Page 63 of 85 FirstFirst ... 1338535455565758596061626364656667686970717273 ... LastLast
Results 1,241 to 1,260 of 1797

Thread: Guide to the ETW map

Hybrid View

Previous Post Previous Post   Next Post Next Post
  1. #1
    Artifex
    Patrician

    Join Date
    Oct 2008
    Location
    London, UK
    Posts
    1,332

    Default Re: Guide to the ETW map

    Quote Originally Posted by PietroMicca View Post
    Code:
    <ary type="boundaries">
        <boundaries unknown="2266038529 (87110101)" path_id="64" vertex_index="5"/>
        <boundaries unknown="2 (00000002)" path_id="65472" vertex_index="12"/>
    </ary>
    If path_id is treated as signed int, then it's 64 and -64 - seemed too much to be a coincidence
    (code to display it as signed uploaded).

    -64 and -63 are the only two "negative" numbers, everything else is nonnegative.

    Here are all path_ids from S2TW:

    1267 -64
    140 0
    134 64
    82 128
    181 192
    134 256
    143 320
    64 384
    39 448
    108 512
    110 576
    57 640
    95 704
    85 768
    74 832
    592 896
    10 960
    6 1024
    4 1088
    11 1152
    8 1216
    18 1280
    13 1344
    4 1408
    20 1472
    14 1536
    7 1600
    9 1664
    7 1728
    13 1792
    13 1856
    18 1920
    13 1984
    14 2048

    All these are multiples of 64 as far as I can tell!

    In ETW they're not, but they're most common:


    3281 -64
    2409 -63
    271 0
    42 64
    90 128
    129 192
    93 256
    11 320
    126 384
    20 448
    23 512
    280 576
    23 640
    85 704
    19 768
    115 832
    109 896
    62 960
    118 1024
    97 1088
    110 1152
    152 1216
    162 1280
    47 1344
    121 1408
    93 1472
    154 1536
    287 1600
    193 1601
    128 1664
    109 1728
    43 1792
    95 1856
    80 1920
    32 1921
    79 1984
    100 2048
    4 2049
    49 2112
    13 2113
    26 2176
    224 2177
    38 2240
    119 2241
    38 2304
    52 2305
    9 2368
    55 2369
    7 2432
    134 2433
    180 2497
    122 2561
    120 2625
    357 2689
    378 2753
    2 2817
    3 2881
    100 2945
    117 3009
    72 3073
    1926 3136
    1453 3137
    18 3200
    22 3264
    3 3328
    13 3392
    23 3456
    65 3520
    23 3584
    2 3648
    3 3712
    51 3776
    48 3840
    17 3904
    7 3968
    11 4032
    27 4096
    33 4160
    26 4224
    1 4288
    2 4352
    1 4416
    27 4480
    20 4544
    51 4608
    4 4672
    34 4736
    2 4800
    4 4864
    11 4928
    6 4992
    20 5056
    32 5120
    11 5184
    24 5248
    5 5312
    26 5376
    2 5440
    25 5504
    2 5568
    21 5632
    6 5696
    2 5760
    3 5824
    5 5888
    16 5952
    21 6016
    3 6017
    4 6080
    26 6144
    35 6145
    33 6208
    9 6209
    4 6272
    4 6336
    5 6400
    5 6401
    12 6465
    28 6529
    24 6593
    1 6657
    11 6721
    18 6785
    22 6849
    3 6913
    14 6977
    3 7041
    19 7105
    12 7169
    2 7233
    54 7297
    2 7361
    39 7425
    1 7489
    16 7553
    3 7617
    45 7681
    1 7745
    34 7809
    20 7873
    1 7937
    34 8001
    69 8065
    55 8129

    Maybe we divided this <u> wrong?

    Is there any reason why path_ids should have this form?

  2. #2
    Ordinarius
    Join Date
    Jul 2009
    Location
    Italy
    Posts
    737

    Default

    Quote Originally Posted by taw View Post
    Maybe we divided this <u> wrong?
    I also have doubts about path ids. But I can state that vertex_index is certainly correct. All boundaries are evidence of this. There is a perfect match.

    Quote Originally Posted by taw View Post
    Is there any reason why path_ids should have this form?
    there's a strange coincidence

    if we consider path id as unsigned integers 65472/64=1023

    1023 is a value occurring many times in <u2> tags in empty cells

    could be path id an uint16?
    Last edited by Ishan; April 03, 2012 at 10:54 PM. Reason: Double Post

  3. #3

    Default

    Hey everybody! As soon as you guys figure this junk out, if you want to help realize something that were all thinking, then come here, to realize a dream.

    http://www.twcenter.net/forums/showt...6#post11269496
    Last edited by Ishan; April 03, 2012 at 10:52 PM. Reason: merged

  4. #4
    Dynamo11's Avatar Domesticus
    Join Date
    Mar 2009
    Location
    England
    Posts
    2,209

    Default Re: Guide to the ETW map

    Quote Originally Posted by OldDrunkBastard View Post
    Hey everybody! As soon as you guys figure this junk out, if you want to help realize something that were all thinking, then come here, to realize a dream.

    http://www.twcenter.net/forums/showt...6#post11269496
    Not to sound like a but how's about leaving that stuff out until we've actually cracked the map.


