[opensuse-project] openSUSE Strategy Discussion: The Linux Distribution Platform Strategy
Hi all! Today we continue with public discussions about strategy proposals submitted by you, our beloved community. The first one is the "Linux Distribution Platform Strategy": ---8<------8<------8<------8<------8<------8<--- === openSUSE, the Linux distribution platform === == Goals == The goal of openSUSE as a project is to provide a platform for distributing Linux and software running on Linux to a wide range of users. This platform consists of tools for creating software distributions, the openSUSE distribution as base and reference implementation, and the community supporting the tools and the distribution. On top of the platform the openSUSE universe consists of more specific distributions, which make use of openSUSE infrastructure and technology. Examples are SLES, MeeGo, openSUSE Education, KDE and GNOME live systems, and could also be for example developer or cloud oriented distributions and more. So openSUSE provides a way for subteams to address specific user groups and needs. openSUSE also provides means to distribute software independent of the environment of the user to spread Linux based software and make software easily available for use in the openSUSE distribution and systems based on it. The openSUSE distribution acts as a reference distribution, providing an environment for testing the used technology, a stabilizing ground for common components, and a real-life use case for applying technology and distributing Linux software. It's targeted at technically interested users, including programmers and system administrators. It has a focus on good user experience and making technology available to end users. It doesn't target users with highly specific technical needs. == Activities == = Essentials = * Provide stable set of supported core packages distributions can build on * Broad hardware support of components and platforms * Provide tools for package and distribution building and testing (e.g. openSUSE Build Service) * Provide common building blocks for distribution, e.g. installer, configuration tool, maintenance tools, development tools for web, native and other applications, and more * Provide home for overall community and specific openSUSE teams, e.g. bug tracker, wiki, mailing lists, collaboration tools * Create the official openSUSE distribution as reference implementation * Enable, support, and collaborate with specific teams to create their own distributions * Enable and support, and collaborate with upstream developers to build and distribute their software on openSUSE = Good to have = * Provide wide variety of packages for further use * Community for user support * Work on standards which make it easier to mix components, e.g. free desktop standards * Collaboration with other distribution platforms = No focus = * Directly providing a polished distribution for non-technical end users * Bleeding edge technology ---8<------8<------8<------8<------8<------8<--- -- Best Regards / S pozdravom, Pavol RUSNAK SUSE LINUX, s.r.o openSUSE Boosters Team Lihovarska 1060/12 PGP 0xA6917144 19000 Praha 9 prusnak[at]opensuse.org Czech Republic -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
Pavol Rusnak wrote:
Hi all!
Today we continue with public discussions about strategy proposals submitted by you, our beloved community. The first one is the "Linux Distribution Platform Strategy":
FYI, it is the height of the European summer holidays, it might not be the best time to start such a discussion. -- Per Jessen, Zürich (15.4°C) -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Monday 26 July 2010 22:00:55 Pavol Rusnak wrote:
Hi all!
Today we continue with public discussions about strategy proposals submitted by you, our beloved community. The first one is the "Linux Distribution Platform Strategy": [...]
All in all a great proposal, I have some clarifying questions and comments.
- Broad hardware support of components and platforms
So, what is the stance on Sparc, HPPA, Arm - do we want all of these as core part of the project?
- Provide tools for package and distribution building and testing (e.g. openSUSE Build Service)
- Provide common building blocks for distribution, e.g. installer, configuration tool, maintenance tools, development tools for web, native and other applications, and more
- Provide home for overall community and specific openSUSE teams, e.g. bug tracker, wiki, mailing lists, collaboration tools
We currently do not provide a complete home, e.g. there's no source repository. I don't expect that the proposal wants to change this but like to make it clear that we cannot provide a "complete home".
- Create the official openSUSE distribution as reference implementation
The "as reference implementation" strikes me as a bit odd and I'd like to see this a bit better explained. Reference for what? Also, isn't the distribution more than just a reference? It's a usable distribution... Andreas -- Andreas Jaeger, Program Manager openSUSE, aj@{novell.com,opensuse.org} Twitter: jaegerandi | Identica: jaegerandi SUSE LINUX Products GmbH, GF: Markus Rex, HRB 16746 (AG Nürnberg) Maxfeldstr. 5, 90409 Nürnberg, Germany GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563C C272 A126
On Tuesday 27 July 2010 12:50:48 Andreas Jaeger wrote:
- Broad hardware support of components and platforms
So, what is the stance on Sparc, HPPA, Arm - do we want all of these as core part of the project?
I think if somebody is doing the work to support one of these platforms it should be welcomed by the project, but I don't think that any of these platforms should be an explicit part of the strategy. We shouldn't rule out any platforms either, of course.
We currently do not provide a complete home, e.g. there's no source repository. I don't expect that the proposal wants to change this but like to make it clear that we cannot provide a "complete home".
Yes, the home for handling the distribution parts is important, but a full upstream code hosting platform is out of the scope of openSUSE, I think.
- Create the official openSUSE distribution as reference implementation
The "as reference implementation" strikes me as a bit odd and I'd like to see this a bit better explained. Reference for what?
Also, isn't the distribution more than just a reference? It's a usable distribution...
It's a reference for what you can do with openSUSE, so it's of course a usable distribution, but it's not the only possible distribution you can build using the tools, packages and other infrastructure of openSUSE. Reference is meant as a standard example what the openSUSE platform is capable to do. -- Cornelius Schumacher <cschum@suse.de> -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Tuesday 27 July 2010 15:01:28 Cornelius Schumacher wrote:
On Tuesday 27 July 2010 12:50:48 Andreas Jaeger wrote:
- Broad hardware support of components and platforms
So, what is the stance on Sparc, HPPA, Arm - do we want all of these as core part of the project?
I think if somebody is doing the work to support one of these platforms it should be welcomed by the project, but I don't think that any of these platforms should be an explicit part of the strategy. We shouldn't rule out any platforms either, of course.
We currently do not provide a complete home, e.g. there's no source repository. I don't expect that the proposal wants to change this but like to make it clear that we cannot provide a "complete home".
Yes, the home for handling the distribution parts is important, but a full upstream code hosting platform is out of the scope of openSUSE, I think.
So it should focus on integrating very well with platforms like gitorious, sourceforge and launchpad?!?
- Create the official openSUSE distribution as reference implementation
The "as reference implementation" strikes me as a bit odd and I'd like to see this a bit better explained. Reference for what?
Also, isn't the distribution more than just a reference? It's a usable distribution...
It's a reference for what you can do with openSUSE, so it's of course a usable distribution, but it's not the only possible distribution you can build using the tools, packages and other infrastructure of openSUSE. Reference is meant as a standard example what the openSUSE platform is capable to do.
So the focus does not lay here anymore, but on developing the tools to let developers build and distribute their software. The fact openSUSE has a bunch of reference implementations is secondairy. -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Wednesday 28 July 2010 10:46:32 Jos Poortvliet wrote:
On Tuesday 27 July 2010 15:01:28 Cornelius Schumacher wrote:
Yes, the home for handling the distribution parts is important, but a full upstream code hosting platform is out of the scope of openSUSE, I think.
So it should focus on integrating very well with platforms like gitorious, sourceforge and launchpad?!?
This definitely is an important part. As Rajko said in another email in this thread this is an area which currently lacks and should be improved.
It's a reference for what you can do with openSUSE, so it's of course a usable distribution, but it's not the only possible distribution you can build using the tools, packages and other infrastructure of openSUSE. Reference is meant as a standard example what the openSUSE platform is capable to do.
So the focus does not lay here anymore, but on developing the tools to let developers build and distribute their software. The fact openSUSE has a bunch of reference implementations is secondairy.
I wouldn't call it secondary. I think having a strong distribution still should be a primary focus of the openSUSE community as a whole. This is where it's proven, if all the platform work actually works, so we definitely need it. -- Cornelius Schumacher <cschum@suse.de> -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
Andreas Jaeger wrote:
- Create the official openSUSE distribution as reference implementation
The "as reference implementation" strikes me as a bit odd and I'd like to see this a bit better explained. Reference for what?
From wikipedia: In the software development process, a reference implementation is the standard from which all other implementations, with their attendant customizations, are measured, and to which all improvements are added.
Also, isn't the distribution more than just a reference? It's a usable distribution...
ntpd is the reference implementation of the ntp protocol, but it is also a very useful product. -- Per Jessen, Zürich (15.4°C) -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
Per Jessen wrote:
Andreas Jaeger wrote:
- Create the official openSUSE distribution as reference implementation
The "as reference implementation" strikes me as a bit odd and I'd like to see this a bit better explained. Reference for what?
From wikipedia: In the software development process, a reference implementation is the standard from which all other implementations, with their attendant customizations, are measured, and to which all improvements are added.
The reference implementation is the measuring stick that everybody else uses. "Are we as good as or better than openSUSE?" To me it means it means we need to excellent in <whatever we define as core>, good in <what most users would like to do with opensuse> and reasonably suitable for <everything else>. I understand that there is no current standard definition to strive towards, but that is our opportunity to set it! -- Per Jessen, Zürich (15.4°C) -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Tuesday 2010-07-27 19:41, Per Jessen wrote:
- Create the official openSUSE distribution as reference implementation
The "as reference implementation" strikes me as a bit odd and I'd like to see this a bit better explained. Reference for what?
From wikipedia: In the software development process, a reference implementation is the standard from which all other implementations, with their attendant customizations, are measured, and to which all improvements are added.
