You are correct that the explanation begins with the presumption that spatial is simpler. I was attempting to explain the thinking process of the software's authors based on what I'm sure they believe to be true.
Though it wasn't the point of my post, I do think spatial is a simpler concept. I think you would agree that people who understand browser-type file management don't find spatial complicated per se, they just don't like it (understandably). Of course, this is just opinion which is why, being too lazy to go conduct proper usability studies of my own, I have not focused on whether spatial is better. My points are only that the people who say it obviously sucks for everyone have no hard evidence, and that the maintainers have behaved in an internally consistent fashion.
You'll note that it's two separate issues:
1) "is spatial easier for new users to understand?"
2) "if it is easier, should it be the default?"
#2 only comes into the picture if the answer to #1 is yes. If the answer is no, the only reason to make it the default is if you are purposely trying to make life difficult for new users:-).
I don't think we are disagreeing here, except that I will correct your last sentence: "I can tell you [some] people don't like spatial.";-)
> as the new Gnome overrides the users existing setup and forces the new behavior on them (requiring > them to explicitly disable it) the burden of proof is on the Gnome devs.
That's a good point and is true for many UI changes where the old functionality is kept around. I think, however, that in this particular situation the reasoning is something like this:
- If the "advanced" option is the default, "advanced" users don't need to change any options to get the desired functionality.
- If the simpler option is the default, normal humans don't need to change any options to get the functionality that is believed to best fit them.
- Advanced users know what they like and are capable of editing preferences. Basic users will take what is given.
So the basic idea is that making spatial mode something other than default would put it out of reach of exactly the group of users that it is intended to help.
> As painful as it seems for the Gnome guys to either test this out or believe anyone who says so, > most users disable spatial Nautilus. This should be done by default. However otherwise the Gnome on > FC3 feels great, particularly the file associations and launcher editing tools.
I'll bite. Why should they believe that? Because you have asserted that it is true? Has anyone (meaning you) done any actual research demonstrating that that is the case? I know four normal-human gnome users and none (0) of them would even know what spatial/browser modes are let alone change it. Does that prove anything?
Here are some fun quotes.
"Proof is not anecdotal. Proof is scientific. Show me the research." - Mark Levine, DC
"Anecdotes are useless precisely because they may point to idiosyncratic responses."
"The plural of anecdote is not data." - Roger Brinner
> How is it that we can build reliable, accountable systems to handle power, money, and everything else > in our society, but somehow it's fundamentally impossible to expect that it could be done with voting.
Those sorts of systems have some pretty obvious checks built into them. If my power company charges me twice my normal power bill, I will know that there's and error and call them. If my bank says that there was a large withdrawal that I know should not have ocurred, I will call them.
Both of those situations have happened to me more than once in more than one state with more than one bank/power company.
As for "everything else in our society" let's see... grocery store - I've had problems, restaurants - problems, hospitals & doctors - problems.
We don't actually seem to be able to build a reliable *anything*.:-/
Oh, I got this one. It's because then people would think that those Ubuntu people were really annoying for always bringing it up in unrelated threads. (Ha, ha, only serious)
> That'll be because no-one targets the Mac with > spyware or viruses, because Windows is a soft > enough target and has vastly more market share; > it's not worth their while to yet.
You don't seem to be disagreeing that Mac users are safer. Can you explain why it should matter to them what the reason for that safety is?
Let's say two people have the same weak locks on their doors and windows, but one lives in a crime-ridden neighborhood, while the other lives in a relatively safe one. Should the guy who has been robbed multiple times console himself with the fact that he is only robbed because there are so many more criminals in his area?
> One camp proclaims why their system is good, while the other listens to how it could be better.
You're really not paying any attention here are you? How can anyone listen to what isn't being said? Saying "product X was disappointing" is not the same as "doing Y will make it better." Read my post again. I say that actually telling what that Y is would be helpful. There is no intelligent way to argue that specificity is not helpful, I think you have demonstrated that.
> And I belive ALL feedback is important, even if you have to work to translate it into something useful.
Nope. Lots of feedback is noise.
