Contributor Apiman
Redesign Nautilus: Remove Browser Buttons
Written by pubsbin the 31 May 09 at 03:04.
Related project: Nautilus .
New
A File Manager isn't used, most of the time, as a browser.
I do know about network transparency which allows you to manage files in a remote computer. But the most of the users (as i understand) only use it with their personal and local files.
It is really and waste of space and user attention to have those buttons that are most of the time disabled or have no real effect.
Solution #2:
Combine stop and reload button into one button
Written by
lavinog the 31 May 09 at 16:12.
Since stop is only enabled during a refresh/reload operation, why would you need the reload button during a refresh.
Since stop is only enabled during a refresh/reload operation, why would you need the reload button during a refresh.
Solution #3:
Create a "Custumize toolbar" just like firefox
On firefox, there is a custumize toolbar( right click on the navigation toolbar + custumize)
A new window is created where the user can drag and drop buttons.
My idea is create the same for Nautilus
On firefox, there is a custumize toolbar( right click on the navigation toolbar + custumize)
A new window is created where the user can drag and drop buttons.
My idea is create the same for Nautilus
Solution #4:
Make browser buttons smaller
You can already remove the browser buttons by hiding the main toolbar. Since the forward, back, and reload buttons are actually used when file browsing, it would be better to reduce the space that these buttons take up.
You can already remove the browser buttons by hiding the main toolbar. Since the forward, back, and reload buttons are actually used when file browsing, it would be better to reduce the space that these buttons take up.
Solution #5:
Put controls in title bar
Written by
ctbeiser the 24 Jun 09 at 05:29.
Put the controls in the title bar for the Nautilus window.
Put the controls in the title bar for the Nautilus window.
Drag & drop with Atl + Tab
Ubuntu
In :
Priority : Undefined
Definition : New (Needs guidance)
Implementation : Unknown
Assignee :
Written by netiauga the 28 Feb 08 at 19:05.
Global category: System.
New
I'm so frustrated. I can't use Drag & drop with Atl + Tab to move files.
Keep current GNOME interface, instead of using GNOME Shell
Written by Linux-user the 7 Jun 09 at 16:39.
Related project: Gnome .
New
The developers of GNOME are thinking about changing their interface. They want to replace the current interface (top panel and bottom panel) with something they call GNOME Shell. This new interface will have a bar on the top called "Activities". The old menu called "Applications" will be gone and you'll have to type the name of the application to start this application.
Screenshots:
http://live.gnome.org/GnomeShell/Screenshots
I really don't like this new interface and I've seen many other people complaining about this new interface.
Solution #1:
Keep the current panels
Why does GNOME has to start developing a completely new interface? Let them first finish the current one. Let them first solve those thousands of bugs which are in GNOME for more than several years (to give some examples: icons on the desktop are still overlapping each other, in Nautilus it's still impossible to lasso files in List View, in Nautilus it's still impossible to create a new directory from the right mouse button in List View if there are more items in the directory than fit on the screen).
Those guys keep on adding new features and now they want to introduce a completely new interface. Finish the something before starting something new. Fix bugs before adding new features.
Why does GNOME has to start developing a completely new interface? Let them first finish the current one. Let them first solve those thousands of bugs which are in GNOME for more than several years (to give some examples: icons on the desktop are still overlapping each other, in Nautilus it's still impossible to lasso files in List View, in Nautilus it's still impossible to create a new directory from the right mouse button in List View if there are more items in the directory than fit on the screen).
Those guys keep on adding new features and now they want to introduce a completely new interface. Finish the something before starting something new. Fix bugs before adding new features.
Solution #2:
Allow the user to decide - add as menu/appearance option
Written by
tuxxy the 7 Jun 09 at 22:10.
In future GNOME releases users should be able to choose either the GNOME shell design or be able to revert back to the standard panel GNOME layout. This new design feature could be added as a menu or appearance option to accommodate the users who prefer the old standard GNOME layout.
Not providing this option could alienate some users and force them to adopt a new desktop environment.
In future GNOME releases users should be able to choose either the GNOME shell design or be able to revert back to the standard panel GNOME layout. This new design feature could be added as a menu or appearance option to accommodate the users who prefer the old standard GNOME layout.
Not providing this option could alienate some users and force them to adopt a new desktop environment.
Solution #3:
gnome shell should take profit from wide-screen displays
Written by
yzarc the 8 Jun 09 at 00:12.
the screens is getting wider and wider but gnome seems to don't care about it and gnome shell looks like is in the same way. two horizontal bars also in the gnome shell and even harder to customize.
gnome should profit the opportunity of a brand new interface concept to improve the use of wide-screen. Let the top and button area free and use the side parts (optionally), it is impossible with the current gnome interface, nothing work properly.
the screens is getting wider and wider but gnome seems to don't care about it and gnome shell looks like is in the same way. two horizontal bars also in the gnome shell and even harder to customize.
gnome should profit the opportunity of a brand new interface concept to improve the use of wide-screen. Let the top and button area free and use the side parts (optionally), it is impossible with the current gnome interface, nothing work properly.
Solution #4:
Use Gnome Shell, but make things more discoverable
Written by
Endolith the 11 Jun 09 at 16:26.
Gnome Shell looks like an improvement. Searching for activities or documents is better and faster than menus if you know what you're looking for. But searching only works if you know the name of the thing you're searching for. The traditional hierarchical navigation is better suited for when you know what you want to do, but don't know what program does it.
There should still be categories, and you should be able to see them in the search results and navigate through them if you type their names. Applications should be assigned to multiple categories as appropriate, like Totem could be in both "Audio" and "Video".
Searching should work on both the application name and the program's description, as well as synonyms, so you can find Firefox by searching for "web browser", for instance.
With an empty search box, something needs to be shown to help the user get started searching for apps and realize what it's capable of.
Gnome Shell looks like an improvement. Searching for activities or documents is better and faster than menus if you know what you're looking for. But searching only works if you know the name of the thing you're searching for. The traditional hierarchical navigation is better suited for when you know <i>what</i> you want to do, but don't know what program does it.
There should still be categories, and you should be able to see them in the search results and navigate through them if you type their names. Applications should be assigned to multiple categories as appropriate, like Totem could be in both "Audio" and "Video".
Searching should work on both the application name <i>and</i> the program's description, as well as synonyms, so you can find Firefox by searching for "web browser", for instance.
With an empty search box, something needs to be shown to help the user get started searching for apps and realize what it's capable of.
Solution #5:
Make the transition smooth
Lobby the folks at Gnome to make the transition as smooth as possible.
1. Take small steps towards the new UI rather than one big leap. Every step should involve a small change.
2. The UI must be intuitive at every step.
3. Do NOT force all the users to use the new UI. Instead, make every change OPTIONAL.
Bottom line is that those who wish to stick to the classic Gnome interface should be allowed to do so until they're ready to move on.
Lobby the folks at Gnome to make the transition as smooth as possible.
1. Take small steps towards the new UI rather than one big leap. Every step should involve a small change.
2. The UI must be intuitive at every step.
3. Do NOT force all the users to use the new UI. Instead, make every change OPTIONAL.
Bottom line is that those who wish to stick to the classic Gnome interface should be allowed to do so until they're ready to move on.
Solution #6:
Take more time for the transition
Written by
xfuser4 the 2 Jul 09 at 09:22.
