Bedrock

[Nikolai_D]

whatcha guys think about it?
Anyone used already?

[joshuapurcell]

Sounds awesome even if it does have a more involved installation process. Here's a link for those who aren't familiar:

http://opensource.osu.edu/~paradigm/bedrock/

There's alot of good info on the website, but I'm interested to see more results from various scenarios (multiple distros being used together to work through common issues, etc.). I know I'm not the only one who has to deal with multiple distros (and the incompatibilities between them) on a daily basis.

[crazycheese]

I don't like that multitude of features are achieved by compexity (of mashing everything together). I view it highest as a research project.

[joshuapurcell]

I think the same general complaint could have been made for many different technologies before they became standard.

Take for instance web services: why not just have the two applications talk to each other directly? Why have this intermediate layer of complexity (i.e. web server, service application on client and server sides, HTTP/REST, etc.) handle the communication between these two applications?

This approach that Bedrock is taking can be looked at in the same way. Various Linux distros do the same thing in different ways, and there is no hope of moving towards a single way of doing some task (installing apps, displaying a window, etc.) because there is simply no one best way to do that task. But that task is repeated regardless of how differently it's done across all Linux distributions. The approach that Bedrock is taking is the way to handle that problem without demanding that any change take place in the client distributions that it handles.

I think this will actually lead to better future Linux distributions, and not because they will adopt the approach of Bedrock. Instead what could happen is more Linux users will be able to easily handle generic Linux system tasks in a variety of ways that were not easily done on one system prior to this approach. We'll be able to more easily compare tools that handle a specific job by further removing the issues around distribution incompatibilities.

This approach could also help further Linux adoption, because it wouldn't matter as much which distribution a company or organization targets for their project.

The big issue I see at this point is the complexity that Bedrock adds in, but I think more work can be done towards hiding this complexity so that it is easier to get up and running with multiple functioning distros.