If you have an actual criticism about gconf, or anything else for that matter, expressing it well is a good start to getting it resolved.
There are things about gconf to criticize, and those criticisms don't seem to be ignored. It's just that the popular sorts of criticisms (OMG it's the registry!! Gnome hax0rs are teh eevil!!) are so tired and inaccurate as to be a complete waste of time.
The complaints about gconf seemed pretty useless to me. What gconf is really about is providing a nice library to encapsulate preferences storage/updates. the Gconf editor is not meant to be something that you use on anything resembling a regular basis.
Declaring it difficult to use, compared to the alternative (your text editor of choice) seems a strong enough claim that it should have been backed up by more description.
I think there are different measures of literacy being employed here. Go read some of the Federalist Papers which were written for the "common man" and see if you think the average "literate" high school graduate could comprehend it. Lowering the standards will definitely lead to more people passing.
"As a Libertarian candidate, I frequently face the "wasted vote" syndrome. People tell me that I'm a good candidate. They believe in what I stand for, but they can't bring themselves to vote for me because they don't want to waste their vote. If you were in prison, and you had a 50% chance of lethal injection, a 45% chance of going to the electric chair, and only a 5% chance of escape, are you likely to vote for lethal injection because that is your most likely outcome? Your survival depends on voting for escape even if that's only a 5% chance. If you continue to vote for the Democrats or the Republicans, you are committing political suicide. The only chance we have of saving our constitutional republic is to vote Libertarian, even if that's only a 5% probability of getting into office. We have to demonstrate that we are not satisfied with the status quo. Voting for the lesser of two evils and your candidate wins and you still get evil."
You can draw GTK+2 apps using Qt widgets but that doesn't magically give the applications DCOP interfaces
When D-Bus is adopted in future versions of KDE and Gnome they will. I think there is already a DCOP D-Bus bridge. Merging KIO and Gnome-vfs (not to mention mozilla's necko) is probably a looong way down the road
Fedora is still struggling to reimpliment Debian's community, and is still making it difficult for "outsiders" to have as much say as RedHat folks. </hearsay>
Debian and UserLinux have almost the reverse of the relationship between Fedora and RHEL. RHEL starts with Fedora and adds various things. UserLinux starts with the HUGE number of well-tested packages in Debian and whittles it down to a manageable core that can be depended on to be there. It's sort of a "Debian standard base" if you will.
One excellent feature is that instead of relying on stuff like Dag Wieers excellent, but still 3rd party, set of packages for Fedora, nearly every open source application of note is packaged in debian (and has been checked against Debian's very strict policy) and will be easy to install on a UserLinux system. When the next Debian and UserLinux releases come out, the upgrade path for those "add-on" packages will also have been well tested.
So, long story short (too late), what really makes UserLinux valuable is that it _is_ Debian, and has all the strength and experience of Debian behind it.
every so often, unstable just breaks. It's not like it's planned, but the fact is, with so many package maintainers,
something's bound to go wrong - and it usually does every few months. At that point, you've got to go and uninstall and reinstall packages to make dpkg not complain about weird circular dependency problems - an irony for a distribution that so many claim is the answer to "dependency hell".
The "unstable" moniker refers to the state of the interdependencies between packages. There is no irony whatsoever in the version explicitly labelled "unstable" having unstable dependencies. IMO the fact that the problems happen so rarely, and that the reasons for them are generally well known (gcc 3, gnutls, libpng problems were all known about well in advance) is a testament to the quality of the work the debian developers do.
To a certain extent, you can test the waters before installing by using apt-listbugs and knowing how to roll back versions of packages.
That said, debian is not for everybody. I like it because it Works For Me, it has just about every package that I am interested in and is constantly being updated. It works for my wife because I admin her system for her (and I always apt-get upgrade my own system before touching hers). For a lot of people, stable plus backports would be a great fit if it were better documented (or blessed) how to do that.
In the end though, this isn't Highlander, there can be more than one distro. I like a lot of the other distros, I just like debian more, I ain't mad at you if you run fedora or suse (now, lindows... maybe).
