Talk:PARAM.SFO

From PS3 Developer wiki
Revision as of 16:49, 12 February 2019 by Ada L0ve Lace (talk | contribs)
Jump to navigation Jump to search

Unknown or new parameters

This page is an index/worklog/notepad with sections refered to the front page. All the sections in this page are things related with PARAM.SFO that are not completly documented, and the final purpose is to move the info when completed to the front page. This also will help to keep the front page free of speculation.

CATEGORY (new content formats)

  • CATEGORY VF - (HDD content, NOT in game column) ---> unknown name, used as a valid category filter in the XML configuration file for XMB game column in path: dev_flash/vsh/resource/explore/xmb/category_game.xml in key="gameDir"
    • key="gameDir" is an XMB item that allows to order the contents by "rating" or "timeCreated" the command used contains a list of categories that are excluded using the operator "An" = An not equal (VF is at the end of the list excluded with An+Game:Game.category VF)
    • Similar than MN, 1P?
  • CATEGORY XI - Confirmed ---> System File for Extra Content Item (or eXtra Item in short form) Used on blu ray structure?

Indi games has a specific license, if XMB differences it it will be using ATTRIBUTE flags (similar identification than neogeo and pcengine because has a special license)

  • PlayStation Mobile Implemented around 4.21 firmware, asociated with a text string in explore_plugin_full\Text\English.xml (inside explore_plugin_full.rco). See [PS3 and PSVita Cross Functions]
<Text name="msg_applications_psm">Applications (PlayStation®Mobile)</Text>

ATTRIBUTE (flags brainstorming)

XMB In-Game System Chat Voice Disable (Flag Nº7) : Value and name description are related to this flag but need to be confirmed from PS3 side / not used on bootable but as patches?.

Function to disable system chat voice utility (xmb/friend/start new chat appear when PSN connected) (implemented directly on FW or game or as flag into param.sfo for short period?) e.g. of this function: PlayStation Home 1.4 after people reported chat voice disturbance(?) , LittleBigPlanet 2.

Kept as flag for convenience (to disable per game through update game and not FW) but can be "system reminiscent" and not working anymore from param.sfo?

Boot logo Disable ? <-- this is the "license text hidden" flag ? FW max: 2.80 for ps3 fat? on patches?

Some flags may only be effective with some hardwares (PSP/PSVITA/ Torne /Nasne...), FW, Games, or Category and/or in a cumulative way and others conditions.

Unkown flag nº21 (not tested by myself btw, this is my understanding after decrypting ada hints :P --Sandungas (talk) 22:01, 4 February 2014 (EST))

  • Previously proposed to be named "XMB In-Game for Software"... but "About This Game Disabled" looks a more intuitive name (note some of the other flags in the third byte of ATTRIBUTE like "Game Purchase Enabled" and "License Logo Disabled" enables/disables an option in side menu). This 3 flags could be grouped together because "About This Game" is another option in side menu, are under the third byte, and there is another unknown flag at his side that could work in a similar way (changing options in side menu)
  • When activated, inside the game, and inside XMB In-Game, then trying to boot other content the warning text message is different (it mentions "software" and the variable %1 instead of "game"). In the next examples (copyed from a RCO) the first line is the normal error message and the second one is what appears with this flag enabled
<Text name="msg_need_quit_game_to_operate_ask">To perform this operation, you must quit the game. Do you want to quit the game now?</Text>
<Text name="msg_need_quit_to_operate_ask">To perform this operation, you must quit [%1]. Do you want to quit the software now?</Text>

ATTRIBUTE for other CATEGORY's

subcategories from X0 to X7

Subcategories

X category appear (listed on game_ext_plugin.prx) as HG (sub cat)but with different parameters and need cumulative flag (n22 at least+ 4th byte: half byte) The 4th byte is not used as binary (only?) but as Hex value for half byte (right nibble, left nibble is ignored?) to set X category. If the value is 0 or end up with 0, and the X maker flag is set under right SFO X parameters, X0 (PCEngine) is read by the system.

