]> https://gitweb.dealii.org/ - dealii.git/log
dealii.git
6 years agoMerge pull request #6048 from drwells/rectangular-full-matrix
Wolfgang Bangerth [Fri, 16 Mar 2018 23:00:54 +0000 (17:00 -0600)]
Merge pull request #6048 from drwells/rectangular-full-matrix

Rectangular full matrix

6 years agoMerge pull request #6047 from masterleinad/fix_64bit_dof
David Wells [Fri, 16 Mar 2018 17:48:58 +0000 (13:48 -0400)]
Merge pull request #6047 from masterleinad/fix_64bit_dof

Fix 64-bit dof type in test

6 years agoUpdate tests with new output. 6048/head
David Wells [Fri, 16 Mar 2018 17:25:22 +0000 (13:25 -0400)]
Update tests with new output.

Since the default test matrix is rectangular, most FullMatrix tests now
have different output.

6 years agoUpdate tests with identical output.
David Wells [Fri, 16 Mar 2018 17:21:14 +0000 (13:21 -0400)]
Update tests with identical output.

The FullMatrix tests now default to using rectangular matrices, but this
does not make sense for every test. This commit modifies tests so that
they use the old (square) test matrix.

6 years agoUpdate tests with truncated output.
David Wells [Fri, 16 Mar 2018 17:18:05 +0000 (13:18 -0400)]
Update tests with truncated output.

6 years agoUse rectangular matrices in FullMatrix tests.
David Wells [Fri, 16 Mar 2018 17:22:09 +0000 (13:22 -0400)]
Use rectangular matrices in FullMatrix tests.

This catches more bugs.

6 years agoFix some assertions in FullMatrix.
David Wells [Fri, 16 Mar 2018 16:03:23 +0000 (12:03 -0400)]
Fix some assertions in FullMatrix.

6 years agoFix some (rectangular) FullMatrix methods.
David Wells [Thu, 15 Mar 2018 20:03:12 +0000 (16:03 -0400)]
Fix some (rectangular) FullMatrix methods.

6 years agoFix 64-bit dof type in test 6047/head
Daniel Arndt [Fri, 16 Mar 2018 11:00:21 +0000 (12:00 +0100)]
Fix 64-bit dof type in test

6 years agoMerge pull request #6023 from bangerth/parallelize
Daniel Arndt [Tue, 13 Mar 2018 18:40:33 +0000 (19:40 +0100)]
Merge pull request #6023 from bangerth/parallelize

Parallelize FETools::compute_projection_matrices()

6 years agoMerge pull request #6044 from luca-heltai/coupling-distributed
Daniel Arndt [Tue, 13 Mar 2018 16:09:28 +0000 (17:09 +0100)]
Merge pull request #6044 from luca-heltai/coupling-distributed

Distributed coupling.

6 years agoUpdated documentation for parallel triangulations. 6044/head
Luca Heltai [Tue, 13 Mar 2018 09:06:46 +0000 (10:06 +0100)]
Updated documentation for parallel triangulations.

6 years agoAdded distributed test.
Luca Heltai [Tue, 13 Mar 2018 09:03:45 +0000 (10:03 +0100)]
Added distributed test.

6 years agoMerge pull request #6043 from Rombur/xsdk
Jean-Paul Pelteret [Tue, 13 Mar 2018 06:17:12 +0000 (07:17 +0100)]
Merge pull request #6043 from Rombur/xsdk

Do not use macros in the cuda matrix-free

6 years agoMerge pull request #6042 from MFraters/add_change_entry_for_json_parser_for_input_files
Wolfgang Bangerth [Tue, 13 Mar 2018 02:16:33 +0000 (20:16 -0600)]
Merge pull request #6042 from MFraters/add_change_entry_for_json_parser_for_input_files

add change entry for the add_json_parser_for_input_files pull request.

6 years agoDo not use macros in the cuda matrix-free 6043/head
Bruno Turcksin [Tue, 13 Mar 2018 00:58:31 +0000 (20:58 -0400)]
Do not use macros in the cuda matrix-free

6 years agoadd change entry for the add_json_parser_for_input_files pull request. 6042/head
MFraters [Mon, 12 Mar 2018 13:26:41 +0000 (14:26 +0100)]
add change entry for the add_json_parser_for_input_files pull request.

6 years agoMerge pull request #6040 from bangerth/doc-update-3
Daniel Arndt [Mon, 12 Mar 2018 21:54:44 +0000 (22:54 +0100)]
Merge pull request #6040 from bangerth/doc-update-3

Improve readability of doxygen-generated pages.

6 years agoMerge pull request #6031 from bangerth/update-step-14
Daniel Arndt [Mon, 12 Mar 2018 20:29:45 +0000 (21:29 +0100)]
Merge pull request #6031 from bangerth/update-step-14

Don't use SynchronousIterator in step-9, 14

6 years agoMerge pull request #6036 from luzpaz/misc-typos
David Wells [Mon, 12 Mar 2018 16:31:46 +0000 (12:31 -0400)]
Merge pull request #6036 from luzpaz/misc-typos

Misc. typos

6 years agoMerge pull request #6028 from kronbichler/improve_hermite_poly
Bruno Turcksin [Mon, 12 Mar 2018 15:02:21 +0000 (11:02 -0400)]
Merge pull request #6028 from kronbichler/improve_hermite_poly

Make Hermite-like polynomials represent constant by (1,1,...1)

6 years agoDistributed coupling now working.
Luca Heltai [Mon, 12 Mar 2018 13:56:05 +0000 (14:56 +0100)]
Distributed coupling now working.

6 years agoMerge pull request #5999 from luca-heltai/coupling
Wolfgang Bangerth [Mon, 12 Mar 2018 12:10:10 +0000 (06:10 -0600)]
Merge pull request #5999 from luca-heltai/coupling

Non-matching sparsity and mass matrix.

6 years agoChange weight in Hermite-like polynomials to make constant represented by (1,1,1... 6028/head
Martin Kronbichler [Sat, 10 Mar 2018 10:41:10 +0000 (11:41 +0100)]
Change weight in Hermite-like polynomials to make constant represented by (1,1,1,...1).

6 years agoSource typos 6036/head
luz.paz [Mon, 12 Mar 2018 10:56:36 +0000 (06:56 -0400)]
Source typos

6 years agoFound more typos
luz.paz [Mon, 12 Mar 2018 05:03:33 +0000 (01:03 -0400)]
Found more typos

6 years agoRevision
luz.paz [Mon, 12 Mar 2018 04:31:11 +0000 (00:31 -0400)]
Revision

6 years agoMisc. typos
luz.paz [Sun, 11 Mar 2018 17:32:36 +0000 (13:32 -0400)]
Misc. typos

Found via `codespell`

6 years agoMerge pull request #6032 from bangerth/doc-comm-world
Martin Kronbichler [Mon, 12 Mar 2018 11:33:32 +0000 (12:33 +0100)]
Merge pull request #6032 from bangerth/doc-comm-world

Document the use of MPI_COMM_WORLD in MPI_InitFinalize.

6 years agoMerge pull request #6041 from masterleinad/allow_valgrind
Wolfgang Bangerth [Mon, 12 Mar 2018 11:28:01 +0000 (05:28 -0600)]
Merge pull request #6041 from masterleinad/allow_valgrind

Add missing file used in #6035

6 years agoMerge pull request #6038 from bangerth/doc-update-2
Martin Kronbichler [Mon, 12 Mar 2018 10:56:29 +0000 (11:56 +0100)]
Merge pull request #6038 from bangerth/doc-update-2

Fix doxygen markup.

6 years agoMerge pull request #6033 from bangerth/doc-update
Martin Kronbichler [Mon, 12 Mar 2018 10:51:48 +0000 (11:51 +0100)]
Merge pull request #6033 from bangerth/doc-update

Document why using MPI_COMM_WORLD is ok in one place.

6 years agoMerge pull request #5993 from MFraters/add_json_parser_for_input_files
Daniel Arndt [Mon, 12 Mar 2018 10:00:35 +0000 (11:00 +0100)]
Merge pull request #5993 from MFraters/add_json_parser_for_input_files

Add json parser for input files

6 years agoAdd missing file used in #6035 6041/head
Daniel Arndt [Mon, 12 Mar 2018 09:35:01 +0000 (10:35 +0100)]
Add missing file used in #6035

6 years agoMerge pull request #6035 from masterleinad/allow_valgrind
Matthias Maier [Mon, 12 Mar 2018 06:24:10 +0000 (01:24 -0500)]
Merge pull request #6035 from masterleinad/allow_valgrind

Allow to run the testsuite with valgrind

6 years agoFix a couple of places in the documentation of a function. 5993/head
Wolfgang Bangerth [Mon, 12 Mar 2018 04:20:23 +0000 (22:20 -0600)]
Fix a couple of places in the documentation of a function.

6 years agoImprove readability of doxygen-generated pages. 6040/head
Wolfgang Bangerth [Mon, 12 Mar 2018 03:21:40 +0000 (21:21 -0600)]
Improve readability of doxygen-generated pages.

6 years agoFix doxygen markup. 6038/head
Wolfgang Bangerth [Mon, 12 Mar 2018 03:03:25 +0000 (21:03 -0600)]
Fix doxygen markup.

6 years agoDocument why using MPI_COMM_WORLD is ok in one place. 6033/head
Wolfgang Bangerth [Sun, 11 Mar 2018 07:06:29 +0000 (00:06 -0700)]
Document why using MPI_COMM_WORLD is ok in one place.

6 years agoMerge pull request #6025 from masterleinad/fix_clang_6_performance
Timo Heister [Sun, 11 Mar 2018 18:45:26 +0000 (14:45 -0400)]
Merge pull request #6025 from masterleinad/fix_clang_6_performance

Fix llvm-6 clang-tidy performance findings

6 years agoAllow to run the testsuite with valgrind 6035/head
Daniel Arndt [Sun, 11 Mar 2018 16:02:40 +0000 (17:02 +0100)]
Allow to run the testsuite with valgrind

6 years agoMerge pull request #6034 from bangerth/add-test
Daniel Arndt [Sun, 11 Mar 2018 13:14:09 +0000 (14:14 +0100)]
Merge pull request #6034 from bangerth/add-test

Add a test that works on only a subset of MPI processes.

6 years agoMerge pull request #5998 from bangerth/update-example-output
Denis Davydov [Sun, 11 Mar 2018 10:42:42 +0000 (16:12 +0530)]
Merge pull request #5998 from bangerth/update-example-output

Update a couple of things on how we generate output in the example steps

6 years agoMerge pull request #6022 from bangerth/version-string
Daniel Arndt [Sun, 11 Mar 2018 09:38:46 +0000 (10:38 +0100)]
Merge pull request #6022 from bangerth/version-string

Provide a function that returns the version string.

6 years agoFix llvm-6 clang-tidy performance findings 6025/head
Daniel Arndt [Sat, 10 Mar 2018 05:44:28 +0000 (06:44 +0100)]
Fix llvm-6 clang-tidy performance findings

6 years agoAdd a test that works on only a subset of MPI processes. 6034/head
Wolfgang Bangerth [Sun, 11 Mar 2018 07:26:47 +0000 (00:26 -0700)]
Add a test that works on only a subset of MPI processes.

The test is intended to verify that we don't have stray references to MPI_COMM_WORLD.

6 years agoDocument the use of MPI_COMM_WORLD in MPI_InitFinalize. 6032/head
Wolfgang Bangerth [Sun, 11 Mar 2018 06:57:37 +0000 (23:57 -0700)]
Document the use of MPI_COMM_WORLD in MPI_InitFinalize.

6 years agoUpdate a couple of things on how we generate output in the example steps. 5998/head
Wolfgang Bangerth [Sun, 4 Mar 2018 08:23:50 +0000 (01:23 -0700)]
Update a couple of things on how we generate output in the example steps.

6 years agoAvoid the use of SynchronousIterator in step-9. 6031/head
Wolfgang Bangerth [Sat, 10 Mar 2018 16:39:41 +0000 (09:39 -0700)]
Avoid the use of SynchronousIterator in step-9.

While there also mark a solution vector as a reference to avoid copying it.

6 years agoAvoid the use of SynchronousIterator in step-14.
Wolfgang Bangerth [Sat, 10 Mar 2018 16:39:10 +0000 (09:39 -0700)]
Avoid the use of SynchronousIterator in step-14.

6 years agoPoint to the only tutorial program that uses this class.
Wolfgang Bangerth [Sat, 10 Mar 2018 16:23:47 +0000 (09:23 -0700)]
Point to the only tutorial program that uses this class.

6 years agoAdd a changelog entry. 6022/head
Wolfgang Bangerth [Sat, 10 Mar 2018 05:35:18 +0000 (22:35 -0700)]
Add a changelog entry.

6 years agoProvide a function that returns the version string.
Wolfgang Bangerth [Sat, 10 Mar 2018 05:34:58 +0000 (22:34 -0700)]
Provide a function that returns the version string.

6 years agoMerge pull request #6029 from bangerth/update-step-14
Daniel Arndt [Sat, 10 Mar 2018 20:30:51 +0000 (15:30 -0500)]
Merge pull request #6029 from bangerth/update-step-14

Avoid re-init'ing the same vector twice in step-14.

6 years agoAvoid re-init'ing the same vector twice in step-14. 6029/head
Wolfgang Bangerth [Sat, 10 Mar 2018 15:54:44 +0000 (08:54 -0700)]
Avoid re-init'ing the same vector twice in step-14.

6 years agoMerge pull request #6027 from bangerth/doc-update
David Wells [Sat, 10 Mar 2018 13:25:42 +0000 (08:25 -0500)]
Merge pull request #6027 from bangerth/doc-update

Minor doc update to step-35.

6 years agoMerge pull request #6003 from masterleinad/relaxation_block_set_kernel
Martin Kronbichler [Sat, 10 Mar 2018 10:27:45 +0000 (11:27 +0100)]
Merge pull request #6003 from masterleinad/relaxation_block_set_kernel

Allow for setting the size of the kernel instead of a threshold in RelaxationBlock

6 years agoMinor doc update to step-35. 6027/head
Wolfgang Bangerth [Sat, 10 Mar 2018 10:13:51 +0000 (03:13 -0700)]
Minor doc update to step-35.

6 years agoMerge pull request #6024 from kronbichler/remove_mpi_comm_world
Martin Kronbichler [Sat, 10 Mar 2018 10:04:54 +0000 (11:04 +0100)]
Merge pull request #6024 from kronbichler/remove_mpi_comm_world

Do not use MPI_COMM_WORLD in MG Transfer

6 years agoRemove use of MPI_COMM_WORLD in serial Trilinos vector. 6024/head
Martin Kronbichler [Sat, 10 Mar 2018 06:22:05 +0000 (07:22 +0100)]
Remove use of MPI_COMM_WORLD in serial Trilinos vector.

6 years agoDo not use MPI_COMM_WORLD.
Martin Kronbichler [Sat, 10 Mar 2018 06:10:18 +0000 (07:10 +0100)]
Do not use MPI_COMM_WORLD.

6 years agoRun the computation of projections from children in parallel. 6023/head
Wolfgang Bangerth [Sat, 10 Mar 2018 05:47:10 +0000 (22:47 -0700)]
Run the computation of projections from children in parallel.

6 years agoBreak the main part of a loop out of the loop and put it into a lambda.
Wolfgang Bangerth [Sat, 10 Mar 2018 05:38:06 +0000 (22:38 -0700)]
Break the main part of a loop out of the loop and put it into a lambda.

6 years agoMerge pull request #6020 from bangerth/doc-update
Bruno Turcksin [Fri, 9 Mar 2018 23:00:13 +0000 (18:00 -0500)]
Merge pull request #6020 from bangerth/doc-update

Improve documentation of FETools::compute_projection_matrices().

6 years agoMerge pull request #6014 from kronbichler/improve_hermite_like_interp
Bruno Turcksin [Fri, 9 Mar 2018 22:58:58 +0000 (17:58 -0500)]
Merge pull request #6014 from kronbichler/improve_hermite_like_interp

Extend Hermite-like interpolation to case degree=2

6 years agoMerge pull request #6021 from tjhei/cuda_documentation_fixes
Bruno Turcksin [Fri, 9 Mar 2018 22:45:37 +0000 (17:45 -0500)]
Merge pull request #6021 from tjhei/cuda_documentation_fixes

CUDA: doxygen fixes

6 years agoCUDA: doxygen fixes 6021/head
Timo Heister [Fri, 9 Mar 2018 19:32:26 +0000 (14:32 -0500)]
CUDA: doxygen fixes

- disable DEAL_II_ALWAYS_INLINE and __device__ in doxygen documentation
- avoid @ingroup for member functions because they will be moved not
copied to the group

6 years agoImprove documentation of FETools::compute_projection_matrices(). 6020/head
Wolfgang Bangerth [Fri, 9 Mar 2018 15:09:17 +0000 (08:09 -0700)]
Improve documentation of FETools::compute_projection_matrices().