  5. #5

    Default Re: Guide to the ETW map

    Quote Originally Posted by Dynamo11 View Post
    Not to sound like a but how's about leaving that stuff out until we've actually cracked the map.
    Ya i know just pre-advertising. eager i guess. I feel as if we're on the edge of a breakthrough. sry

  6. #6

    Default Re: Guide to the ETW map

    Some of my observations.

    As the impassible trade theatres only have a path_id of -64 or 0 I'm guessing that a path_id of 1-5445 refers to a passable region.

    Given that every path_id that's -64 to -59 has a 'boundary unknown' of 2 or 3 I believe that they're all the same type of entry, possible an impassible area. Alternatively they could be used to move an off map unit (in a ship or building) back onto the map.

    I don't know what a path_id of 0 is meant to represent. Its structure is similar to that of path_id 1-5445 but it's present in theatres with no passable areas. It may be the sea or the coast.

    The path_id is often repeated and sometimes repeated several times in the same grid_cell. The vertex_index is repeated fewer times and never with the same path_id. Most likely both are to link various cells together to form a road or boundary.

    If a gird_cell has a lot of boundaries with similar path_ids it's may be a settlement with the path_ids referring to the various buildings.
    Last edited by uanime5; April 05, 2012 at 09:59 AM.
    Morning Sun (adds Korea and China to the Shogun 2 map)
    http://www.twcenter.net/forums/forum...28-Morning-Sun

    Expanded Japan mod (97 new regions and 101 new factions)
    http://www.twcenter.net/forums/showt...ew-factions%29

    How to split a region in TWS2
    http://www.twcenter.net/forums/showt...split-a-region

    Eras Total Conquest 2.3 (12 campaigns from 970-1547)

  7. #7
    Ordinarius
    Join Date
    Jul 2009
    Location
    Italy
    Posts
    737

    Default Re: Guide to the ETW map

    Quote Originally Posted by uanime5 View Post
    As the impassible trade theatres only have a path_id of -64 or 0 I'm guessing that a path_id of 1-5445 refers to a passable region.

    Given that every path_id that's -64 to -59 has a 'boundary unknown' of 2 or 3 I believe that they're all the same type of entry, possible an impassible area. Alternatively they could be used to move an off map unit (in a ship or building) back onto the map.
    Negative values are for sea-land transition areas.

    Quote Originally Posted by uanime5 View Post
    I don't know what a path_id of 0 is meant to represent. Its structure is similar to that of path_id 1-5445 but it's present in theatres with no passable areas. It may be the sea or the coast.
    path_id from 0 to 5445 are regional path and land-land transition areas

    we can easily see in India theatre each path_id is multiple of 64. If we divide each path_id for 64, we will obtain an integer indicating a region or a transition areas between regions.

    in particular

    Code:
      1 <!-- unexplorable -->
      2 <!-- ceylon -->
      3 <!-- carnatica -->
      4 <!-- mysore -->
      5 <!-- malabar -->
      6 <!-- hyderabad -->
      7 <!-- bijapur -->
      8 <!-- orissa -->
      9 <!-- ahmadnagar -->
      10 <!-- berar -->
      11 <!-- gujarat -->
      12 <!-- malwa -->
      13 <!-- bengal -->
      14 <!-- rajpootana -->
      15 <!-- sindh -->
      16 <!-- hindustan -->
      17 <!-- baluchistan -->
      18 <!-- punjab -->
      19 <!-- afghanistan -->
      20 <!-- kashmir -->
      21 <!-- wilderness_khiva -->
     
