6 weeks delay proposition

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

6 weeks delay proposition

Lionel Dricot (aka Ploum)
Hello,

I want to discuss about the 6 weeks delay proposed by Mark Shuttleworth :
https://lists.ubuntu.com/archives/ubuntu-devel-announce/2006-March/000094.html

The second meeting to talk about this 6 weeks delay is in a few hours but I want to write my ideas down before.

Don't forget to add "IMHO" everywhere in this text. It's just my opinion but I sometimes forget to write it.

Time based release :
----------------------------

When people have time, they will take it. If there is no deadline, a project will take forever or, at least, take a long time. Not because people are not working but because they always want something more polished then break something else, ...
The release of Debian Sarge is a perfect example.
That's why a lot of project are now using/switching to a time based release.


Ubuntu 6 months process :
------------------------------------

Ubuntu is now quite famous. The 6 months process is, IMHO, a good part of this success : it improves the global quality, it allows people to use stable but up-to-date softwares (nearly the bleeding edge) but, more, it's predicable ! You can easily schedule upgrade, testing. You can also decide to wait a few weeks before migrating a bunch of computers.

Non-techies users are already looking at screenshots and asking us : "Where will we have them ? April ?"  or "You tell me that this annoying bug is fixed. When will I have the fix ? April ?".

Some LUG and Ubuntu communities have also scheduled an event every 6 months. The traditionnal install parties are becoming install/upgrade parties and that's good. 

At first glance, 6 weeks delay seems bad.


Dapper Drake Enterprise :
---------------------------------

As Mark said, Dapper Drake is very special because it's the first Enterprise ready Ubuntu edition (with DVD collector, director's cut edition, isn't it ?). It seems good to add more polish and more testing since it will have a longer support life.

At first glance, 6 weeks delay seems good.


Just adding a 6 week delay ?
-------------------------------------

What if we just add a 6 weeks delay ? It would be, IMHO, wrong.
1) Just adding 6 weeks will not add that much quality. A lot of contributor will think : "Ok, so I'm cool for 6 weeks more before the rush". There will be lot of discussion for 6 weeks more (Must we add NM with WPA support ? With the new theme, my desktop glows in the dark ! ...)
2) It would be a big deception for a big part of the community. People that are waiting Dapper to install it on their desktop, scheduled install parties, ...
3) There would be no more beta-testers than usual. Only usual contributors and testers will install and report bugs against the intermediate flight CD.
The biggest problem with the Breezy release was hardware oddities. With only a 6 weeks delay, we will not expand the testing field.

But :
1) Dapper obviously needs more polishing/testing thant Breezy and Hoary.


What can we choose ?
-----------------------------

As usual in this kind of problem, the solution is nor black or white - insert here random Confucius quote - but in the middle.
So I suggest the following :

- Let release Dapper as usual, without any delay. Community will be very happy.
- Call this Dapper : "Community Edition" or something like that.
- Leave Dapper frozen for 6 weeks.
- In this timeframe, collect all the new bugs from install parties, new users, users with strange hardware and so one.
- Choose carefully wich bugs must be really solved in Dapper.
- 6 weeks later, release : "Dapper Enterprise edition" and print CDs.

I don't know if this kind of solution is realistic or not, but it would make nearly everyone happy, IMHO.  More, it would make a lot of people beta-testers of the Enterprise edition.

But, in respect of the community, it seems very important to me to release this "Community edition" as a normal release, not a beta one. It will not warn you against installing it or possible data loss. It could be installed on a beginner's computer. It *is* a release. The special would not be this community edition but the enterprise one, the same release with just a little more love.


Thank you for reading, any comment is welcome. See you at the meeting :-)


Lionel  (ploum on IRC)

--
ubuntu-devel mailing list
[hidden email]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel
Reply | Threaded
Open this post in threaded view
|

Re: 6 weeks delay proposition

John Augustine-2
Lionel Dricot (aka Ploum) wrote:
Hello,

I want to discuss about the 6 weeks delay proposed by Mark Shuttleworth :
https://lists.ubuntu.com/archives/ubuntu-devel-announce/2006-March/000094.html

The second meeting to talk about this 6 weeks delay is in a few hours but I want to write my ideas down before.