6 years agoMerge pull request #6012 from tjhei/cuda_warnings
Bruno Turcksin [Fri, 9 Mar 2018 14:09:38 +0000 (09:09 -0500)]
Merge pull request #6012 from tjhei/cuda_warnings

CUDA: fix cmake 3.10 and remove warning

6 years agoMerge pull request #6005 from jppelteret/fix_vector_from_trilinos_vector
Martin Kronbichler [Fri, 9 Mar 2018 03:39:26 +0000 (04:39 +0100)]
Merge pull request #6005 from jppelteret/fix_vector_from_trilinos_vector

Fix Vector::operator=(TrilinosWrappers::MPI::Vector)

6 years agoExtend Hermite-like interpolation to case degree=2. 6014/head
Martin Kronbichler [Thu, 8 Mar 2018 23:25:36 +0000 (00:25 +0100)]
Extend Hermite-like interpolation to case degree=2.

6 years agoMerge pull request #6013 from tjhei/lapack_warnings
Jean-Paul Pelteret [Thu, 8 Mar 2018 20:24:10 +0000 (21:24 +0100)]
Merge pull request #6013 from tjhei/lapack_warnings

remove LAPACK warning about uninitialized info variable

6 years agoremove LAPACK warning about uninitialized info variable 6013/head
Timo Heister [Thu, 8 Mar 2018 17:21:40 +0000 (12:21 -0500)]
remove LAPACK warning about uninitialized info variable

6 years agoCUDA: fix cmake 3.10 and remove warning 6012/head
Timo Heister [Thu, 8 Mar 2018 17:11:14 +0000 (12:11 -0500)]
CUDA: fix cmake 3.10 and remove warning

6 years agoMerge pull request #6009 from bangerth/doc-update-13
Matthias Maier [Thu, 8 Mar 2018 17:11:07 +0000 (11:11 -0600)]
Merge pull request #6009 from bangerth/doc-update-13

Better document one place.

6 years agoMerge pull request #6008 from bangerth/cleanup-2
Matthias Maier [Thu, 8 Mar 2018 17:10:35 +0000 (11:10 -0600)]
Merge pull request #6008 from bangerth/cleanup-2

Better document a piece of code.

6 years agoMerge pull request #6007 from bangerth/cleanup
Matthias Maier [Thu, 8 Mar 2018 17:10:05 +0000 (11:10 -0600)]
Merge pull request #6007 from bangerth/cleanup

Simplify code.

6 years agoMerge pull request #6011 from tamiko/add-gitignore
Wolfgang Bangerth [Thu, 8 Mar 2018 16:50:13 +0000 (09:50 -0700)]
Merge pull request #6011 from tamiko/add-gitignore

Add a .gitignore file

6 years agoAdd a .gitignore file 6011/head
Matthias Maier [Thu, 8 Mar 2018 16:46:16 +0000 (10:46 -0600)]
Add a .gitignore file

Let's ignore the "code-gallery" directory and all directories starting
with "build*".

6 years agoAdd test for the read json parameter handler.
MFraters [Thu, 8 Mar 2018 15:59:32 +0000 (16:59 +0100)]
Add test for the read json parameter handler.

6 years agoadd json parser for input files.
MFraters [Sun, 4 Mar 2018 02:25:43 +0000 (03:25 +0100)]
add json parser for input files.

6 years agoBetter document one place. 6009/head
Wolfgang Bangerth [Thu, 8 Mar 2018 14:54:25 +0000 (07:54 -0700)]
Better document one place.

6 years agoBetter document a piece of code. 6008/head
Wolfgang Bangerth [Thu, 8 Mar 2018 14:41:43 +0000 (07:41 -0700)]
Better document a piece of code.

6 years agoSimplify code. 6007/head
Wolfgang Bangerth [Thu, 8 Mar 2018 14:37:19 +0000 (07:37 -0700)]
Simplify code.

The existing code seemed to suggest that something different had to happen if n_levels()>0.
But it was really that there was nothing to do if the triangulation is empty. Reflect this
in both the code and a comment.

6 years agoMerge pull request #6006 from tjhei/cuda_doc
Bruno Turcksin [Thu, 8 Mar 2018 13:37:25 +0000 (08:37 -0500)]
Merge pull request #6006 from tjhei/cuda_doc

create CUDA doxygen module

6 years agoMoved some header files to .cc 5999/head
Luca Heltai [Thu, 8 Mar 2018 07:46:37 +0000 (08:46 +0100)]
Moved some header files to .cc

6 years agoAllow for setting the size of the kernel instead of a threshold 6003/head
Daniel Arndt [Tue, 6 Mar 2018 04:39:24 +0000 (05:39 +0100)]
Allow for setting the size of the kernel instead of a threshold

6 years agocuda doxygen 6006/head
Timo Heister [Wed, 7 Mar 2018 21:10:43 +0000 (16:10 -0500)]
cuda doxygen

6 years agoFix Vector::operator=(TrilinosWrappers::MPI::Vector) 6005/head
Jean-Paul Pelteret [Wed, 7 Mar 2018 16:54:25 +0000 (17:54 +0100)]
Fix Vector::operator=(TrilinosWrappers::MPI::Vector)

Data was being copied to the local vector directly from the distributed
vector (as opposed to its localised equivalent). This lead to only a
partial copy of data, and all of the data was offset incorrectly in the
returned vector (giving each process a different view of the localised
vector).

6 years agoMerge pull request #6004 from masterleinad/avoid_warnings_spherical
Denis Davydov [Wed, 7 Mar 2018 09:09:06 +0000 (10:09 +0100)]
Merge pull request #6004 from masterleinad/avoid_warnings_spherical

Avoid compiler warnings for Function::Spherical

6 years agoMerge pull request #5995 from bangerth/doc-update
Martin Kronbichler [Wed, 7 Mar 2018 06:41:27 +0000 (07:41 +0100)]
Merge pull request #5995 from bangerth/doc-update

Expand on the discussion in the introduction to step-14.

6 years agoAvoid compiler warnings for Function::Spherical 6004/head
Daniel Arndt [Wed, 7 Mar 2018 04:19:12 +0000 (05:19 +0100)]
Avoid compiler warnings for Function::Spherical

6 years agoMerge pull request #5732 from Jie-Cheng/petsc_mmult_follow_up
Daniel Arndt [Wed, 7 Mar 2018 03:14:51 +0000 (22:14 -0500)]
Merge pull request #5732 from Jie-Cheng/petsc_mmult_follow_up

Follow-up of #5655: add mmult and Tmmult to petsc sparse matrix and petsc parallel sparse matrix

6 years agoFollow-up of #5655: add PETScWrappers::SparseMatrix::mmult and PETScWrappers::MPI... 5732/head
Jie Cheng [Sun, 14 Jan 2018 06:08:47 +0000 (01:08 -0500)]
Follow-up of #5655: add PETScWrappers::SparseMatrix::mmult and PETScWrappers::MPI::SparseMatrix::mmult

6 years agoSplit coupling.h into coupling.cc and coupling.inst
Luca Heltai [Tue, 6 Mar 2018 11:08:09 +0000 (12:08 +0100)]
Split coupling.h into coupling.cc and coupling.inst

6 years agoAdded SPARSE_MATRICES to explicit instantiations.
Luca Heltai [Tue, 6 Mar 2018 11:07:37 +0000 (12:07 +0100)]
Added SPARSE_MATRICES to explicit instantiations.

6 years agoMerge pull request #6000 from masterleinad/changelog_5997
David Wells [Mon, 5 Mar 2018 17:02:19 +0000 (12:02 -0500)]
Merge pull request #6000 from masterleinad/changelog_5997

Changelog entry for #5997

6 years agoChangelog.
Luca Heltai [Mon, 5 Mar 2018 15:40:27 +0000 (16:40 +0100)]
Changelog.

6 years agoChangelog entry for #5997 6000/head
Daniel Arndt [Mon, 5 Mar 2018 14:57:31 +0000 (15:57 +0100)]
Changelog entry for #5997

6 years agoNon matching sparsity and mass matrix.
Luca Heltai [Mon, 5 Mar 2018 14:34:20 +0000 (15:34 +0100)]
Non matching sparsity and mass matrix.


In the beginning the Universe was created. This has made a lot of people very angry and has been widely regarded as a bad move.

Douglas Adams


Typeset in Trocchi and Trocchi Bold Sans Serif.