Bay 12 Games Forum

Newest version doesn't seem to be working.

Loading the xml file in gives a popup saying "There was an error loading this XML file! Do you wish to attempt a repair?"


Selecting yes gives a Microsoft .NET Framework Exception: https://pastebin.com/5eTxq3EQ


If you select continue the program does nothing and never loads the legends file.

The XML files I used along with my world gen parameters can be found here if you need them: http://dffd.bay12games.com/file.php?id=14913

The ZIP you provided only includes the two XML files.
To look into your problem I need all of these files:

regionx-xxxxx-xx-xx-legends.xml
regionx-xxxxx-xx-xx-legends_plus.xml (if you have this file)
regionx-xxxxx-xx-xx-world_history.txt
regionx-xxxxx-xx-xx-world_map.bmp
regionx-xxxxx-xx-xx-world_sites_and_pops.txt
Did you give your dwarves custom names that include special characters? If that is the case, your problem could be this bug.

47.03 world.
Up until this world I hadn't noticed much loading speed difference. I've made quite a few 47.03 worlds now. This was the first one I tried with Dfhack though since 44.12.

If you provide the zipped export files, I can try to look into what causes the long loading times.

Ups, the link to DFFD still pointed to the old 1.20.01 version.

Fixed now 1.20.02!
Download here!

By the way, another modded species *does* appear as civilized and shows up normally.

Have you read my previous answer to you?

It is exactly the same problem as stated in my reply to Shonai_Dweller. I have no reliable way to map the Civilized World Population list to the race identifiers in the XML file and I have no information about tags like [INTELLIGENT] or any others in the exported files.
For Example:

Code: [Select]

Civilized World Population

9385 Dwarves
17448 Humans
11963 Elves
75109 Goblins
278 Kobolds

Total: 114183

From this part of the export is my information about which races are "civilized".
The problem is that these race names are in plural and in the XML a definition of a historicalfigure looks like this:

Code: [Select]

<historical_figure>
<id>93</id>
<name>etur</name>
<race>DWARF</race>
<caste>FEMALE</caste>
        ...
So i have to connect that DWARF is related to Dwarves, which is simple for a human but is not so trivial in code.
And as soon as there are races like CAMEL_1_HUMP_MAN, it is impossible to automatically connect to a human readable string (that is additionally in plural) without hardcoding every relation.
That is why I filed this issue in the bugtracker and hope that Toady puts the races with singular/plural and identifier somewhere in the exported XML.

To clarify:
I have an algorithm that tries to connect the RaceId (CAMEL_1_HUMP) to the human readable plural string (one-humped camels).

This algrorithm succeeds in >90% of the cases if it can follow simple rules like in the case of humans:
HUMAN -> human -> humans (add an "s")
DWARF -> dwarf -> dwarves (if it ends with "f", replace with "ves")
ELF -> elf -> elves (same rule as with dwarves)

But if the RaceId can not be brought into the plural form with a simple rule, the algorithm fails.
CAMEL_1_HUMP -> camel one hump -> camel one humps (it replaces underscores with spaces and single numbers with the word for this number)

So it looks as if one of your modded species can easily be transformed to the correct plural form and the other not, given that both appear in the "Civilized World Population" list of the regionx-xxxxx-xx-xx-world_sites_and_pops.txt file.

Hey I wanted to ask what the percentages after certain entries mean. For example stuff like Related Entities: The Holy Coal of  The Boat of Sweltering (Member 1%) or Worshipped Deities: the dwarf deity The Lavender Mother, Deity (99%). I assume it has something to do with the closeness of the relationship, but where is the percentage coming from exactly?

It is the strength of the relation, a value between 0-100.
Because of the range between 0 and 100, I chose to write the percentage symbol. Maybe not the best choice?
For worshipping deities, legends mode translates the value to a human readable string like "an ardent worshipper of The Lavender Mother", but I have not had the time to look into which subrange belongs to which string.
For membership in an entity, legends mode does not display the value at all.

What it exactly means and what the implications in the game are, you have to ask Toady.

What do you think about this? I simply exported every creature_raw with id, names, and all flags ('cause it's easy to iterate over all of them, and it saves the task of picking the ones needed, and adding missed ones later).
I placed it before sites as the data is sort of world gen ("sort of", as I suspect failed experiments are actually generated during history).

