Dag Wieers wrote:
But my (outside) stance on the LiveCD is that it should give the best achievable experience possible for people when trying CentOS. To me that includes adding drivers that are available in other repositories (which are missing from upstream, including wireless firmware, etc...)
Sounds good. Are there any legal issues around the wireless firmware files ? If its a case of just talking to the providers of these firmware to get permission - thats something which we could potentially do.
I do agree that as soon as we leave the deliver-what-upstream-has path, we may open a can of worms (do we also want to fix known bugs ? replace upstream software ? legality ?), so we have to decide what is desirable, what is possible and where the project's effort ends.
This is an interesting subject to me, and I think we ( dag and I ) have spoken about this many times in different venues. Exactly where does the line get drawn and exactly what is considered 'CentOS brandable' and what isnt. Since the topic is now being actively discussed I'll start a new thread just focused on this one idea and we can thrash it out there.
What are the various goals of the LiveCD project ?
Rather than goals, can we narrow it down to 'use cases'. I think there is only one goal - to build either one or many livecd's based on and around the CentOS distro. If we focus on use-case, it allows wider and more elaboate cases to come in, and we can then look at mechanisms to solve those issues in some sort of a priority order.
Perhaps I am being pedantic, but it will help in the long run.
I am sure we can list more applicable use-cases for this project. And in the end we have to agree which fit the LiveCD project (eg. whether we make an official CentOS LiveCD and a seperate CentOS-plus LiveCD) and what the main priorities are.
Another thing I'd like to propose is that we seperate out the livecd 'project' into upstream for the CentOS provided livecd(s). That would allow a separate entity and therefore a different crew to focus only on the tooling components and perhaps work with parallel branches.[1]
Also, to stimulate the LiveCD development there should be no reason why we couldn't release a LiveCD 5.3.1 or LiveCD 5.3.2 if there are compelling reasons to make a new respin (and update the mirrors).
Nope, but it would be 5.3-1 and 5.3-2! So as to not confuse with upstreams 5.3.1 and 5.3.2 etc. Although most people who dont really know the whole story will get confused anyway
Where do we want to go with this and what is realistically achievable ?
Where I would like to see this go, and this is my personal take on this is:
- Send out a call for user-cases - Setup a feedback process, email / lists work well - Nominate a person ( Patrice ? ) to be the 'lead' on the efforts - Setup whatever infrastructure we need to make this happen - Build foo!