The reference implementation is the measuring stick that everybody else uses. "Are we as good as or better than openSUSE?"
The reference implementation is what's obsolete after a few years usually. Can't let that happen to openSUSE, happened with Debian, sort of. -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
Jan Engelhardt wrote:
On Tuesday 2010-07-27 19:41, Per Jessen wrote:
- Create the official openSUSE distribution as reference implementation
The "as reference implementation" strikes me as a bit odd and I'd like to see this a bit better explained. Reference for what?
From wikipedia: In the software development process, a reference implementation is the standard from which all other implementations, with their attendant customizations, are measured, and to which all improvements are added.
The reference implementation is the measuring stick that everybody else uses. "Are we as good as or better than openSUSE?"
The reference implementation is what's obsolete after a few years usually. Can't let that happen to openSUSE, happened with Debian, sort of.
It can't be obsolete and be a refimpl at the same time, it's as easy as that. When the standard(s) move, the refimpl has to lead the way. Anyway, because we don't actually have a standard, we have to set it - if we can't meet our own standards, we've lost the game before it starts :-( -- Per Jessen, Zürich (15.4°C) -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On top of the platform
On Monday 26 July 2010 22:00:55 Pavol Rusnak wrote: the openSUSE universe consists of more specific
distributions, which make use of openSUSE infrastructure and technology. Examples are SLES, MeeGo, openSUSE Education, KDE and GNOME live systems, and could also be for example developer or cloud oriented distributions and more.
This sounds a description of the status quo. What other derivatives do you have in mind that would actually represent a growth in overall openSUSE-derived distributions? How would you measure the potential success of this strategy? The only really popular derivative distributions that I'm aware of are the legally dubious ones that include all the binary drivers, codecs and firmware out of the box. I'm consciously omitting Debian-Ubuntu from this because the enormous amount of work invested to make Debian into Ubuntu makes this unlikely to happen to a reference openSUSE, unless MeeGo bases completely on us. If that's seen as a realistic aim for this proposal, we should be honest about it and rename it 'Hook Up With MeeGo' :). Will -- Will Stephenson, openSUSE Team SUSE LINUX Products GmbH - Nürnberg - AG Nürnberg - HRB 16746 - GF: Markus Rex -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
Am Mittwoch, 28. Juli 2010, 08:50:44 schrieb Will Stephenson:
On Monday 26 July 2010 22:00:55 Pavol Rusnak wrote:
On top of the platform
the openSUSE universe consists of more specific
distributions, which make
use of openSUSE infrastructure and technology.
Examples are SLES, MeeGo,
openSUSE Education, KDE and GNOME live
systems, and could also be for
example developer or cloud oriented
distributions and more.
This sounds a description of the status quo. What other derivatives do you have in mind that would actually represent a growth in overall openSUSE-derived distributions? How would you measure the potential success of this strategy?
The only really popular derivative distributions that I'm aware of are the legally dubious ones that include all the binary drivers, codecs and firmware out of the box. I'm consciously omitting Debian-Ubuntu from this because the enormous amount of work invested to make Debian into Ubuntu makes this unlikely to happen to a reference openSUSE, unless MeeGo bases completely on us. If that's seen as a realistic aim for this proposal, we should be honest about it and rename it 'Hook Up With MeeGo' :).
Will
I think Pavol was thinking more of a 'Spin' way of doing things as Fedora does, but in opensuse one could use the leverage of hazzlefree added obs repos without having it to put right into openSUSE:Factory or making an external repository with hard to maintain consitency to the mainline distribution. As Edu does it right now I think. For example Fedora tries besides the obvious desktop spins are hardware design and emulation platform: http://spins.fedoraproject.org/fel/ For more see http://spins.fedoraproject.org/ Problem with Fedora is you need to push these into the official repos afaik, with obs this could be done much more quickly, dump your special package group into it's own repo and use kiwi or suse studio to produce a livecd. But these pieces are mostly already in place imho, there is nothing much to do here beside get people to create these derivates. btw this sounds like a rehash of http://lists.opensuse.org/opensuse- project/2010-06/msg00694.html except it sounds more like keeping the current opensuse state as a useful bundle by itself already for regular desktop users. Karsten -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Wednesday 28 July 2010 08:50:44 Will Stephenson wrote:
On top of the platform
On Monday 26 July 2010 22:00:55 Pavol Rusnak wrote: the openSUSE universe consists of more specific
distributions, which make use of openSUSE infrastructure and technology. Examples are SLES, MeeGo, openSUSE Education, KDE and GNOME live systems, and could also be for example developer or cloud oriented distributions and more.
This sounds a description of the status quo. What other derivatives do you have in mind that would actually represent a growth in overall openSUSE-derived distributions? How would you measure the potential success of this strategy?
The only really popular derivative distributions that I'm aware of are the legally dubious ones that include all the binary drivers, codecs and firmware out of the box. I'm consciously omitting Debian-Ubuntu from this because the enormous amount of work invested to make Debian into Ubuntu makes this unlikely to happen to a reference openSUSE, unless MeeGo bases completely on us. If that's seen as a realistic aim for this proposal, we should be honest about it and rename it 'Hook Up With MeeGo' :).
If I understand the strategy correctly, this one is not about providing a base to develop spins on (like the 'base for deriviates' strategy) but a platform to distribute software with and on for a variety of distributions, targeting developers. So very much focussing on techies, software developers, thus, and not, as the base for deriviates-strategy, focussing on endusers (through the spins). In this strategy the spins are just reference implementations, which might be great in themselves but aren't the core focus of the community.
Will
--
Will Stephenson, openSUSE Team SUSE LINUX Products GmbH - Nürnberg - AG Nürnberg - HRB 16746 - GF: Markus Rex
-- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
Heya, I like this idea even though to me it seems like a rehash of the 'base for derivates' idea. Still I am wondering what will happen with the desktops? As a base you sure need to offer a desktop environment in these times, and people sure want to mix in gnome, kde and other gui applications, so will these be kept inside or outside of the 'reference'? Then most packages aren't really disjunct, so where do I put a library that is used by two project at once? Will I have to link my project to like 5 other projects so I can pull in all the dependencies? Will I just _link to the specific library and thus result in serious obs load? So if we keep it the current way inviting more 'spins' I'd be fine with that as these can mostly be kept completely seperate from the factory tree, but 'core packages distributions can build on' doesn't sound that way to me. Regards, Karsten Am Montag, 26. Juli 2010, 22:00:55 schrieb Pavol Rusnak:
Hi all!
Today we continue with public discussions about strategy proposals submitted by you, our beloved community. The first one is the "Linux Distribution Platform Strategy":
---8<------8<------8<------8<------8<------8<---
=== openSUSE, the Linux distribution platform ===
== Goals ==
The goal of openSUSE as a project is to provide a platform for distributing Linux and software running on Linux to a wide range of users. This platform consists of tools for creating software distributions, the openSUSE distribution as base and reference implementation, and the community supporting the tools and the distribution.
On top of the platform the openSUSE universe consists of more specific distributions, which make use of openSUSE infrastructure and technology. Examples are SLES, MeeGo, openSUSE Education, KDE and GNOME live systems, and could also be for example developer or cloud oriented distributions and more. So openSUSE provides a way for subteams to address specific user groups and needs. openSUSE also provides means to distribute software independent of the environment of the user to spread Linux based software and make software easily available for use in the openSUSE distribution and systems based on it.
The openSUSE distribution acts as a reference distribution, providing an environment for testing the used technology, a stabilizing ground for common components, and a real-life use case for applying technology and distributing Linux software. It's targeted at technically interested users, including programmers and system administrators. It has a focus on good user experience and making technology available to end users. It doesn't target users with highly specific technical needs.
== Activities ==
= Essentials =
- Provide stable set of supported core packages distributions can build on
- Broad hardware support of components and platforms
- Provide tools for package and distribution building and testing (e.g. openSUSE Build Service)
- Provide common building blocks for distribution, e.g. installer, configuration tool, maintenance tools, development tools for web, native and other applications, and more
- Provide home for overall community and specific openSUSE teams, e.g. bug tracker, wiki, mailing lists, collaboration tools
- Create the official openSUSE distribution as reference implementation
- Enable, support, and collaborate with specific teams to create their own distributions
- Enable and support, and collaborate with upstream developers to build and distribute their software on openSUSE
= Good to have =
- Provide wide variety of packages for further use
- Community for user support
- Work on standards which make it easier to mix components, e.g. free desktop standards
- Collaboration with other distribution platforms
= No focus =
- Directly providing a polished distribution for non-technical end users
- Bleeding edge technology
---8<------8<------8<------8<------8<------8<---
-- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
Heya,
I like this idea even though to me it seems like a rehash of the 'base for derivates' idea. Right, both proposals were pretty close and we merged them into this one. So,
On Wednesday 28 July 2010 09:26:33 Karsten König wrote: there is no derivates proposal anymore. best M
Still I am wondering what will happen with the desktops? As a base you sure need to offer a desktop environment in these times, and people sure want to mix in gnome, kde and other gui applications, so will these be kept inside or outside of the 'reference'? Then most packages aren't really disjunct, so where do I put a library that is used by two project at once? Will I have to link my project to like 5 other projects so I can pull in all the dependencies? Will I just _link to the specific library and thus result in serious obs load? So if we keep it the current way inviting more 'spins' I'd be fine with that as these can mostly be kept completely seperate from the factory tree, but 'core packages distributions can build on' doesn't sound that way to me.
