| 1 |
$Header: /u/gcmpack/MITgcm/doc/tag-index,v 1.53.6.3 2003/01/24 05:00:36 dimitri Exp $ |
| 2 |
|
| 3 |
Notes on tags used in MITgcmUV |
| 4 |
============================== |
| 5 |
|
| 6 |
Branch release1_coupled contains code for coupling |
| 7 |
to UCLA atmospheric model: |
| 8 |
|
| 9 |
updated useSingleCpuIO option in mdsio_writefield.F |
| 10 |
to work with multi-field files, e.g., the pickup files |
| 11 |
|
| 12 |
Added lower bound diffusivity to kpp_routines.F (see !!!NOTE!!!). |
| 13 |
|
| 14 |
cvs tag ucla1 |
| 15 |
Branched off tag release1_beta1 |
| 16 |
Modified eesupp/src/eeboot_minimal.F, eedie.F, exch_jam.F, |
| 17 |
global_max.F, global_sum.F, gsum_jam.F, ini_procs.F, and main.F |
| 18 |
model/src/dynamics.F, forward_step.F, freeze.F, ini_parms.F, |
| 19 |
set_defaults.F, the_main_loop.F, the_model_main.F, timestep.F, |
| 20 |
write_state.F, and mdsio_writefield.F |
| 21 |
pkg/timeave/TIMEAVE_STATV.h, timeave_statv_write.F, |
| 22 |
timeave_statvars.F, and model/inc/PARAMS.h, |
| 23 |
Added model/src/do_coupled_ucla.F and verification/coupled_ucla |
| 24 |
NOTES: |
| 25 |
MPI_COMM_WORLD -> MPI_COMM_MODEL in |
| 26 |
exch_jam.F, global_max.F global_sum.F, and gsum_jam.F |
| 27 |
Added useSingleCpuIO: ini_procs.F, PARAMS.h, ini_parms.F, |
| 28 |
set_defaults.F, mdsio_writefield.F, and subroutine |
| 29 |
GATHER_2D in do_coupled_ucla.F |
| 30 |
|
| 31 |
================================================================= |
| 32 |
|
| 33 |
checkpoint43 |
| 34 |
o added two new advection tests |
| 35 |
o added global 4x4 degree ocean test |
| 36 |
o added test of H&S94 with free surface on cube |
| 37 |
o some final mods for gradient check, adjoint,... |
| 38 |
|
| 39 |
checkpoint42 |
| 40 |
Fixed AD-related problems: |
| 41 |
o Store directives up-to-date with re-arranged Adams-Bashforth |
| 42 |
(mainly thermodynamics.F) |
| 43 |
o New store directives for multi-dim. advection schemes |
| 44 |
* new CPP flag ALLOW_MULTI_DIM_ADVECTION |
| 45 |
* new common block and key passkey |
| 46 |
(mainly gad_advection.F) |
| 47 |
o Modified store directives for split of dynamics/thermodynamics |
| 48 |
for the case ALLOW_KPP |
| 49 |
o Cleaned argument list for timestep_tracer.F |
| 50 |
|
| 51 |
checkpoint41 |
| 52 |
o re-formatted/added comments in prologues of key subroutines |
| 53 |
to work with protex |
| 54 |
|
| 55 |
checkpoint40 |
| 56 |
o moved calc_gs, calc_gt and calc_tr1 back to model/src |
| 57 |
o added FLT (floats) package from Arne along with demo. experiment |
| 58 |
o re-arranged Adams-Bashforth scheme for tracers to allow easier |
| 59 |
implementation of forward-in-time forcing terms |
| 60 |
o more mods for non-linear free-surface including new variable |
| 61 |
o modified multi-dim method to work on cube |
| 62 |
o bug fix in KPP (from DM) |
| 63 |
|
| 64 |
checkpoint40pre9 |
| 65 |
o added DST advection schemes |
| 66 |
o fix MPI bug on SGI's |
| 67 |
o modifications for non-linear free-surface |
| 68 |
o fixed conditionals for fixed flow/passive tracer only mode |
| 69 |
o added tension/strain form of stress tensor |
| 70 |
o changes to commenting of arguments for TAF/TAMC |
| 71 |
o added/updated many matlab scripts |
| 72 |
o added multi-dimensional advection method for tracers |
| 73 |
|
| 74 |
checkpoint40pre8 |
| 75 |
o new package rw/ |
| 76 |
Added method for dumping intermediate local arrays: |
| 77 |
mdsio_writetile - same as mdsio_writefield except works from inside |
| 78 |
bi,bj loop |
| 79 |
