Difference between revisions of "Release Details"

From Einstein Toolkit Documentation
Jump to: navigation, search
(add comment on common machines,)
(update bluewaters status)
Line 51: Line 51:
 
<tr><td colspan="2">Test all examples in the gallery (http://einsteintoolkit.org/about/gallery/) with the current thornlist, and update the thornlist mentioned on the page</td><td></td><td></td><td></td></tr>
 
<tr><td colspan="2">Test all examples in the gallery (http://einsteintoolkit.org/about/gallery/) with the current thornlist, and update the thornlist mentioned on the page</td><td></td><td></td><td></td></tr>
 
<tr><td rowspan="29">(Re)run Testsuites</td><td>datura  </td><td>as necessary</td><td>  </td><td>?</td></tr>
 
<tr><td rowspan="29">(Re)run Testsuites</td><td>datura  </td><td>as necessary</td><td>  </td><td>?</td></tr>
<tr>                                        <td>bluewaters </td><td>as necessary</td><td>Roland</td><td>good</td></tr>
+
<tr>                                        <td>bluewaters </td><td>as necessary</td><td>Roland</td><td>balsara4_2d passes, test_simpson has convergence order 4.0+/-1e-3, SphericalHarmonicRecon/regression_test likely accesses uninitialized data, the remaining tests are likely looking at data too close to singularities</td></tr>
 
<tr>                                        <td>forge    </td><td>as necessary</td><td>      </td><td>?</td></tr>
 
<tr>                                        <td>forge    </td><td>as necessary</td><td>      </td><td>?</td></tr>
 
<tr>                                        <td>hopper  </td><td>as necessary</td><td> </td><td>good</td></tr>
 
<tr>                                        <td>hopper  </td><td>as necessary</td><td> </td><td>good</td></tr>

Revision as of 23:34, 28 April 2014

Schedule for ET_2014_05

  • 2014-04-21: Slight freeze: no new major additions to the toolkit (new thorns or tools, major patches)
  • 2014-05-05: Feature freeze: no new features allowed anymore, extensive testing begins
  • 2014-05-12: Release branches are created, development version un-frozen again
  • 2014-05-15: Total freeze of release branches, preparation of tar-balls etc. begins
  • 2014-05-19: Target release data (or earlier if "ready")

Tools/Links

  • The release branch name will be ET_2014_05
  • The release name is not yet fixed
  • The list of proposed names for last release was
    • Lorentz
    • Tesla
    • Wheeler
    • DeWitt
    • Feynman
    • Hilbert
    • Turing
    • Brahe
    • Planck
  • Testsuite Status page (add the testsuite *.log files created by scripts below to the repository at [1] to have them automatically included)
  • Instructions for running the tests using SimFactory: Testsuite Machines
  • Release announcements: short and long
  • Release announced to: users@einsteintoolkit.org, {news|users}@cactuscode.org, Jennifer Claudet <jennifer@cct.lsu.edu> for AllCCT, http://hyperspace.aei.mpg.de/, http://astro-sim.org/
  • TODO update: List of commits since the last release: Einstein Toolkit ChangeLog
  • ChangeLog text
  • I would like to suggest that testing is performed on the latest version of common operating systems, for example Ubuntu, Fedora, Scientific Linux, Mac OS etc. People often want to build the ET on their laptop or workstation. We should provide somewhere instructions for which packages etc are needed. This information is probably already available in the minds of the current developers who use these systems already, and this should be written up more formally somewhere. This information could also be provided on the Cactus website. We could have simfactory configurations to go with this. (Ian) I usually (at irregular intevals) test the testsuites on Debian (stable), Ubuntu (latest) and Fedora (latest though that requires changes to simfactory I think) machines. This tests debian.cfg, ubuntu.cfg and fedora.cfg so that they work with the simplified instructions. I do not test on OSX (since I hardly use it). (Roland)

Release Plan

TaskDue dateResponsibleStatus
Regenerate all Kranc thorns (McLachlan, EinsteinExact and WeylScal4)
Test all examples in the gallery (http://einsteintoolkit.org/about/gallery/) with the current thornlist, and update the thornlist mentioned on the page
(Re)run Testsuitesdatura as necessary ?
bluewaters as necessaryRolandbalsara4_2d passes, test_simpson has convergence order 4.0+/-1e-3, SphericalHarmonicRecon/regression_test likely accesses uninitialized data, the remaining tests are likely looking at data too close to singularities
forge as necessary ?
hopper as necessary good
jenkins as necessary ?
lonestar as necessary
orca as necessary good
pandora as necessary ?
philip as necessary ?
queenbee as necessary updated 20121101
stampede as necessary ?
stampede-mic as necessary cannot run testsuite (file system is unreliable)
supermikeas necessary ?
tezpur as necessary updated 20121101
trestles as necessary updated 20121026
supermuc as necessary ?
vesta as necessary cannot run testsuite (job submission is weird)
zwicky as necessary
(private) amarok as necessaryPeter updated 20121101
(private) bethe as necessaryRoland
(private) cygnus as necessaryMatt ?
(private) newhorizonsas necessaryYosef?
(private) redshift as necessaryErik good
(private) spine as necessaryFrank ?
(private) titan as necessaryErik good
(private) topf as necessaryFrank √?
Fix Testsuites some thorn someone
Check & Improve Documentation (www, wiki)everyone
Write announcement texts (short and long)Frank
Create release branchsvn.einsteintoolkit.org Frank?
svn.cactuscode.org Frank?
carpetcode.org Frank?
aei.mpg.de Frank?
svn.cct.lsu.edu Frank?
github.com/ianhinder/Kranc.gitFrank?
github.com:gridaphobe/CRL.git Frank?
github.com:rhaas80/Outflow.git Frank?