I don't think that its a bad idea to make a "hard" transition between Gnome 2 and Gnome 3.
But I think, that the Gnome people are hurrying too much. It would be better to take enough time to design Gnome 3.
- It would be important to use (paied?) user interface specialists to design Gnome Shell
- It would be important to make great API designs and provide great development tools for Gnome 3
I don't think that its a bad idea to make a "hard" transition between Gnome 2 and Gnome 3.
But I think, that the Gnome people are hurrying too much. It would be better to take enough time to design Gnome 3.
- It would be important to use (paied?) user interface specialists to design Gnome Shell
- It would be important to make great API designs and provide great development tools for Gnome 3
Easier to choose an application
Written by caryb the 29 Feb 08 at 04:59.
Global category: System.
New
How many times have you clicked an application to find *buntu doesn't know what to do with it? what happens next is real handy, it drops you to your home folder to pick a application! This works for us seasoned Linux folks but what about the poor newbies?
Solution #1:
Keep track of applications
Written by
caryb the 29 Feb 08 at 04:59.
A data file with possible applications for tasks that could be upgraded when needed. If the application is not installed perhaps a link to synaptic or adept etc to install and select the missing application.
A data file with possible applications for tasks that could be upgraded when needed. If the application is not installed perhaps a link to synaptic or adept etc to install and select the missing application.
Solution #2:
Choose an application using the current gnome-menu structure
Read the desktop files, sort them into categories, just like the Gnome menu does.
Read the desktop files, sort them into categories, just like the Gnome menu does.
Solution #3:
Sort programs alphabetically
Written by
Yfrwlf the 26 Jun 09 at 13:29.
One way to make choosing a program to open a file easier would be to alphabetically sort the available programs. See http://ubuntuforums.org/showthread.php?t=1197544 for screen shot.
Solution #4:
Show suggested programs first, then all in alphabetical order
Written by
jarko_ the 30 Jan 09 at 11:00.
Suggest programs first. Like in mp3 the suggested programs-group could contain Rhythmbox, totem, vlc...
Suggest programs first. Like in mp3 the suggested programs-group could contain Rhythmbox, totem, vlc...
Solution #5:
Show mime-base/magic solution first then frequently used then suggested..
I like Solution #2 but I think it's better to list first item resulting after a parsing against magic/mime-type, the resulting list will be processed so that will list as the frequently used.
After this initial statistic results there will be the Solution #2 (show suggested program first and then all alphabetical order).
The result will be that first an order will be build based on real statistical usage and then based on generic progression.
Using magic and mime we can track in some way the content of the file and based on this information we can build a list of the frequently used associated application (statistical order), then at the bottom of this list we can continue with the escluded items ordered by suggested (with some criteria) and the alphabetical (this involve understanding how to handle localizations)
I like Solution #2 but I think it's better to list first item resulting after a parsing against magic/mime-type, the resulting list will be processed so that will list as the frequently used.
After this initial statistic results there will be the Solution #2 (show suggested program first and then all alphabetical order).
The result will be that first an order will be build based on real statistical usage and then based on generic progression.
Using magic and mime we can track in some way the content of the file and based on this information we can build a list of the frequently used associated application (statistical order), then at the bottom of this list we can continue with the escluded items ordered by suggested (with some criteria) and the alphabetical (this involve understanding how to handle localizations)
Solution #6:
Group programs by type and then alphabetically
This is similar to Solution#2, but when listing programs, categorize them by type (Accessories, Education, Games, Graphics, etc), as done in the Applications menu. And within each category, sort them alphabetically.
This is similar to Solution#2, but when listing programs, categorize them by type (Accessories, Education, Games, Graphics, etc), as done in the Applications menu. And within each category, sort them alphabetically.
Solution #7:
Make it possbile to change sorting
Written by
zooounds the 25 Feb 09 at 15:31.
For example alphabetical could be default but sorting by category could be easily selected.
For example alphabetical could be default but sorting by category could be easily selected.
Solution #8:
Keep programs like Comix out of the right-click menu
Written by
WDC the 25 Feb 09 at 15:20.
When I right-click on any file, regardless of type, I have Comix and a couple other programs that are pinned by default. I don't think that programs should be able to pin themselves to the menu.
When I right-click on any file, regardless of type, I have Comix and a couple other programs that are pinned by default. I don't think that programs should be able to pin themselves to the menu.
Solution #9:
Create Default Action Types and Add Previous apps to those
Written by
doctormo the 25 Feb 09 at 18:20.
Instead of having one generic 'Open' action, we should change the magic so that the application's interaction with a file is in context. For instance opening an svg file in Everice or Firefox is just for viewing, so 'View with Filefox', 'View with Document Viewer' and then a section for editors 'Edit with Inkscape', 'Edit with Scribus'
This does mean added extra flexibility into the opening system, but it should allow the menu to be more clear and allow settings such as 'Duoble click means View, Alt+Click means Edit' and for that to make sense to nautilus.
Instead of having one generic 'Open' action, we should change the magic so that the application's interaction with a file is in context. For instance opening an svg file in Everice or Firefox is just for viewing, so 'View with Filefox', 'View with Document Viewer' and then a section for editors 'Edit with Inkscape', 'Edit with Scribus'
This does mean added extra flexibility into the opening system, but it should allow the menu to be more clear and allow settings such as 'Duoble click means View, Alt+Click means Edit' and for that to make sense to nautilus.
Solution #1:
Use the new Ubuntu Notifications
We could use the new Ubuntu notifications to do provide this alert. Just something simple like "USB Webcam detected" with an icon of a webcam.
We could use the new Ubuntu notifications to do provide this alert. Just something simple like "USB Webcam detected" with an icon of a webcam.
Solution #2:
Make them optional
Since I already have hotplug scripts that execute when a new device is detected, I don't need these notifications. In this case, detection notifications would be a nuisance.
Since I already have hotplug scripts that execute when a new device is detected, I don't need these notifications. In this case, detection notifications would be a nuisance.
Solution #3:
Use HAL notify script
Written by
DnaX the 6 Jun 09 at 00:11.
An implementation of solution #1: There is this python script that notify new devices discovered by HAL. Work fine.
https://code.launchpad.net/~dnax88/+junk/hal-notify
Solution #4:
Only notify about problematic devices
I expect when I plug in a new device it will be detected and configured and ready for my use within 10 seconds or so. A notification can be displayed if the device is NOT usable for some reason or isn't ready within the 10 seconds. (2 different notification messages).
The old equation: silence = success
I expect when I plug in a new device it will be detected and configured and ready for my use within 10 seconds or so. A notification can be displayed if the device is NOT usable for some reason or isn't ready within the 10 seconds. (2 different notification messages).
The old equation: silence = success
Solution #5:
Green popup=working hardware / Red popup=not supported, extra attention...
Written by
walterav the 7 Jun 09 at 22:37.
It might give a "false assumption" that the hardware is also supported and directly working with ubuntu.
My suggestion would be that it gives a notification that is green/if the hardware directly works, it might fade away!
Other wise make the notification "red" with a extra dialog box that say's this hardware is not supported, or needs the following procedure, or something with cancel.
This idea can be combined with solution 1 / 3
It might give a "false assumption" that the hardware is also supported and directly working with ubuntu.
My suggestion would be that it gives a notification that is green/if the hardware directly works, it might fade away!