SFO TEST Tool from Deroad can boot and return to XMB even under X category (when MINIS don t and are stuck after being X categorized, error occurred during the start operation (80010006))

Folder position in XMB ordered by format (in fw 3.55 with hand-made PARAM.SFO's):

Bootable contents installed in game column: [triangle]>[group by]: (works with square too in any item of game column)

*from dev_flash/vsh/resource/explore/xmb/registory.xml
-format <------------- GameList:GameList.type %xCB_LIST_TYPE_GAME_FORMAT
-tag/album <--------- GameList:GameList.type %xCB_LIST_TYPE_GAME_TAGS
-all <---------------- long list of "conditionals" excluding CATEGORY's that doesnt belongs to XMB game column in the format: An+Game:Game.category **

By pressing square, group content format position (the 8 X category folders can be displayed at the same time in XMB):

  • Mini icon: PS3 icon Playstation 3
  • Mini icon: PSP Remaster icon PSP Remaster, on 3.55, Mini icon: not present (-) ,not accessible (with titles inside).
  • Mini icon: PSP Minis icon minis
  • mini icon: PSP iconPSP (Playstation portable): category PP
  • Mini icon: PS2 icon Playstation 2
  • mini icon: PS1 iconPlaystation
  • X1 Mini icon: not present (-)

Attribute used (01 00 60 01)on EBOOT.PBP SFO for Minis (group content format on PSP: minis)

  • X2 Mini icon: not present (-)

When Attribute used on EBOOT.PBP SFO on PSP: NEOGEO

  • X0 Mini icon:PCEngine icon PCEngine

Enables X category. Introduced initially as PCEngine.

Jul 15, 2009 Works alone and can be used with any "HG" category content, its not clear what does this flag other than the content has his own "tiny grey icon" with a "PCE" logo and are placed inside his own sub-folder when ordering contents (with square button) in the XMB (3.55, 3.41?). Same attributes on PSP

  • X3 Mini icon: not present (-) ?
  • X4 Mini icon:NeoGeo icon NEOGEO

Dec 21, 2010 (Few days after 3.55) Can be used with any "HG" category content. It has his own "tiny grey icon" with a "NEOGEO" logo and are placed inside his own subfolder when ordering contents (with square button) in the XMB (3.55, 3.41?).

  • X5 (same for X6 and X7) Mini icon: not present (-)

(X3 might have been used already, X5,X6 and X7 not used yet?)

Group content order is similar on PSP for the available icons (PS3 folder being PSP and no "PS2"). See Talk:Eboot.PBP#Others:_PARAM.SFO_attribut_.28PSP.29

Other app with category not implemented yet on 3.55 FW such WT or unknown as VF on 4.21 can also just not "appear" on XMB and not being on the "Unknown" folder as Corrupted/Unsupported data. To access/delete from 3.55 XMB your PSP remaster folder s titles, group content it (not by format).
  • The existence of the weird flags X0 (for PCEngine) and X4 (for NeoGeo) and we called them "subcategories" (refered to the fact that are an extension of HG category) seems to be related with the fact that those emulators are running under the PSP emulator. PC Engine and NeoGeo are part of the PSP firmware

APP_VER, TARGET_APP_VER, PATCH_FILE, PS3_SYSTEM_VER, VERSION (Patches)

Lot of things to add here, maybe needs another page, but im not the best one to document how patches works



When a patch for a game (e.g: a disc game) is installed, the game boots from the "patch installation" folder, XMB loads from this folder his Content Information Files, a new PARAM.SFO, and a new EBOOT.BIN

After one of this installations PS3 loads both PARAM.SFO's (the "original.sfo" one from disc + the new from the "patch.sfo" folder)... some values are taken from the "original.sfo" (e.g: BOOTABLE, RESOLUTION, SOUND_FORMAT... are not present in patches)... and others are taken from the "patch.sfo" (e.g: APP_VER increases in every patch)

Patches uses a system of "overwrite flag/param_key" that makes the PS3 ignore the "original.sfo" value and use the value from the "patch.sfo"... this can be used in combination with the "standard flags" (one flag indicates the value must be taken from the patch... and the other flag enables the feature)