Don't forget to add "IMHO" everywhere in this text. It's just my opinion but I sometimes forget to write it.

Time based release :
----------------------------

When people have time, they will take it. If there is no deadline, a project will take forever or, at least, take a long time. Not because people are not working but because they always want something more polished then break something else, ...
The release of Debian Sarge is a perfect example.
That's why a lot of project are now using/switching to a time based release.


Ubuntu 6 months process :
------------------------------------

Ubuntu is now quite famous. The 6 months process is, IMHO, a good part of this success : it improves the global quality, it allows people to use stable but up-to-date softwares (nearly the bleeding edge) but, more, it's predicable ! You can easily schedule upgrade, testing. You can also decide to wait a few weeks before migrating a bunch of computers.

Non-techies users are already looking at screenshots and asking us : "Where will we have them ? April ?"  or "You tell me that this annoying bug is fixed. When will I have the fix ? April ?".

Some LUG and Ubuntu communities have also scheduled an event every 6 months. The traditionnal install parties are becoming install/upgrade parties and that's good. 

At first glance, 6 weeks delay seems bad.


Dapper Drake Enterprise :
---------------------------------

As Mark said, Dapper Drake is very special because it's the first Enterprise ready Ubuntu edition (with DVD collector, director's cut edition, isn't it ?). It seems good to add more polish and more testing since it will have a longer support life.

At first glance, 6 weeks delay seems good.


Just adding a 6 week delay ?
-------------------------------------

What if we just add a 6 weeks delay ? It would be, IMHO, wrong.
1) Just adding 6 weeks will not add that much quality. A lot of contributor will think : "Ok, so I'm cool for 6 weeks more before the rush". There will be lot of discussion for 6 weeks more (Must we add NM with WPA support ? With the new theme, my desktop glows in the dark ! ...)
2) It would be a big deception for a big part of the community. People that are waiting Dapper to install it on their desktop, scheduled install parties, ...
3) There would be no more beta-testers than usual. Only usual contributors and testers will install and report bugs against the intermediate flight CD.
The biggest problem with the Breezy release was hardware oddities. With only a 6 weeks delay, we will not expand the testing field.

But :
1) Dapper obviously needs more polishing/testing thant Breezy and Hoary.


What can we choose ?
-----------------------------

As usual in this kind of problem, the solution is nor black or white - insert here random Confucius quote - but in the middle.
So I suggest the following :

- Let release Dapper as usual, without any delay. Community will be very happy.
- Call this Dapper : "Community Edition" or something like that.
- Leave Dapper frozen for 6 weeks.
- In this timeframe, collect all the new bugs from install parties, new users, users with strange hardware and so one.
- Choose carefully wich bugs must be really solved in Dapper.
- 6 weeks later, release : "Dapper Enterprise edition" and print CDs.

I don't know if this kind of solution is realistic or not, but it would make nearly everyone happy, IMHO.  More, it would make a lot of people beta-testers of the Enterprise edition.

But, in respect of the community, it seems very important to me to release this "Community edition" as a normal release, not a beta one. It will not warn you against installing it or possible data loss. It could be installed on a beginner's computer. It *is* a release. The special would not be this community edition but the enterprise one, the same release with just a little more love.


Thank you for reading, any comment is welcome. See you at the meeting :-)


Lionel  (ploum on IRC)
I agree with this for the most part.  Delaying Dapper for 6 weeks to release an Enterprise version of an OS that is, for the most part, an end user OS.

I totally agree with releasing a "Community Edition", however, this should simply be Dapper, no special names, and 6 weeks later, the "Dapper Enterprise" edition should be released.

