Re: [ros-users] best practices for ROS package repositories?

Top Page
Attachments:
Message as email
+ (text/plain)
Delete this message
Reply to this message
Author: Bill Morris
Date:  
To: ros-users
Subject: Re: [ros-users] best practices for ROS package repositories?
On Thu, 2010-08-19 at 15:04 -0500, Michael Quinlan wrote:
> a) Each package should have keywords associated with it. For example
> if I wanted to look at all the possible vision algorithms I could do
> so easily, currently you have to search for various phrases such as
> vision or image or camera etc to discover the complete list.


Keywords would be nice to have in the manifest

> b) Similar to the sandbox discussion, we should have a way of not
> listing 'lab specific' packages on the main packages page, maybe a
> 'Don't list' Tag in the manifest?


We use <review status="experimental" notes="Under active development"/>
in the manifest to denote code that is in the "sandbox". Perhaps there
could be some standardization for the meaning of status codes.

Another idea is that there could be an element that could be added to
the manifest such as <index update="ignore"> and something like
<index update="86400"> to ask that your repo be indexed once a day.

or perhaps a .rosindexer_ignore file would be sufficient

> > This is a case where repository maintainers are going to have to tell
> > WG what you want us to support. One of the things we like about ROS is
> > that lets you exercise a lot of freedom in organizing your code, and
> > it also has a low barrier to entry for distributing code. We'd rather
> > not tamper with that by telling people how they have to organize their
> > code and raising the barrier to having your code distributed.

...
> > - Ken


While I appreciate the flexibility, I sometimes wish there were better
defaults or a suggested approach. This would not limit flexibility but
provide a suggested setup for people who don't want to spend significant
time thinking about the implications of their repository/stack/package
layout.

-Bill