning:
-Frequently people have their personal data (internet content, projects, documents under construction), system data (installed programs, configuration files, etc.) and some shared data (like music, video, documents, etc.). If you have more than one system you should use partitions to organize it. If you have only one system you may use partitions to separate the contents from each other. This technique permits, for example, one to erase the whole system data preserving your personal and shared data and allowing you to install a new system with your old content.

[....]

A choice to install without all the applications and game bloat.

I dont want
- calculator
- tomboy
- all the games
- gimp
- open office
- ekiga
- evolution
- pidgin
- transmission
- movie player

well maybe i do want one or 2 of those apps. but hey, there is synaptic..

Just the OS and some admin tools, gedit, firefox, altho firefox could also be installed after 1st boot.

Basicbuntu

This is a small thing. In system>preferences>keyboard shortcuts, the windows key cannot be used as a modifier. It can only be used as a stand alone key. I think it should be allow to be a modifier, as when I use compiz, I like to set cube rotation to windows+#. When I turn off compiz, I'd like to make metacity have a seamless transition in terms of hot keys, but that is only possible if the windows key works as a modifier, not a stand alone key.

A separate partition for /home has been proposed for a long time in the forums. It implies some risks, though, so based on disk size Ubiquity should estimate the amount of space that should be left for / or whether a separate partition should be made at all. Then...

-The first time an user installs Ubuntu, he is given the option to set a separate /home. This option is selected by default, with a size for each partition based on a sane guess:

http://img155.imageshack.us/img155/7958/firstinstallaro2.png

-Of course, he can just choose not to set a separate /home. This option will be selected by default if the results of the system test suggest that's the best thing to do.

http://img186.imageshack.us/img186/6498/firstinstallbfs6.png

-Manual install is also possible. Selecting it greys out everything related to separate /home, since it's implied that the user doesn't want to be guided.

http://img177.imageshack.us/img177/7976/firstinstallcvc2.png

-If the user set a separate /home, the next time he installs Ubuntu a new option appears and is selected by default, prompting to use the existing /home partition. All other options are still available, though.

http://img155.imageshack.us/img155/9034/secondinstalliq1.png

[....]

Create a place where we linux users can shout out to the world "WE WANT TO PLAY ON UBUNTU LINUX"

Let us all sign and show the big companies that we exist and we are a potential marked that grows more and more every day!

Instead of asking the ubuntu team to make games work with ubuntu linux (which is quite stupid, hows that suposed to work and how much waste of development would that be?) make the gaming companies see that we exist and want support! Dont let all our requests get scattered and ignored over the www, make them all meet and bundle at one official ubuntu page.

There are games done for windows, mac, wii, ps3, psp and all other plataforms, why not ubuntu linux?

If the gaming world is able to visit a page and see "xxxxxxx ubuntu linux users want to play games!" why shouldnt they start doing it? It should be much easyer to develop games for a WORKING OS then for windows, with all its bugs and problems. That would mean less work for them, less support needed and more fun for us!

Please ubuntu team! Help us shout out!

How many mouse clicks you need to find out the CPU type of your laptop?
In XP it is 3
In Ubuntu it is 4

You will find a rational and coherent discussion on this topic at http://digg.com/microsoft/Why_I_Still_Use_Windows_Despite_the_Peer_Pressure?t=1 6396532#c16404645

I request Ubuntu team to prioritize the USER desired features and Profile the number of mouse clicks required for those features.

I am sure I am not the first one to make such a suggestion, but still I'll make my point here: the _only_ thing that Ubuntu lacks in comparison to Mandriva and OpenSuSE is a graphical system center like YaST or DrakConf. So, Ubuntu really needs something to manage hardware, network and other issues without involving the command line. This center (like drakconf) should also be able to install necessary packages, when user asks for, say, webcam support or things like that. Please consider this thing as soon as possible, since it's one of the biggest Ubuntu's flaws.

I've been a linux user since RH 5.0. In that prehistoric time, screens were most of the time 800x600 (14" and 15"). Now, I don't have precise numbers and stats but I do believe the vast majority of us own at least a 19" on desktop and most laptop owners run a 1280x800 rez.
I read billions times despite the grow of desktop managers, linux users should learn and don't forget CLI, linux commands since we are lucky enough to have a wonderful and powerful tool. The problem is when I switch on tty1 on every distro (ok I haven't tested every distro out there but I still believe I'm right) I have an ugly thing making my monitor look like shit: 640x480 (80x25). It's not only ugly as hell it's not appealing at all for new users and certainly not encourage to use CLI.
So I believe it might be possible to detect somehow (sorry I'm not an X developer) the monitor settings or resolution picked for the desktop during install and adapt something approaching (I didn't say the same, I said 'approaching') to have something which dosn't make look my monitor like an ugly thing coming from 30 years ago.

And yes I know it's possible to change that in menu.lst but what about beginners, new users or just people who want something nice and usable out of the box without doing everything by hand afterward and sometimes not succeeding?

Give us automatically a nice resolution for tty to use fully our 2008 (!) monitors and encourage people to use linux commands on consoles and not only on terminals.

Thank you for reading.

Brainstorm Bucks
================

The current voting system suffers from the following problems.

1. Unlimited number of votes. Voting yes to all the ideas is the same as not voting at all. Voting yes to most ideas is almost as ineffective as voting yes to all of them.