      0 <!-- separator -->
     
      2 3, 4 <!-- carnatica, mysore-->
      2 4, 5 <!-- mysore, malabar-->
      2 3, 6 <!-- carnatica, hyderabad-->
      3 4, 5, 7 <!-- mysore, malabar, bijapur-->
      2 4, 7 <!-- mysore, bijapur-->
      2 4, 6 <!-- mysore, hyderabad-->
      3 3, 4, 6 <!-- carnatica, mysore, hyderabad-->
      2 3, 8 <!-- carnatica, orissa-->
      2 5, 7 <!-- malabar, bijapur-->
      2 6, 8 <!-- hyderabad, orissa-->
      3 3, 6, 8 <!-- carnatica, hyderabad, orissa-->
      2 6, 7 <!-- hyderabad, bijapur-->
      2 7, 9 <!-- bijapur, ahmadnagar-->
      3 6, 7, 9 <!-- hyderabad, bijapur, ahmadnagar-->
      2 6, 9 <!-- hyderabad, ahmadnagar-->
      2 6, 10 <!-- hyderabad, berar-->
      3 6, 8, 10 <!-- hyderabad, orissa, berar-->
      2 8, 10 <!-- orissa, berar-->
      2 9, 11 <!-- ahmadnagar, gujarat-->
      3 6, 9, 12 <!-- hyderabad, ahmadnagar, malwa-->
      2 6, 12 <!-- hyderabad, malwa-->
      2 9, 12 <!-- ahmadnagar, malwa-->
      3 6, 10, 12 <!-- hyderabad, berar, malwa-->
      2 10, 12 <!-- berar, malwa-->
      3 9, 11, 12 <!-- ahmadnagar, gujarat, malwa-->
      2 11, 12 <!-- gujarat, malwa-->
      2 8, 13 <!-- orissa, bengal-->
      2 11, 14 <!-- gujarat, rajpootana-->
      3 11, 12, 14 <!-- gujarat, malwa, rajpootana-->
      2 12, 14 <!-- malwa, rajpootana-->
      2 10, 13 <!-- berar, bengal-->
      2 11, 15 <!-- gujarat, sindh-->
      3 10, 12, 13 <!-- berar, malwa, bengal-->
      2 12, 13 <!-- malwa, bengal-->
      3 11, 14, 15 <!-- gujarat, rajpootana, sindh-->
      2 12, 16 <!-- malwa, hindustan-->
      3 12, 13, 16 <!-- malwa, bengal, hindustan-->
      2 14, 15 <!-- rajpootana, sindh-->
      2 14, 16 <!-- rajpootana, hindustan-->
      3 12, 14, 16 <!-- malwa, rajpootana, hindustan-->
      2 13, 16 <!-- bengal, hindustan-->
      2 15, 17 <!-- sindh, baluchistan-->
      3 14, 15, 18 <!-- rajpootana, sindh, punjab-->
      2 14, 18 <!-- rajpootana, punjab-->
      2 15, 18 <!-- sindh, punjab-->
      3 15, 17, 18 <!-- sindh, baluchistan, punjab-->
      2 17, 19 <!-- baluchistan, afghanistan-->
      2 17, 18 <!-- baluchistan, punjab-->
      3 17, 18, 19 <!-- baluchistan, punjab, afghanistan-->
      3 14, 16, 18 <!-- rajpootana, hindustan, punjab-->
      2 18, 19 <!-- punjab, afghanistan-->
      2 16, 18 <!-- hindustan, punjab-->
      3 16, 18, 20 <!-- hindustan, punjab, kashmir-->
      2 16, 20 <!-- hindustan, kashmir-->
      2 18, 20 <!-- punjab, kashmir-->
      2 19, 20 <!-- afghanistan, kashmir-->
      3 18, 19, 20 <!-- punjab, afghanistan, kashmir-->
      2 19, 21 <!-- afghanistan, wilderness_khiva-->
    Now if we consider each row as an element of an array, unexplorable is in position 0, ceylon 1 and so on. In the end of listed regions we have to add a generic sea region (wrapping all sea regions inside that theatre) always in n position. Then we can continue we the land-land transition areas: <!-- carnatica, mysore-->; <!-- mysore, malabar--> and so on. But 64 multiple rules is not respected in other major theaters. Why? Because in code there are several signs of a different original plan. For instance Crete as a separated path and a separated path_id. This means that originally Crete was planned as a separated region, but finally included in Greece. Crete path_id is 896 (/64= 14) and Greece path_id is 897 (896+1).

    Quote Originally Posted by uanime5 View Post
    The path_id is often repeated and sometimes repeated several times in the same grid_cell.
    path_id obviously repeated: pathfinding of a regions is compound by many traits (hundreds some times).

    Quote Originally Posted by uanime5 View Post
    The vertex_index is repeated fewer times and never with the same path_id. Most likely both are to link various cells together to form a road or boundary.
    Did you find vertex_index repeated in the same theatre? Where?

    Anyway vertex_index indicates where the trait in <u4_ary> in pathfidning_areas is. So vertex_index joins boundaries to a particular segment/polygons.

  8. #8
    Artifex
    Patrician

    Join Date
    Oct 2008
    Location
    London, UK
    Posts
    1,332

    Default Re: Guide to the ETW map

    8 : 24 bit division (that is - path_id / 256) is plainly wrong - all vertex_index numbers become crap:

    <boundaries unknown="2 (00000002)" path_id="255" vertex_index="12583037"/>
    <boundaries unknown="219217921 (0d110001)" path_id="12" vertex_index="4194435"/>

    10 : 22 bit division (that is - path_id / 64) may or may not be correct, but at least isn't obviously wrong, so I put that in the code.

    Check if it make sense.

  9. #9
    Ordinarius
    Join Date
    Jul 2009
    Location
    Italy
    Posts
    737

    Default Re: Guide to the ETW map

    uanime try the new update version of taw's converter

    look at possible mistakes

    at first glance everything looks perfect now

  10. #10

    Default Re: Guide to the ETW map

    Okay I've tried the new converter.