Spoiler (click to show/hide)

Code: [Select]

:
    file:write("<rivers>\n")
    for riverK, riverV in progress_ipairs(df.global.world.world_data.rivers, 'river') do
        file:write("\t<river>\n")
        file:write("\t\t<name>"..escape_xml(dfhack.df2utf(dfhack.TranslateName(riverV.name, 1))).."</name>\n")
        file:write("\t\t<path>")
        for pathK, pathV in progress_ipairs(riverV.path.x, 'river section') do
            file:write(pathV..","..riverV.path.y[pathK]..",")
            file:write(riverV.flow[pathK]..",")
            file:write(riverV.exit_tile[pathK]..",")
            file:write(riverV.elevation[pathK].."|")
        end
        file:write("</path>\n")
        file:write("\t\t<end_pos>"..riverV.end_pos.x..","..riverV.end_pos.y.."</end_pos>\n")
        file:write("\t</river>\n")
    end
    file:write("</rivers>\n")

    file:write("<creature_raw>\n")
    for creatureK, creatureV in ipairs (df.global.world.raws.creatures.all) do
        file:write("\t<creature>\n")
        file:write("\t\t<creature_id>"..creatureV.creature_id.."</creature_id>\n")
        file:write("\t\t<name_singular>"..creatureV.name[0].."</name_singular>\n")
        file:write("\t\t<name_plural>"..creatureV.name[1].."</name_plural>\n")
        for flagK, flagV in ipairs (df.creature_raw_flags) do
            file:write("\t\t<"..flagV:lower()..">"..tostring(creatureV.flags[flagV]).."</"..flagV:lower()..">\n")
        end
        file:write("\t</creature>\n")
    end
    file:write("</creature_raw>\n")
:


The complete file is too large to be posted, so I just provide the section before (unchanged) and the proposed new stuff.

Wow, really cool! Is there already a version of the script that works with dfhack alpha0 and does not crash?

What could cause my modded species to be displayed under "Other Populations" instead of "Civilized Populations" in their own civ screens? They have [INTELLIGENT] and [BENIGN] and all.

It used to be that only the civ building races showed up there, while animal peoples, gorlaks, etc. showed up under other. I haven't tried the later versions (because I haven't played at all), but assume that to remain the case.

Note that "I've modded stuff and now things don't work as I expect" questions are better asked to modders than to tool providers who may have enough trouble keeping up with vanilla, and then you'd presumably have to provide the whole modded part so they have a chance to find the mistake/issue.
Once you and/or your friendly modder support has figured out a probably reason for why it doesn't work and you think that's a tool bug/deficiency, you can come back to the tool provider with an exact description of what doesn't work (and why it ought to, if it isn't obvious to non modders as well).

It is exactly the same problem as stated in my reply to Shonai_Dweller. I have no reliable way to map the Civilized World Population list to the race identifiers in the XML file and I have no information about tags like [INTELLIGENT] or any others in the exported files.
For Example:

Civilized World Population

   9385 Dwarves
   17448 Humans
   11963 Elves
   75109 Goblins
   278 Kobolds

   Total: 114183

From this part of the export is my information about which races are "civilized".
The problem is that these race names are in plural and in the XML a definition of a historicalfigure looks like this:

   <historical_figure>
      <id>93</id>
      <name>etur</name>
      <race>DWARF</race>
      <caste>FEMALE</caste>
        ...

So i have to connect that DWARF is related to Dwarves, which is simple for a human but is not so trivial in code.
And as soon as there are races like CAMEL_1_HUMP_MAN, it is impossible to automatically connect to a human readable string (that is additionally in plural) without hardcoding every relation.
That is why I filed this issue in the bugtracker and hope that Toady puts the races with singular/plural and identifier somewhere in the exported XML.

Is the lack of Experiment names (Hfexp58691E Hum 1, etc.) something only DFhack can fix?
Also, Night Trolls (and I guess Bogeymen in there somewhere?) are referred to as "night creature" in the historical figures filter list. But experiments are "night creatures" too and they are listed separately (under code names mentioned above). Maybe want to rename this category?

I already filed an issue about that. DFHack definetly makes things with the races clearer, but I hope Toady exports race names (singular and plural) in a separate section of the XML.

Finally, "hired as a scout" is missing for historical figures (requires DFHack?).

I think that comes from the additions to the "battle" event collection. DF Legends Mode has event entries related to battles that are not exported as typical events.
But the information for these entries is in the collection, so I think I will implement some custom events that are not parsed from the XML but created directly from the "battle" event collection.

Oh. did I mention, I think I did, that Legends Viewer is my absolute favorite DF util? Just thought I'd better mention that again as I start listing criticisms.

Thanks! Always glad to hear that!