Hi,
this is a reminder about our weekly ET phone call on Monday (today), 10am central time.
We will use the usual phone line:
Each participant of the teleconference will call 578-4942 (225-578-4942 from off-campus or 866-573-0 359 toll free).
They will hear: Welcome to Voice Plus Conferencing. Please enter your Conference ID. Press the # key when complete.
Participants will press 118682# for Conference Line one.
They will hear: Welcome to Voice Plus Conferencing. For conference chairperson services, press the # key. All other participants please wait.
“All other participants” will wait on the line until the chairperson has ‘entered’ the teleconferenc e.
Frank
Hi,
this is a (late) reminder about our weekly ET phone call on Monday (today), 10am central time, in about 10 minutes.
Main topic will be the release this week.
Frank
We will use the usual phone line:
Each participant of the teleconference will call 578-4942 (225-578-4942 from off-campus or 866-573-0359 toll free).
They will hear: Welcome to Voice Plus Conferencing. Please enter your Conference ID. Press the # key when complete.
Participants will press 118682# for Conference Line one.
They will hear: Welcome to Voice Plus Conferencing. For conference chairperson services, press the # key. All other participants please wait.
“All other participants” will wait on the line until the chairperson has ‘entered’ the teleconference.
On Mon, Jun 14, 2010 at 09:50:56AM -0500, Frank Loeffler wrote:
Main topic will be the release this week.
The most important page to coordinate the release will be
http://docs.einsteintoolkit.org/et-docs/Release_Details
and links mentioned there. We will discuss their use during the call.
Frank
Hello all,
here are the notes of today's meeting:
The main topics were how to split the remaining work running the test suites for the upcoming release of the Einstein Toolkit as well as how to split work for MHD.
Present were Ian Hinder, Eloisa, Frank, Scott, Tanja, Roland, Yoseph, Christian Ott, Bruno Mundim
The initial part of the call was taken up discussing the Einstein Toolkit release plan and assigning machines for running test suites to maintainers. See http://docs.einsteintoolkit.org/et-docs/Release_Details for results
* Eloisa provided a status update for the failing testsuites in QuasiLocalMeasures, which she can make work up to issues that probably stem from AHFinderDirect since they differ when running on a single or multiple processors
* the string used to identify the release branch will be of the form ET_YYYY_MM (without an explicit release number). This name will continued to be used even if we fix a critical bug months later.
* we agreed to run the testsuites until Wednesday, then create the release branch, and test the release branch again on a small number of machines
* Ian has some testsuites for McLachlan about which there will be a separate discussion on the mailing list
* we will have a short call on Wednesday to sort out final issues before the release
Concerning MHD:
* a table outlining the split of responsibilities is at http://docs.einsteintoolkit.org/et-docs/MHD_implementation_details_and_discu...
* for now we will stick to the second-order polynomial approximation to the eigenvalues used in the HLL solver
* the MHD developers will have a discussion as to whether to have an explicit GRHydro::Bcom variable or to re-use HydroBase::Bvec for this purpose together with a grid scalar to indicate which state Bvec is in
* we agreed to create a wiki page for MHD
Yours, Roland
On Mon, Jun 14, 2010 at 01:23:50PM -0400, Roland Haas wrote:
- we will have a short call on Wednesday to sort out final issues
before the release
We will not have that call today, nor will we have a release today. I will send an update a bit later today.
In short, the missing 'must be there' at the moment are: - We need to create branches in all repositories (today) to be able to test that - We need branch support for GetComponents (also today)
The testsuites are in good shape (although not excellent still), so that while it might be nice to improve here, this is not a release blocker anymore. Please update the results nevertheless if there are changes happening.
As written - a more detailed plan will be send out later.
Frank
On 16 Jun 2010, at 18:22, Frank Loeffler wrote:
On Mon, Jun 14, 2010 at 01:23:50PM -0400, Roland Haas wrote:
- we will have a short call on Wednesday to sort out final issues
before the release
We will not have that call today, nor will we have a release today. I will send an update a bit later today.
In short, the missing 'must be there' at the moment are:
- We need to create branches in all repositories (today) to be able to test that
I have just added an ET_2010_06 branch to Kranc, and Barry has added one to AEIThorns/AEILocalInterp, which is the only AEI repository in the Einstein Toolkit thornlist.
Hi,
the current plan is as following:
There will be a 'hard freeze' tomorrow, June 17th at 2pm CDT. That means by then nothing can go into any ET repository (everything inside the ET thornlist) unless approved by me - which will only happen if that change is really necessary for the release.
We will have a evo (?) meeting at that time: 2pm, talking about the remaining issues with the release.
After that we will create the remaining release branches and test this one a couple (not all) machines to make sure this worked and GetComponents can handle this. Note that AEIThorns and Kranc already have those branches: please refrain from commits to those already now.
The test cases are in good enough shape for a release, but it would be nice if the results on a couple of machines could be updated.
Please indicate any issue / progress with looking at the documentation.
Frank
users@lists.einsteintoolkit.org
-
Frank Loeffler -
Ian Hinder -
Roland Haas