mdsio_writelocal - same as mdsio_writetile except works for local |
| 80 |
arrays |
| 81 |
write_local_r? - higher-level wrapper for mdsio_writelocal |
| 82 |
Controlled by diagFreq. Defaults to zero (ie. no dumps) |
| 83 |
Example given at end of mom_vecinv.F that dumps some local arrays. |
| 84 |
o replaced calc_mom_rhs.F by mom_fluxform.F/mom_vecinc.F to |
| 85 |
control flux-form versus vector invariant code. |
| 86 |
Switch is runtime flag vectorInvariantMomentum (defaults to false) |
| 87 |
or genmake -DISABLE ... |
| 88 |
o TAMC-related update to cope with mom_fluxform.F/mom_vecinc.F |
| 89 |
(additional storing is now needed for case vecinv; |
| 90 |
currently nothing implemented to switch this storing off |
| 91 |
in case fluxform is used since no CPP flag available). |
| 92 |
|
| 93 |
checkpoint40pre7 |
| 94 |
o bug fix in dynamics.F |
| 95 |
calc_viscosity called before k=1,Nr loop. |
| 96 |
o added some #ifdef's ALLOW_PASSIVE_TRACER |
| 97 |
o cleaned initialisations in ctrl_map_ini |
| 98 |
related to TAF |
| 99 |
|
| 100 |
checkpoint40pre6 |
| 101 |
o split dynamics into thermodynamics & dynamics |
| 102 |
o corresponding split of calc_diffusivity into |
| 103 |
calc_diffusivity, calc_viscosity |
| 104 |
(kpp_calc_diff, kpp_calc_visc). |
| 105 |
o Added diffkr, kapgm to control variables |
| 106 |
o bug fix in gmredi_slope_limit for ldd97 |
| 107 |
|
| 108 |
checkpoint40pre5 |
| 109 |
o proper initialization of Adams-Bashforth time stepping |
| 110 |
+ uses forward step for very first step of model (iter=0) |
| 111 |
o re-generated *all* output since almost all output is modified by this |
| 112 |
change (sorry but we've been putting this one off for about 2 years) |
| 113 |
+ natl_box, aim.5l_Equatorial_Channel and aim.5l_zon-ave were |
| 114 |
*not* regenerated since they are already out of date and I don't |
| 115 |
want to create the impression that they are suddenly working. |
| 116 |
o removed KAP stuff for DEC's |
| 117 |
o modified/added many .cvsignore files in */code and */input |
| 118 |
o added new expt "solid-body.cs-32x32x1" |
| 119 |
+ this uses an alternative form of ini_curvilinear_grid.F which |
| 120 |
is not necessarily any better than the current one... |
| 121 |
|
| 122 |
checkpoint40pre4 |
| 123 |
o added re-initialisations of utrans,vtrans,rtrans in |
| 124 |
calc_common_factors which are crucial for TAF. |
| 125 |
o extended iMin,jMin range for calc_common_factors, |
| 126 |
calc_diffusivity |
| 127 |
o added option ALLOW_PASSIVE_TRACER |
| 128 |
o a few minor updates and fixes |
| 129 |
|
| 130 |
checkpoint40pre3 |
| 131 |
o created calc_gtr1.F for gad package |
| 132 |
o split the_main_loop and re-introduced forward_step |
| 133 |
o some recomputations of nIter0 for TAMC |
| 134 |
|
| 135 |
checkpoint40pre2 |
| 136 |
o Updated to latest branch of AJA, JMC |
| 137 |
o Added gradient check package |
| 138 |
o Added routines for passive tracer handling |
| 139 |
o Added carbon sequestration sensitivity experiment |
| 140 |
|
| 141 |
checkpoint39 |
| 142 |
o Modifications and fixes to support TAMC |
| 143 |
(mainly missing or wrong directives) |
| 144 |
o Corrected common blocks when fooling TAMC to produce adjoint I/O |
| 145 |
in addummy_in_stepping |
| 146 |
o Updated verification/vero/ (bug fixes in code/, correct pickups in input/) |
| 147 |
o Enabled "sliced" (xz-,yz-) exchanges, needed for obcs. |
| 148 |
o Included following packages: |
| 149 |
cal: calendar |
| 150 |
exf: external forcing |
| 151 |
ecco: essential routines to control use of cal, exf |