Regards, Karsten
Am Montag, 26. Juli 2010, 22:00:55 schrieb Pavol Rusnak:
Hi all!
Today we continue with public discussions about strategy proposals submitted by you, our beloved community. The first one is the "Linux Distribution Platform Strategy":
---8<------8<------8<------8<------8<------8<---
=== openSUSE, the Linux distribution platform ===
== Goals ==
The goal of openSUSE as a project is to provide a platform for distributing Linux and software running on Linux to a wide range of users. This platform consists of tools for creating software distributions, the openSUSE distribution as base and reference implementation, and the community supporting the tools and the distribution.
On top of the platform the openSUSE universe consists of more specific distributions, which make use of openSUSE infrastructure and technology. Examples are SLES, MeeGo, openSUSE Education, KDE and GNOME live systems, and could also be for example developer or cloud oriented distributions and more. So openSUSE provides a way for subteams to address specific user groups and needs. openSUSE also provides means to distribute software independent of the environment of the user to spread Linux based software and make software easily available for use in the openSUSE distribution and systems based on it.
The openSUSE distribution acts as a reference distribution, providing an environment for testing the used technology, a stabilizing ground for common components, and a real-life use case for applying technology and distributing Linux software. It's targeted at technically interested users, including programmers and system administrators. It has a focus on good user experience and making technology available to end users. It doesn't target users with highly specific technical needs.
== Activities ==
= Essentials =
Provide stable set of supported core packages distributions can build
on
Broad hardware support of components and platforms
Provide tools for package and distribution building and testing
(e.g. openSUSE Build Service)
Provide common building blocks for distribution, e.g. installer,
configuration tool, maintenance tools, development tools for web, native and other applications, and more
Provide home for overall community and specific openSUSE teams, e.g.
bug tracker, wiki, mailing lists, collaboration tools
Create the official openSUSE distribution as reference implementation
Enable, support, and collaborate with specific teams to create their
own distributions
Enable and support, and collaborate with upstream developers to build
and distribute their software on openSUSE
= Good to have =
Provide wide variety of packages for further use
Community for user support
Work on standards which make it easier to mix components, e.g. free
desktop standards
Collaboration with other distribution platforms
= No focus =
- Directly providing a polished distribution for non-technical end users
- Bleeding edge technology
---8<------8<------8<------8<------8<------8<---
-- Michael Löffler, Product Management SUSE LINUX Products GmbH - Nürnberg - AG Nürnberg - HRB 16746 - GF: Markus Rex -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
All in all, seems like a good proposal, but a couple of comments that immediately come to mind for me are below: On Mon, 26 Jul 2010 22:00:55 +0200, Pavol Rusnak wrote:
The openSUSE distribution acts as a reference distribution, providing an environment for testing the used technology, a stabilizing ground for common components, and a real-life use case for applying technology and distributing Linux software. It's targeted at technically interested users, including programmers and system administrators. It has a focus on good user experience and making technology available to end users. It doesn't target users with highly specific technical needs.
To me, "good user experience" isn't very specific, and in some ways might come across as contradicting the "no focus" area of "non-technical users" - not that technical users can't have a good user experience, but that there's an implication of 'polish' that's more associated with a non- technical audience. I hope that makes sense, because it seems at the moment to not be easy to explain what I mean.
== Activities ==
= Essentials = [...]
- Broad hardware support of components and platforms
I agree with the comments Andreas and Cornelius made with regards to other architectures. It might be good to spell that out a bit more clearly that contributions for non-x86/x86_64 platforms (and whatever else is considered "standard") are welcome but not guaranteed to be provided (or something to that effect).
= Good to have = [...]
- Community for user support
I would consider this essential. Without a community for user support, there's no feedback mechanism to improve the reference distribution. It also seems odd to not have this listed as "essential" since providing the home for the community is considered essential. Put another way, if having a user support community isn't essential, then why invest the time and resources in providing the home for that community (not that I think providing the home isn't essential - I provide the counterexample to demonstrate this point only).
= No focus =
- Directly providing a polished distribution for non-technical end users
I think we do this now with the main distribution, and that to an extent, this is an essential goal as well (or at least "good to have") because it provides a basis for the derivatives to provide that polish. I guess this perhaps needs to be more specifically defined for me as to what isn't/wouldn't be included. I couldn't even say what I would take as implied as not being provided by this statement.
- Bleeding edge technology
With the pace of OSS development, what's considered bleeding edge today may not be in a month. It might be helpful to have some examples presented in this discussion to help clarify what might not be included and how a determination would be made as to when something was no longer considered "bleeding edge" and would be considered for inclusion. Of course, I'm thinking in terms a a general guideline, not a hard and fast rule, because obviously everything's going to be a little different. For example, some things that may have been included in the past that might be considered bleeding edge are things like ext4, btrfs, the original updater that evolved from red-carpet (the first one that was implemented in Mono back in the 10.x days - back when the resolver spiked CPU utilization for extended periods of time for many users), beagle, and even KDE4 (not in its current state, but when it was added to 11.2). Those technologies might at their inclusion have been considered "technology previews". Most of these technologies are no longer considered bleeding edge by the majority of users (btrfs might be the only exception to that). I mention them not to "open old wounds" but as examples where the project has included technologies in a release that may have been considered bleeding edge at the time. I'm not making any judgments here about the technologies themselves - but part of the issue I have with this statement is that 'bleeding edge' in and of itself can be in the eye of the beholder. For example, while some users may consider KDE4 at the 11.2 release to have been bleeding edge, it's very possible that the dev team didn't consider it to be because they'd been using it regularly - in other words, it's can be a matter of perspective, and the perspective used to apply this label needs to be defined in order for the statement to have any real meaning. It might make more sense instead of saying there's no focus on bleeding edge technology to clearly mark anything that's a technology preview as such and include it with those caveats. That way, that newer technology can be tested by those who want to do so, but they know that they are essentially alpha- or beta- testing. (Of course, it could be argued that those who do that should be using Factory instead of a numbered release - and perhaps it will be.) Jim -- Jim Henderson Please keep on-topic replies on the list so everyone benefits -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Wednesday 28 July 2010 09:53:32 Jim Henderson wrote:
On Mon, 26 Jul 2010 22:00:55 +0200, Pavol Rusnak wrote:
The openSUSE distribution acts as a reference distribution, providing an environment for testing the used technology, a stabilizing ground for common components, and a real-life use case for applying technology and distributing Linux software. It's targeted at technically interested users, including programmers and system administrators. It has a focus on good user experience and making technology available to end users. It doesn't target users with highly specific technical needs.
To me, "good user experience" isn't very specific, and in some ways might come across as contradicting the "no focus" area of "non-technical users" there's an implication of 'polish' that's more associated with a non-
- not that technical users can't have a good user experience, but that
technical audience. I hope that makes sense, because it seems at the moment to not be easy to explain what I mean.
Good user experience isn't very specific, that's right. But the important part is that it is a focus. That we actually care about the experience of the user and take this into account when taking decisions about what to do and how to do it. What this means in details of course has to be worked out, but if this is a serious direction, this is just a natural part of development. I don't think it's a contradiction to not focusing on non-technical end users. These are orthogonal issues. Focus on good user experience is one thing, which users are our target group is another thing. The "polish" will be different dependent on the target group, but it needs to be done to provide a good experience in any case.
= Good to have = [...]
- Community for user support
I would consider this essential. Without a community for user support, there's no feedback mechanism to improve the reference distribution. It also seems odd to not have this listed as "essential" since providing the home for the community is considered essential. Put another way, if having a user support community isn't essential, then why invest the time and resources in providing the home for that community (not that I think providing the home isn't essential - I provide the counterexample to demonstrate this point only).
Sounds reasonable to me.
= No focus =
- Directly providing a polished distribution for non-technical end users
I think we do this now with the main distribution, and that to an extent, this is an essential goal as well (or at least "good to have") because it provides a basis for the derivatives to provide that polish. I guess this perhaps needs to be more specifically defined for me as to what isn't/wouldn't be included. I couldn't even say what I would take as implied as not being provided by this statement.
You could also read that non-focus as: We are not Ubuntu. For the definition of what a non-technical end-user is, I like to think of this as as somebody who never bought a computer magazine. What that then implies for the actual products, needs to be worked out, but it won't involve problems like explaining how to use a mouse or what a harddisk is.
- Bleeding edge technology
It might make more sense instead of saying there's no focus on bleeding edge technology to clearly mark anything that's a technology preview as such and include it with those caveats. That way, that newer technology can be tested by those who want to do so, but they know that they are essentially alpha- or beta- testing. (Of course, it could be argued that those who do that should be using Factory instead of a numbered release - and perhaps it will be.)
You could also read that non-focus as: We are not Fedora. But I agree with your point that including newer technology as some kind of labeled previews make sense, and of course sometimes new technology is also the right way to go. But openSUSE's strategy should not be to focus on always having the latest technology. -- Cornelius Schumacher <cschum@suse.de> -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Wed, 28 Jul 2010 10:17:54 +0200, Cornelius Schumacher wrote:
Good user experience isn't very specific, that's right. But the important part is that it is a focus. That we actually care about the experience of the user and take this into account when taking decisions about what to do and how to do it. What this means in details of course has to be worked out, but if this is a serious direction, this is just a natural part of development.
