3 ms·

I have a severe (self-diagnosed :P ) case of Not Invented Here. It prevents me from hacking on Somebody Else's Code. It's why I use Ubuntu instead of Plan9 or H

I have a severe (self-diagnosed :P ) case of Not Invented Here. It prevents me from hacking on Somebody Else's Code. It's why I use Ubuntu instead of Plan9 or HURD or Movitz. I really want to move toward one of those, but I just can't make myself make the time to get used to them to the point where I can add enough to them to feel comfortable that I'll be able to do anything I need to. Is there a tutorial for people like me to learn how to get involved with such a big undertaking as contributing to a partially-completed OS written by someone else?

jff 16y ago
The three OS examples you give are all at varying stages of completeness: Movitz is just a text-mode Lisp prompt; HURD has been in development for over 20 years and still hasn't had a stable release; Plan 9 has had multiple commercial releases in those same 20 years, each a significant advance over the other, and now the whole system is free. I've worked with Plan 9 for several years now. The advantages for a developer are that you can understand the entirety of the kernel with a little effort, there is a centralized location for user-contributed code, the tools are excellent, and the community is small enough that you can really get to know everyone. Linux has 30 different applications for every task, and most of them suck hard in at least one major point, but you get the feeling that there's nothing left to write. In Plan 9, there's a wide open field of useful things to create. It would be nice to have some decent calendaring tools, for instance. While we only really have one windowing system, rio, there's nothing at all to stop you from making a different one, say with title bars and buttons--and people would use it, especially clueless disgruntled newbies :) So really, the tutorial for Plan 9 is "read the wiki, play with the system, write some code when you're ready."
- calcnerd256 16y ago
  I guess it's time for me to try Plan 9 in a virtual machine again soon. Last time I tried it, I wasn't able to get it up to the point that I could use the Web browser. That's kind of necessary these days, if one wants to stick with an OS.

jff 16y ago

The three OS examples you give are all at varying stages of completeness: Movitz is just a text-mode Lisp prompt; HURD has been in development for over 20 years and still hasn't had a stable release; Plan 9 has had multiple commercial releases in those same 20 years, each a significant advance over the other, and now the whole system is free. I've worked with Plan 9 for several years now. The advantages for a developer are that you can understand the entirety of the kernel with a little effort, there is a centralized location for user-contributed code, the tools are excellent, and the community is small enough that you can really get to know everyone. Linux has 30 different applications for every task, and most of them suck hard in at least one major point, but you get the feeling that there's nothing left to write. In Plan 9, there's a wide open field of useful things to create. It would be nice to have some decent calendaring tools, for instance. While we only really have one windowing system, rio, there's nothing at all to stop you from making a different one, say with title bars and buttons--and people would use it, especially clueless disgruntled newbies :) So really, the tutorial for Plan 9 is "read the wiki, play with the system, write some code when you're ready."

calcnerd256 16y ago
I guess it's time for me to try Plan 9 in a virtual machine again soon. Last time I tried it, I wasn't able to get it up to the point that I could use the Web browser. That's kind of necessary these days, if one wants to stick with an OS.