    The vertex_index now matches the highest entries in pathfinding_areas <u4> and the path_id for all sea and impassible areas is -1 (impassible regions, seas, rivers, and theatres are all set to -1 in the regions.esf; while the main theatres are 1, 2, or 3).

    One odd thing is that the path_id is 1 too low for the region. So region 23 has path_id 22. For example this entry:
    Code:
      <rec type="grid_cells">
       <bin6>ab ff ab ab ; 75 83 88 ab</bin6>
       <ary type="boundaries">
        <boundaries unknown="2952798480 (b0002110)" path_id="22" vertex_index="70752"/>
        <boundaries unknown="2 (00000002)" path_id="-1" vertex_index="70762"/>
        <boundaries unknown="1695633440 (65115020)" path_id="70" vertex_index="70770"/>
        <boundaries unknown="220340224 (0d222000)" path_id="24" vertex_index="70780"/>
       </ary>
      </rec>
    This seems to be for regions 22 (wilderness_tejas) and 24 (cherokee_territory), and boundary 70. However boundary 70 is between regions 23 (georgia_usa) and 25 (carolinas) and there isn't a boundary between wilderness_tejas and cherokee_territory because they aren't next to each other. So it seems that path_id is not the same as the region number because path_id starts at 0, while the region number starts at 1. So to get the region number you have to add 1 to the path_id, for example path_id="22" + 1 = region 23.

    This is going to be confusing but it does seem to be correct.

    Still no luck getting ETW to accept my new entry without crashing when I click on anything.
    Last edited by uanime5; April 05, 2012 at 05:00 PM.
    Morning Sun (adds Korea and China to the Shogun 2 map)
    http://www.twcenter.net/forums/forum...28-Morning-Sun

    Expanded Japan mod (97 new regions and 101 new factions)
    http://www.twcenter.net/forums/showt...ew-factions%29

    How to split a region in TWS2
    http://www.twcenter.net/forums/showt...split-a-region

    Eras Total Conquest 2.3 (12 campaigns from 970-1547)

  11. #11
    Ordinarius
    Join Date
    Jul 2009
    Location
    Italy
    Posts
    737

    Default Re: Guide to the ETW map

    Quote Originally Posted by uanime5 View Post
    This seems to be for regions 22 (wilderness_tejas) and 24 (cherokee_territory), and boundary 70. However boundary 70 is between regions 23 (georgia_usa) and 25 (carolinas) and there isn't a boundary between wilderness_tejas and cherokee_territory because they aren't next to each other.
    You have done some mistakes.

    22 georgia_usa
    24 carolinas
    70 georgia_usa, carolinas

    this is the correct list for America pathfinding

    Code:
    -1	sea-land transition areas
    0	new_grenada
    1	french_guyana
    2	dutch_guyana
    3	new_andalusia
    4	panama
    5	trinidad_tobago
    6	the_caribbean_sea
    7	guatemala
    8	curacao
    9	windward_islands
    10	new_spain
    11	leeward_islands
    12	hispaniola
    13	jamaica
    14	cuba
    15	wilderness_mexico
    16	bahamas
    17	florida
    18	tejas
    19	new_mexico
    20	lower_louisiana
    21	wilderness_tejas
    22	georgia_usa
    23	cherokee_territory
    24	carolinas
    25	upper_louisiana
    26	wilderness_great_plains
    27	kaintuck_territory
    28	virginia
    29	maryland
    30	pennsylvania
    31	michigan_territory
    32	algonquin_territory
    33	new_york
    34	new_england
    35	great_plains
    36	ontario
    37	iroquois_territory
    38	maine
    39	acadia
    40	new_france
    41	huron_territory
    42	newfoundland
    43	wilderness_hudsonsbay
    44	wilderness_canada
    45	western_plain
    46	beaver_territory
    47	ruperts_land
    48	labrador
    49	northwest_territories
    50	sea
    51	french_guyana, dutch_guyana
    52	new_grenada, new_andalusia
    53	new_grenada, panama
    54	panama, guatemala
    55	guatemala, new_spain
    56	new_spain, wilderness_mexico
    57	new_spain, tejas
    58	wilderness_mexico, new_mexico
    59	wilderness_mexico, new_mexico, wilderness_tejas
    60	new_mexico, wilderness_tejas
    61	tejas, wilderness_tejas
    62	tejas, lower_louisiana
    63	florida, lower_louisiana
    64	florida, georgia_usa
    65	lower_louisiana, cherokee_territory
    66	florida, cherokee_territory
    67	florida, lower_louisiana, cherokee_territory
    68	georgia_usa, cherokee_territory
    69	tejas, lower_louisiana, wilderness_tejas
    70	georgia_usa, carolinas
    71	lower_louisiana, wilderness_tejas
    72	upper_louisiana, wilderness_great_plains
    73	lower_louisiana, upper_louisiana
    74	cherokee_territory, kaintuck_territory
    75	carolinas, kaintuck_territory
    76	cherokee_territory, upper_louisiana
    77	wilderness_tejas, wilderness_great_plains
    78	carolinas, virginia
    79	kaintuck_territory, virginia
    80	virginia, maryland
    81	upper_louisiana, kaintuck_territory
    82	maryland, pennsylvania
    83	upper_louisiana, michigan_territory
    84	virginia, maryland, pennsylvania
    85	pennsylvania, algonquin_territory
    86	kaintuck_territory, new_york
    87	kaintuck_territory, virginia, new_york
    88	virginia, new_york
    89	virginia, pennsylvania
    90	michigan_territory, new_york
    91	algonquin_territory, new_england
    92	pennsylvania, new_york
    93	wilderness_great_plains, great_plains
    94	algonquin_territory, iroquois_territory
    95	new_york, ontario
    96	new_york, iroquois_territory
    97	new_england, maine
    98	new_england, iroquois_territory
    99	iroquois_territory, new_france
    100	maine, acadia
    101	ontario, new_france
    102	ontario, huron_territory
    103	iroquois_territory, maine
    104	iroquois_territory, maine, new_france
    105	maine, new_france
    106	wilderness_great_plains, great_plains, wilderness_hudsonsbay
    107	wilderness_great_plains, wilderness_hudsonsbay
    108	ontario, wilderness_canada
    109	ontario, new_france, wilderness_canada
    110	new_france, wilderness_canada
    111	great_plains, wilderness_hudsonsbay
    112	ontario, huron_territory, wilderness_canada
    113	huron_territory, wilderness_canada
    114	acadia, new_france
    115	huron_territory, ruperts_land
    116	huron_territory, wilderness_canada, ruperts_land
    117	wilderness_canada, ruperts_land
    118	huron_territory, wilderness_hudsonsbay
    119	huron_territory, wilderness_hudsonsbay, ruperts_land
    120	wilderness_hudsonsbay, ruperts_land
    121	wilderness_canada, labrador
    122	wilderness_hudsonsbay, northwest_territories