  • Million dollars questions in relationship with ATTRIBUTE:
    • Why ATTRIBUTE flags for patches (used to overwrite other ATTRIBUTE flags or param_key's) are placed at the third byte of ATTRIBUTE instead of first byte ?. Being an special group under CATEGORY GD they could be placed anywhere, all the other groups of flags are placed starting from top of the table represented in wiki (matching chronologicall usage/implementation of them), this is the only group of flags that doesnt starts with his first (or oldest) flag with a value of 00000000000000000000000000000001
      • Is because the first and second byte of ATTRIBUTE for patches (16 flags) was used before ? (so are reserved and never was used, or used but unknown, or used but only internally) ?
      • Is because the overwrite flags for patches needed to be placed in the third byte because every byte is managed by different process ? (third byte is marked in frontpage as "flags for game & software features").
      • Is posible every byte is managed in a different level ? (e.g: one for vsh, other for lv2, etc...)
      • Other theories ?... This is a good hint of how ATTRIBUTE flags are managed, wich part of the firmware is responsible of managing them, and if are managed as separated bytes
    • When a game is booted from a disc and there is a game patch installed for this game in internall hdd, there are 2 SFO's readed (one in the disc, and another in internall hdd)... some areas from the SFO in the disc are ignored and taken from the SFO in internall hdd
      • Who is the responsible of loading parts sellectivelly from each of them ?, is just by the presence of the same param_key's in the internall SFO ?. In this case the third byte of ATRIBUTE should be an exception of this rule because the ATTRIBUTE flags in the patch (overwrite flags) uses the third byte, but ATTRIBUTE flags in the disc in third byte are readed too because contains features that are not overwrited (as game purchase enabled, install packages, etc...)
      • XMB database stores a copy of all PARAM.SFO's. Is posible the 2 files (SFO from disc, and SFO from his patch) are merged before indexed in database ? (or some of his values, more specifically the resulting of applying the overwrite flags from the patch)
    • The process of adding or removing a flag from a previous value of ATTRIBUTE is simply a math operation with numberic values, some of the flags for patches overwrites ATTRIBUTE, but there is one that allows to overwrite Remote Play (an ATTRIBUTE flag) and RESOLUTION, SOUND_FORMAT (param_key's)... The overwriting of param_key's is a different process (not a math calculation) where the new value replaces the old value. What all them has in common is all overwrited data types (math calculations with ATTRIBUTE values or overwriting of RESOLUTION and SOUND_FORMAT) are "integer 4 bytes"
    • Why overwrite Remote Play (an ATTRIBUTE flag) and RESOLUTION, SOUND_FORMAT (param_key's) are grouped together ?... This is a bit weird

CONTENT_ID, NP_COMMUNICATION_ID, NPCOMMID (Network and purchase features)

Needs to be explained in another page

ACCOUNT_ID, PARAMS, PARAMS2, SAVEDATA_FILE_LIST, SAVEDATA_PARAMS, PADDING (security, integer special data type)

All this parameters uses the data type = 0400 (in the SFO) named "utf8 Special Mode" in the wiki page (not used in SFX so we dont have an accurate name, consequently this "special mode" is an invented name)... seems to have the particularity of returning a "right" or "wrong" when comparing the value, so it can be said that the difference between the simple "utf8" and "utf8 Special Mode" is only from the point of view of how the system manipulate this data. There is no real explain yet why this data is marked differently and who is the responsible of processing it.

  • If the "scale" used for the second byte follows the proportion of 2^X (e.g. 1, 2, 4, 8, 16, 32, etc...) then we are missing a "04 01" valid data type in the list

PARAMS

This table is a collection of examples of some "rare" savedatas, probably will be needed to erase this section (replaced by a resume in front page) when all areas are identifyed, but by now is usefull to have an overview of how this process works and to find good candidates for tests, feel free to add any other examples you consider interesting ;)

The examples in the table are considered rare because most of the other games uses the values: 1,5,1,2,2 (2 last pointers pointing to the same counter slot 2 at 0x0C) or 1,5,1,3,3 (2 last pointers pointing to the same counter slot 3 at 0x10), so dont add this kind of games to the table because are pointless (by now the only identifyed pointers are the values at ffset 0x03 and 0x04 and the interesting examples are when this pointers are different, or when they changes his values to different numbers like 2,3 or 3,2, or any other combination of 2 different numbers, this means the SFO has several counters enabled), also the table doesnt includes the "user/console" areas that starts at offset 0x18 because are identifyed and explained in front page