| 152 |
(NB: this is not the full ECCO environment) |
| 153 |
o disabled some packages by default in genmake |
| 154 |
|
| 155 |
checkpoint38 |
| 156 |
o see notes_c37_adj.txt |
| 157 |
o corrected wrong layer index in external_forcing_t |
| 158 |
o replaced the_main_loop/forward_step by the_main_loop |
| 159 |
o removed forceoutput from write_state |
| 160 |
(replaced criteria for initial/final state) |
| 161 |
o added runtime flag for pre-C35 pickups: usePickupBeforeC35 |
| 162 |
|
| 163 |
checkpoint37 |
| 164 |
o mdsio.F routine split and packaged in pkg/mdsio |
| 165 |
(mdsio.F and mdsio_gl.F still exist in eesupp/src but are not |
| 166 |
used unless genmake -disable=mdsio) |
| 167 |
o allow a more precise surface-geopotential gradient for the atmosphere |
| 168 |
taking into account the change in space of Alpha (=1/rho) at the surface. |
| 169 |
o variable "etaN" (replace cg2d_x) is the surface R-anomaly [r unit] |
| 170 |
(ocean: sea surface height [m] / atmos: surface pressure anomaly [Pa]) |
| 171 |
o Change units: Phi_Hydrostatic (=phiHyd), Phi_Surface(=cg2d_x) and |
| 172 |
Phi_NonHydrostatic (=cg3d_x), all have now the dimension of a potential |
| 173 |
[m2/s2] (pressure/rhoConst in the ocean ; Geo-Potential in the atmosphere). |
| 174 |
o Time-Average diagnostics defined as a pseudo package "timeave". |
| 175 |
o pickup files : old pickup (before c35) did not work with new code (c35 |
| 176 |
and after) unless activating 2 commented lines in initialise-varia. |
| 177 |
o Non-Hydrostatic pickup file Pb fixed. |
| 178 |
o bug with Rigid-lid version (since c35) fixed. |
| 179 |
|
| 180 |
checkpoint36 |
| 181 |
o implement Crank-Nickelson barotropic time-stepping |
| 182 |
o recover Time-Average diagnostics |
| 183 |
o fix Pb in checkpoint (PR#19) |
| 184 |
o AIM become a standard package ; clean unused CPP_kees |
| 185 |
|
| 186 |
checkpoint35 |
| 187 |
o subtantial rewrite of dynamics.f |
| 188 |
o allows staggered time-stepping |
| 189 |
o packaged obcs, aim (Atmospheric Physics), zonal_filt, shap_filt |
| 190 |
o added bottom drag (linear + quadratic) |
| 191 |
|
| 192 |
checkpoint34 |
| 193 |
o Modified exchange routines for correct reverse simulation. |
| 194 |
o Added verification output to comply with bug fix in impldiff.F |
| 195 |
as of c33 (exp2, natl_box). |
| 196 |
o Corrected store directives in dynamics.F |
| 197 |
o Corrected array boundaries in impldiff.F |
| 198 |
o Corrected array dimensions in gmredi. |
| 199 |
Added temporary scalars to avoid storage by TAMC (gmredi, kpp). |
| 200 |
o Added routine GLOBAL_SUM_INT to global_sum.F |
| 201 |
o Added mdsio_gl.F for ECCO purposes. |
| 202 |
|
| 203 |
checkpoint33 |
| 204 |
o Fixed bug in lower BC in impldiff.F |
| 205 |
o Fixed confusion about units of forcing arrays in FFIELDS.h |
| 206 |
namely Fu,Fv,Qnet,Qsw,EmPmR: |
| 207 |
- Removed verification/natl_box/code/external_fields_scale.F |
| 208 |
(did not differ from that in model/src) |
| 209 |
- Changed units of fu,fv,Qnet,Qsw,EmPmR back to proper units |
| 210 |
(see FFIELDS.h for description) |
| 211 |
- Scale fu,fv,Qnet,Qsw,EmPmR when used in external_forcing_surf.F, |
| 212 |
kpp_calc.F and kpp_transport_t.F |
| 213 |
- Removed model/src/external_fields_scale.F and calls to it |
| 214 |
- verification/natl_box uses flux data with "atmospheric" sign so |
| 215 |
a special version of external_fields_load.F is used to |
| 216 |
change the data as it's read in. This way, the arrays |
| 217 |
