Jump to content

Dragonblood

Members
  • Posts

    126
  • Joined

  • Last visited

Reputation Activity

  1. Thanks
    Dragonblood reacted to zilav in [WIPz] TES5Edit   
    You clearly don't understand how plugins work and what does it mean for a plugin to be a master for another one. Unfortunately it'll take too much time to teach you which I don't have. I suggest you to use the existing mod merging tools, or join xEdit discord and ask people around there to teach you, maybe someone will have time.
  2. Thanks
    Dragonblood reacted to lmstearn in [WIPz] TES5Edit   
    @Dragonblood: There's this old thread which kind of explains stuff. It recommends the CK. Any good to you?
  3. Thanks
    Dragonblood reacted to zilav in [WIPz] TES5Edit   
    I understand what you said - you want an easy solution to merge plugins, and there are: merge plugins standalone tool and it's predecessor, the xEdit merging script, which will merge 2 esps just fine in one click.
    However when you are asking for a way to copy records from one esp into another without adding said esp as a master, now this is what shows that you have a vague understanding of how plugin masters work and that such request is simply impossible to implement without breaking the data. I told you already the only possible (correct, safe, working) way of removing masters, but seems you didn't like it.
  4. Thanks
    Dragonblood reacted to zilav in [WIPz] TES5Edit   
    No, you have been asking for how to copy as override records from esp1 to esp2 without adding esp1 as a master. I told you it is impossible and you should copy and then remove master the proper way, and described how to do so. You didn't like the answer and continued to insist on adding an option to xEdit of copying records without adding masters because you've been doing it wrong for 5 years.
    If you really want to know how to properly merge mods manually, then this is covered in the xEdit Training Manual (the Help button in the right top corner, and linked on every xEdit's description page on Nexus, but who reads nowadays). The latest xEdit builds made it easier:
    Right click on B, click Add Masters and chose A Right click on B and select "Inject Forms Into Master..." menu, chose A. It will ask to optionally preserve ObjectIDs (the YYYYYY part in FormIDs XXYYYYYY numbers), in your case it doesn't matter when simply merging two mods. Expand B, select all record groups, right click, hold Ctrl+Shift and click "Deep copy as override into", chose A. Say Yes to overwrite any changes if asked. Close and save A only. Now you have A+B properly merged into A.
  5. Like
    Dragonblood reacted to zilav in [WIPz] TES5Edit   
    From the video it is clearly visible that LODs are not from vanilla trees, they are scott pines from SFO probablly. Vanilla game has only ordinary pines and aspens.
  6. Like
    Dragonblood reacted to zilav in [WIPz] TES5Edit   
    Those are not vanilla trees, so he MUST use DynDoLOD/TES5LODGen with proper billboards. The issue is not a wrong Z height, but wrong LODs. The patcher script itself doesn't even touch Z values, only X and Y.
  7. Like
    Dragonblood reacted to zilav in [WIPz] TES5Edit   
    I don't see why it wouldn't. There are no changes between Skyrim and FO4 regarding how the localization works.
     
     
    That is a very specific task. In 8 years of xEdit history this is the first time I hear that someone ran out of FormIDs space, probably because one of plugins used very high values for it's records for some reason which leaved huge empty unused range(s). I prefer that you'll implement that in your merger directly since it is relative to merging task. And/or if you can write a script for that, I'lll include it with xEdit.
  8. Like
    Dragonblood reacted to zilav in [WIPz] TES5Edit   
    xEdit parses records on demand, so keeping track of additional errors objects per file is like another form of reference building with all the update code overhead on every operation and increased memory usage. Doesn't worth it for the purposes of xEdit since showing errors to the user is more than enough. I can see it being useful for your task, but you already solved that
  9. Like
    Dragonblood reacted to zilav in [WIPz] TES5Edit   
    Those are structural data errors along with unused and insufficient data, they are reported through wbProgressCallback() so intercept them in that handler.
    You don't need to build references to check for errors, none of the current custom checking functions use them and probably never will.
  10. Like
    Dragonblood reacted to zilav in [WIPz] TES5Edit   
    Not yet. xEdit should pass 2 tests before it can be even considered stable
    1. Check For Errors shoudn't report false positives on master file
    2. Copy As Override should work on all vanilla master records and always produce ITM data, even copying full worldspaces if there is enough memory,
     
    I'm now on the first test only and it is almost working as intended.
  11. Like
    Dragonblood reacted to zilav in [WIPz] TES5Edit   
    Uploaded build that fixes attack data issue in Skyrim.
    You can now use "Check for Errors" on Fallout 4 plugins since 95% of definitions are done.
  12. Like
    Dragonblood reacted to zilav in [WIPz] TES5Edit   
    That was a bitch to catch. First I created a new plugin with that record from Dawnguard using CK, then the same one using xEdit. All meaningful fields were identical but still the xEdit's one didn't work. Then I right clicked and mark modified a record from CK's plugin in xEdit and saved. This caused the whole record to be reconstructed using xEdit's definitions and recompressed, and this one still worked fine when loaded in CK, so at least xEdit was not losing any values that are already present in a plugin.
    I replaced all attack data subrecords with hex arrays and compared both plugins, sure there was a single bit difference otherwise I would start to pull my hairs. Turns out that when I was decoding Skyrim, I copied flags definition code from somewhere else in the source code to not type all those hex bit masks over and over again myself, but didn't spot that it has a parameter to ignore undefined flags. So there is an extra flag set in Attack Data which xEdit was hiding as commanded by definitions and was not copying itself, that's why plugins looked identical but worked differently. However as I mentioned it doesn't remove that flag either if it is already present, so saving CK made plugins in xEdit was safe in this case.
    I'll update that in the next build, stay tuned
  13. Like
    Dragonblood reacted to zilav in [WIPz] TES5Edit   
    Uploaded new build to fix those errors in DLC.
     
