Branch cleanup

[QuLogic]

There seem to be several extra branches that should be deleted or merged somehow:

• Xie_Zhinan: I feel like the topics mentioned in the commits were merged to master, but the branch does not appear to be. @xiezhinan, can you clarify?
• new_branch_for_Xie_Zhinan: This is a completely disjoint branch containing 56 commits which are all empty. @eheien: did something go wrong with the conversion here, or did you just remove all the directories affected by these commits?
• BIOT & BIOT@15505: These are two empty commits next to each other. However, the "following" four commits on master (65ad9c0 to a518aa0) are merging commits. In other words, this branch should have been merged, but the SVN end was not done correctly.
• SEM_tangente_branch: Not sure? The initial commit claims it's "private", so maybe it shouldn't be here?
• SEM_2D_nicolas_unstruct: This appears to be merged, but also contains the third copy of the same commits below.

These are merged and irrelevant. They should be deleted:

• trunk@8873: Merging of branch below.
• trunk@8872: Merging of branch below.
• SEM_2D_nicolas_unstruct@8759: Continued on SEM_2D_nicolas_unstruct. However, this entire code line does not branch from somewhere on master, and duplicates most of the commits at the same time. I believe this stems from the fact that this branch was a subdirectory of the main code, which was then turned into a "real" branch (insomuch as that exists in SVN).

[QuLogic]

The last commit for the branches directory in SVN (20322) says: "removed branches/Xie_Zhinan, which was unused". Does that mean we can remove the Xie_Zhinan branch from git entirely?

[QuLogic]

Ah, I have figured out why some commits appear in triplicate. A long time ago, some development was done in a subdirectory, which eventually replaced the top-level directory. Because of this, git followed development back through history and ended up with commits on <toplevel>/<files> plus equivalent commits on <toplevel>/<subdirectory>/<files>. I will see if I can remove the redundant commits.

[eheien]

new_branch_for_Xie_Zhinan: This is a completely disjoint branch containing 56 commits which are all empty. @eheien: did something go wrong with the conversion here, or did you just remove all the directories affected by these commits?

It looks like most (all?) of the commits on this branch were in the examples directory, which I removed and put on the CIG servers for space reasons. So I don't believe anything went wrong, the changes were just made in a directory no longer in the Github repository.

[jedbrown]

For future reference, you can use the --prune-empty argument to git
filter-branch (which I assume is what you used to remove those
directories).

[eheien]

Thanks for the tip.

On Nov 13, 2013, at 4:23 PM, Jed Brown notifications@github.com wrote:

For future reference, you can use the --prune-empty argument to git
filter-branch (which I assume is what you used to remove those
directories).
—
Reply to this email directly or view it on GitHub.

[komatits]

Hi,

That branch was created by mistake and never used, you can safely delete
it (no need to keep its history either)

Thanks,
Dimitri.

On 11/14/2013 01:16 AM, Eric Heien wrote:

new_branch_for_Xie_Zhinan: This is a completely disjoint branch
containing 56 commits which are all empty. @eheien
https://github.com/eheien: did something go wrong with the conversion
here, or did you just remove all the directories affected by these commits?

It looks like most (all?) of the commits on this branch were in the
examples directory, which I removed and put on the CIG servers for space
reasons. So I don't believe anything went wrong, the changes were just
made in a directory no longer in the Github repository.

—
Reply to this email directly or view it on GitHub
#1 (comment).

Dimitri Komatitsch
CNRS Research Director (DR CNRS), Laboratory of Mechanics and Acoustics,
UPR 7051, Marseille, France http://komatitsch.free.fr

[komatits]

Dear Elliott,

Since David Luet and Eric Heien are also cleaning the official repo on
their side, please make sure you contact them first to see how to
proceed to clean the branches (maybe emailing them your suggestions
instead of working on another repo).

Thanks,
Dimitri.

On 11/14/2013 12:50 AM, Elliott Sales de Andrade wrote:

Ah, I have figured out why some commits appear in triplicate. A long
time ago, some development was done in a subdirectory, which eventually
replaced the top-level directory. Because of this, git followed
development back through history and ended up with commits on
|/| plus equivalent commits on
|//|. I will see if I can remove the
redundant commits.

—
Reply to this email directly or view it on GitHub
#1 (comment).

Dimitri Komatitsch
CNRS Research Director (DR CNRS), Laboratory of Mechanics and Acoustics,
UPR 7051, Marseille, France http://komatitsch.free.fr

[QuLogic]

I actually started this yesterday and finished it already. It's available at https://github.com/QuLogic/specfem2d-clean as soon as it finishes uploading.

The far history was pretty messy, but it turns out that the triplicate commits are all redundant (I verified this by comparing all commits that were duplicated). I just dropped the split section, and that saved about 400 commits.

This cleanup drops the number of commits from 1517 to just shy of 1000 and the .git size from ~106M to ~37M, which is a nice savings.

[eheien]

Thanks for working on this Elliott. If it’s OK with David and the rest of the SPECFEM team I’d be fine with replacing the existing SPECFEM2D/3D Github repositories with your changes. If there are any problems, at worst we can revert to the original based on the current SVN repository. Does anyone have any concerns about this?

On Nov 13, 2013, at 9:42 PM, Elliott Sales de Andrade notifications@github.com wrote:

I actually started this yesterday and finished it already. It's available at https://github.com/QuLogic/specfem2d-clean as soon as it finishes uploading.

The far history was pretty messy, but it turns out that the triplicate commits are all redundant (I verified this by comparing all commits that were duplicated). I just dropped the split section, and that saved about 400 commits.

This cleanup drops the .git size from ~106M to ~37M, which is a nice savings.

—
Reply to this email directly or view it on GitHub.

[xiezhinan]

Hi Eric and Jed,

Yes, please just delete that, because all those early empty comments from
me are useless.
I am sorry to bring the trouble.

Thank you so much.

Best regards,
Zhinan

2013/11/14 Eric Heien notifications@github.com

new_branch_for_Xie_Zhinan: This is a completely disjoint branch containing
56 commits which are all empty. @eheien https://github.com/eheien: did
something go wrong with the conversion here, or did you just remove all the
directories affected by these commits?

It looks like most (all?) of the commits on this branch were in the
examples directory, which I removed and put on the CIG servers for space
reasons. So I don't believe anything went wrong, the changes were just made
in a directory no longer in the Github repository.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/1#issuecomment-28447921
.

[komatits]

Fine with me.

Dimitri.

On 11/14/2013 06:53 AM, Eric Heien wrote:

Thanks for working on this Elliott. If it’s OK with David and the rest
of the SPECFEM team I’d be fine with replacing the existing SPECFEM2D/3D
Github repositories with your changes. If there are any problems, at
worst we can revert to the original based on the current SVN repository.
Does anyone have any concerns about this?

On Nov 13, 2013, at 9:42 PM, Elliott Sales de Andrade
notifications@github.com wrote:

I actually started this yesterday and finished it already. It's
available at https://github.com/QuLogic/specfem2d-clean as soon as it
finishes uploading.

The far history was pretty messy, but it turns out that the
triplicate commits are all redundant (I verified this by comparing all
commits that were duplicated). I just dropped the split section, and
that saved about 400 commits.

This cleanup drops the .git size from ~106M to ~37M, which is a nice
savings.

—
Reply to this email directly or view it on GitHub.

—
Reply to this email directly or view it on GitHub
#1 (comment).

Dimitri Komatitsch
CNRS Research Director (DR CNRS), Laboratory of Mechanics and Acoustics,
UPR 7051, Marseille, France http://komatitsch.free.fr

[luet]

If Dimitri is ok with it then that's what we should do.
Matthieu has spent an afternoon building the issues list. I wonder if there
is a way to save them.
Thanks,
David

On Thu, Nov 14, 2013 at 5:57 AM, komatits notifications@github.com wrote:

Fine with me.

Dimitri.

On 11/14/2013 06:53 AM, Eric Heien wrote:

Thanks for working on this Elliott. If it’s OK with David and the rest
of the SPECFEM team I’d be fine with replacing the existing SPECFEM2D/3D
Github repositories with your changes. If there are any problems, at
worst we can revert to the original based on the current SVN repository.
Does anyone have any concerns about this?

On Nov 13, 2013, at 9:42 PM, Elliott Sales de Andrade
notifications@github.com wrote:

I actually started this yesterday and finished it already. It's
available at https://github.com/QuLogic/specfem2d-clean as soon as it
finishes uploading.

The far history was pretty messy, but it turns out that the
triplicate commits are all redundant (I verified this by comparing all
commits that were duplicated). I just dropped the split section, and
that saved about 400 commits.

This cleanup drops the .git size from ~106M to ~37M, which is a nice
savings.

—
Reply to this email directly or view it on GitHub.

—
Reply to this email directly or view it on GitHub
<#1 (comment)
.

Dimitri Komatitsch
CNRS Research Director (DR CNRS), Laboratory of Mechanics and Acoustics,
UPR 7051, Marseille, France http://komatitsch.free.fr

—

Reply to this email directly or view it on GitHubhttps://github.com//issues/1#issuecomment-28475282
.

David Luet
Linux Administrator/Software & Programming Analyst
Princeton University
Department of Geosciences & PICSciE
PICSciE (9 am - 1 pm): 609-258-8194
Geosciences (1 - 5 pm): 609-258-7945

[eheien]

Only the code will be changed, the changes to issues, wiki, etc. will remain untouched. I’ll go ahead and bring Elliott’s repositories over today.

-Eric

On Nov 14, 2013, at 10:10 AM, David Luet notifications@github.com wrote:

If Dimitri is ok with it then that's what we should do.
Matthieu has spent an afternoon building the issues list. I wonder if there
is a way to save them.
Thanks,
David

On Thu, Nov 14, 2013 at 5:57 AM, komatits notifications@github.com wrote:

Fine with me.

Dimitri.

On 11/14/2013 06:53 AM, Eric Heien wrote:

Thanks for working on this Elliott. If it’s OK with David and the rest
of the SPECFEM team I’d be fine with replacing the existing SPECFEM2D/3D
Github repositories with your changes. If there are any problems, at
worst we can revert to the original based on the current SVN repository.
Does anyone have any concerns about this?

On Nov 13, 2013, at 9:42 PM, Elliott Sales de Andrade
notifications@github.com wrote:

I actually started this yesterday and finished it already. It's
available at https://github.com/QuLogic/specfem2d-clean as soon as it
finishes uploading.

The far history was pretty messy, but it turns out that the
triplicate commits are all redundant (I verified this by comparing all
commits that were duplicated). I just dropped the split section, and
that saved about 400 commits.

This cleanup drops the .git size from ~106M to ~37M, which is a nice
savings.

—
Reply to this email directly or view it on GitHub.

—
Reply to this email directly or view it on GitHub
<#1 (comment)
.

Dimitri Komatitsch
CNRS Research Director (DR CNRS), Laboratory of Mechanics and Acoustics,
UPR 7051, Marseille, France http://komatitsch.free.fr

—

Reply to this email directly or view it on GitHubhttps://github.com//issues/1#issuecomment-28475282
.

David Luet
Linux Administrator/Software & Programming Analyst
Princeton University
Department of Geosciences & PICSciE
PICSciE (9 am - 1 pm): 609-258-8194
Geosciences (1 - 5 pm): 609-258-7945
—
Reply to this email directly or view it on GitHub.

[eheien]

I just copied the SPECFEM2D and 3D clean repositories from Elliott into the official CIG organization repos at:

https://github.com/geodynamics/specfem2d
https://github.com/geodynamics/specfem3d

Please take a look at them and let us know if there are any problems. I still have copies of the original repositories so we can restore them quickly if there are any problems. Thanks to Elliott for doing the hard work of cleaning everything up.

On Nov 13, 2013, at 9:42 PM, Elliott Sales de Andrade notifications@github.com wrote:

I actually started this yesterday and finished it already. It's available at https://github.com/QuLogic/specfem2d-clean as soon as it finishes uploading.

The far history was pretty messy, but it turns out that the triplicate commits are all redundant (I verified this by comparing all commits that were duplicated). I just dropped the split section, and that saved about 400 commits.

This cleanup drops the .git size from ~106M to ~37M, which is a nice savings.

—
Reply to this email directly or view it on GitHub.

[QuLogic]

Looks good to me; the refs all seem to match what I have. We will need to send a message to the mailing list for everyone to recreate all their forks and clones.

[QuLogic]

This repo's cleanup looks complete.