taskSavannah Administration - Tasks: task #4564, Submission of Hurd's AlteRnative...

 
 

task #4564: Submission of Hurd's AlteRnative Development

Submitter:  Sergio López <k0ro>
Submitted:  Thu 25 Aug 2005 07:08:35 PM UTC
   
 
Should Start On:  Wed 24 Aug 2005 10:00:00 PM UTC Should be Finished on:  Sat 03 Sep 2005 10:00:00 PM UTC
Category:  Project Approval Priority:  * 5 - Normal
Status:  Cancelled Privacy:  Public
Assigned to:  kickino Open/Closed:  Closed
* Mandatory Fields

Add a New Comment Rich Markup
   

Jump to the original submission

  Spam posted by anonymous
  Spam posted by anonymous
  Spam posted by anonymous
  Spam posted by anonymous
  Spam posted by anonymous
  Spam posted by anonymous
  Spam posted by anonymous
  Spam posted by anonymous
  Spam posted by anonymous
  Spam posted by anonymous
Sat 27 Aug 2005 05:05:14 PM UTC, comment #15: 

Hello.

I tried to be objected and wanted to discussed this registration with Alfred directly.

I don't like or dislike this project, but it's always better to work together on the original project than try to fork it.


And Savannah's primary goal is to support the official GNU software but if you fork such a project and other maintainers work with you on the nonGNU code, it's hard to put them back into the official GNU project.

I tried to found out a way for your project, but if you would like to cancel this registration, I had to accept.

Regards,

Just for the logs; the ##hurd-log is avaiable at
http://web.walfield.org/~deride/%23%23hurd-20050826

Sebastian Wieseler <kickino>
Sat 27 Aug 2005 04:56:51 PM UTC, comment #14: 

Sergio, you still don't get it, ams-branch is not my branch.
Nobody has to talk to me to get something commited there.

The only rude one here is being you by quoting totally irrelevant
information that isn't even related to this.  Keep it technical, please.

Yes, it took 2 years for the ext2fs patch to get in, but since
there exists no released version of the Hurd, and everyone uses
CVS HEAD for their daily use, we cannot commit something that big
without extensive testing, which it got during the 2 years.  Claiming
that the ext2fs patch only touches libpager is quite naive.
If a design issue is `not suitable', it can be fixed.  Infact,
most of your messages are just `Here is a patch, it does this, it is
really experimental', not a concrete question.

I hoped that one could discuss with you in a reasonable tone, but
it seems that you are not and never were interested in cooperating
with the Hurd project.  And that is sad.

Alfred M. Szmidt <ams>
Sat 27 Aug 2005 04:44:59 PM UTC, comment #13: 

From ##hurd IRC channel of freenode. (kickino is Sebastian):

<kickino@12:10> But which options do we have? Reject it, and he register somewhere else? Or let him approve and we can 'controll' the project on Savannah. :/

Ok, that's all. This is too much for me, I can't get it. I just wanted to keep in touch with the GNU project (read the fucking project's description!!!), and Savannah seems to be the better way. I was wrong, it's obiously that it isn't.

Please CLOSE the registration process.

One more thing. Remember this words Alfred, you're doing much more harm than good to the Hurd project. Not for this, but for your everyday attitude.

Bye.

Sergio López <k0ro>
Sat 27 Aug 2005 04:31:29 PM UTC, comment #12: 

"Sergio, sorry, but I must call you a liar. You know perfectly well that the branch in question is NOT my branch"

You can claim what you want, but the facts are that, if you want that something enters in "ams-branch", you must talk with Alfred. And, sincerely, Alfred is a extremly rude person, that doesn't listen to nobody except himself. Sorry, I can't work with him. (For the record, please check this http://web.walfield.org/~deride/##hurd-20050824, but anyway, you can look to a random day, and you'll read ams bitching someone with 75% of probability).

"Any type of interface changes should be thought about hard, since they break things."

And what do you think I'm trying here (http://lists.gnu.org/archive/html/bug-hurd/2004-12/msg00242.html), here
(http://lists.gnu.org/archive/html/bug-hurd/2005-07/msg00278.html) and here
(http://lists.gnu.org/archive/html/bug-hurd/2005-08/msg00069.html)?
Claiming that a patch must enter into CVS? No! I'm asking design issues. Do you suggest that I must wait a little more? I'm waiting since December 2004!

"I actually applied a patch that changed the
interface in libpager in ams-branch a few days ago, the ext2fs >2gb patch to be exact."

1) That patch has took 2 years to enter into some official branch!

2) That patch is a great work, but it's changes on the interface are minimal compared to which I propouse.

"You are asking that we should apply patches blindly."

No, I'm not (here you're the liar). I'm asking to know what changes are right and what are not. Nobody answers. So I have 3 options:

1) Decide the design issues myself, invest a significant amount of work alone, make it stable (!) and have the risk that the changes you've made aren't right to maintainer's eyes (so time lose).

2) Wait for a reply

3) Do nothing and let the Hurd die silently (this is mostly the same that "2").