The table differences 2 types of savedata, one for "profile" and other for standard "savedata", the reason is because are treated different by the system, after the first analysis it looks the profile support more features, it can evolve by changing his type, and it can store the counters from the other savedatas when SFO's are merged (this happens after completing the game)

  • Notes
    • Iris manager (and all his forks) has an hexeditor inside his archive manager, just navigate to the SFO press "X" and scroll down to offset 0x558 (where PARAMS area usually starts), is easy to recognize because always starts with this 1,5,1,2,2 bytes, at the next line are the 4 counters, 4 bytes lenght each
    • To know exactly wich files has been updated inside the savedata folder is needed to look at his timestamps, iris manager doesnt displays timestamps, but you can use the FTP feature to see then from a FTP client running in a PC connected to the PS3, be sure the FTP client doesnt fool you with the timestamps, usually after a disconection/conection the old timestamps are displayed, is needed to "refresh" the files (with F5 in filezilla) or close/open the FTP client again to force it to refresh timestamps
    • The system considers the savedata has been updated when the timestamp of the SFO is updated (and consequently the PFD has been updated too). But sometimes the PFD is the only file updated in the folder, in this cases the system doesnt considers the savedata has been updated. The number of times the SFO has been updated is stored in one of the counters inside PARAMS, most games loads and updates the SFO when the game boots (to be able to update his timestamp in XMB and display the last time the game was booted)
Game Title Savedata type PARAMS areas (by offset) Notes
0x0 0x01 0x02 0x03 0x04 0x05 0x06 0x07 0x08 0x0C 0x10 0x14
untold legends 1 1 1 2 2 0 0 0 00000000 40000000 00000000 01000000 old game, after completing the game
venetica profile 1 1 1 2 3 0 0 0 00000000 01000000 02000000 01000000 old game, after completing the game
yakuza 4 SYSTEM 1 3 1 2 2 0 0 0 00000000 E1010000 00000000 01000000
yakuza 4 CLEAR 1 1 1 2 2 0 0 0 00000000 01000000 00000000 01000000 after completing the game, needed to start a game+
need for speed most wanted 1 5 1 2 3 0 0 0 00000000 7A010000 64000000 01000000
motogp 13 1 5 1 3 3 0 0 0 00000000 00000000 16000000 01000000
tintin 1 5 2 2 2 0 0 0 00000000 AC010000 00000000 01000000
prince of persia sands of time 1 5 1 2 2 0 0 0 00000000 FF000000 00000000 01000000
prince of persia sands of time 1 5 2 2 2 0 0 0 00000000 E6000000 00000000 01000000
the last of us profile 1 7 1 3 3 0 0 0 00000000 00000000 01000000 01000000 new profile savedata, no previous savedatas
the last of us profile 1 8 1 3 3 0 0 0 00000000 00000000 02000000 01000000 boot game, quit game (see examples below)
the last of us profile 1 7 1 3 3 0 0 0 00000000 00000000 97000000 01000000 after completing the game
the last of us save 1 2 1 3 3 0 0 0 00000000 00000000 21000000 01000000 after completing the game
tomb raider profile 1 5 1 3 3 0 0 0 00000000 00000000 01000000 01000000 new profile savedata, no previous savedatas
tomb raider save 1 5 1 3 3 0 0 0 00000000 00000000 01000000 01000000 new game savedata, no previous savedatas
tomb raider profile 1 5 1 2 3 0 0 0 00000000 91000000 14000000 01000000 after completing the game
  • Tests examples
    • The last of us ---> http://pastebin.com/raw.php?i=SAiDzjLN (overall save behaviour with the profile, identifying update counter)
      • the game loads and updates the profile in every boot (by only updating the SFO timestamp and consequently the PFD), and changes the value at offset 0x01 to 8
      • when the user changes some option or forces the save the value at offset 0x01 is changed to 7
      • the profile is updated 1 time for every submenu inside "options" (there are 5 submenues under options)
      • every time the SFO is updated, the counter at 0x10 increases 1 unit
    • tomb raider ------> http://pastebin.com/raw.php?i=S13yrq9R (overall save behaviour with profile and save, update counter, timestamps, failed attempt to increase the game completion counter at offset 0x564)
      • every time the game boots, only the PFD is updated (no other file is updated)
      • When the savedata is updated all the files inside the folder are updated (included ICON0.PNG timestamp)
      • the profile is updated 1 time for every submenu under options, and another time when exiting options (3 submenues + 1 when exiting options for a total of 4 updates)
      • the profile is updated 1 time when exiting extras
      • every time the SFO is updated, the counter at 0x10 increases 1 unit
      • when the game is completed a new counter is added to the profile savedata, and the pointer at offset 0x03 changes to a new counter slot (from slot 3 to slot 2), the new value stored in counter slot 2 is partially unknown but seems to be another counter taken from the SFO of the save used to finish the game