I'm curious what makes you think that purchasers of Windows or Solaris are automatically indemnified. All the Sun and MS licenses that I have seen state that the software is offered AS-IS with NO WARRANTY for any purpose including merchantibility or fitness for any purpose.
Are you suggesting that if I pay for GPLed software that I would receive the same protection? Or that if I was given proprietary software for free that I wouldn't be indemnified? Or is it only the combination of proprietary, paid for that gives the recipient special legal standing?
One tactic that I have used successfully for some time is to "sanitize"[0] potentialy destructive attachments on incoming emails.
This means that.exe files get renamed to whatever.exe.bin and the content type gets changed to application/binary. This way a user has to really want to run that executable, and know how. I also have it dig into zip and tnef files and do the same there.
Now that I think of it, this is sort of a poor-man's executable bit. It doesn't actually prevent execution, it just adds another step (that isn't just an "are you sure?" dialog) to the process.
You might want to try jhbuild or garnome then. Jhbuild will let you build either a specific version of Gnome, or CVS HEAD, garnome is more release-oriented. There is very little to be gained by building "by hand" (typing./configure && make && make install). I haven't done much modifying of jhbuild, but garnome makes it easy to pass options to configure/make etc.
I use 1024x768 too (on a laptop, no other choice) and the gnome toolbars are fine... you do know that you can change the size of the toolbar right? Also, hopefully you do know that Fedora's version of Gnome is not representative of "real" gnome. They set their toolbar quite large (64px I think).
That's an interesting idea, though it seems to be a solution looking for a problem.
There does not appear to be much that it does that a "Drafts" mechanism in a mail client and authenticated SMTP cannot also do.
Also, the grandparent's point was not that one should not use smtp: "Client machines should be talking to their SMTP server in an authenticated manner using one of the ports like tcp/465 and the like"
-Mark
Re:Okay, so what do you do?
on
Are You Annoying?
·
· Score: 3, Funny
> And you could never send him an email without getting it back, grammer corrected. What a prick.
That's *grammar* not grammer! Can you follow this??
Slashdot Mirror
This mercedes interior really reminds me of this 1998 land cruiser... It's almost as if the designers wanted the cars to all be similar. What unoriginal S.O.B.s
Seriously, is your point: "The open file dialog is not innovative" ?
-Mark
You are correct that the explanation begins with the presumption that spatial is simpler. I was attempting to explain the thinking process of the software's authors based on what I'm sure they believe to be true.
:-).
;-)
Though it wasn't the point of my post, I do think spatial is a simpler concept. I think you would agree that people who understand browser-type file management don't find spatial complicated per se, they just don't like it (understandably). Of course, this is just opinion which is why, being too lazy to go conduct proper usability studies of my own, I have not focused on whether spatial is better. My points are only that the people who say it obviously sucks for everyone have no hard evidence, and that the maintainers have behaved in an internally consistent fashion.
You'll note that it's two separate issues:
1) "is spatial easier for new users to understand?"
2) "if it is easier, should it be the default?"
#2 only comes into the picture if the answer to #1 is yes. If the answer is no, the only reason to make it the default is if you are purposely trying to make life difficult for new users
I don't think we are disagreeing here, except that I will correct your last sentence: "I can tell you [some] people don't like spatial."
-Mark
> as the new Gnome overrides the users existing setup and forces the new behavior on them (requiring
> them to explicitly disable it) the burden of proof is on the Gnome devs.
That's a good point and is true for many UI changes where the old functionality is kept around. I think, however, that in this particular situation the reasoning is something like this:
- If the "advanced" option is the default, "advanced" users don't need to change any options to get the desired functionality.
- If the simpler option is the default, normal humans don't need to change any options to get the functionality that is believed to best fit them.
- Advanced users know what they like and are capable of editing preferences. Basic users will take what is given.
So the basic idea is that making spatial mode something other than default would put it out of reach of exactly the group of users that it is intended to help.