"It is possible to maintain large changes outside the tree, the ext2fs patch was maintained like that for a long long long time."

I repeat again, is extremly difficult to make clean patches when you're touched the interface. Ext2fs patch only touches libpager a little ! Think about changes in GNUMach/MiG and even Glibc!

Sergio López <k0ro>
Sat 27 Aug 2005 03:52:00 PM UTC, comment #11: 

Sergio, sorry, but I must call you a liar.  You know perfectly well that the branch in question
is NOT my branch, it is called ams-branch, but anyone with
commit access is free to commit any kind of change they feel is
approriate.  The reason why ams-branch is called ams-branch
is simply because that is what Roland wanted, I wanted to call
it devel-branch (see bug-hurd archives, I had to correct the
announcment).

Any type of interface changes should be thought about hard, since
they break things.  I actually applied a patch that changed the
interface in libpager in ams-branch a few days ago, the ext2fs >2gb patch
to be exact.

New interfaces should also be thought about hard, since if they
are wrong, and things start using them, it will be a pita to fix.

You are also refering to a extremely experimental patch, this cannot
go into the trunk, it won't go into ams-branch either since it
is completely untested, and nobody has looked at it at all.  You are asking
that we should apply patches blindly.  You haven't even signed
the copyright assignment papers that are needed to get any kind
of leagally significat patch into the Hurd, and still ask that
we should apply them when we cannot do so without hurting the whole
Hurd and GNU project.

It is possible to maintain large changes outside the tree, the ext2fs
patch was maintained like that for a long long long time.

If there is a need for a very experimental branch (I do not see
such a need right now), then it can be created.

Alfred M. Szmidt <ams>
Sat 27 Aug 2005 03:42:37 PM UTC, comment #10: 

Oh! I forgot. I've already signed and sent assginment papers for GNU Mach, Glibc and the Hurd, but I don't know if they have arrived to the FSF yet.

Sergio López <k0ro>
Sat 27 Aug 2005 03:38:23 PM UTC, comment #9: 

Sorry, typo (s/that get/that you get/):

Anyway, Sebastian, I don't want that you get into trouble. If you think this thing is getting too hard, please close the registration (I'll try to open the project elsewhere, where no one complains about it :-).

Sergio López <k0ro>
Sat 27 Aug 2005 03:36:27 PM UTC, comment #8: 

As you probably know, the Hurd is composed by 3 components (GNU Mach, Glibc and the Hurd itself) that need to properly comunicate between them. To do this, they use a common interface, and this means that if you change something in any of those components 1) you must also change the others 2) Your changes will render the entire package incompatible with previous versions.

Alfred is complaining to me about not sending more patches to bug-hurd, but he doesn't know how difficult is to generate a patch that can be used by a user when you've already changed the interface. He's also bravely announcing his (please note this) branch, but he perfecly knows that is not possible to apply changes that change the interface (that could render the entire package incompatible).

It's easy to say "the patches will enter when they're stable", so I must do all the work of design, programming and testing before they can enter into a CVS? It's a huge task!

Anyway, Sebastian, I don't want that get into trouble. If you think this thing is getting too hard, please close the registration (I'll try to open the project elsewhere, where no one complains about it :-).

Thanks for your time!

P.D: The unreleased code is the one related with this http://lists.gnu.org/archive/html/bug-hurd/2005-07/msg00278.html . I can't make a proper "diff" because I need to reimplement it in a clean tree before :-/

Sergio López <k0ro>
Sat 27 Aug 2005 02:52:09 PM UTC, comment #7: 

Logged in now.  I might note that other Hurd developers will use
this new branch for their code.  Sergio, I hope that you will
not follow through with this fork, it will only hurt the community.

Alfred M. Szmidt <ams>
Sat 27 Aug 2005 02:31:04 PM UTC, comment #6: 

Hello.

We cannot simply close this registration without asking Sergio about it.

But if the patches (if they are usefull and stable) can be inlcuded in the main project it's better than fork it, I think.


By the way, Alfred, could you please be so kind to login at Savananh before posting any comments? Thanks.


Sergio? We need your dicision. Would you like we remove your project or do you still want to go on with the registration process?

Regards,

Sebastian Wieseler <kickino>
Sat 27 Aug 2005 02:01:08 PM UTC, comment #5: 

I might note that Sergio has only sent one patch that is
even worth including at this point, the other patch he sent was
a experimental patch.  From the looks, Sergio has written more
patches, but hasn't bothered sending them to -email is unavailable- for
inclusion.

I do not know of any other patches that are really important
that are pending at this time that would warrent a fork
of the Hurd on Savannah.

Sergio's patches touch a lot of areas in the Hurd, and should
NOT be included into HEAD untill they have been proven stable.
Branches are used for code that touches lots of areas, and this
is exactly what Sergio's patches do.  So it takes time to understand
and test them.  Sergio has not bothered waiting for anyone to do
any kind of testing by users of his patches, nor has he bothered
with getting a second opinion by people who hack on the Hurd, of which
all are voluenteers, so it takes time to get such a response.