That makes sense to me. I may be thinking a little further down the road beyond just the strategy, which does mean that the specifics are not necessarily - "relevant" isn't the right word, because the specifics are relevant - maybe "as critical at this stage of the discussion".
I don't think it's a contradiction to not focusing on non-technical end users. These are orthogonal issues. Focus on good user experience is one thing, which users are our target group is another thing. The "polish" will be different dependent on the target group, but it needs to be done to provide a good experience in any case.
OK, that makes sense to me as well. Needing to define more specifically that target group (beyond the 'high level' description here) is something that probably falls later in the discussion.
= No focus =
- Directly providing a polished distribution for non-technical end users
I think we do this now with the main distribution, and that to an extent, this is an essential goal as well (or at least "good to have") because it provides a basis for the derivatives to provide that polish. I guess this perhaps needs to be more specifically defined for me as to what isn't/wouldn't be included. I couldn't even say what I would take as implied as not being provided by this statement.
You could also read that non-focus as: We are not Ubuntu.
For the definition of what a non-technical end-user is, I like to think of this as as somebody who never bought a computer magazine. What that then implies for the actual products, needs to be worked out, but it won't involve problems like explaining how to use a mouse or what a harddisk is.
The Ubuntu comparison makes this clearer in my mind. Though arguably our current userbase does include both technical and non-technical users (I see this in the demographics in the forums, for example - it often makes for interesting interactions between members of the two groups) - so we might want to consider that there is a decent size non-technical user following already that may feel left out with this direction (though it sounds like perhaps the direction hasn't been previously defined).
- Bleeding edge technology
It might make more sense instead of saying there's no focus on bleeding edge technology to clearly mark anything that's a technology preview as such and include it with those caveats. That way, that newer technology can be tested by those who want to do so, but they know that they are essentially alpha- or beta- testing. (Of course, it could be argued that those who do that should be using Factory instead of a numbered release - and perhaps it will be.)
You could also read that non-focus as: We are not Fedora.
But I agree with your point that including newer technology as some kind of labeled previews make sense, and of course sometimes new technology is also the right way to go. But openSUSE's strategy should not be to focus on always having the latest technology.
The key differentiator here would thus be that we might include new technology if it makes sense, but the focus is on that ultimate usability as a reference implementation for re-spins and for a technical audience. Do you think that captures/restates the idea here? Or put another way, "always" is the key here - as in "not always, but perhaps sometimes, when it makes sense" - which would apply to the examples I provided earlier, or also possibly the inclusion of XEN virtualisation. Jim -- Jim Henderson Please keep on-topic replies on the list so everyone benefits -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Wednesday 28 July 2010 10:49:49 Jim Henderson wrote:
The Ubuntu comparison makes this clearer in my mind. Though arguably our current userbase does include both technical and non-technical users (I see this in the demographics in the forums, for example - it often makes for interesting interactions between members of the two groups) - so we might want to consider that there is a decent size non-technical user following already that may feel left out with this direction (though it sounds like perhaps the direction hasn't been previously defined).
Right. It's still about focus, so targeting just everybody is not possible or likely to yield good results. Not focusing the openSUSE distribution on non- technical users also doesn't mean that somebody else couldn't jump in and fill that gap. It actually might be a very good and promising project to take openSUSE and add some more end-user polishing. The last openSUSE survey seemed to indicate that most users actually are quite technical users, so I'm not sure how big the group would be we would leave out with this direction.
The key differentiator here would thus be that we might include new technology if it makes sense, but the focus is on that ultimate usability as a reference implementation for re-spins and for a technical audience. Do you think that captures/restates the idea here?
Yes. -- Cornelius Schumacher <cschum@suse.de> -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
-----Original Message----- From: Cornelius Schumacher [mailto:cschum@suse.de] Sent: 28 July 2010 09:18 To: opensuse-project@opensuse.org Subject: Re: [opensuse-project] Re: openSUSE Strategy Discussion: The Linux Distribution Platform Strategy
On Wednesday 28 July 2010 09:53:32 Jim Henderson wrote:
On Mon, 26 Jul 2010 22:00:55 +0200, Pavol Rusnak wrote:
The openSUSE distribution acts as a reference distribution, providing an environment for testing the used technology, a stabilizing ground for common components, and a real-life use case for applying technology and distributing Linux software. It's targeted at technically interested users, including programmers and system administrators. It has a focus on good user experience and making technology available to end users. It doesn't target users with highly specific technical needs.
To me, "good user experience" isn't very specific, and in some ways might come across as contradicting the "no focus" area of "non-technical users" there's an implication of 'polish' that's more associated with a non-
- not that technical users can't have a good user experience, but that
technical audience. I hope that makes sense, because it seems at the moment to not be easy to explain what I mean.
Good user experience isn't very specific, that's right. But the important part is that it is a focus. That we actually care about the experience of the user and take this into account when taking decisions about what to do and how to do it. What this means in details of course has to be worked out, but if this is a serious direction, this is just a natural part of development.
I don't think it's a contradiction to not focusing on non-technical end users. These are orthogonal issues. Focus on good user experience is one thing, which users are our target group is another thing. The "polish" will be different dependent on the target group, but it needs to be done to provide a good experience in any case.
I agree - I interpret "good user experience" mostly as "it works out of the box" and "you can do normal configuration using a GUI". Technical experts will always want to edit the config files, but most people (I included) like to be able to install most of the stuff I use without error messages and without it breaking other stuff. Some of the stuff I focus more attention on and delve into the internals, but only some. One aspect of this which is important is stability - it worked yesterday and it works (the same) today and I expect it to also work tomorrow, even if I've updated / installed security patches etc. That's a big factor in encouraging commercial companies (e.g. ATI) to provide drivers or ported software (e.g skype) - it costs them money if they have to do extensive retesting and incompatibility fixing every few weeks. David -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
Administrator wrote:
I agree - I interpret "good user experience" mostly as "it works out of the box" and "you can do normal configuration using a GUI".
I like that interpretation, but I would add something to the effect of "a minimum of small annoying QA bugs". Not sure quite how to phrase it, but these are the small irritating things that can easily screw up the overall picture coz' they give the user an uneasy feeling of "this isn't quite finished". Personally, I dislike changing keyboard accelerator keys or the same key used more than once. -- Per Jessen, Zürich (19.0°C) -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
Jim Henderson wrote:
- Bleeding edge technology
For example, some things that may have been included in the past that might be considered bleeding edge are things like ext4, btrfs, the original updater that evolved from red-carpet (the first one that was implemented in Mono back in the 10.x days - back when the resolver spiked CPU utilization for extended periods of time for many users), beagle, and even KDE4 (not in its current state, but when it was added to 11.2). Those technologies might at their inclusion have been considered "technology previews".
The one thing that made ext4 bleeding edge instead of a "technology preview" was making it default. Not being bleeding edge doesn't mean excluding bleeding edge software, but it does mean not using it by default or even depending on it. -- Per Jessen, Zürich (18.6°C) -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
- Bleeding edge technology
For example, some things that may have been included in the past that might be considered bleeding edge are things like ext4, btrfs, the original updater that evolved from red-carpet (the first one that was implemented in Mono back in the 10.x days - back when the resolver spiked CPU utilization for extended periods of time for many users), beagle, and even KDE4 (not in its current state, but when it was added to 11.2). Those technologies might at their inclusion have been considered "technology previews".
The one thing that made ext4 bleeding edge instead of a "technology preview" was making it default. Not being bleeding edge doesn't mean excluding bleeding edge software, but it does mean not using it by default or even depending on it.
Doesn't this argue for having more "repositories" (or collections of packages) and dependencies between repositories? These "technology previews" could be kept out of the core repository (and so not be seen by inexpert users) but easy to add. Having inter-repository dependencies would allow the solver to select packages from the right place (i.e. create a combination that has been tested) for additional packages required as dependencies and ignore the alternatives offered from other repositories, even if they are newer or in some other way "look better". David -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Wed, 28 Jul 2010 12:50:50 +0200, Per Jessen wrote:
The one thing that made ext4 bleeding edge instead of a "technology preview" was making it default. Not being bleeding edge doesn't mean excluding bleeding edge software, but it does mean not using it by default or even depending on it.
Makes sense to me. :-) Jim -- Jim Henderson Please keep on-topic replies on the list so everyone benefits -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
Le 26/07/2010 22:00, Pavol Rusnak a écrit :
= No focus =
- Directly providing a polished distribution for non-technical end users
with this I wont be openSUSE anymore... jdd -- http://www.dodin.net http://pizzanetti.fr -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Wednesday 2010-07-28 19:09, jdd wrote:
Le 26/07/2010 22:00, Pavol Rusnak a écrit :
= No focus =
- Directly providing a polished distribution for non-technical end users
with this I wont be openSUSE anymore...
Indeed that is simply - unacceptable. -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
* Jan Engelhardt <jengelh@medozas.de> [2010-07-28 19:42]:
On Wednesday 2010-07-28 19:09, jdd wrote:
Le 26/07/2010 22:00, Pavol Rusnak a écrit :
= No focus =
- Directly providing a polished distribution for non-technical end users
with this I wont be openSUSE anymore...
Indeed that is simply - unacceptable.