Other wise make the notification "red" with a extra dialog box that say's this hardware is not supported, or needs the following procedure, or something with cancel.
This idea can be combined with solution 1 / 3
Solution #6:
Menu
Give a menu that gives some information such as:
*Status
*Compatibility
*Type of Device
*Programs which use the device (So give Nautilus/Dolphin for a USB Flash Drive, Network Manager for a WiFi adapter, etc.)
The menu would fade away and would not be obtrusive, but would give the user information about the device and give options on what to do.
Give a menu that gives some information such as:
*Status
*Compatibility
*Type of Device
*Programs which use the device (So give Nautilus/Dolphin for a USB Flash Drive, Network Manager for a WiFi adapter, etc.)
The menu would fade away and would not be obtrusive, but would give the user information about the device and give options on what to do.
Solution #7:
Solution 1 + icon that provides configuration
Written by
DaVince the 15 Jun 09 at 22:25.
It would probably be a good idea to have an icon pop up while a notification is shown, so that accessibility to configuration of this little tool is available. Anyone who doesn't like the notifications or wants to disable them for certain hardware will be able to do so by clicking this icon (a special configuration window will pop up).
The icon will automatically disappear shortly after the notification was shown.
It would probably be a good idea to have an icon pop up while a notification is shown, so that accessibility to configuration of this little tool is available. Anyone who doesn't like the notifications or wants to disable them for certain hardware will be able to do so by clicking this icon (a special configuration window will pop up).
The icon will automatically disappear shortly after the notification was shown.
Provide a proper, understandable descriptions of directories
Written by weblordpepe the 27 May 09 at 11:56.
Related project: Nautilus .
New
----
Ever tried explaining what all that stuff was under 'Filesystem'?
----
At the moment, the Nautilus file browser shows almost everything it sees as either a default Folder icon, unless it is a registered file, filetype or customized folder. When the user selects the Filesystem icon in Nautilus, they are presented with a lot of folder icons for everything under /
Nautilus doesn't know what /dev and /usr and /etc are, so neither does the user. The default browsing view has ZERO smarts. By no means does the file browser indicate what the majority of main, key items are.
Gigabytes of resources occupy this realm. Very useful resources for both end-users, developers, and any other person. These need to be useful to people IF THEY CHOOSE TO LOOK. This will not be solved by 'hiding' the filesystem structure. It is there, full of resources, and should be made easy to use by not just programs, but people too.
This idea is about adjusting Nautilus, not altering the files names.
Solution #1:
Differentiated icons
Use the built-in feature for 'customized folder icons' to show unique icons for /usr, /etc, /home, /var and so on. Have this done in the default config.
Everything else looks like a regular folder/item.
Perhaps allow Nautilus to understand MIME types of folders, or some other intelligent categorizing.
<br><br><br>
<table style="width:500px;"><tr><td><img width=400px src="http://silenceisdefeat.com/~weblordpepe/nautilussuggestion2.png"/></td></tr></table><br><br>
Use the built-in feature for 'customized folder icons' to show unique icons for /usr, /etc, /home, /var and so on. Have this done in the default config.
Everything else looks like a regular folder/item.
Perhaps allow Nautilus to understand MIME types of folders, or some other intelligent categorizing.
Solution #2:
Use differentiated icons AND friendly-wording as Primary description
Using both customized icons (Solution #1) and human-friendly descriptions as the main label for the item in Nautilus:
For example:
The /etc directory can appear as:
Configuration (etc)
The /var directory can appear as:
Logs (var)
The /home directory can appear as:
User profiles (home)
Here is a screenshot. Of course the theme, icons, fonts are just suggestions.
Thanks Shady3D for the picture-embedding tip
<br><br><br>
<table style="width:400px;"><tr><td><img width=400px src="http://silenceisdefeat.com/~weblordpepe/nautilusproposal.png"/></td></tr></table><br>
Using both customized icons (Solution #1) and human-friendly descriptions as the main label for the item in Nautilus:
For example:
The /etc directory can appear as:
Configuration (etc)
The /var directory can appear as:
Logs (var)
The /home directory can appear as:
User profiles (home)
Here is a screenshot. Of course the theme, icons, fonts are just suggestions.
Thanks Shady3D for the picture-embedding tip
Solution #3:
Option to show friendly-wording as "secondary" description
Written by
Ssdg the 27 May 09 at 14:41.
Using both customized icons (Solution #1) and human-friendly descriptions as the secondary label for the item in Nautilus:
The problem of #2 is that non GTK apps (Qt, Tk, console,...) won't be aware of the real names, showing this description faded at the bottom of the real name would be less confusing.
(idea by Ssdg. picture by weblordpepe)
<br><br><br>
<table style="width:400px;"><tr><td><img width=400px src="http://silenceisdefeat.com/~weblordpepe/nautilusidea3.png"/></td></tr></table><br>
Using both customized icons (Solution #1) and human-friendly descriptions as the secondary label for the item in Nautilus:
The problem of #2 is that non GTK apps (Qt, Tk, console,...) won't be aware of the real names, showing this description faded at the bottom of the real name would be less confusing.
(idea by Ssdg. picture by weblordpepe)
Solution #4:
Additional column in details view
How about an additional column of some kind, with a description. Similar to Windows Service Management console where it tells you what each service does.
How about an additional column of some kind, with a description. Similar to Windows Service Management console where it tells you what each service does.
Solution #6:
Text Help Files in each directory
Written by
RobOrr the 27 May 09 at 21:23.
With the way Ubuntu works for the basic user who just uses simple apps like a music player, internet browser and so forth, they shouldn't really have to mess with these too much, just like a basic user in XP doesn't know anything about the WINDOWS folder. I suggest a simple text file in each folder detailing what it contains and what actions the files in this folder usually do. It would take up very little hard drive space, and also be usefully informative to anyone who wanted to know a little more. the problem would be informing the user that these text files existed...
With the way Ubuntu works for the basic user who just uses simple apps like a music player, internet browser and so forth, they shouldn't really have to mess with these too much, just like a basic user in XP doesn't know anything about the WINDOWS folder. I suggest a simple text file in each folder detailing what it contains and what actions the files in this folder usually do. It would take up very little hard drive space, and also be usefully informative to anyone who wanted to know a little more. the problem would be informing the user that these text files existed...
Solution #7:
Hide files by default.
Written by
nloewen the 28 May 09 at 00:29.
By default hide system folders, but make them easily accessible through 'view -> show hidden files'. Also make it clear when they are unhidden that these files are system files and could cause undesired operation if you mess with them. I'm not talking about hiding it away forever. I'm talking about hiding it to keep things simpiler and less cluttered.
And when a user chooses/needs to go there it should be easy and simple to understand what is there why its there and how it works. Hide it from people who don't want it and make it easily available to those that do.
To all of you voting down, remember that this is what OSX does and it is considered on of the most user friendly OS's around. (with one exception, the files are a lot easier to get at in this solution). The average user doesn't care how there computer works, only that is does and if they see files sitting around that they didn't put there they get curious, and (occasionally) screw things up.
By default hide system folders, but make them easily accessible through 'view -> show hidden files'. Also make it clear when they are unhidden that these files are system files and could cause undesired operation if you mess with them. I'm not talking about hiding it away forever. I'm talking about hiding it to keep things simpiler and less cluttered.