SAVEDATA_LIST_PARAM

Example of SAVEDATA_LIST_PARAM containing the "folder size in sectors" in the gamesaves generated by the game "The Orange Box: BLES00153, BLES00171, BLES00172, BLUS30055"

Game Selector Boot Process

When the disc boots, the EBOOT.BIN loads APP_CHOOSER.SELF, is a "game selector" and contains 5 games: half life 2, half life 2ep1, half life 2ep2, Portal, and Team fortress 2

Each of this games has his own .SELF and his own SAVECONTENTPS3 folder (each game uses different Content Information Files when generating his savedata). There is another savedata folder generated/loaded by the selector that is common. In some games (e.g: portal 1) there are several savedata folders to store: the "game settings", "autosave", and "manual save"

All the PARAM.SFO's generated in this savedata folders contains SAVEDATA_LIST_PARAM with a param_len of 7 bytes, and his values are strings of 6 numbers + a zero (NULL terminated)

The value stored by this game in SAVEDATA_LIST_PARAM is the "size of the folder on disk". This size is dependant of the filesystem (512 bytes cluster size in UFS2), and how many clusters are taken as the "minimun chunk" (that seems to be 2 clusters)

Because the writing process standards (and from the point of view of available clusters for the next writing operation) when a single byte is used in a cluster, this whole cluster is considered "not available for next writing process of a new file". This means e.g: that a file with a size 513 bytes uses 2 complete clusters (512 bytes in the first cluster and 1 byte in the second cluster, both clusters are reserved for this file)

To calculate the number of clusters used by a folder... is needed to calculate the number of clusters used by each file inside the folder, this includes the files that uses a number of clusters that is not even and needs to be "rounded-up". e.g: a file with size 1 byte (that is supposed to fit in 1 cluster) uses 2 clusters because the number "1" is not even. In few words, the number of clusters used by a file needs to be "rounded-up to next even number"

  • This save game is a good example because always contains the same values in SAVEDATA_LIST_PARAM (the second one from euro region): BLES00153-PORCONF-0 [Link 1] [Link 2]
File name real file size (in bytes) theorical disk size (in clusters) cluster round-up to even number real disk size (in bytes)
ICON0.PNG 12.461 12461/512=24,3... = 25 25+1= 26 26*512= 13.312
PARAM.PFD 32.768 32768/512= 64 64*512= 32.768
PARAM.SFO 2.736 2736/512=5,3... = 6 6*512= 3.072
PIC1.PNG 89.268 89268/512=174,3... = 175 175+1= 176 176*512= 90.112
POR-0.CFG 893 893/512=1,7... = 2 2*512= 1.024
POR-0.CHK 16 16/512=0,03... = 1 1+1= 2 2*512= 1.024
Folder contents total
138.142 273 clusters 276 clusters 141312 (SAVEDATA_LIST_PARAM)
  • The process is:
    • 1.- Divide by the cluster size
    • 2.- Round-up cluster number (if needed)
    • 3.- Multiply by the same cluster size (to know the amount of bytes reserved for this clusters)

This can be simplifyed by using 1024 (2 clusters)... because this way there is no need of step 2 (the "round-up")... but i have no idea if naming this "chunk" of 2 clusters a "page"... or why UFS2 filesystem takes them 2 by 2... i prefer to left this example as is, feel free to correct me

Also, there is a simple "trick" that can be used to display the files divided in clusters of the size you want (maybe are others but this is the faster way i found). By formatting a usb pendrive in advanced mode, using the cluster size your need, then moving the folder there (by using FAT32 with 1024 bytes cluster the value reported by your operative system is the same one stored inside SAVE_DATA_LIST)

  • PARAM.PFD relationship ?