I second that, openSUSE has a large non-technical userbase as one can easily see by browsing through the opensuse users list or the forums. In fact, back in the day S.u.S.E. had a reputation of being easy to install and use (before Mandrake and then Ubuntu existed). Apart from that, some contributors might actually be motivated to create a polished product which is also usable by "non-technical" end-users. This proposal smells a lot like Fedora (and its "Spins"), just without the "bleeding edge" focus. -- Guido Berhoerster -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
= No focus =
- Directly providing a polished distribution for non-technical end
users
with this I wont be openSUSE anymore...
Indeed that is simply - unacceptable.
I second that, openSUSE has a large non-technical userbase as one can easily see by browsing through the opensuse users list or the forums. In fact, back in the day S.u.S.E. had a reputation of being easy to install and use (before Mandrake and then Ubuntu existed). Apart from that, some contributors might actually be motivated to create a polished product which is also usable by "non-technical" end-users. This proposal smells a lot like Fedora (and its "Spins"), just without the "bleeding edge" focus.
I think there may be a problem because of the word "polished". We need a clear definition. For me the minimum standard is that: 1) the UI all works - the buttons & controls all do what they look like they should do 2) the UI is "obvious" - that if you have something you want to do and which this package says it does, then you can see easily where to get to the right functions 3) the UI is consistent - same colours & fonts everywhere, things line up & behave the same way (e.g. shading unavailable controls, round corners, border style) 4) the main / most common setup & config options are defaulted sensibly for openSUSE and are changeable through the UI 5) it installs out of the box & will run after install without error 6) errors are caught & reported rather than being silently ignored or causing crashes 7) it integrates nicely, e.g. "funny" languages (e.g. utf8 characters, right-to-left), file open dialogues, keyboard layout, sound system I'm sure others can provide an expanded set of stuff which has been wrong with packages they've used... None of this is new but many people forget to check when launching a package upon the "public". David -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Wednesday 28 July 2010 19:42:38 Jan Engelhardt wrote:
On Wednesday 2010-07-28 19:09, jdd wrote:
Le 26/07/2010 22:00, Pavol Rusnak a écrit :
= No focus =
- Directly providing a polished distribution for non-technical end users
with this I wont be openSUSE anymore...
Indeed that is simply - unacceptable.
Note that the platform strategy doesn't exclude a non-technical end user distribution. On the contrary, it supports it. It just moves it from the primary focus of the overall community to the focus of a dedicated subgroup. The platform strategy proposal is also about plurality and providing a space for different interests, goals and target audiences to exist next to each other and still be part of an overall ecosystem. The openSUSE surveys seem to indicate that a majority of our users actually is technically interested, so it seems to be a pretty natural step to focus on them, especially as I think openSUSE already excels in that area. Putting the primary focus of openSUSE on a non-technical end user distribution puts us in direct heads-on competition with Ubuntu and others on a market, where we are clearly not the leader. It might not be impossible to change that, but from a strategically point of view it's certainly easier to act from a position of strength, and I clearly see the strength of openSUSE in the area of technical users right now. -- Cornelius Schumacher <cschum@suse.de> -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
Cornelius Schumacher wrote:
The openSUSE surveys seem to indicate that a majority of our users actually is technically interested, so it seems to be a pretty natural step to focus on them, especially as I think openSUSE already excels in that area.
*or* of those (from *all* our world wide users) who learned of the poll and decided to participate in a non-random, self-selecting, non-scientific poll, in English the majority are technically interested and it is those users who participated who are the ones we care the most about, so it seems to be a pretty natural step to focus on them and their needs and wants. except, of course it is not a rational to make such conclusions from such a poll.. DenverD -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Thursday 2010-07-29 12:05, Cornelius Schumacher wrote:
On Wednesday 28 July 2010 19:42:38 Jan Engelhardt wrote:
On Wednesday 2010-07-28 19:09, jdd wrote:
Le 26/07/2010 22:00, Pavol Rusnak a écrit :
= No focus =
- Directly providing a polished distribution for non-technical end users
with this I wont be openSUSE anymore...
Indeed that is simply - unacceptable.
Note that the platform strategy doesn't exclude a non-technical end user distribution. On the contrary, it supports it. It just moves it from the primary focus of the overall community to the focus of a dedicated subgroup.
What bothers me about all this "We will focus on this" and "We won't focus on that anymore" is that it's completely blurry. Consider this hypothetical thought: Let end users be the primary focus and technical users be a secondary target, with a work shift from Base:System towards KDE. Suppose that there is now a loose group of individuals that bring back the life to Base:System, perhaps even more than is done to KDE. The work being done now does not match with the strategy anymore. What do you do? Who determines where a distribution is headed, what its primary targets are? What if the strategy reads end users, but suddenly every developer loses interest in KDE and GNOME? My point is, you say primary focus is here or there, and I am questioning whether you (pl.) are even in the position to declare such a statement when it's not even clear that the end-user area receives the most love right now. There is one strategy declaration that would actually be universally applicable: Our focus is on what our contributors submit.
Putting the primary focus of openSUSE on a non-technical end user distribution puts us in direct heads-on competition with Ubuntu and others on a market, where we are clearly not the leader.
Sometimes it's better to be better rather than being a leader. -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Jul 29, 10 13:02:51 +0200, Jan Engelhardt wrote:
On Thursday 2010-07-29 12:05, Cornelius Schumacher wrote:
On Wednesday 28 July 2010 19:42:38 Jan Engelhardt wrote:
On Wednesday 2010-07-28 19:09, jdd wrote:
Le 26/07/2010 22:00, Pavol Rusnak a écrit :
= No focus =
- Directly providing a polished distribution for non-technical end users with this I wont be openSUSE anymore... Indeed that is simply - unacceptable.
If that is meant to give usabiltiy some thought, I am all for it. Because that would help everybody, be they techies or not.
Note that the platform strategy doesn't exclude a non-technical end user distribution. On the contrary, it supports it. It just moves it from the primary focus of the overall community to the focus of a dedicated subgroup.
What bothers me about all this "We will focus on this" and "We won't focus on that anymore" is that it's completely blurry.
If it is blurry, it is not in focus. That simple. :-) The concept of primary focus and 'focus of a dedicated subgroup' is irritating, indeed. It is like having two eyes staring into different directions.
My point is, you say primary focus is here or there, and I am questioning whether you (pl.) are even in the position to declare such a statement when it's not even clear that the end-user area receives the most love right now.
We can, and we should declare goals. But nothing should prevent a volunteer to contribute outside of the declared goals or focii. Let's assume I am a contributor and I have a plan for myself (like, e.g. I hack this specific application until it behaves like it should). In that case, I'd decide for myself, how to resolve conflicting project goals with my own plans. Sure thing such goals would have influence. Maybe, I'd speak up and challenge the declared goals, or maybe I silently try to adapt, or maybe the conflict isn't that much of a conflict in the end. If I am undecided, what next, then declared goals from the project are exactly what I need. They provide guidelines, where to go.
There is one strategy declaration that would actually be universally applicable: Our focus is on what our contributors submit.
Yes, except for one important aspect: Whenever there is a gap, like something that is urgently needed, but contributors don't pick it up by themselves. Then (and maybe only then) we need a person who finds that gap and makes it well known, and explains the difference it would make, ... That person I'd call The Manager. cheers, JW- PS: when it comes to focus, it is the *things* that should be aligned so that it is easy to see a broader picture in one view. It is not the *people* who's view is to be adjusted. -- o \ Juergen Weigert paint it green! __/ _=======.=======_ <V> | jw@suse.de back to ascii! __/ _---|____________/ \ | 0911 74053-508 __/ (____/ /\ (/) | _____________________________/ _/ _ vim:set sw=2 wm=8 SUSE LINUX Products GmbH, GF: Markus Rex, HRB 16746 (AG Nuernberg) -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Thursday 2010-07-29 13:44, Juergen Weigert wrote:
Note that the platform strategy doesn't exclude a non-technical end user distribution. On the contrary, it supports it. It just moves it from the primary focus of the overall community to the focus of a dedicated subgroup.
What bothers me about all this "We will focus on this" and "We won't focus on that anymore" is that it's completely blurry.
If it is blurry, it is not in focus. That simple. :-) The concept of primary focus and 'focus of a dedicated subgroup' is irritating, indeed. It is like having two eyes staring into different directions.
That would at least match the chameleon physiology... :D
My point is, you say primary focus is here or there, and I am questioning whether you (pl.) are even in the position to declare such a statement when it's not even clear that the end-user area receives the most love right now.
We can, and we should declare goals. But nothing should prevent a volunteer to contribute outside of the declared goals or focii. Let's assume I am a contributor and I have a plan for myself (like, e.g. I hack this specific application until it behaves like it should).
The idea was to shed light in the area as to what happens when $contributors (@contributors even) as a whole do, in total, more work outside the goals. Then we've missed the goal -- or the goal has long been reached hehe. -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Thursday 29 July 2010 13:02:51 Jan Engelhardt wrote:
There is one strategy declaration that would actually be universally applicable: Our focus is on what our contributors submit.
That's not a strategy. A strategy is supposed to give future direction and a guideline for taking decisions. Examples always have their flaws, but let's still try to make two to illustrate it: So for example if we would have decide to drop packages because we run in some media limit, how would you decide, if you would drop Frozen Bubble or Emacs. With a strategy focused on non-technical end-users the decision is clear, we would drop Emacs, with a strategy focused on technical users the decisions also would be clear, we would drop Frozen Bubble. With "focus on what our contributors submit" we would end up with a random decision or a conflict. Or as another example, how to take a decision, if to give a terminal application a prominent place in the desktop menu. If we focus on non- technical end users we wouldn't do it, if we focus on technical users we would do it. With the "focus on what our contributors submit" there is no real way to decide that, and we might end up with a distribution, which has Emacs, but not a terminal, which wouldn't serve anybody anymore. I would love to see openSUSE as the distribution, where I can be sure that I always have easy access to a terminal and Emacs (and vim, and others of course). That's what we are good at, and what many people want.
Sometimes it's better to be better rather than being a leader.
(I'll refrain from making comparisons to Video 2000 here ;-) -- Cornelius Schumacher <cschum@suse.de> -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Friday 2010-07-30 11:36, Cornelius Schumacher wrote:
On Thursday 29 July 2010 13:02:51 Jan Engelhardt wrote:
There is one strategy declaration that would actually be universally applicable: Our focus is on what our contributors submit.
That's not a strategy. A strategy is supposed to give future direction and a guideline for taking decisions.
Examples always have their flaws, but let's still try to make two to illustrate it: So for example if we would have decide to drop packages because we run in some media limit, how would you decide, if you would drop Frozen Bubble or Emacs. With a strategy focused on non-technical end-users the decision is clear, we would drop Emacs, with a strategy focused on technical users the decisions also would be clear, we would drop Frozen Bubble. With "focus on what our contributors submit" we would end up with a random decision or a conflict.
In practice, hell will be freezing over before I can get you to drop emacs from the media to make room for frozen-bubble, no matter how focused you are on end users :p -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
* Jan Engelhardt <jengelh@medozas.de> [2010-07-30 12:15]:
In practice, hell will be freezing over before I can get you to drop emacs from the media to make room for frozen-bubble, no matter how focused you are on end users :p
Accoding to this proposal we'll have an emacs and a vi spin. -- Guido Berhoerster -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Friday 2010-07-30 12:53, Guido Berhoerster wrote:
- Jan Engelhardt <jengelh@medozas.de> [2010-07-30 12:15]:
In practice, hell will be freezing over before I can get you to drop emacs from the media to make room for frozen-bubble, no matter how focused you are on end users :p
Accoding to this proposal we'll have an emacs and a vi spin.
The discussion was about the openSUSE proper, was it not? -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
Torsdag den 29. juli 2010 13:02:51 skrev Jan Engelhardt:
What bothers me about all this "We will focus on this" and "We won't focus on that anymore" is that it's completely blurry.
Consider this hypothetical thought: Let end users be the primary focus and technical users be a secondary target, with a work shift from Base:System towards KDE. Suppose that there is now a loose group of individuals that bring back the life to Base:System, perhaps even more than is done to KDE. The work being done now does not match with the strategy anymore. What do you do? Who determines where a distribution is headed, what its primary targets are? What if the strategy reads end users, but suddenly every developer loses interest in KDE and GNOME?
My point is, you say primary focus is here or there, and I am questioning whether you (pl.) are even in the position to declare such a statement when it's not even clear that the end-user area receives the most love right now.
I guess the idea is that when members have voted for, and selected a strategy, they will respect it, and "abide" by it. And it will also affect what type of new contributors would come to the project. E.g. noone would join Ubuntu, who's interested in a high-tech, geeky distro, and noone craving ultimate stability would join Fedora. But I share your concern. I think the proposed strategies can only expect to get minimal support because they are all way too far away from the status quo and past of SUSE - which is the reason why most people are here to begin with. Furtermore, the strategy is seen as only affecting volunteers (and boosters I guess), with apparently no commitment from Novell to the openSUSE strategy whatsoever. So I too fear that regardless what ends up being decided, it'll just be empty words, that won't change a thing in real life. Without wholehearted support from (most) volunteers and employees, the strategy will be dead in the water. -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Monday 26 July 2010 15:00:55 Pavol Rusnak wrote:
- Provide home for overall community and specific openSUSE teams, e.g. bug tracker, wiki, mailing lists, collaboration tools
It was a lot of words about distribution as software platform, but only above sentence about infrastructure that is base for all other activities. The base you build upon sets limits what you can and can not do. Discussing what we want has touch with reality only if we know what we have now. We get software from upstream, integrate it in distribution and publish results. Problem are many places in a workflow that require manual intervention on, essentially, communication tasks: - from communication channels to bug reporting, or feature, facility - from local bug reporting to upstream and vice verse - from one part of communication infrastructure to the other - from bugzilla to documentation - from documentation to various forms of document presentation and so on. The problem is software that is not designed to work as one system forcing us to: - Move and reformat whole messages by hand. - Search for information in various places. - Duplicate efforts because it is often faster to slap some info to user asking for it, then to find already written. The basic reason is that not many have understanding of the whole process, those that have are not connected to other and sharing knowledge is more a wish then reality. For instance, do we have any flow diagram that presents path software takes from origin to user and how user feedback flows in the other direction? No that I know. How we can create plans about system that we don't know exactly how it works and what we have to do to improve it? Having list of problems is not enough if we don't know how solutions will interact with system. I guess that recent problems with a wiki are good illustration how otherwise good plans can turn ugly if applied isolated from the rest of the workflow. -- Regards, Rajko -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
Mandag den 26. juli 2010 22:00:55 skrev Pavol Rusnak:
== Goals ==
The goal of openSUSE as a project is to provide a platform for distributing Linux and software running on Linux to a wide range of users. This platform consists of tools for creating software distributions, the openSUSE distribution as base and reference implementation, and the community supporting the tools and the distribution.
I must protest this proposal in the strongest terms :-) Being a good platform for appliances/derivatives is fine and all, but it cannot be the officially stated primary purpose/goal of the project. I don't think existing nor potential contributors will be very motivated by the idea of working to create a platform for others to "leach" on. While openSUSE being a good everyday system for themselves is an "accident" or after-thought. I think the very small minority of people who are actually interested in creating derivates and appliances will: a) tend to use their personal everyday distro anyway (probably one that focuses on being a good general purpose distro) b) be "leaches", contributing little or nothing to the project, if they _do_ choose openSUSE as their platform. If a person asks me "what is openSUSE about? Why should I use it?" And I'd have to respond: "Oh, openSUSE is mainly about being a good base/reference for derivatives/appliances", well.. I'd hate to see the expression on his face. I'm very much against _any_ strategy that doesn't have as its main focus openSUSE being the best possible everyday OS for <enter some significant group of users here>. -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Friday 2010-07-30 14:15, Martin Schlander wrote:
Mandag den 26. juli 2010 22:00:55 skrev Pavol Rusnak:
== Goals ==
The goal of openSUSE as a project is to provide a platform for distributing Linux and software running on Linux to a wide range of users. This platform consists of tools for creating software distributions, the openSUSE distribution as base and reference implementation, and the community supporting the tools and the distribution.
I must protest this proposal in the strongest terms :-)
Being a good platform for appliances/derivatives is fine and all, but it cannot be the officially stated primary purpose/goal of the project. [...] If a person asks me "what is openSUSE about? Why should I use it?" And I'd have to respond: "Oh, openSUSE is mainly about being a good base/reference for derivatives/appliances", well.. I'd hate to see the expression on his face.
I concur. That would feel a bit like Windows -- no office, no compiler, just the pure base. -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Friday 30 July 2010 14:15:28 Martin Schlander wrote:
I don't think existing nor potential contributors will be very motivated by the idea of working to create a platform for others to "leach" on. While openSUSE being a good everyday system for themselves is an "accident" or after-thought.
That's not what the proposal says. The openSUSE distribution itself is a prime component of this strategy. It's not about derivatives, it's about a great platform, and a great distribution built on this platform. If you don't read that in the proposal, we should adapt the text to make it clear, that the openSUSE distribution is a primary focus, not an after- thought. -- Cornelius Schumacher <cschum@suse.de> -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Friday 30 July 2010 08:33:38 Cornelius Schumacher wrote:
If you don't read that in the proposal, we should adapt the text to make it clear, that the openSUSE distribution is a primary focus, not an after- thought.
Hi Cornelius, Can someone with experience take time and depict that with a flow diagram ? Why I'm asking this? They tend to lower or remove language comprehension barrier from equation. They describe relations between elements in very concise form, that otherwise require a lot of text. Why I'm not doing this? I have no distribution creation experience. The best I can provide is some kind of rough sketch that everybody already knows, but even that I can't provide in a better form then some odp file that can be stored on the wiki. Why is this important? We need strong consent on what to do, and we will have no or weak consent if many of us have some reservations, ie. "I support idea, but not this part." Sum of all reservations can be small, but they create dissatisfaction and get disproportionally more attention in our minds then points that we agree with. Conversely it takes a lot more energy to handle relatively small amount of disagreements at every step of planing. I believe that diagram can help to understand current situation and reasons for proposals to everybody that doesn't live everyday 8 hours inside the distribution vendor. Sincerely, it may help even you guys to get better overview where we are then pure texts. I don't believe that text of reasonable size can provide amount of information in one flow diagram. -- Regards, Rajko -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Friday 30 July 2010 20:18:51 Rajko M. wrote:
Can someone with experience take time and depict that with a flow diagram ?
What exactly would you like to see in this flow diagram? -- Cornelius Schumacher <cschum@suse.de> -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Monday 02 August 2010 05:10:14 Cornelius Schumacher wrote:
On Friday 30 July 2010 20:18:51 Rajko M. wrote:
Can someone with experience take time and depict that with a flow diagram ?