And when a user chooses/needs to go there it should be easy and simple to understand what is there why its there and how it works. Hide it from people who don't want it and make it easily available to those that do.
To all of you voting down, remember that this is what OSX does and it is considered on of the most user friendly OS's around. (with one exception, the files are a lot easier to get at in this solution). The average user doesn't care how there computer works, only that is does and if they see files sitting around that they didn't put there they get curious, and (occasionally) screw things up.
Solution #8:
Mouse over it information
Written by
Rodrigo the 30 May 09 at 11:31.
Well I thought that the info could be provided when you move the mouse over the directory in Nautilus.
The option could be switched on or off, as a helping aid.
(I have made a mockup, sorry I am not that good with Gimp, I tried to do my best. If you want more info let me know)
From
ubuntu
Well I thought that the info could be provided when you move the mouse over the directory in Nautilus.
The option could be switched on or off, as a helping aid.
(I have made a mockup, sorry I am not that good with Gimp, I tried to do my best. If you want more info let me know)
<table style="width:auto;"><tr><td><a href="http://picasaweb.google.com/lh/photo/vqEz0sBAAo-7Vkxirizu0A?feat=embedwebsite"><img src="http://lh6.ggpht.com/_nve8T1z0v-4/SiEbPsMOxXI/AAAAAAAAAc8/l3GPwatANTM/s144/1.jpg" /></a></td></tr><tr><td style="font-family:arial,sans-serif; font-size:11px; text-align:right">From <a href="http://picasaweb.google.com/rodrigodonado/Ubuntu?feat=embedwebsite">ubuntu</a></td></tr></table>
Solution #9:
Dynamic 'drill down' or 'Search query' or 'database' portal to the files
The / filesystem categorized, and indexed. Then provide a user's portal (seperate program??) to the resources available indexed by ANY property. System-updates and 'file' utility do 90% of this job already.
EG:
* GUI to select type of resource. User clicks 'Image' resources.
* Then GUI provides options for: image size (scalable), image format (icons?), etc etc
Search results appear as files/folders which match the query. (e.g. /usr/share/pixmaps and /usr/local/shared/pixmaps)
Each time the user selects a more detailed search parameter, the results appear as folders + files. The GUI in question can even use the folders in the search-results to dynamically create more search criteria. For example:
* User searches for images
* Search results provides a few folders, some with descriptions of 32x32, 16x16, scalable, and some with application titles.
* GUI dynamically creates search-criteria called 32x32, 16x16, scalable, and the application titles.
This index can be built on the fly using 'what files does XYZ package use' utility, and the 'what type of file is this?' utility (file).
Think of Windows Vista's customizable/savable search-queries which appear as folders. But completely dynamic.
The / filesystem categorized, and indexed. Then provide a user's portal (seperate program??) to the resources available indexed by ANY property. System-updates and 'file' utility do 90% of this job already.
EG:
* GUI to select type of resource. User clicks 'Image' resources.
* Then GUI provides options for: image size (scalable), image format (icons?), etc etc
Search results appear as files/folders which match the query. (e.g. /usr/share/pixmaps and /usr/local/shared/pixmaps)
Each time the user selects a more detailed search parameter, the results appear as folders + files. The GUI in question can even use the folders in the search-results to dynamically create more search criteria. For example:
* User searches for images
* Search results provides a few folders, some with descriptions of 32x32, 16x16, scalable, and some with application titles.
* GUI dynamically creates search-criteria called 32x32, 16x16, scalable, and the application titles.
This index can be built on the fly using 'what files does XYZ package use' utility, and the 'what type of file is this?' utility (file).
Think of Windows Vista's customizable/savable search-queries which appear as folders. But completely dynamic.
Solution #10:
Go GOBO Style: Friendly Directory Names to begin with
http://www.gobolinux.org/index.php?page=at_a_glance
~] cd /
/] ls
Programs
Users
System
Files
Mount
Depot
Solution #11:
Introduce a sidebar, where additional information or previews are shown
Written by
Brunni the 10 Jun 09 at 09:41.
Use a sidebar to schow additional information to the selected item.
e.g.
- Size, Datetime, Permission
- File Preview (image/video)
...
and of course:
Is it a system-folder. What is it used for. Where can you get additional informatioen e.g. wiki etc.
Use a sidebar to schow additional information to the selected item.
e.g.
- Size, Datetime, Permission
- File Preview (image/video)
...
and of course:
Is it a system-folder. What is it used for. Where can you get additional informatioen e.g. wiki etc.
Solution #12:
Use simple status bar/file properties menu descriptions
Written by
DaVince the 15 Jun 09 at 22:04.
People seem to be overthinking this solution; the simplest to me seems to just have a short description in the status bar when selecting the folder and in the file properties window.
People seem to be overthinking this solution; the simplest to me seems to just have a short description in the status bar when selecting the folder and in the file properties window.
Solution #13:
Show the description under the folder name & friendly icons
Written by
dysmann the 17 Jun 09 at 12:43.
Just show the name as usually (home), and show the description (Users Documents) under the name, in italic, gray
Just show the name as usually (home), and show the description (Users Documents) under the name, in italic, gray
Integrate Google Wave into Ubuntu
Written by suneej the 30 May 09 at 15:31.
Related project: XChat-GNOME IRC Chat .
New
Google Wave is open-source and will be released later this year. If it could be integrated into Ubuntu then it could greatly improve the productivity and communication of networked Ubuntu users. For example more than one person can work on a GIMP or OpenOffice document simultaneously in real time etc...
Solution #1:
Integrate Google Wave into Ubuntu
Written by
suneej the 30 May 09 at 15:31.
Google Wave is open-source and will be released later this year. If it could be integrated into Ubuntu then it could greatly improve the productivity and communication of networked Ubuntu users. For example more than one person can work on a GIMP or OpenOffice document simultaneously in real time etc...
It can be integrated into the Ubuntu desktop....because its open-source it can easily be modified
Google Wave is open-source and will be released later this year. If it could be integrated into Ubuntu then it could greatly improve the productivity and communication of networked Ubuntu users. For example more than one person can work on a GIMP or OpenOffice document simultaneously in real time etc...
It can be integrated into the Ubuntu desktop....because its open-source it can easily be modified
Solution #2:
Application Plugins to make use of Google Wave features
The best way to accomplish this task is to write plug-ins for the existing applications, to integrate new features based on Google Wave into the existing applications like GIMP or OpenOffice.
Plug-ins are OPTIONAL for user and are easy to develop, since they can be written without modifying the actual application code.
The best way to accomplish this task is to write plug-ins for the existing applications, to integrate new features based on Google Wave into the existing applications like GIMP or OpenOffice.
Plug-ins are OPTIONAL for user and are easy to develop, since they can be written without modifying the actual application code.
Solution #3:
Use wave for the ubuntu community
Written by
arand the 1 Jun 09 at 10:17.
Use wave to bring together all scattered pieces of the ubuntu community: Forums, mailing lists, launchpad, wiki, brainstorm, irc, etc., and hence create a common, transparent interface applicable to all things ubuntu. Old services will remain, extensions will congregate content into the wave interface.
Use wave to bring together all scattered pieces of the ubuntu community: Forums, mailing lists, launchpad, wiki, brainstorm, irc, etc., and hence create a common, transparent interface applicable to all things ubuntu. Old services will remain, extensions will congregate content into the wave interface.