-Mark
> As painful as it seems for the Gnome guys to either test this out or believe anyone who says so,
> most users disable spatial Nautilus. This should be done by default. However otherwise the Gnome on
> FC3 feels great, particularly the file associations and launcher editing tools.
I'll bite. Why should they believe that? Because you have asserted that it is true? Has anyone (meaning you) done any actual research demonstrating that that is the case? I know four normal-human gnome users and none (0) of them would even know what spatial/browser modes are let alone change it. Does that prove anything?
Here are some fun quotes.
"Proof is not anecdotal. Proof is scientific. Show me the research." - Mark Levine, DC
"Anecdotes are useless precisely because they may point to idiosyncratic responses."
"The plural of anecdote is not data." - Roger Brinner
-Mark
> How is it that we can build reliable, accountable systems to handle power, money, and everything else
:-/
> in our society, but somehow it's fundamentally impossible to expect that it could be done with voting.
Those sorts of systems have some pretty obvious checks built into them. If my power company charges me twice my normal power bill, I will know that there's and error and call them. If my bank says that there was a large withdrawal that I know should not have ocurred, I will call them.
Both of those situations have happened to me more than once in more than one state with more than one bank/power company.
As for "everything else in our society" let's see... grocery store - I've had problems, restaurants - problems, hospitals & doctors - problems.
We don't actually seem to be able to build a reliable *anything*.
-Mark
Oh, I got this one. It's because then people would think that those Ubuntu people were really annoying for always bringing it up in unrelated threads. (Ha, ha, only serious)
-Mark
FYI, the Ubuntu installer _is_ the Debian (Sarge) installer.
-Mark
> That'll be because no-one targets the Mac with
> spyware or viruses, because Windows is a soft
> enough target and has vastly more market share;
> it's not worth their while to yet.
You don't seem to be disagreeing that Mac users are safer. Can you explain why it should matter to them what the reason for that safety is?
Let's say two people have the same weak locks on their doors and windows, but one lives in a crime-ridden neighborhood, while the other lives in a relatively safe one. Should the guy who has been robbed multiple times console himself with the fact that he is only robbed because there are so many more criminals in his area?
-Mark
> One camp proclaims why their system is good, while the other listens to how it could be better.
You're really not paying any attention here are you? How can anyone listen to what isn't being said? Saying "product X was disappointing" is not the same as "doing Y will make it better." Read my post again. I say that actually telling what that Y is would be helpful. There is no intelligent way to argue that specificity is not helpful, I think you have demonstrated that.
-Mark
> And I belive ALL feedback is important, even if you have to work to translate it into something useful.
Nope. Lots of feedback is noise.
If you have an actual criticism about gconf, or anything else for that matter, expressing it well is a good start to getting it resolved.
There are things about gconf to criticize, and those criticisms don't seem to be ignored. It's just that the popular sorts of criticisms (OMG it's the registry!! Gnome hax0rs are teh eevil!!) are so tired and inaccurate as to be a complete waste of time.
-Mark
The complaints about gconf seemed pretty useless to me. What gconf is really about is providing a nice library to encapsulate preferences storage/updates. the Gconf editor is not meant to be something that you use on anything resembling a regular basis.
Declaring it difficult to use, compared to the alternative (your text editor of choice) seems a strong enough claim that it should have been backed up by more description.
-Mark
You are sooooo good looking
-Mark
I think there are different measures of literacy being employed here. Go read some of the Federalist Papers which were written for the "common man" and see if you think the average "literate" high school graduate could comprehend it. Lowering the standards will definitely lead to more people passing.
-Mark
source
-Mark
When D-Bus is adopted in future versions of KDE and Gnome they will. I think there is already a DCOP D-Bus bridge. Merging KIO and Gnome-vfs (not to mention mozilla's necko) is probably a looong way down the road
-Mark
Fedora is still struggling to reimpliment Debian's community, and is still making it difficult for "outsiders" to have as much say as RedHat folks.
</hearsay>
Debian and UserLinux have almost the reverse of the relationship between Fedora and RHEL. RHEL starts with Fedora and adds various things. UserLinux starts with the HUGE number of well-tested packages in Debian and whittles it down to a manageable core that can be depended on to be there. It's sort of a "Debian standard base" if you will.
