Talk:Plan of Record-2008/Draft 2: Difference between revisions

From OLPC
Jump to navigation Jump to search
No edit summary
No edit summary
 
Line 2: Line 2:


# OLPC should and does recognize that its ability to act as a field marshal is limited. Even for the direct employees, people should and do work on the problems that they know how to solve, and implicitly make decisions about what they personally see as important. Also of course, simply piling people on a software problem famously [http://en.wikipedia.com/wiki/The_Mythical_Man_Month doesn't help]. Therefore, I think that setting more than one general priority per release would not be very meaningful.
# OLPC should and does recognize that its ability to act as a field marshal is limited. Even for the direct employees, people should and do work on the problems that they know how to solve, and implicitly make decisions about what they personally see as important. Also of course, simply piling people on a software problem famously [http://en.wikipedia.com/wiki/The_Mythical_Man_Month doesn't help]. Therefore, I think that setting more than one general priority per release would not be very meaningful.
:Clarification: I think this lack of perfectly unified direction is a good thing. A basic roadmap is what we need, not individual marching orders. [[User:Homunq|Homunq]] 19:23, 27 April 2008 (EDT)
#:Clarification: I think this lack of perfectly unified direction is a good thing. A basic roadmap is what we need, not individual marching orders. [[User:Homunq|Homunq]] 19:23, 27 April 2008 (EDT)
# I think that if there is just one priority for the next release, it should be the datastore. I have seen several issues - some of which I can't reproduce and have not yet filed into trac, sorry - with that component, and it is absolutely key to the user experience. I also know that there is a grand plan which some believe will make a qualitative change in its stability as well as breaking API - I think that should happen as soon as possible if it is going to happen.
# I think that if there is just one priority for the next release, it should be the datastore. I have seen several issues - some of which I can't reproduce and have not yet filed into trac, sorry - with that component, and it is absolutely key to the user experience. I also know that there is a grand plan which some believe will make a qualitative change in its stability as well as breaking API - I think that should happen as soon as possible if it is going to happen.



Latest revision as of 23:26, 27 April 2008

I am liking how this document is evolving. I have only two points to add:

  1. OLPC should and does recognize that its ability to act as a field marshal is limited. Even for the direct employees, people should and do work on the problems that they know how to solve, and implicitly make decisions about what they personally see as important. Also of course, simply piling people on a software problem famously doesn't help. Therefore, I think that setting more than one general priority per release would not be very meaningful.
    Clarification: I think this lack of perfectly unified direction is a good thing. A basic roadmap is what we need, not individual marching orders. Homunq 19:23, 27 April 2008 (EDT)
  2. I think that if there is just one priority for the next release, it should be the datastore. I have seen several issues - some of which I can't reproduce and have not yet filed into trac, sorry - with that component, and it is absolutely key to the user experience. I also know that there is a grand plan which some believe will make a qualitative change in its stability as well as breaking API - I think that should happen as soon as possible if it is going to happen.

Homunq 19:05, 17 April 2008 (EDT)