What are the goals....

[Douglas Wilson]

I hope no one gets mad at me for asking such a question. I have seen as an observer all the good and bad with unity, I was there almost at the beginning of this great endeavor of Unity Linux and here when a new way was chosen by necessity. I knew or thought I knew the goal was to create a new distro, than a new base were others could build on and may did packages for Unity and some left as is inevitable in endeavors such as these.  

Now Unity uses Mandriva packages which seem more broken then working from watching the list. The last round of email correspondence though concerned me because it sounds that Unity is just going to in the future be absorbed into the Mandriva community spin off because of RPM5. So what are the goals of Unity Linux? Is Unity just in the end become a part of Mandriva Community or is Unity trying to blaze it's own path?

I know everyone is busy with life and it is not my intention to make my friends here mad for my email, I am just curious is all.

Doug Wilson
Former Editor in Chief of PCLOS Magazine :)

--
Use Linux

[Tobias Gerschner]

Hi Douglas,


I think you are spot on. Below a draft that I kept around, hesitating to send it off as a response to the initial email that you're referring to as well :) .

May I just ask, what are the set goals of unity?
And are these goals in line with what you are currently trying to achieve?
Are the steps and the arrangements that you currently have in place the best you can do to achieve these goals?

From my own observations it's necessary -in order to sustain contributing to an open source project for a long time - that you are egoistic enough to focus on what's most important to you. And also acknowledge that this choice will also have its downsides - and learn to live with it.

I've been down the road of trying to please everyone who came along and it burns you out quickly. It's also not very rewarding, even if you're completely altruistic - you simply cannot please everyone. So, start with making yourself happy, because that work is most likely your best and <after> having achieved that see what you can offer along the way.

Either way, good thing to see that this long overdue discussion taking place.

Best Regards


--

Tobias

--
Tobias Gerschner
Yoper Linux - www.yoper-linux.org

Knowing is not enough; we must apply. Willing is not enough; we must do.

[Matthew Dawkins]

On Sat, Mar 24, 2012 at 9:27 PM, Tobias Gerschner <tobias.g...@gmail.com> wrote:

Hi Douglas,


I think you are spot on. Below a draft that I kept around, hesitating to send it off as a response to the initial email that you're referring to as well :) .

May I just ask, what are the set goals of unity?

My personal goals have been day one to create a distro based on community driven tools. Ie:
rpm5.org not rpm.org
smart not urpmi, apt, zypper, yum
createrepo not sythesis, createbase

The last two I have kind relaxed on, but I still like smart more than any other package manager out there.
 

And are these goals in line with what you are currently trying to achieve?

I'd say somewhat so. I'm still helping out with rpm5 releases and keeping development alive around the largest rpm5 package base.
Smart devel maybe a little on hiatus, but honestly I don't have too many complaints about it anymore.
 

Are the steps and the arrangements that you currently have in place the best you can do to achieve these goals?

Yes, we are not doing dual development anymore and I would say both Unity and Mandriva have benefited from this.
 

From my own observations it's necessary -in order to sustain contributing to an open source project for a long time - that you are egoistic enough to focus on what's most important to you. And also acknowledge that this choice will also have its downsides - and learn to live with it.

This is kinda where I'm at. I don't wanna do pure maintenance anymore. If I touch a package, I make sure I leave it in the best possible condition that I can. I try to make things easier to maintain and extend in the future.

I've been down the road of trying to please everyone who came along and it burns you out quickly. It's also not very rewarding, even if you're completely altruistic - you simply cannot please everyone. So, start with making yourself happy, because that work is most likely your best and <after> having achieved that see what you can offer along the way.

I think Jman helped me learn this. So like you say. Look at the sub 100MB CLI. I like it. Unity is fast, small and I'm not maintaining an unmaintainable list of packages :D
 

Either way, good thing to see that this long overdue discussion taking place.

I agree.
So how can I get you all on board?
 

Best Regards


--

Tobias

Matt
 

[Tobias Gerschner]

Well, I think we're both on paths that are very near to each other and also show kinda in the same direction. Personally I believe that <we> need better tools. So my pure and sole focus has been on those tools - neglecting everything else. With the little time I have, I have to be as precious as I can.

And at some points in the future those paths may become one. Personally I am looking at getting at getting a build system in place, that allows reusing rpm recipes between a larger number of distros while maintaining the ability to follow different paths goal etc. Unity is very far away from that particular (personal) goal.

[Matthew Dawkins]

On Sat, Mar 24, 2012 at 10:12 PM, Tobias Gerschner <tobias.g...@gmail.com> wrote:

Well, I think we're both on paths that are very near to each other and also show kinda in the same direction.

