• Log graph

Commits in vsio/vsio:master

• Tuesday October 19 2010

  • 12:14

    

    guillemborrell committed 254f59d

    Now I even have a makefile

• Monday October 18 2010

  • 21:05

    

    guillemborrell committed 7bb6e36

    Trying to push a makefile that doesn't work yet

• Thursday October 14 2010

  • 16:22

    

    guillemborrell committed ce19739

    Print I/O speeds too

  • 16:11

    

    guillemborrell committed 831619f

    increased and variable size in the test program

• Monday August 23 2010

  • 11:18

    

    guillemborrell committed eb92e9d

    Added the project files for publishing

  • 11:04

    

    guillemborrell committed a88075f

    Changed the whole tree to be more logic friendly

• Friday August 20 2010

  • 17:53

    

    guillemborrell committed 301da41

    Despite noone uses this header, I have updated it to reflect the

  • 17:48

    

    guillemborrell committed 7f92f63

    Now vsio_typed correctly turns the comm and info handles to C. I don't

  • 12:35

    

    guillemborrell committed a4d8b4c

    Everything works correctly. Need for iso_c_binding outside the API

• Thursday August 19 2010

  • 19:23

    

    guillemborrell committed 3fe9bea

    Parallel serial work in fortran. Started parallel parallel in fortran

• Wednesday August 18 2010

  • 17:14

    

    guillemborrell committed 269d6a6

    h5dump_dds works with 1 dimensional arrays.

  • 15:17

    

    guillemborrell committed 1456dc8

    Compiles ok, but not tested yet.

  • 14:38

    

    guillemborrell committed 1ed7fc6

    Now much clearer. Every dummy argument is a pointer unless it is

  • 13:08

    

    guillemborrell committed 325e3d0

    It works for the const data. I dont't know how the value attribute

  • 12:40

    

    guillemborrell committed 3978bb0

    I get to create the dataset but the data doesn't find the pointer.

  • 11:03

    

    guillemborrell committed 3769154

    I start typed version to avoid fortran headaches. Library renamed

• Tuesday August 17 2010

  • 17:18

    

    guillemborrell committed 8726104

    Bug in serial writers solved. Just changing the order for allowing

  • 12:25

    

    guillemborrell committed 6a475bd

    Higher dimensions work for serial and parallel but I think I

  • 10:45

    

    guillemborrell committed 9695119

    Corrected a bug in dataset creation

  • 10:25

    

    guillemborrell committed 8cc5330

    load_var_parallel is able to read the file from dump_var_serial!!!

  • 09:30

    

    Guillem Borrell i Nogueras committed 15d2b10

    Parallell file creation seems to work

• Monday August 16 2010

  • 20:02

    

    guillemborrell committed 57d7fbe

    First prototype of parallel readers and writers

  • 14:49

    

    guillemborrell committed 0a4cd48

    Now the load var serial also works. I only need a parallel version of

• Friday August 13 2010

  • 17:08

    

    guillemborrell committed cb19b33

    The serial writers version is almost finished. You can change

  • 16:40

    

    guillemborrell committed 5f71828

    Seems that the serial version of the writer works properly. Now I can

  • 16:36

    

    guillemborrell committed f745503

    Only staging, but mpi does not work

• Tuesday August 10 2010

  • 19:01

    

    guillemborrell committed ae89238

    Fixed mpi thing in opensuse. Should be changed in other distributions

  • 17:14

    

    guillemborrell committed 892339d

    Second test passed. Now try to compile in the atom

  • 17:08

    

    guillemborrell committed d6e2541

    Tengo corriendo un ejemplo de la nueva API.

• Friday August 06 2010

  • 20:12

    

    guillemborrell committed 03c2bf0

    It seems that the whole hdf5 think works