One excellent feature is that instead of relying on stuff like Dag Wieers excellent, but still 3rd party, set of packages for Fedora, nearly every open source application of note is packaged in debian (and has been checked against Debian's very strict policy) and will be easy to install on a UserLinux system. When the next Debian and UserLinux releases come out, the upgrade path for those "add-on" packages will also have been well tested.
So, long story short (too late), what really makes UserLinux valuable is that it _is_ Debian, and has all the strength and experience of Debian behind it.
-Mark
Good news! Your post is eligible for a free super-leetness upgrade. Simply replace each occurence of "ascii" with "UTF-8 encoded Unicode"
Impress your friends, be the envy of your enemies.
The "unstable" moniker refers to the state of the interdependencies between packages. There is no irony whatsoever in the version explicitly labelled "unstable" having unstable dependencies. IMO the fact that the problems happen so rarely, and that the reasons for them are generally well known (gcc 3, gnutls, libpng problems were all known about well in advance) is a testament to the quality of the work the debian developers do.
To a certain extent, you can test the waters before installing by using apt-listbugs and knowing how to roll back versions of packages.
That said, debian is not for everybody. I like it because it Works For Me, it has just about every package that I am interested in and is constantly being updated. It works for my wife because I admin her system for her (and I always apt-get upgrade my own system before touching hers). For a lot of people, stable plus backports would be a great fit if it were better documented (or blessed) how to do that.
In the end though, this isn't Highlander, there can be more than one distro. I like a lot of the other distros, I just like debian more, I ain't mad at you if you run fedora or suse (now, lindows... maybe).
-Mark
I'm curious what makes you think that purchasers of Windows or Solaris are automatically indemnified. All the Sun and MS licenses that I have seen state that the software is offered AS-IS with NO WARRANTY for any purpose including merchantibility or fitness for any purpose.
Are you suggesting that if I pay for GPLed software that I would receive the same protection? Or that if I was given proprietary software for free that I wouldn't be indemnified? Or is it only the combination of proprietary, paid for that gives the recipient special legal standing?
Any reputable sources that explain the situation?
-Mark
One tactic that I have used successfully for some time is to "sanitize"[0] potentialy destructive attachments on incoming emails.
.exe files get renamed to whatever.exe.bin and the content type gets changed to application/binary. This way a user has to really want to run that executable, and know how. I also have it dig into zip and tnef files and do the same there.
t y.html
This means that
Now that I think of it, this is sort of a poor-man's executable bit. It doesn't actually prevent execution, it just adds another step (that isn't just an "are you sure?" dialog) to the process.
-Mark
[0] http://www.impsec.org/email-tools/procmail-securi
I was actually thinking of the panel wrt to Fedora but you can change the toolbar prefs globally:
Applications->Desktop Preferences->Menus and Toolbars
Some programs, such as gedit, also allow deviation from the systemwide setting.
-Mark
You might want to try jhbuild or garnome then. Jhbuild will let you build either a specific version of Gnome, or CVS HEAD, garnome is more release-oriented. There is very little to be gained by building "by hand" (typing ./configure && make && make install). I haven't done much modifying of jhbuild, but garnome makes it easy to pass options to configure/make etc.
-Mark
I use 1024x768 too (on a laptop, no other choice) and the gnome toolbars are fine... you do know that you can change the size of the toolbar right? Also, hopefully you do know that Fedora's version of Gnome is not representative of "real" gnome. They set their toolbar quite large (64px I think).
-Mark
That's an interesting idea, though it seems to be a solution looking for a problem.
There does not appear to be much that it does that a "Drafts" mechanism in a mail client and authenticated SMTP cannot also do.
Also, the grandparent's point was not that one should not use smtp:
"Client machines should be talking to their SMTP server in an authenticated manner using one of the ports like tcp/465 and the like"
-Mark
> And you could never send him an email without getting it back, grammer corrected. What a prick.
That's *grammar* not grammer! Can you follow this??
Love,
Steve