Well, ever since I contacted you, gmk, pok, bero and jbj predating the pclos split.... I've always believed that.
 

Personally I believe that <we> need better tools. So my pure and sole focus has been on those tools - neglecting everything else. With the little time I have, I have to be as precious as I can.

I agree wholeheartedly  there. It starts with just accepting the work flow and base of an already established base and embracing and enhancing it. The base is where all of us have differed from the beginning. I tried with caos+yoper if you remember.

And at some points in the future those paths may become one. Personally I am looking at getting at getting a build system in place, that allows reusing rpm recipes between a larger number of distros while maintaining the ability to follow different paths goal etc. Unity is very far away from that particular (personal) goal.

Well the goal of ubuild is a start.
But the rpm recipes is a common goal. I think the mdv-ized implementation is what puts off most non-mdv fans.
I've done a lot of work to push 2 changes in Mdv devel that I thought would attract you Tobias/Yoper.
1st usage of pkgconfig provides
2nd removal .la files

[Raphaël Jadot]

2012/3/25 Douglas Wilson <hist...@gmail.com>:

> So what are the goals of Unity Linux? Is Unity just in the end become a part
> of Mandriva Community or is Unity trying to blaze it's own path?

Well, it was not in my goal to give the impression that Mandriva would
eat UL, I've always introduced UL as an independant project.

What should be kept in mind is that Mandriva SA, the company, will
simply leave Mandriva desktop, the distro. No more no less. Maybe then
it will die, maybe it will become something else, but it will not be
able to eat anything. Mandriva SA will only be a sponsor and give some
infrastructure, but Mandriva SA will not have any more the power to
decide anything.

As I said, UL can simply leave, but it can also be leader in next
distro development and still be independant at the same time. Maybe it
could be the root of the next distro. Maybe... many things. What will
be the future, it will be different, so trying to make things move in
a good direction can be interesting.

--
RJ

[Kaleb]

2012/3/25 Raphaël Jadot <ashle...@hodo.fr>

Well, it was not in my goal to give the impression that Mandriva would
eat UL, I've always introduced UL as an independant project.

What should be kept in mind is that Mandriva SA, the company, will
simply leave Mandriva desktop, the distro. No more no less. Maybe then
it will die, maybe it will become something else, but it will not be
able to eat anything. Mandriva SA will only be a sponsor and give some
infrastructure, but Mandriva SA will not have any more the power to
decide anything.

As I said, UL can simply leave, but it can also be leader in next
distro development and still be independant at the same time. Maybe it
could be the root of the next distro. Maybe... many things. What will
be the future, it will be different, so trying to make things move in
a good direction can be interesting.

I believe you are trying to give us as much opportunity as possible, so first of all I want to thank you for that.

Second, you speak of UL possibly becoming a leader in MDV development. Let me be clear on what I want. I want a stable, reasonably up to date, rolling release distro that is easy to remaster. If MDV can provide that, then fine, I'm in it for the long haul. If not, then I'm not sure what I'll do.

[Raphaël Jadot]

Le dim. 25 mars 2012 19:27:18 CEST, Kaleb a écrit :

> Second, you speak of UL possibly becoming a leader in MDV development. Let
> me be clear on what I want. I want a stable, reasonably up to date, rolling
> release distro that is easy to remaster. If MDV can provide that, then
> fine, I'm in it for the long haul. If not, then I'm not sure what I'll do.
>

I completely understand your anger and weariness. For my part, if
mandriva can't be a stable base, i'll also simply stop using it, and of
course it will be absolutly normal that UL stop to be based on it :)

Nonetheless It's a matter of days (jmcroset told me he want to make a
first annouce this week) before we know what will happen to Mandriva.

As I told, he's also very opened to all your desires, ideas, so

for my part, i think a serious discussion should be initiated about RPM
problems (because it's a RPM problem, so not only a Mandriva problem,
even though the main RPM contrib are working for Mandriva/Rosa) I'm not
a dev, so I don't know all the technical part, but I think it's better
to build together and learn for mistakes (and fix them) before moving
to another solution (LFS or something else).

Anyway, whatever will be the choice of UL team, it will have my strong
support, as much as I can help bringing something.

[dillj]

To achieve this kind of stability, it'd be logical to point our smart channel configs at release repos from mandriva. This could then be updated everytime we release in accordance with the current state of mdv repos. 

As pkg developer, my main interest would be centered on cooker as this is where the development takes place. 

However I think we can combine this nicely. 

My suggestion would be to do releases based on stable repos so that we can work out all bugs and ppl can work with stable software while releasing a dev-isos based on cooker for all 'breaking-edge' folks etc.