The maintainer of the Hurd is also currently swamped, so this leads
to further delays.

Alfred M. Szmidt <ams@gnu.org>, Hurd hacker.

Anonymous
Sat 27 Aug 2005 01:45:20 PM UTC, comment #4: 

Sergio has NOT discussed this with the maintainers.  Infact,
there already exists a experimental branch for exactly this type
of changes.  Please discard this project.  Sergio's patches cannot
be merged into the new branch since he hasn't signed papers yet.

Alfred M. Szmidt <ams@gnu.org>, Hurd hacker

Anonymous
Fri 26 Aug 2005 06:36:40 PM UTC, comment #3: 

Hello.

Is it possible that you include in the *diff files the license/coypright information of them?

In order to release your project properly and unambiguously under the GPL, please place copyright notices and permission-to-copy statements at the beginning of every file of source code.

In addition, if you haven't already, please include a copy of the plain text version of the GPL,
available from http://www.gnu.org/licenses/gpl.txt, into a file named "COPYING".

For more information, see http://www.gnu.org/licenses/gpl-howto.html.

The GPL FAQ explains why these procedures must be followed.
To learn why a copy of the GPL must be included with every copy of the code, for example,
see http://www.gnu.org/licenses/gpl-faq.html#WhyMustIInclude.


And we would like to have a look at these "some more unreleased code", too. Can you please point me to this code?

Regards,

Sebastian Wieseler <kickino>
Thu 25 Aug 2005 07:58:45 PM UTC, comment #2: 

"Do you have got any own source code yet or do you have got only the idea now?"

Yes, some patches ("http://sinrega.org/gnumach-multipage.diff", "http://sinrega.org/libpager.diff", "http://sinrega.org/manage-multithread.diff") and some more unreleased code.

"Then do you want to copy the whole source code from HURD and then work independet from the HURD team on this code?"

Yes.

"And did you discussed this with the HURD team, too?"

Only with a few independent developers, the GNU Maintainers of Hurd are mostly absent from "*-hurd@gnu.org" mailing lists.

Thanks for your time.
Regards,

Anonymous
Thu 25 Aug 2005 07:29:34 PM UTC, comment #1: 

Hi,
I'm evaluating the project you submitted for approval in Savannah.


Do you have got any own source code yet or do you have got only the idea now?

Then do you want to copy the whole source code from HURD and then work independet from the HURD team on this code?

And did you discussed this with the HURD team, too?

Regards,

Sebastian Wieseler <kickino>
Thu 25 Aug 2005 07:08:35 PM UTC, original submission:  

A new project has been registered at Savannah
The project account will remain inactive until a site admin approve or discard the registration.


######### REGISTRATION ADMINISTRATION #########

While this item will be useful to track the registration process, approving or discarding the registration must be done using the specific "Group Administration" page, accessible only to site administrators, effectively logged as site administrators (superuser):

  <https://savannah.gnu.org/admin/groupedit.php?group_id=7909>


######### REGISTRATION DETAILS #########

Full Name:
----------
  Hurd's AlteRnative Development

System Group Name:
-----------------
  hard

Type:
-----
  non-GNU software & documentation

License:
--------
  GNU General Public License V2 or later

Description:
------------
  HARD (Hurd's AlteRnative Development) doesn't need too much explanation. It's simply a fork from the official Hurd's Development, which is mostly stuck. You can think about it as an "independent and experimental development branch of Hurd". In a future, we expect that some of our changes will enter into the official repository (because of this, GCS and copyright assignment is also required for developers).

It will be composed by the following packages:

 - GNU Mach (gnumach-1-branch from official Hurd's CVS - http://www.gnu.org/software/hurd/gnumach.html)
 - Hurd (http://www.gnu.org/software/hurd)
 - (Maybe) Little parts or patches from/for Glibc (http://www.gnu.org/software/libc)

All the changes will retain the original license (GPL). We sincerely think that our work is needed, and that will help in a future to the official branch.

Thanks.


Sergio López <k0ro>

 

(Note: upload size limit is set to 16384 kB, after insertion of the required escape characters.)

Attach Files:
   
   
Comment:
   

No files currently attached

 

Depends on the following items: None found

Digest:
   task dependencies.

 

CC list is empty

 

There are 0 votes so far. Votes easily highlight which items people would like to see resolved in priority, independently of the priority of the item set by tracker managers.

Only logged-in users can vote.

 

Follow 8 latest changes.

Date Changed by Updated Field Previous Value => Replaced by
2005-08-27 kickino Dependencies- task #4575 is dependent
2005-08-27 kickino StatusNeed Info Cancelled
    Open/ClosedOpen Closed
2005-08-27 kickino StatusWait reply Need Info
    Carbon-Copy- Added -email is unavailable-
2005-08-26 kickino StatusNeed Info Wait reply
2005-08-25 kickino StatusNone Need Info
    Assigned toNone kickino

Back to the top

Powered by Savane 3.13-02a9.
Corresponding source code