have the right units and signs at all times tha a user could |
| 218 |
possibly use them. |
| 219 |
o Corrected genmake to delete a temporary file during configuration |
| 220 |
|
| 221 |
checkpoint32 |
| 222 |
o Scaling of forcing fields moved from external_forcing.F to |
| 223 |
external_fields_scale.F, called directly after loading fields. |
| 224 |
o Surface relaxation terms added to surface tendency fields in |
| 225 |
extra subroutine. |
| 226 |
o impldiff.F changed to avoid storing by TAMC of huge intermediate |
| 227 |
fields. |
| 228 |
o Changes in GMRedi to comply with packaging structure. |
| 229 |
o Modifications and bug fixes in KPP package. |
| 230 |
o Verification experiment for KPP added (verification/nat_box/) |
| 231 |
|
| 232 |
checkpoint31 |
| 233 |
o Packages interface brought in line with defined standard |
| 234 |
-> new routines packages_...F |
| 235 |
-> changes in initialise_fixed.F, initialise_varia.F |
| 236 |
-> enables initialisation interface for ECCO package |
| 237 |
o Rescaling of external forcing fields immediately after read |
| 238 |
(removed from external_forcing.F) |
| 239 |
-> new: surface_fields_scale.F |
| 240 |
-> affects ini_foorcing.F, external_fields_load.F |
| 241 |
o Computation of surface tendencies before tendency updates. |
| 242 |
-> new: external_forcing_surf.F |
| 243 |
o Change of variable names useGMRedi, useKPP, useECCO |
| 244 |
o Further changes in dynamics.F, impldiff.F for TAMC |
| 245 |
o Tested for experiments 0, 2, 4. |
| 246 |
- not yet tested for GMRedi. |
| 247 |
- not yet tested for KPP. |
| 248 |
|
| 249 |
checkpoint30 |
| 250 |
o Updates for OBCs |
| 251 |
o New experiment for test OBCs (internal_wave) |
| 252 |
o New output.txt in verification/exp4/results/ |
| 253 |
o Updates for KPP: calculation of viscosity corrected, new variables |
| 254 |
kpp_dumpFreq and kpp_taveFreq |
| 255 |
o CPP'd calls to system routines for TARGET_CRAY_VECTOR (P.H.) |
| 256 |
and entry for T90 in genmake |
| 257 |
|
| 258 |
checkpoint29 |
| 259 |
o packaged KPP and GM/Redi |
| 260 |
|
| 261 |
checkpoint28 |
| 262 |
o Corrected initialization of hFacMin for difference vertical coordinates |
| 263 |
o Modified calculation of partial cell thickness to more robust algorithm |
| 264 |
o config_summary.F: replaced write_1D_... by write_0D_... |
| 265 |
read_write.F: added subroutines write_0D_... |
| 266 |
o routines calc_common_factors, calc_gs, calc_gt, calc_phi_hyd, find_rho: |
| 267 |
included some initialisations required for TAMC. |
| 268 |
o routine calc_isoslopes: modified sqrt computation to make it |
| 269 |
differentiable for zero argument. |
| 270 |
o routines dynamics, impldiff: |
| 271 |
included store directives and tape key computations required for TAMC. |
| 272 |
o ALL TAMC related changes are between ifdef ALLOW_AUTODIFF_TAMC |
| 273 |
|
| 274 |
checkpoint27 |
| 275 |
o fixed calc_isoslopes() to so that cancellation of terms in tensor |
| 276 |
formulation works properly with variable resolution. |
| 277 |
o restructured the calling tree between the_model_main() and dynamics() |
| 278 |
o split initialise() into initialise_fixed() and initialise_varia() |
| 279 |
o introduced initialization file pSurfInitFile for free surface (cg2d_x) |
| 280 |
in new routine ini_psurf() |
| 281 |
|
| 282 |
checkpoint26 |
| 283 |
o moved some #ifdef in optional routines to encompass everything |
| 284 |
except the argument declarations. This was using up some memory. |
| 285 |
o fixed macros like _ADT to work with both versions of Linux cpp |
