Contributor derick.eisenhardt

It would be great if Ubuntu (or other distributions as well) provide an easy to use remote desktop solution. While it's quite easy to connect to a remote X-Server (x over ssh or nx), it's rather hard to get also support for sound (from the remote desktop), printing (using your local printers on the remote desktop) and other things like "tray-icons" / "notification area" (have the Notification-Icons of remote applications in the local "notification area").

Right now, the most likely successful migrations to Ubuntu happen for *nix experts and total novices who have someone install Ubuntu for them, but the most likely people to try Ubuntu are Windows power users--those who like a lot of configuration options but who are also used to using the GUI for tweaking options and not manually editing a text configuration file.

The most popular requests for editing the /boot/grub/menu.lst file involve changing the default boot option (to Windows instead of Ubuntu) and changing the timeout before a boot option is selected. The other less popular ones would be just bonus features, I guess.

Expansion of Add Remove Programs - show user-provided descriptions, feature lists, screenshots, comments, ratings, integration with launchpad. Initial page should be top-rated apps, or newest apps. Managed by a webapp?

Or... leave it as it is, but create a website displaying all the great software available for ubuntu including all the features mentioned above with apturl links to install.

From #4613 (merged):

It should be indicated how much I'll have to download to install something.

We have download limits here and well... quite often I exceed them. It would be useful to indicate how big the download is going to be.

There is currently no possible open source means of properly playing more then a left and right channel by optical cable on speakers. The problem is that to send more channels, we need Dolby digital, DTS, or WMA Pro, none of which we can legally support without paying for licensing (no thanks), and using closed sourced code.

If we expect to attract gamers or anyone who has more then 2 speakers, we need to encourage the development of an open standard. In the long run, this benefits everyone (except Dolby Labs) because anyone could make compatible decoders and speakers without coughing up money for licensing. It also allows us to add support to Linux without closed source dodginess. Canonical could get the ball rolling by suggesting the idea to Ogg Vorbis, and other audio organisations who may support it.


Most decoders do support upscaling from 2.0 to 5.1, but it never sounds right for gaming. And its not true 5.1. Plus, who wants to buy a high end decoder, and not fully utilise it for gaming?

When you plug in something like a USB dongle, it should be mounted synchronously (-o sync) instead of in buffered mode.

For me, this is far preferable to the minute performance gain gotten by mounting the device in buffered mode. You're very unlikely to be using such devices for anything except data storage/retreival, in which case mounting synchronously is a major benefit. Expecting new users (or those used to Windows XP's behaviour) to know you have to right-click, 'unmount', before removing the drive is not a good idea, as well as just being unnecessary if the device is mounted synchronously. It's also irritating for people like me who know you need to do it, but either occasionally forget or don't want to bother.

Mount the thing synchronously by default. This is what Windows XP does, and it allows you to just remove the drive after any file transfers have finished. Yes, problems will occur if you remove it durung a file transfer, but that's gonna happen in ANY mount mode. It's far more obvious to the user that this will cause a problem than removal of the drive before unmounting. Also stop displaying that 'unsafe device removal' message for devices mounted synchronously.

It would be nice to have a feature like: "delete items older than 2 weeks"

i'm reaaly need SUBJ

Ubuntu should include a way to add/remove fonts from a repository just like the add/remove program panel. When a user needs a new font, they could open up the window, and the program would download the latest repository fonts. The user could look through the fonts and choose which ones to install by simply checking a box and clicking "apply." A user could search for font keywords to find certain styles of fonts. For example, someone could search for keywords words "serif","sans serif","cursive", and the GUI would display fonts that match the search. This would be a killer tool for designers.

Furthermore, the GUI could be very helpful. When a user browses a font, a sample image of it could be automatically loaded so that the user can quickly find fonts that are appealing to them. A rating system or popularity gauge could be used so that the best fonts can be highlighted.

A tool like this would create inscentive for font designers to make their designs free/public. Credits could be given to designers in the font descriptions.