  12. #12

    Default Re: Guide to the ETW map

    Good news I've figured out what I need to do to get the pathfinding to work. As each boundary now has the line number for its pathfinding_area entry I was able to find all the boundaries for a farm in georgia_usa, copy them, change path_id and vertex_index for my cherokee farm, and add them to the correct grid_cells.

    The bad news is that this caused some problems elsewhere with pathfinding and ETW crashed when I ended the turn. I suspect this is because I cloned the boundaries from a Georgian farm and that modifying the unknown 1 and 2 may fix this problem.


    For those who are interested in the details the farm in georgia_usa had 8 'vertices counts' in the pathfinding_areas and 8 boundaries in the grid_data (2 per cell). The new cherokee farm I made had the same number of 'vertices counts' so I needed to add 8 new boundaries to 4 cells in the grid_data to make it work. The only requirements for the 4 cells is that they be next to each other and form a square.

    You also need to increase the total number of entries to account for the extra 8 entries. So in the Americas theatre if you add 8 boundaries you need to increase the 63623 to 63631.

    EDIT: changing the unknown 1 and 2 entries just brought back the cannot leave building bug. It seems that they tell ETW what each type of boundary is for. So changing them to something new won't fix this problem.
    Last edited by uanime5; April 05, 2012 at 07:30 PM.
    Morning Sun (adds Korea and China to the Shogun 2 map)
    http://www.twcenter.net/forums/forum...28-Morning-Sun

    Expanded Japan mod (97 new regions and 101 new factions)
    http://www.twcenter.net/forums/showt...ew-factions%29

    How to split a region in TWS2
    http://www.twcenter.net/forums/showt...split-a-region

    Eras Total Conquest 2.3 (12 campaigns from 970-1547)

  13. #13
    alhoon's Avatar Comes Rei Militaris
    Join Date
    Apr 2008
    Location
    Chania, Greece
    Posts
    24,223

    Default Re: Guide to the ETW map

    Quote Originally Posted by uanime5 View Post
    Good news I've figured out what I need to do to get the pathfinding to work.

    ...
    The bad news is that this caused some problems elsewhere with pathfinding and ETW crashed when I ended the turn.
    + rep! ! !
    alhoon is not a member of the infamous Hoons: a (fictional) nazi-sympathizer KKK clan. Of course, no Hoon would openly admit affiliation to the uninitiated.
    "Angry Uncle Gordon" describes me well.
    _______________________________________________________
    Beta-tester for Darthmod Empire, the default modification for Empire Total War that does not ask for your money behind patreon.
    Developer of Causa Belli submod for Darthmod, headed by Hammeredalways and a ton of other people.
    Developer of LtC: Random maps submod for Lands to Conquer (that brings a multitude of random maps and other features).

  14. #14
    Artifex
    Patrician

    Join Date
    Oct 2008
    Location
    London, UK
    Posts
    1,332

    Default Re: Guide to the ETW map

    I split unknown1 into two parts 16:16 - from visual inspection it looks like correct division.

    <ary type="boundaries">
    <boundaries unknown1="0 (0000)" unknown2="2 (0002)" path_id="-1" vertex_index="5"/>
    <boundaries unknown1="3345 (0d11)" unknown2="0 (0000)" path_id="0" vertex_index="13"/>
    </ary>

    It would be entirely possible that division between two 32-bit number is wrong, so path_id overflows to unknown2, but path_id="-1" is incompatible with unknown2="2",
    so at least that theory should be excluded.

    If you want to change things and check if ETW likes them or not, you might need to understand unknown1 / unknown2 as well.

  15. #15
    Artifex
    Patrician

    Join Date
    Oct 2008
    Location
    London, UK
    Posts
    1,332

    Default Re: Guide to the ETW map

    Does it mean we can almost add new buildings? That alone would make Georgia and Dagestan and other minor countries playable and would be ing amazing.

  16. #16

    Default Re: Guide to the ETW map

    Quote Originally Posted by taw View Post
    Does it mean we can almost add new buildings? That alone would make Georgia and Dagestan and other minor countries playable and would be ing amazing.
    Yes but currently it destroys the pathfinding of the whole theatre and causes ETW to crash after 1 turn.

    I'll test how the new converter affects the unknown entries. Using the old converter only changing the unknown 2 stopped the pathfinding from working, changing the unknown 1 didn't effect anything.
    Morning Sun (adds Korea and China to the Shogun 2 map)
    http://www.twcenter.net/forums/forum...28-Morning-Sun

    Expanded Japan mod (97 new regions and 101 new factions)
    http://www.twcenter.net/forums/showt...ew-factions%29

    How to split a region in TWS2
    http://www.twcenter.net/forums/showt...split-a-region

    Eras Total Conquest 2.3 (12 campaigns from 970-1547)

  17. #17
    Ordinarius
    Join Date
    Jul 2009
    Location
    Italy
    Posts
    737

    Default Re: Guide to the ETW map

    I have also good news. Recent news show that there aren't more path than regions/boundaries. So there isn't a separated path for Crete or other regions. Limit for regions is higher. It should be 16^4=65536.

    Quote Originally Posted by uanime5 View Post
    You also need to increase the total number of entries to account for the extra 8 entries. So in the Americas theatre if you add 8 boundaries you need to increase the 63623 to 63631.
    This number is not clear, maybe we need to analyze it better. Maybe there's an hex version of this number more meaningful.

    By the way, a message for taw. Now we can generate a more meaningful path_id.

    First of all, replace separator comment with sea

    Code:
    84 <!-- karelia -->
    85 <!-- iceland -->
     
    0 <!-- sea -->
     
    2 4, 6 <!-- wilderness_arabia, tripoli-->
    So if you put in an array all comments in this tag listed regions, sea and then land-land transition areas, we can obtain this

    Code:
    <rec type="grid_cells">
       <bin6>1f 46 37 1f ; 42 36 2e 3f</bin6>
       <ary type="boundaries">
        <boundaries unknown="2969723168 (b1026120)" path_id="87 (baluchistan, persia)" vertex_index="13138"/>
        <boundaries unknown="2400457216 (8f141200)" path_id="4 (persia)" vertex_index="13145"/>
        <boundaries unknown="1627459600 (61011010)" path_id="2 (baluchistan)" vertex_index="13151"/>
       </ary>
    </rec>
    (I have used an old version of converter)

  18. #18
    Artifex
    Patrician

    Join Date
    Oct 2008
    Location
    London, UK
    Posts
    1,332

    Default Re: Guide to the ETW map

    ~/etwng/esfxml/tmp/pathfinding_esf_dir$ cat grid_data-0001.xml | perl -nle 'print $1 if /unknown1="(.*?)"/' | sort -n | uniq -c
    5707 0 (0000)
    55 257 (0101)
    45 258 (0102)
    4 260 (0104)
    1 1024 (0400)
    42 1040 (0410)
    45 1056 (0420)
    7 1088 (0440)
    5 1152 (0480)
    3 1282 (0502)
    10 1297 (0511)
    1 1298 (0512)
    2 1313 (0521)
    14 1314 (0522)
    1 1316 (0524)
    1 1410 (0582)
    4 3330 (0d02)
    424 3345 (0d11)
    36 3346 (0d12)
    7 3348 (0d14)
    44 3361 (0d21)
    191 3362 (0d22)
    41 3364 (0d24)
    1 3368 (0d28)
    12 3393 (0d41)
    48 3394 (0d42)
    8 3396 (0d44)
    2 3400 (0d48)
    1 3457 (0d81)
    9 3458 (0d82)
    4 3460 (0d84)
    1 3522 (0dc2)
    99 8192 (2000)
    16 8449 (2101)
    28 8450 (2102)
    1 9216 (2400)
    6 9232 (2410)
    42 9248 (2420)
    3 9280 (2440)
    2 9344 (2480)
    1 9505 (2521)
    2 9506 (2522)
    6 11537 (2d11)
    1 11538 (2d12)
    3 11553 (2d21)
    37 11554 (2d22)
    1 11650 (2d82)
    885 24833 (6101)
    1 24835 (6103)
    10 25873 (6511)
    38 25889 (6521)
    1 25921 (6541)
    5 25985 (6581)
    7 27905 (6d01)
    420 27921 (6d11)
    151 27937 (6d21)
    1 27953 (6d31)
    89 27969 (6d41)
    10 28033 (6d81)
    118 32768 (8000)
    16 33025 (8101)
    59 33026 (8102)
    1 33028 (8104)
    9 33808 (8410)
    22 33824 (8420)
    2 33856 (8440)
    1 33920 (8480)
    1 34065 (8511)
    856 34320 (8610)
    21 34577 (8711)
    31 34578 (8712)
    2 34580 (8714)
    19 36113 (8d11)
    4 36114 (8d12)
    2 36116 (8d14)
    43 36130 (8d22)
    4 36162 (8d42)
    1 36226 (8d82)
    863 36625 (8f11)
    275 36626 (8f12)
    1 36627 (8f13)
    107 36628 (8f14)
    2 36630 (8f16)
    1 36632 (8f18)
    3 36657 (8f31)
    1 36690 (8f52)
    1 36753 (8f91)
    33 40960 (a000)
    1 41217 (a101)
    1 41218 (a102)
    1 42000 (a410)
    1 42306 (a542)
    36 42512 (a610)
    1 42576 (a650)
    2 42769 (a711)
    1 42770 (a712)
    2 44322 (ad22)
    1 44324 (ad24)
    1 44354 (ad42)
    20 44817 (af11)
    2 44818 (af12)
    1 44882 (af52)
    826 45056 (b000)
    28 45313 (b101)
    50 45314 (b102)
    5 46096 (b410)
    31 46112 (b420)
    8 46144 (b440)
    1 46370 (b522)
    1 46402 (b542)
    516 46608 (b610)
    1 46640 (b630)
    14 46865 (b711)
    10 46866 (b712)
    21 48401 (bd11)
    4 48402 (bd12)
    2 48417 (bd21)
    43 48418 (bd22)
    1 48420 (bd24)
    2 48450 (bd42)
    378 48913 (bf11)
    142 48914 (bf12)
    2 48915 (bf13)
    16 48916 (bf14)
    2 48945 (bf31)
    62 57601 (e101)
    1 58641 (e511)
    47 59153 (e711)
    14 60689 (ed11)
    4 60705 (ed21)
    2 60737 (ed41)
    563 61201 (ef11)
    2 61203 (ef13)
    1 61233 (ef31)
    1157 61697 (f101)
    1 62721 (f501)
    18 62737 (f511)
    4 62753 (f521)
    2 62785 (f541)
    559 63249 (f711)
    1 63251 (f713)
    1 64769 (fd01)
    392 64785 (fd11)
    56 64801 (fd21)
    1 64805 (fd25)
    1 64817 (fd31)
    125 64833 (fd41)
    1 64865 (fd61)
    1 64897 (fd81)
    1 64899 (fd83)
    56 65297 (ff11)
    21 65299 (ff13)
    1 65301 (ff15)
    9 65329 (ff31)
    1 65331 (ff33)
    8 65361 (ff51)
    4 65425 (ff91)

    This suggests internal structure within unknown1 - it increases every 100 or 400. So it's either 8:8 or 6:10.

    unknown2 has some hints of internal structure - 8:8 bits. Moving low bits of unknown1 to unknown2 makes it even worse, so at least this division seems safe.

    I've tried a different way to divide it into 4 parts - 8:8:8:8 bits, but 6:10:8:8 also seems plausible.

    Tell me if it makes any more sense.

  19. #19
    Artifex
    Patrician

    Join Date
    Oct 2008
    Location
    London, UK
    Posts
    1,332

    Default Re: Guide to the ETW map

    I'm not really convinced about this 8:8:8:8 split either, here are the stats for ETW tutorial pathfinding.esf:

    ~/etwng/esfxml/tmp/pathfinding_esf_dir$ cat grid_data-0001.xml | perl -nle 'print $1 if /unknown1="(.*?)"/' | sort -n | uniq -c
    5707 0 (00)
    104 1 (01)
    100 4 (04)
    32 5 (05)
    833 13 (0d)
    99 32 (20)
    44 33 (21)
    54 36 (24)
    3 37 (25)
    48 45 (2d)
    886 97 (61)
    54 101 (65)
    678 109 (6d)
    118 128 (80)
    76 129 (81)
    34 132 (84)
    1 133 (85)
    856 134 (86)
    54 135 (87)
    73 141 (8d)
    1254 143 (8f)
    33 160 (a0)
    2 161 (a1)
    1 164 (a4)
    1 165 (a5)
    37 166 (a6)
    3 167 (a7)
    4 173 (ad)
    23 175 (af)
    826 176 (b0)
    78 177 (b1)
    44 180 (b4)
    2 181 (b5)
    517 182 (b6)
    24 183 (b7)
    73 189 (bd)
    540 191 (bf)
    62 225 (e1)
    1 229 (e5)
    47 231 (e7)
    20 237 (ed)
    566 239 (ef)
    1157 241 (f1)
    25 245 (f5)
    560 247 (f7)
    579 253 (fd)
    100 255 (ff)

    These at least do not look like bitfield flags.
    Most common 00, 61, 86, 8f, f1.

