- petsc-dev - mailman3.cels.anl.gov

Re: update to BuildSystem-dev for parmetis on Apple
by Victor Eijkhout 28 Jan '07

28 Jan '07

No module named Fiat. I did a pull of the source and buildsystem. Victor. On Jan 25, 2007, at 7:21 PM, Barry Smith wrote: > > I have made an update to BuildSystem-dev that allows the > --download-parmetis option to now work on the Apple. > > Barry > -- Victor Eijkhout, 512 471 5809 (w), 512 499 0504 (h) Texas Advanced Computing Center, The University of Texas at Austin

1 0

update to BuildSystem-dev for parmetis on Apple
by Barry Smith 25 Jan '07

25 Jan '07

I have made an update to BuildSystem-dev that allows the --download-parmetis option to now work on the Apple. Barry

1 0

removal of AOData PETSc object this week
by Barry Smith 05 Jan '07

05 Jan '07

We will be removing the AOData object from petsc-dev in the next week. If you are using it we recommend taking the code from petsc-dev and maintaining it yourself. We appologize for having to make this step but we cannot maintain the code and are developing an alternative approach. Note that the AO object is NOT being removed (nor will it be). Barry

1 0

new location for PETSc mercurial repositories
by Satish Balay 15 Dec '06

15 Dec '06

Folks, We have migrated to a new host for PETSc mercurial repositories. To start using this new location - just edit your .hg/hgrc [for each of the petsc/BuildSystem etc repoistories] - and make the following changes: http://hg.mcs.anl.gov/petsc/petsc-dev to http://petsc.cs.iit.edu/petsc/petsc-dev ssh://[email protected]//home/petsc/hg-repo/petsc-dev to ssh://[email protected]//hg/petsc/petsc-dev [Similarly changes with BuildSystem or other repositories] Satish

1 0

Mercurial 0.9.2 released! (fwd)
by Barry Smith 11 Dec '06

11 Dec '06

The latest release of the Mercurial distributed SCM is available at: http://selenic.com/mercurial/ New features: - merge now follows renames and copies - new layout protects against case-insensitivity issues - new branch and branches commands for managing named branches - push command accepts -r for pushing specified heads or named branches - proper storage of changelog and other metadata in UTF-8 - log, annotate and grep '--follow' follow renames and copies - date parsing is improved and log, update, and revert accept --date ranges - additional command options for log, status, addremove - improved schema for hgweb URLs - bundle can now use '-r' and '--base' removing the need of an base repository - support for git-style extended patches with --git option - new debuginstall command to check for common installation issues New contributions and extensions: - mq - support for quilt-style guards - can import existing changesets into mq ('qimport -r') or commit mq patches as regular changesets ('qdel -f') - edit the log message with 'qrefresh -e' - rename patches with 'qrename' - qheader to display the patch header of particular patches, and '--summary' for qseries, qapplied, qunapplied and qtop - combine patches with 'qfold' - qrefresh supports pattern options to import only a subset of the changes into a patch, to help split changes into multiple patches - patch names for applied patches act like local tags, and can be used in the revision arguments of any hg command - hgk - browse a subset of the history with '--limit' and revision range arguments - shows revision numbers as well as node hashes - churn: graph lines of code changed per user over a range of history - patchbomb: generate inline attachments with '-a' Behavior changes: - hg cat defaults to the working directory revision rather than tip - hg manifest no longer shows internal file revision hashes by default - hg revert now requires the -a flag to revert all files Developer notes: - new high-level API functions in the 'hg' module - new context API simplifies many operations - The changelog can now include arbitrary metadata in key: value form -- Mathematics is the supreme nostalgia of our time. _______________________________________________ Mercurial mailing list Mercurial(a)selenic.com http://selenic.com/mailman/listinfo/mercurial

1 0

recompile petsc-dev after next pull
by Barry Smith 08 Oct '06

08 Oct '06

petsc-dev users, After your next pull on petsc-dev you will need to recompile your PETSc libraries. I have changed the numerical values of the enum MatOption and strange stuff will happen without a recompile. Barry

1 0

New release 2.3.2 of PETSc
by Barry Smith 02 Sep '06

02 Sep '06

---------------------------------------------------------------------------- Portable, Extensible Toolkit for Scientific Computation (PETSc) ---------------------------------------------------------------------------- http://www.mcs.anl.gov/petsc We are pleased to announce the release of the PETSc 2.3.2 parallel software libraries for the implicit solution of PDEs and related problems. There are no MAJOR interface changes, the minor changes are listed at http://www-unix.mcs.anl.gov/petsc/petsc-as/documentation/changes/232.html please check these changes before compiling your codes. You can download from http://www-unix.mcs.anl.gov/petsc/petsc-as/download/index.html and and the installataion instructions are at http://www-unix.mcs.anl.gov/petsc/petsc-as/documentation/installation.html As always, please send bug reports, questions, and requests for new features to petsc-maint(a)mcs.anl.gov. Thanks for your continued support. The PETSc developers, Satish, Matt, Victor, Barry and Hong

1 0

changes to petsc-dev that make break things
by Barry Smith 30 Jul '06

30 Jul '06

I have pushed a set of changes to petsc-dev in 1) how the "AIJ" matrix type is handled (it is no longer a true type and when requested becomes simply a MPIAIJ or SEQAIJ; like BAIJ and SBAIJ have been done) 2) how the MatConvert() figures out how to convert between matrix types including eliminating the MatConvert registration process. Now all convert routines are found by querying the matrix 3) minor simplification of how Superlu, Spooles, etc matrix types are handled as subclasses of their parents. Note: generic conversion between any subclasses to any other is still broken (not broken but slow since it uses the MatConvert_Basic). I think unifying the subclassing process even more is possible. For example, it would be nice to have a common way of saving the "super" methods when the subclass is made, restoring them later and making sure the subclass data is properly destroyed when converting to the superclass or destroying the whole object. Oh well, maybe next year. If we have only one level of implementation inheritence (like we have now) this is pretty easy. Why did I do all this? So I can begin working on allowing setting the matrix type BEFORE setting the matrix sizes. This will allow simplifying a bunch of things. Note: none of the external matrix classes have been tested :-(! Barry

1 0

Fortran users of petsc-dev take heed
by Barry Smith 29 Jul '06

29 Jul '06

I have pushed a set of changes that will require you to run make allfortranstubs before recompiling all libraries; otherwise you will get some undefined symbols like MatPrintHelp(). Everyone should recompile all the libraries on their next pull. Barry

1 0

Re: no file comments for hg
by Yaron Kretchmer 25 Jul '06

25 Jul '06

QT4 has a QT3 "compatability layer". there's even a qt3to4 tool which looks at the source code and mutates all the QT3 calls to QT4 calls. It works quite well Yaron On 7/25/06, Barry Smith <bsmith(a)mcs.anl.gov> wrote: > > > Satish, > > Your proposed solution is ridiculously cumbersome; no sane person > would put up with that nonsense. > > If the authors of the ct would get their heads out of their > asses and update it to Qt4 (which is like, what, a year old) instead > of the outdated Qt3 maybe people who use updated software could use it. > > Barry > > > On Tue, 25 Jul 2006, Satish Balay wrote: > > > On Tue, 25 Jul 2006, Barry Smith wrote: > > > >> On Tue, 25 Jul 2006, Satish Balay wrote: > >> > >>> Yes - not having file comments makes it more difficult to go back in > >>> file history. > >>> > >>> But one way to look at it is - for some of these senarios - it meanas > >>> multiple changesets should be created [with the correct grouping of > >>> changes] instead of the atempted single changeset. > >> > >> Yes, the problem with this is that there is no decent way of > >> making multiple change-sets. > > > > - quit emacs without typing in any comments [so no commit is done] > > - 'hg status' to get the list of modified files > > - hg commit file1 file2 [ to make the appropriate changeset] > > > > I use 'hg ct' though.. > > > >> If I could just edit out the files > >> I did not want when doing the comment for a commit it would make > >> life a MILLION times better (could that be added?). > > > > I'll check on this on the mercurial list.. > > > > Satish > > > >> > >> Barry > >> > >>> > >>> And perhaps some comments can actually go into the code. > >>> > >>> Satish > >>> > >>> On Tue, 25 Jul 2006, Barry Smith wrote: > >>> > >>>> > >>>> Not having file by file comments for changesets is > >>>> starting to annoy me. > >>>> > >>>> Barry > >>>> > >>>> > >>> > >>> > >> > >> > > > > > >

1 0