(old)Puppy Linux Discussion Forum

Meeting is 'tonight' - early morning of Sunday
check your time
http://puppylinux.org/wikka/meeting30May2010

00:00 New York 05:00 London 14:00 Sydney
[Chair - Raffy] [Deputy Chair - Lobster]

Agenda

• Meets, greets, intros
  Reports and insights from active Puppys as available
  Deep Thought 4.2
  Lucid Puppy 5.0
  IRC
  News Site
  Wiki
  Software depositories
  Developer Discussion
  Woof or compile from scratch for Puppy 6
  Code sharing
  Any other business
  creating fault/problem template
  Set next meeting

http://puppylinux.org/wikka/meeting30May2010

Thanks, Lobster, for naming me as "Chair". I can try to moderate the process, with lots of help from IRC regulars. I would leave the discussion of content to the co-chairs.

It would be good to start the discussion here so that only items for consensus-building are taken up during the meeting.

In my view, some of the important concerns are:

1. How an official new release is built and decided - insights are needed from the experiences of "Deep Thought" 4.2 and "Lucid Puppy" 5.0. Playdayz and WhoDo will be good co-chairs here.

2. How a new official release is supported, and (vice-versa) how the usefulness of previous support infrastructure is maximized (despite the new features of the new release). Kirk and Jemimah will be good co-chairs here.

3. How a puplet is built and supported. Ttuuxxx and Tazoc will be good co-chairs here.

4. Files repository. Mark (MU) and Eric (caneri) will be good co-chairs here.

Readers are concerned that the many builds of Puppy can confuse users. The converse can be argued that the many builds of Puppy will be good for users, as they can pick the best choice for their setup. What may be important is how the builds are mapped (like a tree), how well each build is explained/understood, and how each build maximizes the available support infrastructure.

To give you an idea of the website support demands, puppylinux.org is already a business level account, but its usage still exceeds the limits by 150%. With a good mapping of the different builds of Puppy, this usage can be distributed to other existing support facilities.

Of course, a big level of support is provided by this Forum - thanks, John.

Some points are already "given" when I wrote this, like (1) Barry leads the development of the core "Puppy", (2) everyone is free to do development work in his/her own way, and (3) the community does development as well as user support.

Thanks, Lobster, for naming me as "Chair". I can try to moderate the process

Thanks Raffy.
Welcome back to cat herding
. . . truly never been done . . .

I have reduced the meetings to one
Developers welcome
Next meeting will be set here or at the end of the actual meeting

How an official new release is built and decided - insights are needed from the experiences of "Deep Thought" 4.2 and "Lucid Puppy" 5.0.

At the moment Barry decides - based I think on the potential and possibility
So Puppeee is near to becoming an official release being at RC2 stage

5.0.1 is now the official Puppy
and 5.1 is being worked on

Playdayz and WhoDo will be good co-chairs here.

I am happy to second them as chair or deputy chair as they prefer
and step down - or become secretary (secretary still wanted)
I have written to Coolpup to give a report on the Wiki - if he does not
I'll do it - if time or interest

If anyone wants to report on how the IRC channel is now set up
- that would be a useful insight

Raffy / Lobster, may I place this as an agenda discussion Item with the developers?

I want to put it out in advance as i'd like it to have / need some pre-thought.

For those who haven't read this, this first bit is copied from page 8 of Puppy 5.1 to 6.
----------------------------------------
I hope to "attend" if I may, to have a Developer discussion on this:-

I'd also like to expand and put my idea forward that it's time the puppy system grew up and had two component stages.

Puppy 6 may be that time!

The first stage is the core, that part of puppy that should be a "bare-bones of bare-bones". Consisting of the system and basic utilities only, covered under the menu as filesystem, utility, and above, but not including any applications other than a tiny (basic non-invasive) browser. It should be java aware by default. All configuration components should be working towards being interfaced via the browser, then many help directions and messages can be available. I believe there is even a text type browser out there somewhere, that can be used for the pre-x stages.

The second stage consists of applications and packages of applications. The only requirement is that they must be inclusive. In other words, would have no component that is reliant on another package or set of packages. I would only put a restriction on this group of a requirement of having an individual .pet for each application in their .sts package set. This is so those that only want one application can have access to it.

I think you would find that many groups would willing to work to this matrix, as they wouldn't have to supply the core level part, only what is required to meet their version.

i'll add more later

scsijon

------------------------------------------------------------------------------
Continuing:-

My idea is that new users can already download/offcd/buy the standard Puppy as a "look-see and try" and do.

Eventually those that stay (as we all have) will want something slightly different;
some Applications that are in the basic package removed, as they don't use them and perceived that they "are in the way";
some Applications that they have added as pup's or pets, and suit what they want to keep and use for the long run;
or someone else has created a specialist mix .sts package they want to add to trial before permanently adding.

Currently, to "special" a version is still not easy. They can add, but removing "junk" is a problem.

If those at this second stage, can have a "barebones" or basic package plus addon's, they will be able to create a "special version" of their own and to closer to what suits them specifically, and I believe be happy to stay rather than look elsewhere.

It should also make it a little easier to change and update the base when necessary or expedient.

It should also simplify that "specialist .sts package" as the maintainer would have a "standard base" to work with.

And, it all can be done with "minimum of thought" by the adverage user without the need to delve deep in the system.

I believe that they will, spread the word on this capability of Puppy, and that can only be to the good.

The alternative model, as far as I can see, is to have it "mandatory" that each official version created "must" have a barebones version released in parallel. And I don't like the words in quotes in the previous sentance when talking Linux, it's the wrong language!

Anyway, that is what I'd like to "talk" through with the developers.

scsijon

-----------------------------------
added from another post

8-bit wrote:sharibeth,
I agree to a point.
Make a bare-bones Puppy, but then install all the applications from PET packages and from there, make the final ISO.
That way, the ISO/ Puppy would come with it's normal applications, but they could be uninstalled with Puppy Package manager and replaced with ones of ones choosing.
Of course some applications rely on others being there for dependencies. So it would be possible to break Puppy.

Mind you, it is just a thought.

thank you 8bit & sharibeth,

I do like that thought though, it would cover the idea of the structure I am trying to get across.

standard cd for new and happy to work with users;
user modified version with unwanted "bumf" removed for later when you have worked out what you actually want.

scsijon

Raffy / Lobster: Could this be discussed please:

Can there be made available for all a basic set of instructions for those trying a version to help with fault/problem identity simplification.

thanks
scsijon

Have added this to the wiki

Any other business
creating fault/problem template

Meeting is 'tonight' - early morning of Sunday
check your time
http://puppylinux.org/wikka/meeting30May2010

"Barebones" approach sounds good, but in practice quickly becomes a mess...mostly do to dependencies and updating. Staying with well integrated basic apps that all work together has the advantage of introducing apps that the user may not have known about. Pets are then a means for specific usage (I like vim, tcltk, skype, etc). Having a few unused apps isn't a problem with a distro of under 120 mg, even on older computers. Often I've found myself actually using them.

Regarding creating barebones I read this post on Barry's Blog the other day:
http://bkhome.org/blog/?viewDetailed=01620

"Barebones" approach sounds good, but in practice quickly becomes a mess.

before the objective is missed
a bare bones is not for the end user to use or have to use
or even think about using or imagine to use just to be clear
the sky won't fall for the end user because they never see this stage

a stripped down base is what developers could add to the "starting point" test it and
make PACKAGES to update the base between the developers
a happy go between that allows a lot of creativity to select packages
that are customized and not found anywhere else because they are hand crafted

by the people that know what they built and they can test it to perfection before
adding it officially to the base



when good ideas are found in official or special versions "customized versions "

they are added to a base by the developers to make their own flavor

what is happening now and has happened for years is a MESS
everyone adding to an ISO and not having a clue how to make a simple package of their own work to share with the other developers
which is a lazy non productive re inventing the wheel every version dead end
abysmal


1.) for a clear idea of what I am talking about anyone could take any package from any distro slackware ,slitaz ,debian, rpm,(* bsd use their sources) etc.......
linux compatible and make it work on puppy

2.)and I for one can't easily sort out the spaghetti mess of trying to take some features
duct taped super glued into a "customized version" or the latest new thing that went official so sharing is not an option with this behavior
and this is open source linux if anyone has forgotten



so if the developers ever plan to share it has to start with packages (build scripts are the best but wake me up I'm dreaming if puppy would ever take that step)
the package format is trivial at this stage
but important in the building stage

any team work requires a team
being a just a player shows your skills on solo projects
I maintain a solo project so to speak also
and so do others here in the forum
I think there is a great waste of resources
becausee of the mind block called official whatever

if a team idea started nothing could stop it

what I see mostly what type of user puppy attracts is the desktop user
they want to use ubuntu but don't know how to install it and it runs slow on their computer

so they use puppy ,then when they figure out how to install in ubuntu
they leave or they can't leave because it runs too slow

so they decide to bloat puppy and make it run slow

now think about this where are we going





--------------------------------------------
To err is human; to forgive is divine."
-- Alexander Pope


Joe

big_bass wrote: ......hand crafted by the people that know what they built and they can test it to perfection before
adding it officially to the base

sounds good...thanks for the clarification

Lobster-

Not that you asked for my opinion, but ...

I think there are three things that keep this otherwise-excellent Linux distribution from becoming much more useful, and thus much more popular:

1. There are far, far too many different derivatives and versions of Puppy, so many that "Puppy Linux" doesn't mean much. This makes solving problems really difficult, as the exact combination of derivative, version, and hardware is often unique.

2. Finding information on Puppy is really difficult. Forum searches return far too many hits for any common problem, and information outside the forum isn't organized at all. Of course, this is exacerbated by #1.

3. The package repository situation is just abysmal. Finding anything beyond the few packages listed in the package installer involves searching the forum, and then #2 comes into play, exacerbated by #1.

I like Puppy a lot. I've got it running pretty well on my laptop, but it's been a struggle. It deserves to be more popular.

Solving the three problems above doesn't have the glamor of producing a custom-tuned Puppy derivative with one's own name on it, but it is what will make Puppy great.

Talk about that in the developer meeting!

-Shel

Joe

I think Puppy is chaotic
it makes us very innovative
and means there is duplication and divergence

Each Puplet or woof expression has something
I stopped producing Linux Tmxxine
because it was so easy to produce a Puplet
that there was no need for a few different apps
and background to have its own version
http://www.puppylinux.dreamhosters.com/ ... reate.html

Many, many projects fail or are abandoned . . .
Many more invigorate, influence and improve

Some of the presently most creative Puplets and wooflets
(for example Slaxer_pup)
deserve more inter developer co-operation
- that is the reasoning behind meetings
which hopefully we can do once a month or so . . .

We'll see
Hope you can make it

Lobster

I am one hour before New York time so that's 4:00 am
Sunday morning I f I went fishing at that time I could sneak out of the house undetected

I will be able to offer an exchange of ideas
but at a later time in the day

I believe that a balance
is in need most people feel it in the air
thats why meetings are called in the first place
not everyone will or can express themselves
well some steam is usually released at first
because ideas can get bottled up over time

but the first step to fixing a problem
is to know what the problem is

some people focus on problems and get run down

some people focus on solving the problems
and have a better outlook and some wins along the way


the only difference is how you handle the information
sort through it with a lot of salt ( in Mexico with lemon and hot sauce too )

Joe

big_bass wrote:I will be able to offer an exchange of ideas
but at a later time in the day

Don't count on it, Joe. The inmates have taken over the IRC asylum again today. Some nick-swapping loon has been calling himself/herself Lobster, WhoDo, etc and so I gave up... Sad state of affairs. We need a private, invitation only channel to keep the nut cases from spoiling the discussion IMHO.

Can we set something like that up?

This may sound kinda stupid but why not try a private Yahoo group. It beats the hell out of IRC.
At least it did back in my very early years on ebay when things needed to be kept private.
glenco = <> wingrunted

Well, the meeting has started. Guess I'll see how it goes.....

jemimah wrote:Can we set something like that up?

You hung in there well, jemimah. I think a private group is essential for further discussion. At least the trolls can be kept out by requiring a login from the group owner.

Ed, can you organise that please??? Advise devs by PM and accept requests for access by users seeking to give input. We're not trying for secrecy ... just sanity!

here's the notes
ttuuxxx

To answer my own question from my previous post, the meeting was an adventure.......