Upgrade database_stein_watkins to the 2007 data

[cremona]

The original Stein-Watkins database was created in 2003. In 2007, a few revisions and additions were made to the data. In this ticket we upgrade the git optional/huge spkgs database_stein_watkins_mini and database_stein_watkins to version 20070827.

The replacement data files are at ​http://modular.math.washington.edu/ecdb/FIX.tar.

The new upstream tarballs are at:

1. ​http://boxen.math.washington.edu/home/cremona/database_stein_watkins_mini-20070827.tar.bz2
2. ​http://boxen.math.washington.edu/home/cremona/database_stein_watkins-20070827.tar.bz2

[cremona]

New commits:

​2bfd09e	trac #15607: update database_stein_watkins to 2007 data
​9936f24	Where possible, remove optional - database_stein_watkins
​3f48d25	Stein-Watkins database: reviewer patch
​2b283c7	trac 15556: correct package date and upstream tarball name
​6a35c8a	trac 15556: correct typo in previous commit
​38ef1d1	trac 15556: revised spkg-install following review
​747a385	trac 15556: converted database_stein_watkins to git
​9d098a6	trac 15556: converted database_stein_watkins_mini to git

[cremona]

Note: only the last commit (2bd09e) is not on the dependent #15556.

[jdemeyer]

Again, fix the permissions in the upstream tarballs (should be 0644).

[cremona]

Replying to jdemeyer:

Again, fix the permissions in the upstream tarballs (should be 0644).

Oops, sorry. The mini one is done, the big one takes longer...and is now done.

[jdemeyer]

The checksums must be updated now...

[git]

Branch pushed to git repo; I updated commit sha1. New commits:

​4cd8ed3

fixed checksums

[cremona]

I have updated the checksums.

[jdemeyer]

The checksums still do not match the tarballs:

Found local metadata for database_stein_watkins_mini-20070827
Found local sources at /usr/local/src/sage-2217/upstream/database_stein_watkins_mini-20070827.tar.bz2
Checksum: b4a664687ea7f2f77ae63070e7ba620dde06511c vs 4979191668e7cc8490eda75044435519faab0f9c
Invalid checksum for /usr/local/src/sage-2217/upstream/database_stein_watkins_mini-20070827.tar.bz2

Found local metadata for database_stein_watkins-20070827
Found local sources at /usr/local/src/sage-2217/upstream/database_stein_watkins-20070827.tar.bz2
Checksum: 8695b291912c4bbf07c5f6619f133b7985de6c83 vs 3cd5b8f38bc45c48b91229b05cb545801f6babcf
Invalid checksum for /usr/local/src/sage-2217/upstream/database_stein_watkins-20070827.tar.bz2

[jdemeyer]

The sha1sums of the tarballs are indeed:

jdemeyer@boxen:/home/cremona$ sha1sum database_stein_watkins*2007*.tar*
3cd5b8f38bc45c48b91229b05cb545801f6babcf  database_stein_watkins-20070827.tar.bz2
4979191668e7cc8490eda75044435519faab0f9c  database_stein_watkins_mini-20070827.tar.bz2

[cremona]

I discovered that the files on sagemath are not the same as the ones on my machine which explains this discrepancy (since I computed the checksums on my machine). Now I need to work out which files are correct! It is rather tedious since (for the large file) both bzip2/bunzip2 and of course the scp take rather a long time. Be patient...

[git]

Branch pushed to git repo; I updated commit sha1. New commits:

​8ecf61c

trac 15607: fixed checksums again

[cremona]

Try this. I scp's the bz2 files back to home where I reissued the sage-fix-pkg-checksums command. I think that earlier I had changed some files in the tarball on both my machine and sagemath and done the tar/bzip2 again on sagemath to save the file transfer. CLearly that does not work even if the directory contents are the same, perhaps because my username is different. This time I copied the tarballs from sagemath (which you have access to) back before recomputing the checksums.

Not a mistake I'll ever make again, I promise!

[jdemeyer]

Replying to cremona:

Try this. I scp's the bz2 files back to home where I reissued the sage-fix-pkg-checksums command. I think that earlier I had changed some files in the tarball on both my machine and sagemath and done the tar/bzip2 again on sagemath to save the file transfer. CLearly that does not work even if the directory contents are the same, perhaps because my username is different.

Also timestamps are stored in the tarball, that might matter.

[cremona]

Thanks! I don't expect there to be many updates to this one.