What exactly would you like to see in this flow diagram?
Example in: http://lists.opensuse.org/opensuse-project/2010-08/msg00151.html helps a lot to understand problems that strategy should solve, and as I recall some problems in the past, it seems that we were hit not once by not having distro wide defined strategy. -- Regards, Rajko -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
Fredag den 30. juli 2010 15:33:38 skrev Cornelius Schumacher:
On Friday 30 July 2010 14:15:28 Martin Schlander wrote:
I don't think existing nor potential contributors will be very motivated by the idea of working to create a platform for others to "leach" on. While openSUSE being a good everyday system for themselves is an "accident" or after-thought.
That's not what the proposal says. The openSUSE distribution itself is a prime component of this strategy. It's not about derivatives, it's about a great platform, and a great distribution built on this platform.
If you don't read that in the proposal, we should adapt the text to make it clear, that the openSUSE distribution is a primary focus, not an after- thought.
Maybe. I read it as saying that creating a polished product is somewhere between secondary and completely unimportant. But communicating clearly is probably the most important aspect of the strategy. A) Communicate amongst ourselves what we're actually trying to do, so we can all pull in the same general direction. B) Communicate to outsiders what openSUSE is trying to do and not trying to do - and it _must_ be possible to sum this up very clearly in one or two sentences. -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
don't you think: * we are making circles, without progressing; * we are leading in a wrong direction (I know these two statements are somewhat in opposite directions :-) May be we should say: * we love the present status, where openSUSE is a wide open distribution, able to fit servers as well as newbies on desktop; * What are the parts we don't do well enough?? (and discuss how we could do them better) * What are the parts we can forget? (and when/how can we stop working on them) - For example we could say: we can work for old hardware, but no more than XX years old (may be 5 years is good?) - we can promote lxde, but only as a secondary subproject May be our scope should be: *make a polished openSUSE DVD* (meaning: all what is not on the dvd don't need to be polished as neatly) but anyway we have to be a little more practicals jdd -- http://www.dodin.net http://pizzanetti.fr -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
Fredag den 30. juli 2010 21:03:20 skrev jdd:
don't you think:
- we are making circles, without progressing;
- we are leading in a wrong direction
(I know these two statements are somewhat in opposite directions :-)
May be we should say:
- we love the present status, where openSUSE is a wide open distribution, able to fit servers as well as newbies on desktop;
No. Not having any clear direction at all as now, is horrible. It's bad for development and it's bad for marketing. Everybody is completely confused about what openSUSE is or should be. Is it a noob friendly distro? Is it a geeky, distro for experts? Is it a Fedora where you can just dumb all the latest stuff and don't care if any of it works? Is it a desktop-centric distro or is it a server distro? I hear all these conflicting points of views constantly - and a lot of the time people end up being disappointed because openSUSE isn't what they expected. You could argue for any of the above "identities" actually matching the status quo - and that is why openSUSE finally needs to make a decision about what we want to be. And the productive poweruser strategy is an excellent proposal for it ;-) -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
Le 30/07/2010 22:00, Martin Schlander a écrit :
Fredag den 30. juli 2010 21:03:20 skrev jdd:
don't you think:
- we are making circles, without progressing;
- we are leading in a wrong direction
(I know these two statements are somewhat in opposite directions :-)
May be we should say:
- we love the present status, where openSUSE is a wide open distribution, able to fit servers as well as newbies on desktop;
No. Not having any clear direction at all as now, is horrible. It's bad for development and it's bad for marketing. Everybody is completely confused about what openSUSE is or should be.
if it was so horrible, who would be here to discuss? and why do you remove from my post the most relevant part? it's a *fact* that we love the present situation, and we have to know what can be removed I quote myself: * What are the parts we don't do well enough?? (and discuss how we could do them better) * What are the parts we can forget? (and when/how can we stop working on them) it's more effective to start from the present situation and look at the possible changes than to try to create some new beast nobody know what it can do jdd -- http://www.dodin.net http://pizzanetti.fr -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
Fredag den 30. juli 2010 22:52:06 skrev jdd:
Le 30/07/2010 22:00, Martin Schlander a écrit :
Fredag den 30. juli 2010 21:03:20 skrev jdd:
don't you think:
- we are making circles, without progressing;
- we are leading in a wrong direction
(I know these two statements are somewhat in opposite directions :-)
May be we should say:
- we love the present status, where openSUSE is a wide open distribution, able to fit servers as well as newbies on desktop;
No. Not having any clear direction at all as now, is horrible. It's bad for development and it's bad for marketing. Everybody is completely confused about what openSUSE is or should be.
if it was so horrible, who would be here to discuss? and why do you remove from my post the most relevant part?
it's a *fact* that we love the present situation, and we have to know what can be removed
I still don't quite agree. I think most people are here cuz they loved the situation from 3-6 years ago and before that. But I do agree that the present situation should play a very prominent role in the new strategy - and that's what the "productive poweruser" strategy does - and another proposal coming up later takes that approach even further as Jan Engelhardt hinted at :-) -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Friday 30 July 2010 22:52:06 jdd wrote:
Le 30/07/2010 22:00, Martin Schlander a écrit :
Fredag den 30. juli 2010 21:03:20 skrev jdd:
don't you think:
- we are making circles, without progressing;
- we are leading in a wrong direction
(I know these two statements are somewhat in opposite directions :-)
May be we should say:
- we love the present status, where openSUSE is a wide open distribution, able to fit servers as well as newbies on desktop;
No. Not having any clear direction at all as now, is horrible. It's bad for development and it's bad for marketing. Everybody is completely confused about what openSUSE is or should be.
if it was so horrible, who would be here to discuss?
Everyone who cares. People don't leave right away if oS is in a bad spot. It will however make attracting new ppl harder right away and decrease enthusiasm of the current contributors.
and why do you remove from my post the most relevant part?
it's a *fact* that we love the present situation, and we have to know what can be removed
Why is that a fact? Most in the community might still love openSUSE but that doesn't mean they dislike the lack of a clear direction...
I quote myself:
- What are the parts we don't do well enough?? (and discuss how we could do them better)
We have no focus. We need that to make decisions.
- What are the parts we can forget? (and when/how can we stop working on them)
Whatever we decide once we have found a focus most can agree on.
it's more effective to start from the present situation and look at the possible changes than to try to create some new beast nobody know what it can do
Sure, nobody said we should start anew... We must find out what we're good at and what we all want to do, then focus on that.
jdd
Cheers, Jos
Welcome Jos :)
We have no focus. We need that to make decisions.
We can focus first on understanding how software compilations like Linux distributions work. Current multiplication of proposals seems to confirm that this is our primary weak point. -- Regards, Rajko -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
Le 01/08/2010 17:36, Jos Poortvliet a écrit :
it's a *fact* that we love the present situation, and we have to know what can be removed
Why is that a fact? Most in the community might still love openSUSE but that doesn't mean they dislike the lack of a clear direction...
I don't say we don't need better direction, but that we are here because we like openSUSE better than Debian or Red Hat (for example). So we should start from the beginning
- What are the parts we don't do well enough?? (and discuss how we could do them better)
We have no focus. We need that to make decisions.
sorry to say this, but this is plain ridiculous. Nothing is done without focus. May be we didn't sufficiently *describe* our focus, but we have one, and nearly nobody begin to choose a focus, because "a focus", per se, don't mean anything if it's not described in terms of manpower, time and money
- What are the parts we can forget? (and when/how can we stop working on them)
Whatever we decide once we have found a focus most can agree on.
I can't agree. We have to see *first* what we can leave and what we can't. this is *defining* the focus. don't forget the initial goals: "we don't have enough manpower to do all" but nobody said what we don't do well enough! for example we spend some manpower making openSUSE color and logo in any product (for example openoffice), and we could stop wasting manpower there, the distro will still be as stable, good and funny. If I look at the present situation, I find a very good stable and appreciated distro, at least by anybody having tried it. the "stability" problems we sometime have are mostly shared by any distro (nvidia driver for some kind of kernel, who have them before the others?) anyway, I see this discussion go again and again on the same subjects and I don't see any progress in it, and this makes me thing the method may not be good. may be we shoulds weigth as much as possible what is the cost of each option, before saying we want them or not. where is the analysis of the present situation (a brainstorming is not an analysis)? This analysis can only be made internally, because we have no knowledge of the manpower already used for openSUSE * what is the cost of Kde integration? * what is the cost of Gnome integration? .... jdd -- http://www.dodin.net http://pizzanetti.fr -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Friday 30 July 2010 22:52:06 jdd wrote:
[...] it's a fact that we love the present situation, and we have to know what can be removed
It's known that most people hate change and like to keep the status. ;) The problem with the present situation is that it's not documented. Even if we decide to keep the present situation, we should document it so that we all agree on what the present situation is. That one would be already a good step forward - not the step I like to see but still an improvement. Andreas -- Andreas Jaeger, Program Manager openSUSE, aj@{novell.com,opensuse.org} Twitter: jaegerandi | Identica: jaegerandi SUSE LINUX Products GmbH, GF: Markus Rex, HRB 16746 (AG Nürnberg) Maxfeldstr. 5, 90409 Nürnberg, Germany GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563C C272 A126
On Friday 2010-07-30 21:03, jdd wrote:
don't you think:
- we are making circles, without progressing;
- we are leading in a wrong direction
(I know these two statements are somewhat in opposite directions)
In fact they go together. Consider sitting in a carusel; your perceived tangential direction is always forward, and yet you go in circles. :)
- we love the present status, where openSUSE is a wide open distribution, able to fit servers as well as newbies on desktop;
You can express that opinion in one of the next proposals :p -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Friday July 30 2010 15:33:38 Cornelius Schumacher wrote:
On Friday 30 July 2010 14:15:28 Martin Schlander wrote:
I don't think existing nor potential contributors will be very motivated by the idea of working to create a platform for others to "leach" on. While openSUSE being a good everyday system for themselves is an "accident" or after-thought.
That's not what the proposal says. The openSUSE distribution itself is a prime component of this strategy. It's not about derivatives, it's about a great platform, and a great distribution built on this platform.
If you don't read that in the proposal, we should adapt the text to make it clear, that the openSUSE distribution is a primary focus, not an after- thought.
Please do so since reading (your) proposal reads like that to me too. IMHO it mostly sounds like the derivatives ones. My 0,02$ simply are that we already have the tools in place to easily create some derivatives but all / most people using them just "leach" (e.g. just monitor #suse for "Hi, I used the studio for ..."). So, IMHO, it is wrong to put any emphasis on creating derivatives but I would prefer the developers / power users proposal (nevertheless the current state to create derivatives should be at least kept on the same level). If there is any manpower to be spent then please finally make it possible to e.g. simply sync my pim data to my phone since that apparently still is some major pita for most devices. IOW: target the "poweruser" / "developer" and leave "joe six pack" to ubuntu and derivatives to debian (at least mostly) but try to get real world use cases working (yes, I know that is easier typed than done). regards, Stephan -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
Hello, This arguing over focusing openSUSE developement towards "technical users" or "non-technical users" is silly beyond belief. There is one question that needs to be answered before taking any direction at all: What both groups usually want as soon as installation is finished? Answers are mainly the following: 1) Stability 2) Functionality 3) Discoverability - First is obvious but neglected almost universally nowadays when it comes to Linux distros. Much to improve there. - Second is harder to grasp. GNOME with dual-panel setup and Programs/Places/System menus is arguably more functional than KDE kickoff. Functionality = (action+reaction)/time. This also covers generic program behaviour, overlapping with stability. A program that doesn't work as excepted isn't functional. - Third one, discoverability. The better, more describing word for "user friendly". While CLI is the pinnacle of functionality, it takes time to discover how it works. GNOME, on the other hand, has its features right there to click. They are easy to discover. This obviously overlaps with functionality. Getting at least those three common nominators in order should be enough to create something that appeals for everyone. Take it easy, Otso -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Sunday 01 August 2010 11:20:58 Otso wrote:
Hello,
This arguing over focusing openSUSE developement towards "technical users" or "non-technical users" is silly beyond belief. There is one question that needs to be answered before taking any direction at all: What both groups usually want as soon as installation is finished?
Answers are mainly the following:
- Stability
- Functionality
- Discoverability
- First is obvious but neglected almost universally nowadays when it comes to Linux distros. Much to improve there.
That is something that software developers today easily forget, and not only in Linux. The deadline is main goal, not what comes out on that deadline. I really like last news from GNOME that 3.0 is delayed. It is far more responsible then pushing it out just to respect deadline.
- Second is harder to grasp. GNOME with dual-panel setup and Programs/Places/System menus is arguably more functional than KDE kickoff.
Or GNOME with a single panel. The key here is that double panel. After using once upper panel with Programs-Places-System menus it is for me must have and I create my own. The same case is with KDE, but second panel has different functionality then in GNOME, as kickoff provides more functionality then openSUSE GNOME Main Menu (whatever is exact name of Main Menu that appear on installtion).
Functionality = (action+reaction)/time. This also covers generic program behaviour, overlapping with stability. A program that doesn't work as excepted isn't functional.
In that part, interpretation what is "as expected" depends on previous experience.
- Third one, discoverability. The better, more describing word for "user friendly". While CLI is the pinnacle of functionality, it takes time to discover how it works. GNOME, on the other hand, has its features right there to click. They are easy to discover. This obviously overlaps with functionality.
Not always. It has hidden things that surprise even long time computer user.
Getting at least those three common nominators in order should be enough to create something that appeals for everyone.
+1
Take it easy, Otso
-- Regards, Rajko -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Sunday 01 August 2010 18:20:58 Otso wrote:
Getting at least those three common nominators in order should be enough to create something that appeals for everyone.
Stability is a common element for all target users, but what functionality and discoverability acttually mean depends highly on the target audience. For some people a command line tool with a good man page is perfectly functional and discoverable and a great user interface, for other people a GNOME UI or something else works better. There are not many absolutes in user interface design and knowing the audience of the interface is an essential prerequisite for doing it well. -- Cornelius Schumacher <cschum@suse.de> -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On 02/08/2010, Cornelius Schumacher <cschum@suse.de> wrote:
On Sunday 01 August 2010 18:20:58 Otso wrote:
Getting at least those three common nominators in order should be enough to create something that appeals for everyone.
Stability is a common element for all target users, but what functionality and discoverability acttually mean depends highly on the target audience. For some people a command line tool with a good man page is perfectly functional and discoverable and a great user interface, for other people a GNOME UI or something else works better.
There are not many absolutes in user interface design and knowing the audience of the interface is an essential prerequisite for doing it well.
-- Cornelius Schumacher <cschum@suse.de>
Last time I checked (about 5 minutes ago when I opened Konsole for sed), powerful CLI and modern GUI can coexist without any problems. There is absolutely no stepping on advanced users toes if good, easy-to-operate, nice looking and stable GUI with sane defaults is made priority. I used OS X extensively for 2 years 2005-2007, and I was quite happy with it because I could revert to good old *nix command-line wizardy if the GUI couldn't do what I wanted (which was often the case). The existence of pretty GUI didn't interfere with my "powerusing". At all. Using KDE isn't stopping me from doing what I want either, on contrary, having Plasma *not* to crash would be quite helpful. Arguably, it's much easier for advanced users to cope with a easy-to-use GUI that has nice defaults and great stability, than it is for total newbie to tweak everything out of necessity. Powerusers can and will tweak their systems anyway, so concentrating efforts in satisfying our needs (I assume everyone on this mailing list is a poweruser) is hopeless. We have Yast and CLI tools and know our way around them anyway. Take it easy, Otso -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-project+help@opensuse.org
On Monday 02 August 2010 19:54:01 Otso wrote:
On 02/08/2010, Cornelius Schumacher <cschum@suse.de> wrote:
On Sunday 01 August 2010 18:20:58 Otso wrote:
Getting at least those three common nominators in order should be enough to create something that appeals for everyone.
Stability is a common element for all target users, but what functionality and discoverability acttually mean depends highly on the target audience. For some people a command line tool with a good man page is perfectly functional and discoverable and a great user interface, for other people a GNOME UI or something else works better.
There are not many absolutes in user interface design and knowing the audience of the interface is an essential prerequisite for doing it well.
-- Cornelius Schumacher <cschum@suse.de>
Last time I checked (about 5 minutes ago when I opened Konsole for sed), powerful CLI and modern GUI can coexist without any problems. There is absolutely no stepping on advanced users toes if good, easy-to-operate, nice looking and stable GUI with sane defaults is made priority. I used OS X extensively for 2 years 2005-2007, and I was quite happy with it because I could revert to good old *nix command-line wizardy if the GUI couldn't do what I wanted (which was often the case). The existence of pretty GUI didn't interfere with my "powerusing". At all. Using KDE isn't stopping me from doing what I want either, on contrary, having Plasma *not* to crash would be quite helpful.
Arguably, it's much easier for advanced users to cope with a easy-to-use GUI that has nice defaults and great stability, than it is for total newbie to tweak everything out of necessity. Powerusers can and will tweak their systems anyway, so concentrating efforts in satisfying our needs (I assume everyone on this mailing list is a poweruser) is hopeless. We have Yast and CLI tools and know our way around them anyway.
No arguments against it. Still doesn't mean you don't have to make a choice when it comes to where you put your focus. Eg it ain't to crazy to remove all manpages and use slim versions of the GNU tools on a livecd focussing on newbies, while that choice would be ridiculous when aiming for powerusers. And even if you would not have any choices, and powerusers' needs never bump into those of newbies, it helps to make clear who you're focusing on. Be it for attracting developers focused on helping newbies (or powerusers) or simply in your communication towards the users. IOW yes, we must make a choice here. Of course that choice can be to initially optimize for newbies but try not to limit powerusers or anything. It ain't black and white, but there must be a focus SOMEWHERE. If you don't have an agreed-upon strategy, on part of the community can do one thing, the other part the other thing - and thus both will loose out. Of course you could decide upon such a thing, for example: the Gnome team focusses on newbies, the KDE team focuses on powerusers/developers/gamers. Nothing wrong with that, as long as it's clear where the focus lies. Grtz Jos
Take it easy, Otso
participants (19)
-
Administrator -
Andreas Jaeger -
Cornelius Schumacher -
DenverD -
Guido Berhoerster -
Jan Engelhardt -
jdd -
Jim Henderson -
Jos Poortvliet -
Juergen Weigert -
Karsten König -
Martin Schlander -
Michael Loeffler -
Otso -
Pavol Rusnak -
Per Jessen -
Rajko M. -
Stephan Kleine -
Will Stephenson