Other parameters not used in PS3 (only used in other sony consoles)

The only purpose of this table is to keep a record of them to not confuse with the ones used in PS3

Parameter Console Notes
PSP PSV PS4
DISC_ID Yes
DISC_NUMBER Yes Yes
DISC_TOTAL Yes Yes
DISC_VERSION Yes
DRIVER_PATH Yes
GAMEDATA_ID Yes Used in some PSP UMD games, found in game_ext_plugin.sprx
HRKGMP_VER Yes
REGION Yes
USE_USB Yes
ACCOUNT_ID Yes Not the same ACCOUNT_ID than PS3... this one for vita has (fmt:utf8-S, len:0x8, max_len:0x8)
ATTRIBUTE2 Yes
ATTRIBUTE_MINOR Yes
GC_RO_SIZE Yes GameCache ReadOnly Size ?
GC_RW_SIZE Yes GameCache ReadWrite Size ?
INSTALL_DIR_ADDCONT Yes Shared Additional Content, fmt:utf-8, max_len:12
INSTALL_DIR_SAVEDATA Yes Shared Save Data, fmt:utf-8, max_len:12
PARENT_DIRECTORY Yes used in PSvita savedata (fmt:utf8, len:0xB, max_len:0x40)
PSP2_DISP_VER Yes utf-8, max_len:8, example: "00.000"
PSP2_SYSTEM_VER Yes int32
PUBTOOLINFO Yes Yes
SAVEDATA_MAX_SIZE Yes int32
SAVEDATA_VER Yes used in PSvita savedata (fmt:utf8, len:0x6, max_len:0x8)
STITLE Yes App Short Title, utf-8, len:52
STITLE_xx Yes
SUPPORT_URI Yes
APP_TYPE Yes
CONTENT_VER Yes
DISP_LOCATION_x Yes
DOWNLOAD_DATA_SIZE Yes
FORMAT Yes
INSTALL_DIR_SAVEDATA Yes
IRO_TAG Yes
PROVIDER Yes
PROVIDER_xx Yes
PUBTOOL_VERSION Yes
REMOTE_PLAY_KEY_ASSIGN Yes
SERVICE_ID_ADDCONT_ADD_x Yes
SYSTEM_ROOT_VER Yes
SYSTEM_VER Yes
USER_DEFINED_PARAM_x Yes
SUB_TITLE_ID
CONTENT_URL URL of content (not parameter but related to link.xml?)
_PARAM.SFO ? found in game_ext_plugin.sprx (anti exploit/bug prevention ?)
__PARAM.SFO ? found in game_ext_plugin.sprx (anti exploit/bug prevention ?)
/PARAM.SFO ? found in game_ext_plugin.sprx (anti exploit/bug prevention ?)
FOLDER_NAME found in game_ext_plugin.sprx
FOLDER_NAME_00 found in game_ext_plugin.sprx

DISC contents temporal installations (in dev_hdd0/tmp)

/dev_hdd0/tmp/disc_install/

Reserved Folder Names used in HDD or disc contents

L00 up to L05 (License Area subfolders with restricted .SFO's)

  • From dev_bdvd: ?
/dev_bdvd/PS3_GAME/L00/PARAM.SFO (up to L05)
/dev_bdvd/PS3_GAME/C00/L00/PARAM.SFO (up to L05)
  • From dev_hdd0:
/dev_hdd0/game/ABCD12345/Lxx/PARAM.SFO
/dev_hdd0/game/ABCD12345/C01/Lxx/PARAM.SFO ---only working if region setting is L03 country (only other tested: L00 not working)

Speculation: Path used to restrict/use different License in the same region (L03?).

You can quickly check your region area with this sub folder (and inserting content file informations such ICON0). L** path can avoid error some code related to param.sfo/disc.sfo.