    Regarding ONAMs in ESP. No matter what I tried to do in CK (changing scenes, dialogues, moving vanilla refs around, etc.), I couldn't get it to create them in my test plugin. However when I load unofficial patch and save it, CK does create ONAMs. So it is probably a lookup optimization data for changed records triggered to be created when a plugin reaches some size threshold, or whatever.
    Anyway FO4Edit will now always generate ONAMs in any FO4 plugins.
  14. Like
    Dragonblood reacted to zilav in [WIPz] TES5Edit   
    I'm still not convinced that it is not another Beth bug in CK. There are mods made in FO4Edit on nexus that change existing refs and dialogues, they do work just fine without any ONAMs.
     
    Can you test yourself please to check if there any difference between mods with and without ONAM that modify vanilla refs, landscape, navmeshes and dialogues?
  15. Like
    Dragonblood reacted to zilav in [WIPz] TES5Edit   
    http://www.afkmods.com/index.php?/topic/3750-wipz-tes5edit/page-59#entry160258
     
    Wait... what? I'll check that when I get to decoding refs. By the way ONAM now also stores FormIDs of quest children: SCEN, DLBR, DIAL and INFO. This is a pure fun.
  16. Like
    Dragonblood reacted to Kesta in [WIPz] TES5Edit   
    As far as I know, xEdit scripting informations are centralized here: https://www.creationkit.com/index.php?title=TES5Edit_Scripting_Functions
     
    Anything you know know / learned that's not here, feel free to add it there (at least in the discussion page). 
     
    Also tagging @Mator, he had a WIP tutorial at some point, (more like an actual guide, less like a reference). Not sure where it's at now, lot of things happened since then.
  17. Like
    Dragonblood reacted to zilav in [WIPz] TES5Edit   
    Indeed. GMST records should have a valid FormID, however it is unused by the game and xEdit, they override each other by name in Editor ID field. So any mod that wants to set a game settings can either override the one from the game master if it exists there, or just add a new GMST with appropriate name.
  18. Like
    Dragonblood reacted to zilav in [WIPz] TES5Edit   
    No idea. I even not sure that the above method is a good idea since there are occlusion and LOD issues when you start uncombining references judging by comments on nexus.
  19. Like
    Dragonblood reacted to zilav in [WIPz] TES5Edit   
    You can write such script now:
    1) Loop over references of STAT
    2) check if it is precombined
    3) get precombined mesh it is in
    4) remove that mesh/ref from XCRI index of CELL for all refs that are precombined together in the same mesh
     
    This way you are uncombining only refs that needs to be instead of everything when using ini setting.
     
    Better is to recombine meshes that this references are a part of, but that will require mesh work which is out of xEdit's scope for now.
  20. Like
    Dragonblood reacted to zilav in [WIPz] TES5Edit   
    Nope, will fix later.
  21. Like
    Dragonblood reacted to zilav in [WIPz] TES5Edit   
    ARMA is waiting to be decoded, I'll mark it later.
    Precombined entries in CELL were sorted at some point but caused huge lags when copying records because amount of entries in some of them reaches 22000+, so I disabled sorting. xEdit needs some update for internal sort timings and avoid excess ones when copying containers.
    How do you label ALCS? I know that it is an alias, but no idea about it's purpose.
  22. Like
    Dragonblood reacted to zilav in [WIPz] TES5Edit   
    Uploaded new build.
    Updated FO4 definitions and better handling of new plugins.txt format in Beta patch.
    Combined references will be displayed as [Placed Object] in square brackets and mesh name is added into the name
     
    [REFR:000782AA] (places BldWoodPBgRoomCorner03 [sCOL:0001E597] in GRUP Cell Temporary Children of MinutemenHoldingCell "Minutemen Holding Cell" [CELL:000782A8] in Precombined\000782A8_0105A45C_OC.nif)
     
    New option in filter window by precombined status.
  23. Like
    Dragonblood reacted to zilav in [WIPz] TES5Edit   
    What about COMCodsworth [QUST:000BBD94] in Automatron with similar missing data? Or it is intended change?
  24. Like
    Dragonblood reacted to zilav in [WIPz] TES5Edit   
    If those quests and related removed data still work in game with DLCs loaded, then this flag is indeed meant for inheritance similar to how NPCs inherit portions of data from templates. Maybe this is another difference between the public cut version of CK and the real one.
    But if this is indeed a genuine error, then I can't belive how they allowed it to slip through into the relase versions of DLCs because those data removals should be breaking quests.
  25. Like
    Dragonblood reacted to zilav in [WIPz] TES5Edit   
    It is too late now. If I add that behaviour, anyone who at least once opened Options window and clicked Ok has them saved in ini, so master file refs will be built for them automatically since the game master file won't be in that list. I'd need to implement versioning for options to solve this, and that's too much work for non existing problem without any benefits except for a few users.
×
×
  • Create New...