| 286 |
o cleaned up some unused and uninitialized variables |
| 287 |
(helps when debugging with strict compile options) |
| 288 |
o split up CPP_EEOPTIONS.h into options and macros |
| 289 |
(macros are now in eesupp/inc/CPP_EEMACROS.h) |
| 290 |
o patch for 2 processor JAM mode |
| 291 |
|
| 292 |
checkpoint25 |
| 293 |
o updates for OBCs and NH |
| 294 |
o dramatically reduced memory usage |
| 295 |
o added "JAM" routines for use on Artic network |
| 296 |
o parameterization of convection by implicit vertical diffusion |
| 297 |
o added a diagnostic of convective index |
| 298 |
o moved call to calc_divg_hat() from dynamics() to solve_for_pressure() |
| 299 |
(this is was partly for OBCs and NH but also in anticipation of |
| 300 |
an explicit free-surface option) |
| 301 |
o initial state input files for U and V |
| 302 |
|
| 303 |
checkpoint24 |
| 304 |
o Removed some spurious "REAL"s |
| 305 |
o Updated KPP to latest version from JPL |
| 306 |
o Added a COS(latitude) in horizontal viscosity |
| 307 |
|
| 308 |
checkpoint23 |
| 309 |
o Added del^4 diffusion for salt and temperature |
| 310 |
o Multiple minor fixes (implicit none's, arguments to global_sum, etc.) |
| 311 |
|
| 312 |
checkpoint22 |
| 313 |
o Made many small changes: implicit none, format(i5), etc... |
| 314 |
o Introduced KPP mixing scheme (compiles and runs but *not* tested) |
| 315 |
|
| 316 |
checkpoint21 |
| 317 |
o Introduced new I/O package that supports direct-access records (mdsio.F) |
| 318 |
o Split-up mid-level I/O routines |
| 319 |
o Updated input data for experiments 0,1,2,4 and 5 |
| 320 |
o Finally removed the genmake.* scripts (everyone seems happy with genmake) |
| 321 |
|
| 322 |
checkpoint20 |
| 323 |
o Tidied up open-boundaries some more |
| 324 |
o Added non-hydrostatic code |
| 325 |
o Fixed some input data |
| 326 |
o Modified MPI start-up to work in "coupled" context. |
| 327 |
|
| 328 |
checkpoint19 |
| 329 |
o Tidied up directory structure for verification experiments. |
| 330 |
o Added new code for open boundaries. |
| 331 |
o Corrected advective terms near boundaries. |
| 332 |
o Added CPP flag for "old UV" geometry and corrected spherical geometry. |
| 333 |
o Implemented free-slip/no-slip BCs and del^4 dissipation. |
| 334 |
o Split default parameters out of ini_parms.F into set_defaults.F. |
| 335 |
|
| 336 |
checkpoint18 |
| 337 |
o Further changes for optional code generation with |
| 338 |
CPP - ugh! |
| 339 |
|
| 340 |
checkpoint17 |
| 341 |
o Beginning to incorporating modularity |
| 342 |
for supporting atmos. and ocean configs. |
| 343 |
o Small changes for TAMC complinace |
| 344 |
o New ultra simple test case. Barotropic box |
| 345 |
configuration "exp0". |
| 346 |
|
| 347 |
checkpoint16 |
| 348 |
o Minor changes to exchange routines to support |
| 349 |
accumulation in reverse mode. Adjoint of an |
| 350 |
assignment is an addition. |
| 351 |
o Changes to support compilation by TAMC (LR(1) |
| 352 |
syntax) and by g77 under Linux ( 72 character |
| 353 |
limit ). |
| 354 |
|
| 355 |
checkpoint15 |
| 356 |
Rewrote exchange routines to allow |
| 357 |
o General tile <-> tile connectivity. |
| 358 |
o TAMC reverse mode flag ( is this really needed? ) |
| 359 |
o Variable width overlap and exchange region widths. |
| 360 |
o Hooks for sharing data via DMA, Arctic, Memory Channel, |
| 361 |
shmput/shmget, VIA, SYSV shmem and every other cool |
| 362 |
communication method. |
| 363 |
|
| 364 |
checkpoint14 |
| 365 |
