← Back to team overview

lubuntu-desktop team mailing list archive

Re: lubuntu - minimal install or full featured lightweight distro


On Tue, May 19, 2009 at 2:55 PM, C David Rigby <c.david.rigby@xxxxxxxxx> wrote:
> On Tue, 2009-05-19 at 14:20 +0800, Mario Behling wrote:
>> Dear all,
>> I would like to start a discussion about the goals of lubuntu, outline
>> what I believe and get your feedback. This is an important question as
>> we should include packages according to the goals.
>> The usage case I see for lubuntu is a situation like in emerging
>> countries where Internet availability or bandwidth is often limited
>> and the power of CPU etc. is low. I was recently working with the
>> community in Afghanistan. People usually download software through one
>> person, for example in a company overnight, and then share it through
>> a USB stick. So the size of the image is important but not the most
>> important. To install and upload packages on a single computer is more
>> difficult as bandwidth of personal users is limited. It is possible to
>> use Internet, messengers and maybe listen to some music, but
>> downloading larger updates and making net-installations is an issue.
>> Looking at this case, I would like to see lubuntu as a distribution
>> offering more applications than just a minimal install. A lubuntu
>> image could be downloaded by one person and then easily distributed
>> among many people having different usages cases. Currently we already
>> have a minimal install of LXDE in Debian.
>> If here is a choice among different solutions, we should choose the
>> lightweight solutions. However, we also need to look at the support of
>> translations for example to make lubuntu a widely adapted distro. It
>> is important to include applications that support many languages.
>> Defining these goals, it becomes clear - there are many compromises.
>> But as we go on with the project and will release new versions, we are
>> always able to adapt our decisions.
>> The question of lubuntu is:
>> 1. Either we want to make it a minimal install or
>> 2. we want to make it a lightweight alternative to full featured distros.
>> I vouch for the second.
>> What do you think?


many questions and more questions will come up. - It might be a bit
much to solve everything via email. Would be good if the group of
people who were interested at the meetings we had in Singapore get
together and discuss those face to face first as well for example. I
will do the same with people in Germany.

> The lack of ready access to the Internet does make #1 above less useful
> for a desktop user. So, I agree more with #2, given your stated goal.
> Some questions & thoughts come to mind:
> * What is the minimal image (set of packages) that can be called Ubuntu?
> In other words, what functionality of Ubuntu must we have? Related to
> that:
>   * Is it Canonical that decides what the minimum set of packages is in
> order for it to be called an official *buntu derivative?
>   * Is this documented somewhere, in a fashion similar to the Debian
> Policy Manual?
>   I admit to ignorance on the above questions.

I dont know neither. We could also discuss that over time. At the
moment there are many tasks where we dont need immediate discussions
with Canonical, e.g. we need people making new deb-packages of new
LXDE components and people working on artwork.

> * As a stage in development, should we do #1 anyway?

We can, but I am not focusing on this right now.

> If it minimal
> install is defined as Ubuntu server + LXDE or a subset of LXDE's
> components, then that is pretty straight forward. If we are talking
> about reworking Ubuntu desktop, or starting at a more basic level, then
> a solid minimal installation should be the starting point for further
> development. Even if what we all we are doing is subtracting packages,
> and thus functionality, that we decide we do not need, a mistake here is
> painful later when the system lacks something needed by the users.
> I just did a quick web search on 9.10 Karmic Koala, and I see that the
> server will have features like cloud computing. Perhaps that is not
> appropriate to the goals stated above? In that case, we may need to
> carefully define the base on which we will build.
> * Given our lead time, should we start tracking 9.10 as a reference,
> rather than 9.04?

Our work at the beginning is independent of releases - packages and
artwork, which packages. We will see how fast we progress.

> * If we expand our goal to include other situations (e. g., installation
> to a VM, small installation to a USB flash, etc.) then #1 becomes more
> useful in and of itself.

People can just do that. USB stick can be made by anyone from an ISO.
We can think about this later more.

>> Best,
>> Mario
>> PS: The option to have a choice of software to install is a good idea
>> for following releases as well, but requires more competences that we
>> need to build up, I guess.
> The situation you describe makes me wonder about updates to existing
> installations. If we assume that a significant portion of installed
> Lubuntu systems will be used offline, how do we propagate updates to
> them?

People can use a CD for updates/upgrades to the next release. Smaller
updates are usually simply skipped. - It will still work much better
than other OS..

-- parts deleted --



> Regards