Solution #4:
Integrate programs one-by-one into Google Wave via plugins
Written by
kareeser the 3 Jun 09 at 01:56.
Quick example: Evolution.
Both are viable programs on their own, Evolution supports POP and IMAP, and is a general PIM which would benefit greatly from Google Wave as a sync backend. Evolution could even keep their layout, but interface with Google Wave like their own API does.
Quick example: Evolution.
Both are viable programs on their own, Evolution supports POP and IMAP, and is a general PIM which would benefit greatly from Google Wave as a sync backend. Evolution could even keep their layout, but interface with Google Wave like their own API does.
Return Update Notifier for Jaunty
Written by Baggers the 18 Apr 09 at 10:59.
Related project: Update manager .
New
While I have read some mumbles about why this isn't in the notification bar anymore, I can't agree that this was the best course to take.
My three main reasons for this are:
a) Without this notifier many users (especially beginners) will forget to update, I've been using Ubuntu for a couple of years and I find I am updating far less often than I should. Updates are our main security feature, retarding a persons ability to discover new updates is a risk to that security.
b) Getting to updates has moved from a one click process to a 2 click one. While this seems like nothing, any change that makes a useful function more difficult is a regression in my eyes.
c) 3 months down the line more bugs will have been fixed. If a new user comes to Ubuntu they are not going to instinctively know to update immediately after install. This means they may run into bugs fixed month before which will leave a bad impression on the user.
Solution #1:
Restore the Update Notifier
Written by
Baggers the 18 Apr 09 at 10:59.
Simply give back this feature and if need be allow the user to disable it through a GUI interface. We can easily use the new notification system to display the message while keeping the update notifier icon in the panel.
Simply give back this feature and if need be allow the user to disable it through a GUI interface. We can easily use the new notification system to display the message while keeping the update notifier icon in the panel.
Solution #3:
update-notifier as notify-osd bubble with action button&repeat reminder at x min
Written by
macvr the 20 Apr 09 at 16:45.
***as the devs are making a conscious effort to not have the notifications icon***
rather than a pop-under window,display the updates as larger notification bubbles[notify-osd bubbles]
and as devs already have the idea of action buttons in the fallback alerts[
https://wiki.ubuntu.com/NotifyOSD#Fallback%20alert%20boxes ], it would be nice if there was only just have a single button for the user to choose "update now" from the fallback alerts which look like the notify-osd alerts,
and making the notify-osd stay longer for these update/restart notifications...
if the user doesnt wasnt to update the bubble disappears or could hover over the bubble to make it invisible as the notify-osd works now
and have an option for the repeat reminder ,at x mins , which users can choose from in the updates notification settings tab...
so they end up being less intrusive (as the devs hope to achieve) than the pop-under windows...
***as the devs are making a conscious effort to not have the notifications icon***
rather than a pop-under window,display the updates as larger notification bubbles[notify-osd bubbles]
and as devs already have the idea of action buttons in the fallback alerts[ https://wiki.ubuntu.com/NotifyOSD#Fallback%20alert%20boxes ], it would be nice if there was only just have a single button for the user to choose "update now" from the fallback alerts which look like the notify-osd alerts,
and making the notify-osd stay longer for these update/restart notifications...
if the user doesnt wasnt to update the bubble disappears or could hover over the bubble to make it invisible as the notify-osd works now
and have an option for the repeat reminder ,at x mins , which users can choose from in the updates notification settings tab...
so they end up being less intrusive (as the devs hope to achieve) than the pop-under windows...
Solution #4:
Update by default, with easy opt-out
Whenever an application wants to inform the user of something, or get feedback on something, it should use bubbles... that much makes sense. But when it NEEDS something, when something is so urgent that you would go through all this trouble to make absolutely sure a user acts on its notification, then why is it not done automatically?
Of course, I know there is a major problem with this; specifically a large chunk of the user community (myself included!) don't want our system updating automatically, and hate things that change under our feet with a burning passion. But we are all "advanced users". The entire update process (waiting for a time when computer activity is low, clicking a button, watching it contact repositories, reviewing selected updates, watching the downloads and installations) are useless to those people who don't know what any of these things mean. They just want to do work, and have a secure system.
So how about this: under installation settings, there is an advanced button. Under there, add a new checkbox for "Automatically update this computer" that allows power users to easily control and monitor their setup, but keep it checked by default, so that normal users never have to worry about this.
Whenever an application wants to inform the user of something, or get feedback on something, it should use bubbles... that much makes sense. But when it NEEDS something, when something is so urgent that you would go through all this trouble to make absolutely sure a user acts on its notification, then why is it not done automatically?
Of course, I know there is a major problem with this; specifically a large chunk of the user community (myself included!) don't want our system updating automatically, and hate things that change under our feet with a burning passion. But we are all "advanced users". The entire update process (waiting for a time when computer activity is low, clicking a button, watching it contact repositories, reviewing selected updates, watching the downloads and installations) are useless to those people who don't know what any of these things mean. They just want to do work, and have a secure system.
So how about this: under installation settings, there is an advanced button. Under there, add a new checkbox for "Automatically update this computer" that allows power users to easily control and monitor their setup, but keep it checked by default, so that normal users never have to worry about this.
Solution #5:
If the popunder is ignored, show a notification icon
Whether or not a notification icon should be displayed the moment updates are available is still under debate.
I think though that after the popunder window has been shown, a persistent notification icon should definitely be added if the window was ignored or closed after X amount of time.
Whether or not a notification icon should be displayed the moment updates are available is still under debate.
I think though that after the popunder window has been shown, a persistent notification icon should definitely be added if the window was ignored or closed after X amount of time.
Solution #6:
Notify using the new system: bubble + envelope
When there are updates, notify the user with the new-style bubble ("You have N updates available, ..."). This should happen as often as set in Software Sources (every 1/2/7/14 days).
To allow the user to install updates, and to let the user be aware of updates in cases where they missed the bubble, use the envelope icon. It will show green light for some time, and clicking it will reveal a menu containing "Update manager" item. (Just like Pidgin's "user is online" notifications work, but timeout can be set somewhat longer.)
When there are updates, notify the user with the new-style bubble ("You have N updates available, ..."). This should happen as often as set in Software Sources (every 1/2/7/14 days).
To allow the user to install updates, and to let the user be aware of updates in cases where they missed the bubble, use the envelope icon. It will show green light for some time, and clicking it will reveal a menu containing "Update manager" item. (Just like Pidgin's "user is online" notifications work, but timeout can be set somewhat longer.)
Solution #7:
new notification server should support interactions
so we can click on it when we have new updates
so we can click on it when we have new updates
Solution #8:
Provide an update notifier as a panel object
Those who want to be notified about updates straight away could activate the applet through "Add to Panel", those who don't want it could remove it easily and as it wouldn't sit in the notification area, it wouldn't clash with the new notification system.
It could use the same icons as the old notifier and when there are no updates, the applet could either appear blank (a bit like the Disk Mounter Applet) or display an icon which indicates that the system is up to date.
Those who want to be notified about updates straight away could activate the applet through "Add to Panel", those who don't want it could remove it easily and as it wouldn't sit in the notification area, it wouldn't clash with the new notification system.
It could use the same icons as the old notifier and when there are no updates, the applet could either appear blank (a bit like the Disk Mounter Applet) or display an icon which indicates that the system is up to date.
Solution #9:
Add an option in "Software Sources"->"Updates" Tab
in the "Updates" tab of "Software Sources" there should be an option that says: "Show notification icon if updates available" (opt-in) and another option that says: "Automatically open update-manager if updates available" (opt-out).
Optional:
The update-manager it self should also contain a checkbox that reads: "Automatically open this window if updates are available." And an "Advanced"-Button that opens the "Updates" tab in "Software Sources" to configure behaviour.
in the "Updates" tab of "Software Sources" there should be an option that says: "Show notification icon if updates available" (opt-in) and another option that says: "Automatically open update-manager if updates available" (opt-out).
Optional:
The update-manager it self should also contain a checkbox that reads: "Automatically open this window if updates are available." And an "Advanced"-Button that opens the "Updates" tab in "Software Sources" to configure behaviour.
Solution #10:
Display a notify-osd bubble every hour
Written by
jonasfa the 9 May 09 at 03:56.
Display a notify-osd bubble every 1 hour to remind the user about the updates
Display a notify-osd bubble every 1 hour to remind the user about the updates
Solution #12:
Keep current auto launch behavior, show icon if security updates are pending
Written by
vexorian the 12 May 09 at 15:52.
The current behaviour is better at forcing users to update (launching minimized update manager) however, if SECURITY updates are skipped, the old icon should go to the tray notification are as it did before.
The current behaviour is better at forcing users to update (launching minimized update manager) however, if SECURITY updates are skipped, the old icon should go to the tray notification are as it did before.
Solution #13:
Propose Update on logout
Written by
derLars the 16 May 09 at 18:39.
PROPOSAL:
Logging out of the system (which also includes shutdown and reboots) should propose to update the system, if the user is entitled to do so. This could be easily implemented in the current "you'll be logged off in NN seconds"-message by adding a warning icon, the message
"There are pending updates - should they be installed before [logging off|shutting down|rebooting]? The computer will be [shut down|rebooted] afterwards."
and adding a third button "update before [logout|shutdown|reboot]".
PRO:
Although I highly dislike too nervy messages, the casual user does not care much about updates and risks having his computer hacked. Through this solution, the user impact would be minimal and needed reboots of the system wouldn't stress the user, since he did not want use the system anymore for now anyway.
PROBLEMS:
Of course, this proposal doesn't work for users that never log off, so another solution would have to be found for that case.
PROPOSAL:
Logging out of the system (which also includes shutdown and reboots) should propose to update the system, if the user is entitled to do so. This could be easily implemented in the current "you'll be logged off in NN seconds"-message by adding a warning icon, the message
"There are pending updates - should they be installed before [logging off|shutting down|rebooting]? The computer will be [shut down|rebooted] afterwards."
and adding a third button "update before [logout|shutdown|reboot]".
PRO:
Although I highly dislike too nervy messages, the casual user does not care much about updates and risks having his computer hacked. Through this solution, the user impact would be minimal and needed reboots of the system wouldn't stress the user, since he did not want use the system anymore for now anyway.
PROBLEMS:
Of course, this proposal doesn't work for users that never log off, so another solution would have to be found for that case.
Solution #14:
Highlight the "System" menu, with ocasional blinking.
The idea is that for important and persistent notifications, the "System" menu should be highlighted, and show some ocasional blinking. The user would notice that and click on it. A new menu item would be present there, in which a more clear notification is provided. For other interfaces, the "root" menu could be highlighted and blink. For example, the K of the main KDE menu could blink and change color if updates (or other important notifications) are available. When the user clicks on it, a new menu item would be present, in first plane and highlighted, providing the important user information and waiting for feedback. This behaviour seems to be more informative than the notification icon and less intrusive than the popup.
This is mockup for the default gnome interface:
The idea is that for important and persistent notifications, the "System" menu should be highlighted, and show some ocasional blinking. The user would notice that and click on it. A new menu item would be present there, in which a more clear notification is provided. For other interfaces, the "root" menu could be highlighted and blink. For example, the K of the main KDE menu could blink and change color if updates (or other important notifications) are available. When the user clicks on it, a new menu item would be present, in first plane and highlighted, providing the important user information and waiting for feedback. This behaviour seems to be more informative than the notification icon and less intrusive than the popup.
This is mockup for the default gnome interface:
<img src="http://launchpadlibrarian.net/27499653/notification_idea.png" width=50%>
Every program in Ubuntu is using different way of handling the same problem
Written by grofaty the 30 Jan 09 at 19:12.
Global category: Usability.
New
I installed Ubuntu on my friend's computer. She liked it, but asked me one very simple question: Why in Ubuntu every single program is using different way of handling the same simple problem?
For example:
1. Opening new tab:
- Gedit: File | New - Ctrl+N (This is standard way of opening new window not new tab)
- Firefox: File | New tab - Ctrl+T (I like this the most)
- Terminal: File | Open Tab - Shift+Ctrl+T (Open tab instead of New Tab and who can remember this combination of keys)
- Nautilus: File | New Tab and File | New Window - there is standard way to have new window at first position not new tab
2. Closing windows
- OpenOffice 3: File | Exit (What happend with Quit?)
- Terminal: File | Close window (I don't expect that Terminal will exit when I select this option)
- Nautilus: File | Close (I don't expect that program will exit - I just thought "file" will be closed)
3. Order of options in File menu (I am always searching for Close option):
- every program you look into it has a Close button in different position
4. Quit without saving (buttons and order from left to right):
- Firefox: Quit, Cancel, Save & Quit
- OpenOffice Writer: Save, Discard, Cancel
- Gedit: Close without Saving, Cancel, Save
5. Size of icons:
If you like to change the size of windows you got to be genius, because every program has its own way of settings smaller icons. You have to look in depth if you would like to turn this off.
- Firefox: View | Toolbar | Customize,
- OpenOffice: Tools | Options | OpenOffice.org | View
[....]
Solution #1:
One solution per problem in all programs
Written by
grofaty the 30 Jan 09 at 19:12.
Make: One simple problem, one simple solution. Please don't reinvent the wheel and mess it up to the point it is very hard to use.
Make: One simple problem, one simple solution. Please don't reinvent the wheel and mess it up to the point it is very hard to use.
Solution #2:
Rely on Gnome integration
Written by
Arnaudus the 31 Jan 09 at 14:30.
There are solid reasons why Solution #1 is not very realistic. Our aim is is not only to propose ideas, but ideas that are reasonable!
The only way to get a homogeneous interface is to use a set of software developed by a single community. Ubuntu relies on Gnome: let's use as many Gnome programs by default as possible. Development effort upstream already exists (one of the aims of Gnome is to provide a full set of programs, and another one if to make sure that the interface is homogeneous), so instead of "fighting" upstream teams and forcing them into an "interface standard" they don't respect, just make sure that Ubuntu supports Gnome and proposes by default more Gnome programs, including epiphany (the "firefox" equivalent) and the Gnome office suite (abiword/gnumeric...). Of course, Firefox or Openoffice should still be maintained, but the user should have the choice between fully integrated softs (Gnome) or external independent applications, perhaps more powerful, but not as integrated as the native Gnome tools.
Obviously, the problem is exactly the same for Kubuntu and KDE.
PS:
Reasons why solution #1 is very difficult (if not impossible) to implement in practice: there are only a very few ways to get it work:
1) Ask upstream teams to implement the menus and the shortcuts "The Ubuntu way"
2) Patch upstream programs to follow Ubuntu standards, and distribute "firefox-ubuntu", "oowriter-ubuntu" whith a reorganized interface.
I imagine why it is clear for everyone why (1) is dictatorial and will never be accepted upstream, and that (2) would require an enormous investment of Ubuntu devs and a lot of maintenance (fixing the patches for each new upstream version...). And what about Ubuntu/Kubuntu? Should Kubuntu develop a kde-specific version of Firefox? The amount of work seems unreasonable.
Eventually, I just want to raise an important point: I think the ergonomy of the interface is the responsability of upstream. One likes (or dislikes) The Gimp because of its very specific ergonomy for instance. If Ubuntu decides to change the menus and how everything is organized, it breaks the original idea, the specificty of the software. I find a bit dangerous for innovation to restrict the decision power of the upstream devs about the interface of the soft they develop.
There are solid reasons why Solution #1 is not very realistic. Our aim is is not only to propose ideas, but ideas that are reasonable!
The only way to get a homogeneous interface is to use a set of software developed by a single community. Ubuntu relies on Gnome: let's use as many Gnome programs by default as possible. Development effort upstream already exists (one of the aims of Gnome is to provide a full set of programs, and another one if to make sure that the interface is homogeneous), so instead of "fighting" upstream teams and forcing them into an "interface standard" they don't respect, just make sure that Ubuntu supports Gnome and proposes by default more Gnome programs, including epiphany (the "firefox" equivalent) and the Gnome office suite (abiword/gnumeric...). Of course, Firefox or Openoffice should still be maintained, but the user should have the choice between fully integrated softs (Gnome) or external independent applications, perhaps more powerful, but not as integrated as the native Gnome tools.
Obviously, the problem is exactly the same for Kubuntu and KDE.
PS:
Reasons why solution #1 is very difficult (if not impossible) to implement in practice: there are only a very few ways to get it work:
1) Ask upstream teams to implement the menus and the shortcuts "The Ubuntu way"
2) Patch upstream programs to follow Ubuntu standards, and distribute "firefox-ubuntu", "oowriter-ubuntu" whith a reorganized interface.
I imagine why it is clear for everyone why (1) is dictatorial and will never be accepted upstream, and that (2) would require an enormous investment of Ubuntu devs and a lot of maintenance (fixing the patches for each new upstream version...). And what about Ubuntu/Kubuntu? Should Kubuntu develop a kde-specific version of Firefox? The amount of work seems unreasonable.
Eventually, I just want to raise an important point: I think the ergonomy of the interface is the responsability of upstream. One likes (or dislikes) The Gimp because of its very specific ergonomy for instance. If Ubuntu decides to change the menus and how everything is organized, it breaks the original idea, the specificty of the software. I find a bit dangerous for innovation to restrict the decision power of the upstream devs about the interface of the soft they develop.
Solution #3:
Advocate for an industrial UI standard
Written by
yookoala the 2 Feb 09 at 18:52.
Advocate for a industrial guideline (including menu item allocation, hot-key allocation) as part of Linux Standard Base (LSB) like standard.
Communicate with the biggest open source development organization (i.e. Mozilla, OpenOffice.org, Eclipse, Gnome, KDE) and try to build an industrial agreement. Make a standard that everyone would try to follow.
A standard may unify the way people could do things on Linux. This will shorten the learning cycle of Linux beginner. This will also reduce headache of many Linux users and learners.
Advocate for a industrial guideline (including menu item allocation, hot-key allocation) as part of Linux Standard Base (LSB) like standard.
Communicate with the biggest open source development organization (i.e. Mozilla, OpenOffice.org, Eclipse, Gnome, KDE) and try to build an industrial agreement. Make a standard that everyone would try to follow.
A standard may unify the way people could do things on Linux. This will shorten the learning cycle of Linux beginner. This will also reduce headache of many Linux users and learners.
Solution #5:
Implement an Ubuntu HIG for pre-installed apps
Written by
Temposs the 11 Feb 09 at 06:53.
The problem with following GNOME Interface Guidelines is that Ubuntu, while it does use GNOME as an upstream provider, will not always want to follow the same path. Being tied down to GNOME means Ubuntu devs will desire to apply pressure on GNOME devs to go in certain directions. Other distros' devs doing the same thing will then have potential conflict with Ubuntu, and GNOME will be caught in the middle.
Ubuntu should have its own Interface Guidelines, perhaps inspired by/derived from GNOME's, but not identical.
For each application that comes pre-installed with Ubuntu, a UI interface package created by Ubuntu devs should be applied. Then, every pre-installed Ubuntu program will be unified the way the Ubuntu community/Canonical decides it should be. These packages can be uninstalled to get the original interface, if desired.
Once this is done, a new install of Ubuntu will appear rather unified in its interface, and at the same time uniquely Ubuntu-ish, which are important things in presenting itself to a new user.
The problem with following GNOME Interface Guidelines is that Ubuntu, while it does use GNOME as an upstream provider, will not always want to follow the same path. Being tied down to GNOME means Ubuntu devs will desire to apply pressure on GNOME devs to go in certain directions. Other distros' devs doing the same thing will then have potential conflict with Ubuntu, and GNOME will be caught in the middle.
Ubuntu should have its own Interface Guidelines, perhaps inspired by/derived from GNOME's, but not identical.
For each application that comes pre-installed with Ubuntu, a UI interface package created by Ubuntu devs should be applied. Then, every pre-installed Ubuntu program will be unified the way the Ubuntu community/Canonical decides it should be. These packages can be uninstalled to get the original interface, if desired.
Once this is done, a new install of Ubuntu will appear rather unified in its interface, and at the same time uniquely Ubuntu-ish, which are important things in presenting itself to a new user.
Solution #6:
Make exception for terminals
Written by
pabix the 11 Feb 09 at 10:37.
In terminals, Ctrl-C, Ctrl-V, Ctrl-Any_letter or other mappings have a meaning.
Try to use F1–F12 in htop with Gnome-terminal then with Konsole, you'll see that Konsole has an edge here.
In terminals, Ctrl-C, Ctrl-V, Ctrl-Any_letter or other mappings have a meaning.
Try to use F1–F12 in htop with Gnome-terminal then with Konsole, you'll see that Konsole has an edge here.
Solution #7:
Push for configurable shortcuts upstream
Written by
holizz the 12 Feb 09 at 12:56.
If the upstream applications that are included in the base install included configurable shortcuts then these actions could be standardised throughout the Ubuntu default install.
I'm not sure if this is a good solution or not, but it's a possibility.
If the upstream applications that are included in the base install included configurable shortcuts then these actions could be standardised throughout the Ubuntu default install.
I'm not sure if this is a good solution or not, but it's a possibility.
Solution #9:
Let the OS allow keyboard-shortcut remapping and saved themes.
Getting all applications to use the same shortcuts is an impossible task, even if the different development communities -could- agree on the same standards.
Allowing the user to dynamically remap application shortcuts as they wish at the operating system level is much more practical, and themes for keyboard bindings would allow distros to ship with a "firefox keyboard shortcuts" theme, or an "nautilus shortcut theme" that could change the default keyboard bindings and unify the shortcuts for the applications included in the distro by default.
This would probably require a "keybindings layer" for Ubuntu to mediate between what the user puts in, and what behavior the application follows. I imagine it could work similarly to the keybindings that you can set for compiz with the compiz-settings-manager tool (warning and giving choices to resolve potential binding overlaps).
This solution could also be pushed for with either Gnome, Ubuntu, or both.
Getting all applications to use the same shortcuts is an impossible task, even if the different development communities -could- agree on the same standards.
Allowing the user to dynamically remap application shortcuts as they wish at the operating system level is much more practical, and themes for keyboard bindings would allow distros to ship with a "firefox keyboard shortcuts" theme, or an "nautilus shortcut theme" that could change the default keyboard bindings and unify the shortcuts for the applications included in the distro by default.
This would probably require a "keybindings layer" for Ubuntu to mediate between what the user puts in, and what behavior the application follows. I imagine it could work similarly to the keybindings that you can set for compiz with the compiz-settings-manager tool (warning and giving choices to resolve potential binding overlaps).
This solution could also be pushed for with either Gnome, Ubuntu, or both.
Solution #10:
Apps should learn shortcuts from the user, as with quicksilver/gnome-do.
Essentially, instead of applications forcing us to conform to their keyboard-shortcut scheme, we should be able to train applications to follow the keyboard shortcuts that we choose as we start using an app (by text matching actions similar to Gnome-Do and Quicksilver, if people are familiar).
In essence, you would assign a key combo of your choice (Ctrl-space for example) to start matching possible action shortcuts. So Ctrl-space c o p y would show any "copy" keyboard shortcut actions available for the current application. You then hit enter to perform the copy action for that application.
It would also partial match, so ctrl-space c would also allow you to perform the copy action.
Ctrl-space t a b would match any actions with the word "tab" in them.
So instead of you having to conform to the firefox shortcuts for doing it, or the ubuntu shortcuts for doing it, applications would LEARN YOUR WAY of doing it. The system would learn from your previous choices which action to do by default. In this way, you would get to control which shortcuts or actions would trigger from a certain text search. In short order, you could train Ctrl-space t to mean "open a new tab" in all of the applications with tab capability.
I first saw this idea here:
http://www.cimitan.com/blog/2009/01/31/do-ifying-gtk-30/ so comment and encourage that developer if you find this way of doing things compelling.
Essentially, instead of applications forcing us to conform to their keyboard-shortcut scheme, we should be able to train applications to follow the keyboard shortcuts that we choose as we start using an app (by text matching actions similar to Gnome-Do and Quicksilver, if people are familiar).
In essence, you would assign a key combo of your choice (Ctrl-space for example) to start matching possible action shortcuts. So Ctrl-space c o p y would show any "copy" keyboard shortcut actions available for the current application. You then hit enter to perform the copy action for that application.
It would also partial match, so ctrl-space c would also allow you to perform the copy action.
Ctrl-space t a b would match any actions with the word "tab" in them.
So instead of you having to conform to the firefox shortcuts for doing it, or the ubuntu shortcuts for doing it, applications would LEARN YOUR WAY of doing it. The system would learn from your previous choices which action to do by default. In this way, you would get to control which shortcuts or actions would trigger from a certain text search. In short order, you could train Ctrl-space t to mean "open a new tab" in all of the applications with tab capability.
I first saw this idea here: http://www.cimitan.com/blog/2009/01/31/do-ifying-gtk-30/ so comment and encourage that developer if you find this way of doing things compelling.
Solution #11:
HIG and configuration system
Promote something like libelektra
http://www.libelektra.org for generalized configuration information. (Shortcuts, command line parameters, language etc) With some persuasion and developed code it should be possible to produce code that would allow shortcuts to be set for each and every application in one place without needing to duplicate the information. Once the code is changed in the upstream app.
But it would also allow for a central configuration system across the board without needing significant redesign or
Worse case scenario (no coop from upstream) something like libelektra can be used to generate and edit the legacy configuration information.
The fact is most large apps already allow for remapping of keyboard shortcuts but there is no easy way to set them. A properly designed configuration system adopted across the board would significantly change this landscape. I think a libelektra is a good start in this direction.
Promote something like libelektra http://www.libelektra.org for generalized configuration information. (Shortcuts, command line parameters, language etc) With some persuasion and developed code it should be possible to produce code that would allow shortcuts to be set for each and every application in one place without needing to duplicate the information. Once the code is changed in the upstream app.
But it would also allow for a central configuration system across the board without needing significant redesign or
Worse case scenario (no coop from upstream) something like libelektra can be used to generate and edit the legacy configuration information.
The fact is most large apps already allow for remapping of keyboard shortcuts but there is no easy way to set them. A properly designed configuration system adopted across the board would significantly change this landscape. I think a libelektra is a good start in this direction.
Drag and drop needs to be made consistent (GNOME)
Written by wacked_up the 28 May 09 at 09:08.
Related project: Gnome .
New
Right now dragging and dropping items is fickle and inconsistent. It is unclear what exactly will happen when an item is dropped somewhere. Sometimes the expected action happens, sometimes nothing happens, and worse still, sometimes the wrong action ensues.
Solution #1:
Make drag and drop and copy/paste consistent
We should adopt new guidelines for how to interpret the drop.
This entails the following at least:
* Dropping an item somewhere should always result in an action, even if this means displaying an error. (Right now, often no feedback at all is provided, making the user confused about what happened. Since he started an action, an action should ensue.)
* dropping a file onto a taskbar entry or anywhere on a window of an application should open the file in that application, which is a more specific case of this:
* Dropping an item somewhere should result in undertaking the most logical action, or, if more than one action is possible, have the user choose one.
A few examples:
-dropping a piece of selected text in a folder should result in a file with that text in there, or possibly an empty file with that name (this should be asked).
-dropping an image into a text editor should paste the filename/path of the image in the open document.
-dropping a place from the places menu in nautilus should make use of the modifiers that are already in place to indicate the action to be taken (e.g. copy, open, link, etc). Right now this isn't possible.
-alt-tab and workspace switching should be available while dragging
Of course this list in incomplete, but when implemented, it should make dragging and dropping much more powerful and useful for the user.
The same ideas of course apply to copy/paste actions. They should always perform an action as well, and perform the same action as a drag and drop would.
We should adopt new guidelines for how to interpret the drop.
This entails the following at least:
* Dropping an item somewhere should always result in an action, even if this means displaying an error. (Right now, often no feedback at all is provided, making the user confused about what happened. Since he started an action, an action should ensue.)
* dropping a file onto a taskbar entry or anywhere on a window of an application should open the file in that application, which is a more specific case of this:
* Dropping an item somewhere should result in undertaking the most logical action, or, if more than one action is possible, have the user choose one.
A few examples:
-dropping a piece of selected text in a folder should result in a file with that text in there, or possibly an empty file with that name (this should be asked).
-dropping an image into a text editor should paste the filename/path of the image in the open document.
-dropping a place from the places menu in nautilus should make use of the modifiers that are already in place to indicate the action to be taken (e.g. copy, open, link, etc). Right now this isn't possible.
-alt-tab and workspace switching should be available while dragging
Of course this list in incomplete, but when implemented, it should make dragging and dropping much more powerful and useful for the user.
The same ideas of course apply to copy/paste actions. They should always perform an action as well, and perform the same action as a drag and drop would.