2. One vote per idea. Without the ability to vote more than once per idea, a vote can't express the value of the idea.

I propose a currency system. Everyone starts with 50 Brainstorm Bucks. You can spend them all on 1 idea or spread them over many. You can freely move your Bucks between ideas.

The limited currency will force people to prioritize their ideas and at the same time let people express the importance of each idea.

Hide vote count while idea is still new to eliminate the vote count influence.

Apple has introduced MobileMe,
-labeled as "Exchange for the rest of us".
-unclear as of yet whether they will release an API. They may do so to extend adaption by Windows users.

However,
-as it seems that it will be a big player,
-to provide IPhone sync with Linux,
-to avoid the same problems faced with lack of Exchange support for a long time

MobileMe support should be implemented either through Conduit or some other solution.

Lack of standardised linux packaging is letting down Developers, end users, network administrators, everyone, because there's no standard means of dealing with it, and developers don't want to waste their time to release 5 different packages to accommodate every package type.

The time has come to organise a world wide conference, to design a new standard that is common between distro's. I suggest using the following process to take action:

During Transition period:
- Current Packaging formats are still used
- Standard database for tracking installed packages (makes it possible for other package management systems to interact with the database of the current packaging system)
- means of tracking currently installed files from each package (so that uninstallation can be done through other programs).

By the End of Transition:
Have a packaging format (or a few standard ones) which are used amongst all distro's, that covers all the features required of the various packaging systems. At this point of time, RPM, DEB, ebuild's, etc should be depreciated!


This is really the last major problem that faces linux. With packaging standardised, developers can finally easily release their products for any distro, and a significant amount of time is saved by every distro "reinventing the wheel", and actually getting down to business.

The i386 ISOs do not have anything to do with Intel or the 386 cpus specifically just as the amd64 are not specific to AMD processors. Additionally these descriptors are practically meaningless to non-techies.

What matters for the decision between the ISOs is the difference between 32bit and 64bit, not intel and amd, so why not just call them that way, and thus take away the burden of having to be a cpu technician to understand them?

EDIT: As a compromise, to avoid confusion of the x86/IA ISOs with those for other architectures, descriptors like PC32bit and PC64bit might also be acceptable as that is likely the maximum a home user knows about his or her computer.

I know this idea is bizarre but stick with me. A program witha gui written in maybe tkinter that can build very complex cli commands. A scripting language is first chosen like bash or korn. Then a command list is generated and the user can construct advanced commands, piping their results into one another to generate whatever result is desired. A built in debugger could be developed to speed up development. This would be a great tool for newbs, great educational tool for learning how to script.

I have another idea on brainstorm where someone responded and I would like to ask that person for additional feedback. I think brainstorm should have somewhere on a user's profile where I could send a message to them. When someone is critical of an idea they aren't likely to ever return to see how the idea has progressed.

Please leave a comment if you vote this down. I would like to know what people this is wrong with this idea.

There are thousands of commands without a GUI and not everyone is comfortable with the command line. It would be a huge undertaking to create a GUI for all the command line utilities unless you had a way to automate it.

My idea is to create a program that reads man pages and automatically create Glade XML GUI files. Essentially converting one mark-up, nroff to another, XML. The translation would be intelligent and it could generate a HIGified GUI, input validation and provide a file selection dialog when the command accepts a file name as a parameter for example.

Along with the translator, there would be a helper program that would display the GUI and run the commands with the appropriate switches and provide user feedback on the execution of the command.

Man pages could be updated to provide hints to the translator and we could have a GUI translation specification that man page authors would follow for compliant man pages.

Not only would we be able to have a GUI that works exactly as advertised in the man page but it would automatically be documented.

Creating and Formatting Man Pages
http://www.fnal.gov/docs/products/ups/ReferenceManual/html/manpages.html

The Glade project
http://glade.gnome.org/

(H)uman (I)nterface (G)uidelines
http://library.gnome.org/devel/hig-book/stable/

in xorg.conf in the section
Section "InputDevice"
Identifier "Touchpad"
EndSection

if we add this line at the end of section
Section "InputDevice"
Identifier "Touchpad"
...
Option "SHMConfig" "on"
EndSection

and we make a file "/etc/udev/rules.d/01-touchpad.rules"
in which are wrote:
ACTION=="add", SUBSYSTEM=="input", ID_CLASS="mouse", RUN+="/usr/bin/synclient TouchpadOff=1"
ACTION=="remove", SUBSYSTEM=="input", ID_CLASS="mouse", RUN+="/usr/bin/synclient TouchpadOff=0"

and restart X, every time you connect a usb mouse it disable the touchpad and when disconnect usb mouse it enable touchpad

put it in an option in Preferences, disabled by default??

The Ubuntu startup and shutdown splashscreens with the Ubuntu logo and a progress bar underneath it really looks like something from the 90ies (Windows 95 anyone?)

I don't know if there are any hindrances to using images or animations of higher resolutions than what is currently used (for machines that supports it), but I really think Ubuntu can do better here.

This may be an unimportant improvement compared to many other ideas here, but it should be a very quick and easy fix, and it will really help when you try to convince regular folks to make the switch (a graphically stunning desktop is _so_ important when you show it to occasional users!).

I have a very unusual idea: Seperate the controls from the main content of an application. That would allow the user to shift the controls to another monitor. I often have to hold a presentation: I want the controls on my computer while the projector projects the main content to the wall. The audience dows not need to see the controls.

Make a little Check-Box in the Monitor-Manager to enable this option and specify, that each application should be written to support this feature.