Folder name License Area Regions
L00 SCEJ Japan
L01 SCEA US (L01), Canada (North America), Mexico (L01), Central America, South America
L02 SCEE Europe/Middle East/Africa, U.K./Ireland, Australia/New Zealand (Oceania: L02), Russia, Ukraine, India, Central Asia ,
L03 SCEH Singapore/Malaysia (Southeast Asia)?, Taiwan (L03), Hong Kong (L03)
L04 SCEK Korea (South Korea)
L05 SCH China

(Unlinked Note: Button Behavior: for L05 is also O , Oceania is X. Path Pioneer with significant idea about it: Deroad)

You can load different Lxx folder by using Debug Menu Setting Region but be alert about your TV setting display video (NTSC/PAL).

Here is a collection of folders (from L00 to L10) with icons displaying the different numbers, this way is easy to know which one is used by XMB, to make the test just copy the folders next to USRDIR folder inside the installation folder of any app: region subfolders tests.zip (430.22KB) http://www.sendspace.com/file/ikvult

See also: "license area" restrictions used in combination with "parental level" http://www.ps3devwiki.com/wiki/Talk:PARAM.SFO#Lxx:_License_Area

EX_GAME and EX_VIDEO

  • Theory 2:
    • EX_GAME & EX_VIDEO requires SFO Category 'XR' and are emerged into PS3_EXTRA! (EX_GAME/EX_VIDEO are actually leftovers?)

Others

-GE
PS3 DVD / PS2 CD
/dev_bdvd/NORMAL
/dev_bdvd/HYBRID
-GP
/dev_bdvd/PHOTODIR
/dev_bdvd/PS3_CONTENT/PHOTODIR
/dev_bdvd/PS3_CONTENT/PHOTODIR/PARAM.SFO
-GM
/dev_bdvd/MUSICDIR
/dev_bdvd/PS3_CONTENT/MUSICDIR
/dev_bdvd/PS3_CONTENT/MUSICDIR/PARAM.SFO
-GV
/dev_bdvd/VIDEODIR
/dev_bdvd/PS3_CONTENT/VIDEODIR
/dev_bdvd/PS3_CONTENT/VIDEODIR/PARAM.SFO
-GT
/dev_bdvd/THEMEDIR
/dev_bdvd/PS3_CONTENT/THEMEDIR
/dev_bdvd/PS3_CONTENT/THEMEDIR/PARAM.SFO

Todo and Suggestions

  • Parameters requirements table has lot of "yes" that should be "opt" (optional). The "yes" means the value is mandatory, so the table is not fully correct.Also... this table represents the values needed for each content ordered by CATEGORY, but there are different contents that shares the same CATEGORY (like the different types of patches, DLC's, and game installations... alll them uses "GD". The "parameter table" cant contain all this different variations, by now i dont see a way to do it--Sandungas (talk) 12:25, 29 January 2014 (EST)
  • The internal structure example for SFO needs to be replaced by a new one with at least 2 parameters, this will make more obvious how are meassured the relative offsets displacements
  • Is needed an explain of how the file is loaded in stages by the system, how is indexed, and the method to access single areas to update single values, the rebuilding process, etc... this is complex and i ignore (but i imagine) how is made by the PS3, but the source codes i saw that manages SFO files uses different methods and styles (and probably some others i ignore), by now im not skilled enought to discuss this in the page but an open talk about all this could be interesting --Sandungas (talk) 23:41, 15 February 2014 (EST)
  • Overall page cleanup of obsolete or wrong info, grammar corrections, etc...
  • Link each ID to a subpage, so they can be centrally managed (some are also listed here: ID's and labels
    • The affected sections in front page are: CONTENT_ID, NP_COMMUNICATION_ID, NPCOMMID, and TITLE_ID, explaining them in the .SFO page (as i did initially) doesnt looks the best solution, but is needed at least a short description of each, now there are specific pages for them (without much descriptions) and the decission of separating all the IDs in pages makes hard to have a overall view of them... the best solution i see is to use the ID's and labels converted to a template an use it repeatedly in SFO page, but the names are not the same used in SFO, also this doesnt solves the short description needed in SFO page (that should be duplicated, and expanded in his specific pages)--Sandungas (talk) 12:25, 29 January 2014 (EST)