    ~/etwng/esfxml/tmp/pathfinding_esf_dir$ cat grid_data-0001.xml | perl -nle 'print $1 if /unknown2="(.*?)"/' | sort -n | uniq -c
    6785 0 (00)
    2229 1 (01)
    190 2 (02)
    1 3 (03)
    5 4 (04)
    1471 16 (10)
    3859 17 (11)
    507 18 (12)
    27 19 (13)
    134 20 (14)
    1 21 (15)
    2 22 (16)
    1 24 (18)
    140 32 (20)
    305 33 (21)
    333 34 (22)
    44 36 (24)
    1 37 (25)
    1 40 (28)
    1 48 (30)
    17 49 (31)
    1 51 (33)
    20 64 (40)
    231 65 (41)
    57 66 (42)
    8 68 (44)
    2 72 (48)
    1 80 (50)
    8 81 (51)
    2 82 (52)
    1 97 (61)
    8 128 (80)
    17 129 (81)
    12 130 (82)
    1 131 (83)
    4 132 (84)
    5 145 (91)
    1 194 (c2)

    Mostly 00, 01, 10, 11, 20, 21, 22, 41 - bitfield flags?

    ~/etwng/esfxml/tmp/pathfinding_esf_dir$ cat grid_data-0001.xml | perl -nle 'print $1 if /unknown3="(.*?)"/' | sort -n | uniq -c
    7339 0 (00)
    4795 1 (01)
    288 2 (02)
    29 3 (03)
    39 4 (04)
    6 6 (06)
    6 8 (08)
    604 16 (10)
    190 17 (11)
    270 18 (12)
    116 20 (14)
    3 24 (18)
    1 28 (1c)
    343 32 (20)
    897 33 (21)
    259 34 (22)
    1 35 (23)
    262 36 (24)
    1 38 (26)
    19 40 (28)
    33 48 (30)
    34 64 (40)
    13 65 (41)
    13 66 (42)
    4 68 (44)
    213 80 (50)
    53 81 (51)
    222 82 (52)
    21 84 (54)
    8 88 (58)
    14 96 (60)
    31 97 (61)
    128 98 (62)
    7 100 (64)
    1 102 (66)
    5 104 (68)
    8 112 (70)
    17 128 (80)
    1 129 (81)
    7 130 (82)
    1 132 (84)
    10 144 (90)
    1 145 (91)
    12 160 (a0)
    7 161 (a1)
    3 162 (a2)
    10 164 (a4)
    5 168 (a8)
    22 176 (b0)
    1 178 (b2)
    11 192 (c0)
    11 208 (d0)
    2 209 (d1)
    23 210 (d2)
    4 212 (d4)
    1 216 (d8)
    1 224 (e0)
    1 225 (e1)
    4 226 (e2)
    1 228 (e4)
    1 234 (ea)

    Mostly 00, 01, 02, 10, 11, 12, 14, 20, 21, 22, 24, 50, 52 - bitfield flags?

    ~/etwng/esfxml/tmp/pathfinding_esf_dir$ cat grid_data-0001.xml | perl -nle 'print $1 if /unknown4="(.*?)"/' | sort -n | uniq -c
    2424 0 (00)
    973 1 (01)
    4991 2 (02)
    699 3 (03)
    66 4 (04)
    90 5 (05)
    2 6 (06)
    16 7 (07)
    3462 16 (10)
    2026 17 (11)
    15 20 (14)
    23 21 (15)
    1063 32 (20)
    41 33 (21)
    44 36 (24)
    55 37 (25)
    13 39 (27)
    19 48 (30)
    4 49 (31)
    342 64 (40)
    2 65 (41)
    12 68 (44)
    16 69 (45)
    2 71 (47)
    2 96 (60)
    1 97 (61)
    2 101 (65)
    22 128 (80)
    4 132 (84)
    1 192 (c0)
    1 193 (c1)

    Mostly 00, 01, 02, 03, 10, 11, 20, 40 - are these bitfield flags?

    Could it possibly be 6-8 bit id, and 24-26 bits of binary flags?
    All bits except for possibly the highest 8 tend to be 0s much more often than 1s and they don't look sequential ids or anything like it.

    I'm not sure how to represent it any better.

  20. #20
    Ordinarius
    Join Date
    Jul 2009
    Location
    Italy
    Posts
    737

    Default Re: Guide to the ETW map

    taw, have you split unknown exactly in the same way of path_id/vertex_index?

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •