ROS design suggestion
Moderator: Moderator Team
I just have to explain this a little better.
The Logout Loop on X11 did not contain a crash. The Crash restore program was missing a different way a application was loging the user out.
Since the Crash restore did not detect the vaild logout when I login it though the system had crashed and it just did not record it. So the program that did logout each time I tried to login on X11 until I got into my accout and cleared and removed the crash proffing. Thank god for console access on Unix/Linux/BSD based systems.
Yes its ouch.
This might be just my fear taking from a very bad experence.
Just throw this back as a idea.
If we do crash proofing its on a per application base. Ie you can restore one application at a time. Not all. Solve problem above. And crash restore images are backed in case of a restore required up so if you find one dud app no data loss. Really I see this as a sub application system not a gui thing. Ie it take memory snapshots of targeted applications. This will cost application speed but handy with a unstable application to use it for something.
This is still extreamly hard to pull off reason.
If program writes files to disk and require these files to contain X data at X time by the time of restore the file might contain Y so application crashs.
Snapshot system required many subsystem parts to work. Ie filesystem must be tracked interfaces with other applications must be tracked and so on. Well built crash proffing is extreamly hard to make.
Firefox can have its own internal restore system no need for ours. OpenOffice/MS Office Have there own internals MS Office just need a interface to make it simple. Ie you cannot not get simpler tha OpenOffices start the program again do you wish to restore these files.
Winamp/Most media player. Playlists so no point for most users.
Simple crash protection might be just to recored applications in use. To tell user what they were using and to try rerunning now to restore stuff.
This is why I don't see how it fits with the GUI. Ok I could be wrong here.
The Logout Loop on X11 did not contain a crash. The Crash restore program was missing a different way a application was loging the user out.
Since the Crash restore did not detect the vaild logout when I login it though the system had crashed and it just did not record it. So the program that did logout each time I tried to login on X11 until I got into my accout and cleared and removed the crash proffing. Thank god for console access on Unix/Linux/BSD based systems.
Yes its ouch.
This might be just my fear taking from a very bad experence.
Just throw this back as a idea.
If we do crash proofing its on a per application base. Ie you can restore one application at a time. Not all. Solve problem above. And crash restore images are backed in case of a restore required up so if you find one dud app no data loss. Really I see this as a sub application system not a gui thing. Ie it take memory snapshots of targeted applications. This will cost application speed but handy with a unstable application to use it for something.
This is still extreamly hard to pull off reason.
If program writes files to disk and require these files to contain X data at X time by the time of restore the file might contain Y so application crashs.
Snapshot system required many subsystem parts to work. Ie filesystem must be tracked interfaces with other applications must be tracked and so on. Well built crash proffing is extreamly hard to make.
Firefox can have its own internal restore system no need for ours. OpenOffice/MS Office Have there own internals MS Office just need a interface to make it simple. Ie you cannot not get simpler tha OpenOffices start the program again do you wish to restore these files.
Winamp/Most media player. Playlists so no point for most users.
Simple crash protection might be just to recored applications in use. To tell user what they were using and to try rerunning now to restore stuff.
This is why I don't see how it fits with the GUI. Ok I could be wrong here.
Re: Shell Replacement, Anyone?
TOTALLY AGREE WITH YOU !!!Wierd wrote:So.. You dont like the scrollbars, buttons, radio circles, checkboxes, layout of MS window's "windows"...
Shell replacement is your friend!
So i'm not the unique SharpE-lover... but far from its look (that is great, imho), the coolest things are:Wierd wrote:Here is an article some guy made about his experiences using alternative shells. As you can see, simply replacing the shell opens the door to all kinds of new functionalities.
http://shell-shocked.org/article.php?id=48Mike Holland wrote:I was absolutely speechless! I never thought it possible to improve the look and (total) feel of Windows to such a degee that you couldn't even tell that it IS, in fact, Windows. (Is that a yellow brick road I see, Toto?)
1. It's already done (and have a team) !
2. It's open source (LGPL), so it can be adapted or "ported" easily
3. It have many users (that can become ReactOS users), check out the user's gallery
...in my opinion, it should become the standard ROS shell !
Anyway many peoples don't agree with us, so a (GUI)fork would be reasonable.
»Forward Agency NPO
In progress we (always) trust.
In progress we (always) trust.
-
- Posts: 237
- Joined: Fri Nov 26, 2004 10:44 pm
- Location: Fallowfield
Re: Shell Replacement, Anyone?
Making SharpE the standard shell:Patchworks wrote:...in my opinion, it should become the standard ROS shell !
Main advantage: Attract many SharpE users
Main disadvantage: Deter many, many Windows users
Adding SharpE as an option in the package manager:
Main advantage: Attract many SharpE users
Main disadvantage: Patchworks isn't as happy
-
- Posts: 47
- Joined: Fri Oct 21, 2005 4:27 am
You can add your ideas to the Package Manager wiki page(s):Lucractius wrote:2ndary disadvantages related to integration into package manager. we dont exactly have one yet *has plans for one*
http://www.reactos.org/wiki/index.php/R ... ge_Manager
-
- Posts: 47
- Joined: Fri Oct 21, 2005 4:27 am
So was this design not accepted, or does the project just have no formal way of accepting it? I am just curious because it seems as if most people like this design far better than the current one, and this design makes ROS look more windows like, which I thought was a goal. It just seems odd that a do able design change which makes such an aesthetic and functional inprovement would just be ignored.
The problem is basically that lots of people produce ideas, but only a few produce code. And code is the one thing that makes ReactOS tick. Personally I'd rather invest my coding time in functional changes and it seems that is the case for most of the active developers. So I guess until someone comes by to actually start coding on this, the ideas will just remain ideas.
Re: Shell Replacement, Anyone?
Err... why ?counting_pine wrote:Making SharpE the standard shell:
Main disadvantage: Deter many, many Windows users
Windows zealots don't like alternatives ANYWAY, so where's the problem ? And remember that Windows is going to change. I think hat when ROS will be on the 1.0 release Windows will be _radically_ different fom now.
counting_pine wrote:Adding SharpE as an option in the package manager:
Lucractius wrote:we dont exactly have one yet
GvG wrote:The problem is basically that lots of people produce ideas, but only a few produce code.
...another reason to "give away" some sections (as the shell, for example), or better: get something that just lives standalone, as an external indipendent project.
Just my opinion.
»Forward Agency NPO
In progress we (always) trust.
In progress we (always) trust.
Re: Shell Replacement, Anyone?
Err... why ?counting_pine wrote:Making SharpE the standard shell:
Main disadvantage: Deter many, many Windows users
Windows zealots don't like alternatives ANYWAY, so where's the problem ? And remember that Windows is going to change. I think hat when ROS will be on the 1.0 release Windows will be _radically_ different fom now.
counting_pine wrote:Adding SharpE as an option in the package manager:
Lucractius wrote:we dont exactly have one yet
...another reason to "give away" some sections (as the shell, for example), or better: get something that just lives standalone, as an external indipendent project.GvG wrote:The problem is basically that lots of people produce ideas, but only a few produce code.
Just my opinion.
»Forward Agency NPO
In progress we (always) trust.
In progress we (always) trust.
Re: Shell Replacement, Anyone?
Well, having alternative shells is always good, but i would like the default shell to be similar to explorer (though more customizable as i and others have said in other threads).Patchworks wrote:Err... why ?counting_pine wrote:Making SharpE the standard shell:
Main disadvantage: Deter many, many Windows users
Windows zealots don't like alternatives ANYWAY, so where's the problem ? And remember that Windows is going to change. I think hat when ROS will be on the 1.0 release Windows will be _radically_ different fom now.
And as far as a 1.0 release? I don't think Windows will be radically different, all they seem to do is add one or two GUI features (like transparency, translucency) and rearrage those g--d--- menus and icons so i have to relearn it all again.
ReactOS v1.0 (drool)
pax mei amici amorque et Iesus sacret omnia
Re: Shell Replacement, Anyone?
Patchworks, you seem to miss the point that ReactOS is trying to clone Windows as a whole, which means also the defaults. If people want to use an alternatice shell they will be free to do so, the same as if they wanted to use one on Windows.
How would giving away some sections help? If someone wants to work on the current code they are free to do so already.
How would giving away some sections help? If someone wants to work on the current code they are free to do so already.
Re: Shell Replacement, Anyone?
I know, but let me hope to have a better OS than Windows... 8)ScoTTie wrote:Patchworks, you seem to miss the point that ReactOS is trying to clone Windows as a whole, which means also the defaults.
»Forward Agency NPO
In progress we (always) trust.
In progress we (always) trust.
Who is online
Users browsing this forum: Ahrefs [Bot] and 12 guests