Consistent isomorphism chages made. These allow dynamical core |
| 366 |
to support both pressure and meters as a vertical coordinate. |
| 367 |
|
| 368 |
checkpoint13 |
| 369 |
Release which corrects global_max problem and a couple |
| 370 |
of KGM time-averaging diagnostic errors. Also contains |
| 371 |
a double-gyre experiment setup as well as the baseline |
| 372 |
4 degree global template. The double gyre template is |
| 373 |
organised to allow it to be applied automatically to |
| 374 |
the baseline configuration. |
| 375 |
|
| 376 |
branch-point-rdot |
| 377 |
A few tidy-ups have been made bt the real point of this |
| 378 |
check-point (excuse the pun) is to allow a branch to be |
| 379 |
made for the purposes of adding the rDot alterations. |
| 380 |
I don't want to commit the rDot to the main-trunk until |
| 381 |
we've actually tested it... |
| 382 |
|
| 383 |
checkpoint12 |
| 384 |
Tidy up and added generic genmake script (formerly |
| 385 |
configure script). |
| 386 |
Released to testers. |
| 387 |
|
| 388 |
checkpoint11 |
| 389 |
Re-arranged initialise() and introduced ini_vertical_grid() |
| 390 |
and ini_masks_etc(). |
| 391 |
|
| 392 |
checkpoint10 |
| 393 |
Introduced the configure script. |
| 394 |
Separated out the diags package. |
| 395 |
Added template for writing time-averages of tile/intermediate quantities. |
| 396 |
Updated SIZE.h in exp2 and made the data big-endian. |
| 397 |
|
| 398 |
checkpoint9 |
| 399 |
MPI fix. Filename changes (meta/data). salbin*y stuff. |
| 400 |
General tidy-up. |
| 401 |
SST.bin SSS.bin added to verification/exp2 |
| 402 |
|
| 403 |
checkpoint8 |
| 404 |
Inludes time-dependant forcing and time-averaging. |
| 405 |
Works at 4x4 and 2.8125x2.8125 global |
| 406 |
|
| 407 |
checkpoint7-4degree-ref |
| 408 |
Branch from checkpoint7. Released for application to |
| 409 |
global modeling and general modeling - process studies, regional |
| 410 |
model etc.. |
| 411 |
Configured for verification/exp2 - 4 degree global, two process. |
| 412 |
|
| 413 |
checkpoint7 |
| 414 |
Created by cnh |
| 415 |
Simple 4 degree global simulation |
| 416 |
configuration. |
| 417 |
Includes climatological wind, |
| 418 |
hydrography, bathymetry etc |
| 419 |
Does not include Heat Flux and E-P |
| 420 |
|
| 421 |
checkpoint6 |
| 422 |
Created by cnh |
| 423 |
merged checkpoint5 with CD and |
| 424 |
spherical polar code. |
| 425 |
Still some problems with GM Redi |
| 426 |
abd full topography. |
| 427 |
|
| 428 |
checkpoint5 |
| 429 |
Created by AJA |
| 430 |
contains updates to GM Redi |
| 431 |
and reordered timestepping loop |
| 432 |
for Implicit Diffusion. |
| 433 |
|
| 434 |
checkpoint4 |
| 435 |
Created by cnh |
| 436 |
Contains memory saving macros ( _rdxc etc..) for when |
| 437 |
grid variations are limited. |
| 438 |
Contains polynomial coeffs for EOS and knudsen program |
| 439 |
for generation. |
| 440 |
|
| 441 |
checkpoint3 |
| 442 |
Created by cnh |
| 443 |
Added extra flags for runtime options |
| 444 |
Added this file |
| 445 |
Added implicit free-surface |
| 446 |
Added pre-conditioner |
| 447 |
Put compare01 into repository ( compare01/... ) |
| 448 |
Put data and eedata into repository ( verification/exp1 ) |
| 449 |
|
| 450 |
checkpoint2 |
| 451 |
Created by AJA |
| 452 |
Added GM/Redi |
| 453 |
|
| 454 |
checkpoint1 |
| 455 |
Created by AJA |
| 456 |
Restructured kloop in dynamics.F |
| 457 |
Added non-linear EOS |
| 458 |
|
| 459 |
baseline |
| 460 |
Original checked in code |
Parent Directory
| 
Revision Log
| 
Revision Graph