You couldn't have stated it better, that there is no benefit to making configuration in unix more complicated. How about making it simpler?
Forget to tab in syslog.conf? : comments got you down in inittab, typo in passwd locked you out? Guessing at possible values for/etc/system variables got you stumped?
Granted the lack of structure in/etc has allowed unrestrained creativity, but the knowledge required to navigate/etc successfully is significant and -- I'm guessing --, taken for granted by you and many other long time unix admins like myself.
Without dumbing it down, or removing flexibility, I believe a better way to manage the bits of configuration required by each program would be a centrally managed, accessed, API driven repository for config. And there could be a text editor based "window" into it to anyone that cared to work directly with the data. I don't pretend to know the right way to begin to code this up, but I'm tired of explaining to new admins that are looking to change X in unix, that the only way to know how to find the config file -- is to already know where to find the config file.
You've made an interesting assumption in your statement regarding an install gone awry, namely that because one is installing linux and one can see what is going on behind the scenes, one is able to fix a problem. I would argue that it is only with a knowledge of unix in general and the transparency (of scripting and/or config files) in linux that makes this work.
Consider comparing the configuration of a program via a windows wizard vs editing a config file. I'm not convinced that the power is in the config files, compared to say, the windows registry. Rather, it is in the structure that config files lend to the problem at hand (a single broken config file narrows down the problem, a missing registry entry is quite a large domain to search) and the documentation that accompanies the config file (either in comments or in a manual page). If running a wizard style tool in windows resulted in a structured change to a well documented section of the registry, debugging and advanced understanding of the changes facilitated by the GUI wizard would be possible. And simply moving the registry entries to a config file without any man page would require apriori knowledge of the program being configured!
So where does this leave us? To facilitate the rapid adoption of linux by a larger pool of "new users" the wizard style of program configuration could prove quite useful. This model can be adapted however to provide immediate access to the underlying configuration files _and_ manual pages so that one can at once get the machine configured as required and learn a bit too.
Slashdot Mirror
You couldn't have stated it better, that there is no benefit to making configuration in unix more complicated. How about making it simpler?
/etc/system variables got you stumped?
/etc has allowed unrestrained creativity, but the knowledge required to navigate /etc successfully is significant and -- I'm guessing --, taken for granted by you and many other long time unix admins like myself.
Forget to tab in syslog.conf? : comments got you down in inittab, typo in passwd locked you out? Guessing at possible values for
Granted the lack of structure in
Without dumbing it down, or removing flexibility, I believe a better way to manage the bits of configuration required by each program would be a centrally managed, accessed, API driven repository for config. And there could be a text editor based "window" into it to anyone that cared to work directly with the data. I don't pretend to know the right way to begin to code this up, but I'm tired of explaining to new admins that are looking to change X in unix, that the only way to know how to find the config file -- is to already know where to find the config file.
You've made an interesting assumption in your statement regarding an install gone awry, namely that because one is installing linux and one can see what is going on behind the scenes, one is able to fix a problem. I would argue that it is only with a knowledge of unix in general and the transparency (of scripting and/or config files) in linux that makes this work.
Consider comparing the configuration of a program via a windows wizard vs editing a config file. I'm not convinced that the power is in the config files, compared to say, the windows registry. Rather, it is in the structure that config files lend to the problem at hand (a single broken config file narrows down the problem, a missing registry entry is quite a large domain to search) and the documentation that accompanies the config file (either in comments or in a manual page). If running a wizard style tool in windows resulted in a structured change to a well documented section of the registry, debugging and advanced understanding of the changes facilitated by the GUI wizard would be possible. And simply moving the registry entries to a config file without any man page would require apriori knowledge of the program being configured!
So where does this leave us? To facilitate the rapid adoption of linux by a larger pool of "new users" the wizard style of program configuration could prove quite useful. This model can be adapted however to provide immediate access to the underlying configuration files _and_ manual pages so that one can at once get the machine configured as required and learn a bit too.