Ubuntu should develop it's own open source game development libraries. This could push for a unified way to develop games on Linux platforms. Fast ANSI C libs that don't act as an engine, but rather as a easy way to gain access to the systems hardware from languages like python,c++,java... Why should Windows attract all the game developers. If a standardized library is developed for Linux it would certainly make anyone Question the need to use Directx.

it should be improved,i think avast should be supported
in repo,made aware to general public.Avast is most
popular av for linux,win.but Clamav is also good i dont
know which will be people's choice.clamav is good
but doesnt have best support also it doesnt have gui.
Include context menu for files/folders for virus scan.
scan with Clamav/avast wud be gr8.also clamav should be
available in repo.

I think it is important for the update notifier to be color coded to stress the importance of security updates. Red updates could signify that there is an important security updates while all other updates could stay the original orange color.

Metacity has many, many features but it does not have cool bling.
Things like preventing windows popping up when you are doing something else like putting in you password.
It does not loose windows like Compiz woobly shake of death.
It has better edge detection.
It does not blank windows like Compiz does sometimes.
It does not loose the color of an active window like Compiz.
And there should be constancy with keybinding and behavior.
It would be useful if somebody documented all the tweeks that Metacity has so it could be copied.

It seems the commercial repo falls in and out of favour - back in Dapper days it was launched with quite a bit of fanfare with quite a lot of third party/commercial apps. However these quickly became out of date with security fixes (ie Opera) and often broken by kernel updates (ie VMWare) and people resorted back to getting the packages from the third parties themselves.

From memory it took a few months for the repo to reappear in Edgy but it mainly remained empty. Lately a press release went out about offering Parallels Workstation from the repo, with the option to purchase the full product from the Canonical store. Upon trying this recently it seems this package is already broken by recent updates - I've seen a few threads on the forums about it and also a bug is still unconfirmed in Launchpad.

It seems if Ubuntu is wanting to gain more commercial clout it would be good to maintain/monitor this repo more frequently?

I want an easy GUI for adding desktop widgets like wheather, cpu, notice, pictures, flowers, decorations, animals, chat, mail-alert, diary etc. to my desktop. :D

If you install an LTS version, you may chose to stay with it until the next LTS version. Problem is that you will stay with old programs and will not have access to the newest versions of The Gimp, OpenOffice, ...

People running LTS versions should be able to go and install OpenOffice when it comes out via the Ubuntu repo's and not via getdeb.net or via the official OpenOffice site.

Example: Hardy ships with The Gimp 2.4.5 and it will stay 2.4.5, when 2.5.0 is launched and we want to have it we need to go to getdeb or the Gimp site and download it. Same when OpenOffice 3.0 comes out.

This option should be made possible via the update manager.

This function is not necessary for people running other versions because these have a short lifespan and programs will not change a lot in 6-12 months.

Normally, if I have access to some files, my home directory for example, all applications I run also get that access (unless I run them in some jail).

This does not happen when working with remote files ("Places" > "Connect to server..."). Then each application needs to explicitly be given access to each server it wants to use files on. It would be better if remote parts of the file system worked identically to local ones. Just like with sshfs.

One month is simply not long enough for the beta phase of Ubuntu's development cycle. With each new release Ubuntu seems to be accumulating more and more bugs. Even worse are the tons of regressions slipping through until after the final release, sometimes not getting fixed at all. More focus needs to be put on testing and stability (including universe/multiverse too).

Most users are terrified at the idea of running an alpha, so almost no one but developers and hardcore/bleeding-edge types are doing any real testing 5 out of the 6 months that each release is developed. I'd love to see 3 months of development (alpha) followed by 3 months of testing (beta+rc), but that's probably not realistic. So I propose 1 month of beta, followed by 1 month of release candidates; thus giving us 2 months of testing rather than one. The devs and other folks in charge can make the real decision of exactly how long, but 1 month (3 weeks of beta + 1 week rc) just isn't cutting it guys.

Why not to support LTS desktop releases of Ubuntu for a longer time. For example for 5 years that means as long as server editions are supported currently. This should not be a big problem, as most of updates are identical for both versions. Alternatively it can be full support for the first three years, and only crucial security updates for the last two.