I've been testing/using Dapper since Flight 1(the Southbridge of my new MB requires drivers only included in the most up-to-date installation CDs, like Dapper or FC5) and I really haven't seen any reason lately showing this release needs any kind of a delay.

Also, I feel that pushing the release and calling it an "Enterprise" edition, will steer desktop users toward Breezy, or other systems, which may not have the most up-to-date combination of stable software.

All of this is IMHO as a loyal Debian/Ubuntu user.  I want what's best for the comminuty/OS, and I really believe that releasing Dapper(standard desktop edition) on schedule(or a week or so off) is in the best interest of developers and users.  I also must agree that the "Enterprise" edition is an excellent idea because Ubuntu seems to be missing in that department, but this shouldn't conflict with the standard release(IMO).

I'm sure that what is decided upon will be what's truly best for Dapper, but I wanted to put my 2 cents in as a user.  So please don't judge this too harshly.

--
ubuntu-devel mailing list
[hidden email]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel
Reply | Threaded
Open this post in threaded view
|

Re: 6 weeks delay proposition

Joao Inacio
I don't know if this should be discussed at this list, but i just want
to express my very simple opinion:

whatever is decided, i believe a cd should be released in the
scheduled (mid april?) time...

call it ubuntu flight N, RC1, pre-release, or even 6.04 (if the final
is to be 6.05)... just don't release an "enterprise" version.

Anyway, i think this is offftopic for ubuntu-devel, so sorry again.

On 3/14/06, John Augustine <[hidden email]> wrote:

>  Lionel Dricot (aka Ploum) wrote:
> Hello,
>
>  I want to discuss about the 6 weeks delay proposed by Mark Shuttleworth :
> https://lists.ubuntu.com/archives/ubuntu-devel-announce/2006-March/000094.html
>
>  The second meeting to talk about this 6 weeks delay is in a few hours but I
> want to write my ideas down before.
>
>  Don't forget to add "IMHO" everywhere in this text. It's just my opinion
> but I sometimes forget to write it.
>
>  Time based release :
>  ----------------------------
>
>  When people have time, they will take it. If there is no deadline, a
> project will take forever or, at least, take a long time. Not because people
> are not working but because they always want something more polished then
> break something else, ...
>  The release of Debian Sarge is a perfect example.
>  That's why a lot of project are now using/switching to a time based
> release.
>
>
>  Ubuntu 6 months process :
>  ------------------------------------
>
>  Ubuntu is now quite famous. The 6 months process is, IMHO, a good part of
> this success : it improves the global quality, it allows people to use
> stable but up-to-date softwares (nearly the bleeding edge) but, more, it's
> predicable ! You can easily schedule upgrade, testing. You can also decide
> to wait a few weeks before migrating a bunch of computers.
>
>  Non-techies users are already looking at screenshots and asking us : "Where
> will we have them ? April ?"  or "You tell me that this annoying bug is
> fixed. When will I have the fix ? April ?".
>
>  Some LUG and Ubuntu communities have also scheduled an event every 6
> months. The traditionnal install parties are becoming install/upgrade
> parties and that's good.
>
>  At first glance, 6 weeks delay seems bad.
>
>
>  Dapper Drake Enterprise :
>  ---------------------------------
>
>  As Mark said, Dapper Drake is very special because it's the first
> Enterprise ready Ubuntu edition (with DVD collector, director's cut edition,
> isn't it ?). It seems good to add more polish and more testing since it will
> have a longer support life.
>
>  At first glance, 6 weeks delay seems good.
>
>
>  Just adding a 6 week delay ?
>  -------------------------------------
>
>  What if we just add a 6 weeks delay ? It would be, IMHO, wrong.
>  1) Just adding 6 weeks will not add that much quality. A lot of contributor
> will think : "Ok, so I'm cool for 6 weeks more before the rush". There will
> be lot of discussion for 6 weeks more (Must we add NM with WPA support ?
> With the new theme, my desktop glows in the dark ! ...)
>  2) It would be a big deception for a big part of the community. People that
> are waiting Dapper to install it on their desktop, scheduled install
> parties, ...
>  3) There would be no more beta-testers than usual. Only usual contributors
> and testers will install and report bugs against the intermediate flight CD.
>  The biggest problem with the Breezy release was hardware oddities. With
> only a 6 weeks delay, we will not expand the testing field.
>
>  But :
>  1) Dapper obviously needs more polishing/testing thant Breezy and Hoary.
>
>
>  What can we choose ?
>  -----------------------------
>
>  As usual in this kind of problem, the solution is nor black or white -
> insert here random Confucius quote - but in the middle.
>  So I suggest the following :
>
>  - Let release Dapper as usual, without any delay. Community will be very
> happy.
>  - Call this Dapper : "Community Edition" or something like that.
>  - Leave Dapper frozen for 6 weeks.
>  - In this timeframe, collect all the new bugs from install parties, new
> users, users with strange hardware and so one.
>  - Choose carefully wich bugs must be really solved in Dapper.
>  - 6 weeks later, release : "Dapper Enterprise edition" and print CDs.
>
>  I don't know if this kind of solution is realistic or not, but it would
> make nearly everyone happy, IMHO.  More, it would make a lot of people
> beta-testers of the Enterprise edition.
>
>  But, in respect of the community, it seems very important to me to release
> this "Community edition" as a normal release, not a beta one. It will not
> warn you against installing it or possible data loss. It could be installed
> on a beginner's computer. It *is* a release. The special would not be this
> community edition but the enterprise one, the same release with just a
> little more love.
>
>
>  Thank you for reading, any comment is welcome. See you at the meeting :-)
>
>
>  Lionel  (ploum on IRC)
>
>  I agree with this for the most part.  Delaying Dapper for 6 weeks to
> release an Enterprise version of an OS that is, for the most part, an end
> user OS.
>
>  I totally agree with releasing a "Community Edition", however, this should
> simply be Dapper, no special names, and 6 weeks later, the "Dapper
> Enterprise" edition should be released.
>
>  I've been testing/using Dapper since Flight 1(the Southbridge of my new MB
> requires drivers only included in the most up-to-date installation CDs, like
> Dapper or FC5) and I really haven't seen any reason lately showing this
> release needs any kind of a delay.
>
>  Also, I feel that pushing the release and calling it an "Enterprise"
> edition, will steer desktop users toward Breezy, or other systems, which may
> not have the most up-to-date combination of stable software.
>
>  All of this is IMHO as a loyal Debian/Ubuntu user.  I want what's best for
> the comminuty/OS, and I really believe that releasing Dapper(standard
> desktop edition) on schedule(or a week or so off) is in the best interest of
> developers and users.  I also must agree that the "Enterprise" edition is an
> excellent idea because Ubuntu seems to be missing in that department, but
> this shouldn't conflict with the standard release(IMO).
>
>  I'm sure that what is decided upon will be what's truly best for Dapper,
> but I wanted to put my 2 cents in as a user.  So please don't judge this too
> harshly.
>
> --
> ubuntu-devel mailing list
> [hidden email]
> https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel
>
>
>


--
João Inácio
http://www.jcinacio.com

--
ubuntu-devel mailing list
[hidden email]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel
Reply | Threaded
Open this post in threaded view
|

Re: 6 weeks delay proposition

Santiago Roza-2
On 3/14/06, Joao Inacio <[hidden email]> wrote:

> Anyway, i think this is offftopic for ubuntu-devel, so sorry again.


everything is offtopic for ubuntu-devel: bugs, comments, suggestions,
specs, ideas, discussions, and development-related issues in general.
i think sending code is the one and only ontopic content, but i'm not
sure there is one  :)


--
Santiago Roza
Departamento I+D - Thymbra
[hidden email]

--
ubuntu-devel mailing list
[hidden email]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel