User talk:Darklocq

The UESPWiki – Your source for The Elder Scrolls since 1995
Jump to: navigation, search
Feb.–Apr. 2017


Hey there! Could you give me a quick overview of how OpenMW works? I've had a quick read through OpenMW already, so I understand the basics of what the project does, but I'm still wondering about the technical and legal standing of an open source implementation of Bethesda's proprietary game engine, and how it fits with the ZeniMax Media Terms of Service. Thanks! --Enodoc (talk) 16:31, 23 February 2017 (UTC)

It's a complete reverse engineering. They figured out the file formats (which were poorly documented but not totally undocumented, or no one could have written any of the game-editing and -tweaking tools to begin with, other than Bethesda's own people). Effectively, it works just like a game-editing application at the basic level - its reads in and parses the content files and takes action based on the information it finds in them. In this case, it presents the game to the player, by rendering models, applying game mechanics rules, etc. While you could get an all-new game that third parties had created from scratch using the same data file formats (e.g. some sci-fi game about hunting aliens or whatever), the intent of the project is to run Morrowind in particular and its official add-ons with exactly the same game mechanics and other aspects of game play and appearance (aside from being able to render better graphics, including better, TrueType-based, fonts) as Bethesda's original, and to support its modding architecture. OpenMW doesn't come with a copy of the game; one has to obtain that separately, and OpenMW doesn't itself include any of Bethesda's materials. It's basically just an alternative to Bethesda's EXE and DLL files. It's not "an open source implementation of Bethesda's proprietary game engine", it's an implementation of an all-new, non-proprietary game engine with no Bethesda code. A close analogy would be PDF-reading and -editing applications not created by Adobe. They do exactly the same thing: take a file format, parse it, and do something with it, with results that closely match how Adobe intended it and did it. A perhaps closer analogy would be all the applications that are not Adobe's which can support Photoshop and Illustrator plugins. They're not only capable of reading and outputting Adobe's (among other) data formats, they support the original Adobe apps' "modding" framework as well. — Darklocq  ¢ 18:12, 23 February 2017 (UTC)
That's great, thanks! And this is all okay with the Terms of Service? I'm thinking specifically "You may not modify, adapt, reverse engineer or decompile the Software, or otherwise attempt to derive source code from the Software". --Enodoc (talk) 23:26, 23 February 2017 (UTC)
Certainly, or the project could not have continued publicly since 2008. They have not reverse engineered Bethesda's software, they just built new software that reads the same data file formats, like all the mod editing tools do. (I probably shouldn't have described even that as any form of reverse engineering, since the formats aren't secret, otherwise there would be no editing tools other than Bethesda's original; that was imprecise wording on my part). It really is exactly the same process as creating your own .PDF or .PSD editing and viewing app that works with files produced by Adobe's Acrobat and Photoshop tools. Some of these game file formats aren't even Bethesda's to begin with, but created by others and chosen by Bethesda. (I read a write-up of the original sources once, but forget the details. I'm pretty sure .NIF and .BSA were someone else's work, while .ESP/.ESP is Bethesda's, but I may be getting that backwards). But this isn't a reverse-engineered Morrowind.exe, to begin with; it's an all-new rendering engine, using 0.0% of Bethesda's old source code.

Anyway, there are no legal issues that have been raised against OpenMW, or the project would have collapsed instantly; it's all-volunteer, and has no budget for attorneys. UESP itself is in more risk than OpenMW, since it includes Bethesda's "Elder Scrolls" trademark in the long version of its name. Heh. But Bethesda/Zenimax has gone out of its way to encourage fan-built resources for its games, and never rattles any intellectual-property sabers, other than about outright software piracy (and other companies trying to launch competing games with too-similar names [1], and some other business-to-business suits, e.g. against Interplay for not upholding their end of a licensing agreement, unrelated to TES). It has a huge, valuable user community as a result of this hands-off approach (there are over 10,000 released mods for MW alone). I'm sure Zenimax/Bethesda realize OpenMW is a surprise cash cow for them, since the new market of players all need a copy of their original game to use OpenMW. Game companies in general mostly know better than to give their fanbase trouble over things like this, per the Streisand effect (Nintendo/Pokémon Company Intl. seem to be the exception and are pissing a lot of people off [2]).
— Darklocq  ¢ 01:14, 24 February 2017 (UTC)

Brilliant, thanks for all the details! Don't forget to add a link to the project page on Tes3Mod:Mods! (I know it's not technically a mod, but I'm not sure where else you could put it.) --Enodoc (talk) 09:36, 24 February 2017 (UTC)
Right. That was my thinking, too. I did create a Category:Tes3Mod-Engines for that, for TES3MP (no article yet), and for the couple of articles we have about the original Bethesda engine as such. — Darklocq  ¢ 10:02, 24 February 2017 (UTC)
I'd forgotten, but finally to around to adding it to TES3Mod:Mods [3]. — Darklocq  ¢ 20:53, 12 August 2017 (UTC)


Tes3Mod:TES3MP ‎ article is now created. Thank you for making a good category for it. --Testman (talk) 04:21, 29 June 2018 (UTC)

Testman, schweet. I'd thought of working on it myself, but it would have just be a tiny stub rehashing their own "about" page; I haven't actually used TES3MP, though I really hope it does more and more. I've been working through Oblivion and while I like a few "new" (since MW) features of it and the improved graphics, the gameplay, the plot, the interface, the rather cliché, unicultural Western milieu, and so much else is such a step backward. I really want to get back into MW, which is like another home to me. It would be really cool as a multi-player thing, if it wasn't just a constant murder-fest, LOL. — Darklocq  ¢ 00:31, 28 July 2018 (UTC)

Semantic Markup[edit]

[Refactor: Out-dented and split off from previous thread, now archived.]

In truth, I think your points about using semantic markup are perfectly reasonable, and we should be making more use of the appropriate tags in many cases. But, at the same time, I don't think it's necessary to supplant wiki markup with semantic markup as a preferred style just because it's somewhat more appropriate in some instances. In cases like our technical and template documentation, on the other hand, I think it makes a significant amount of sense, and I've spoken with others who agree with that idea. It's not that we intended to reject your ideas out of hand, just that they needed some more discussion, and to find a middle ground where we're keeping the easy/familiarity of wiki editing for many cases, but acknowledging those cases where semantic markup makes a lot of sense. Robin Hood  (talk) 00:41, 15 April 2017 (UTC)

Thank you for addressing this in particular. It's nice to get past some of this he-said-she-said stuff. I don't draw the distinction you draw; if anything, I would suggest that it's more important to use semantic markup in the public-facing material than in internal documentation that few will see or care about. But any progress on this is a step forward. The major hurdle is probably simply that the average person doesn't know the <em></em> element exists, and is convinced that <i></i> is for emphasis, and will not read the specs which say otherwise. I suspect I should not get involved further in the discussion, at least not for a long time, because toes have been stepped on, on both sides, and it's probably better to wait until they're not sore any longer.

PS: I don't think anyone is "bad" here. Rather, I think a well-known effect is in play, that of topical forums, wikis, and other groups with narrow interests and a small number of regulars becoming increasingly hidebound as their active participant pool shrinks and they all become intimately familiar with each other while resistant to newcomers and their influence. The exact same effect is observable in work groups, political coalitions, AA meetings, specific MMORPG servers, military units, Facebook discussion boards, etc., etc. It's something that has to be recognized and worked against, or it just inexorably happens as part of human small-group dynamics. Wikis in particular, like open source development projects, are usually a bit more resistant to this effect because of their "we want everyone to volunteer and help out" nature, but the more narrow the focus and pool of people, the less resistant it will be. At least without prodding, which is why I've prodded. Wikis are worth prodding in this regard, because their long-term viability depends on suppression of this trend. There's also a certain level of "FIFO" to most small groups, and I'll eventually FO to go do something else if I don't seem to FI, even if I grumble about it.
— Darklocq  ¢ 03:11, 15 April 2017 (UTC)

This being a wiki, I doubt if most people think in terms of <i> (or <b> or <table> or whatever else) at all. Wiki markup is assumed to be the default, and tends to be, even if it's largely just an alternate choice of markup for essentially the same underlying HTML code. Yes, technically, <em> is the semantically better choice for emphasis in dialogue, but what purpose would semantic markup serve here for that particular element? I don't see how most (if any) of the traditional benefits would apply to our gamespaces. It's unlikely we have a great many users who are using screen-readers, after all, and I think most search software is well aware that semantic markup isn't in widespread use on a lot of sites.
Using things like <kbd> and <var>, on the other hand, makes a lot more sense in our technically oriented namespaces, since you're often referring to very specific things, and using tags consistently for the same types of elements makes it easier to identify those elements visually (and format them to your liking). Of course, that's not without its own set of issues. For instance, should a template parameter be inside a variable tag, since it acts sort of like a variable, or should it be in a keyboard tag, since it's something you type in? HTML lacks a <param> tag (at least in this actually does have one for other purposes), which would be ideal here, though CSS classes can readily perform a comparable function. But that's a whole other set of issues that are really more for those involved in documenting in the relevant namespaces to figure out for themselves. My larger point here is just that there's more reason to use proper semantic markup in our technical writing.
I would also note that, despite your efforts on WP, neither {{em}} nor {{strong}} have gained much acceptance there, so we're hardly alone in using traditional wiki markup in preference to semantic HTML markup. Robin Hood  (talk) 08:05, 15 April 2017 (UTC)
"This being a wiki ...": Yes; that's why to do it with templates, which are part of "native" wiki markup style, instead of with bare HTML. The principal argument for semantic markup on a site like this (which, I agree, won't have many visually impaired readers) is user stylesheets, either the MediaWiki local kind, or the in-browser kind, the main approach to which these days seems to be scripting addons like GreaseMonkey. For example, you can switch to a different font for semantic emphasis if you want, to highlight it as definitely intended emphasis, not typographic italics for some other purpose (like book titles, phrases in another language, heading decoration, or mistaken italicization of quotations for no apparent reason, a style virtually unknown in professional print publications but common on blogs, mysteriously). A more general purpose is content reuse; the content can be parsed differently by tools depending on the markup, e.g. use a script to build a list of all pages with semantically emphasized content as likely to contain important tips. Keep in mind also that use of <i></i> and <b></b> (and everyone's un-favorite, <u></u>) for emphasis has been sharply deprecated, even to the point that it took a lot of fighting for these elements to be retained in HTML5 at all, rather that thrown out in favor of CSS applied to <span></span>. Semantic markup is more future-proof. The semantic markup stuff is in Wikpedia's MoS; as on this wiki, the average editor does not actually read much less memorize and assiduously abide by the style guide there. Wiki style guides exist almost entirely for consistent post hoc cleanup efforts by "gnomes", and for resolution/prevention of annoying style bickering.

On the markup tags (it may help to view source on the following):

  • <code></code> is generally used to wrap entire blocks of code, and inline snippets thereof, including wiki template parameters, and the other "computerese" semantic markup tags can be used within or without <code></code>. Technically, a large block of code marked up with <pre></pre> for formatting reasons should actually be done as <code><pre></pre></code> but people often forget (and on a few sites where <code></code> has been "over-styled" with CSS, that doesn't work well; en.Wikipedia is among them). Should always be monospaced.
  • <var></var> is for markup of actual variables, in sentences (as in "Given x Enchantment points, the casting cost to price ratio ..."), in formulae not already marked up with MathML or the like, and in code samples for variables or examples not meant to be taken as literal strings (as in "|text=Your message here"). Most browsers render it in an italic font (inherited from mathematics practice, I think), but as with <em></em>, that's determined by site CSS and ultimately by user stylesheets. It should inherit the CSS font-family of the parent element, so it will not lose the monospacing if used inside <code></code>, nor force monospacing if used in running text. The two below should always be monospaced, except for special applications; see later reply.
  • <kbd></kbd> is used for representing keystrokes, as in "press the Ctrl-Alt-Del key combination to get to the Task Manager in Windows." It can technically also be used for parameter values ("user input", even if from a fixed list of valid values) that are literal and not variable examples, e.g. "The |game=Morrowind parameter and value will produce different output than |game=Oblivion". I rarely see people bother with that, since the visual display of <code></code> and <kbd></kbd> is the same on many sites. People have also argued without resolution for years about what the distinctions between "code" and "input" are.
  • <samp></samp> is used for output, as in the second monospaced part of "Typing make love on a Unix command line produces the amusing response Don't know how to make love. Stop." It's also convenient to use it for filenames and extensions (i.e. partial output of a directory listing, so it's semantically correct markup), especially on a wiki that puts a background shade on <code></code>, which looks poor in running text, but does not do this to <samp></samp>: "There's one practical difference between .esp and .esm files ...."
The exact range of intents for these elements has never been entirely clear or agreed upon, from what I know (and I've been doing Web development since around 1993). As with most style matters, the useful approach is to set a house style, for well-thought-out reasons, and stick to it, because some aspects of this sort of thing are simply arbitrary, especially the exact definition of what is "input" or "output" versus "code", an interpretation matter that often varies widely by context (e.g. scripting versus programming, commandline versus in a file, etc., etc.).
— Darklocq  ¢ 11:30, 15 April 2017 (UTC)  [Updated. — Darklocq  ¢ 23:06, 15 April 2017 (UTC)]
I'd forgotten about the <samp> tag. It's not something we'd use often here, but now that I'm reminded of it, I'll probably find uses for it. And yes, to the other point in that paragraph, I've long been of two minds about our styling of <code> tags. On the one hand, the coloured background stands out, which makes it easy to spot and differentiate code samples visually; on the other hand, it sticks out. But hey, that's what User CSS is for, if people feel the need; being of two minds, I've never changed it for myself.
Touching briefly on some of your other points, we've also chosen an italicized style for dialogue, at least in the later games—I honestly don't remember if we applied that same styling consistently to earlier games. We should probably make that into a template with a CSS style in it at some point, though it would be a massive project to switch it all over. While my bot could potentially do a lot to help out, I can see where it might end up converting things it shouldn't, too, or get confused about where to start and end such a template. We'd probably be better off doing that kind of conversion by hand. Either that or let the bot convert it all and accept that it might generate a few undesirable conversions, not to mention undoubtedly missing a bunch due to incorrect formatting or abnormal situations that required different formatting. That would probably make it easier to see if there were enough remaining instances of '', and the rare <i> to warrant any additional handling or conversion to <em> (whether by direct code or a template). I suspect not, though. Oh and yes, the semantic markup is in Wikipedia's style manual, but only because you put it there and I suspect nobody really noticed or cared enough to debate the issue. Clearly, it's not seeing widespread use on Wikipedia, and your complaint on the Em talk page there that a bot was converting <em> to '' just goes to show that not everyone agrees with the idea. That said, I don't disagree that it's the direction HTML5 is going, but I think part of the point of wiki markup is that, even if it's applied against HTML5 recommendations, it isolates users from needing to worry about that sort of thing.
While I see your point about possibly using JavaScript add-ons (be they fully external or based on User JS), I'm not sure the wiki at large would agree. There are precious few here who get into scripting and CSS changes at all, much less scripting/CSS that would actually target <em> and ''/<i> separately.
Side-tracking, I also want to convert our use of <pre> for template examples to use a template and CSS at some point as well, for much the same reasons you mention for semantic markup, not to mention the possibility of just changing the style site-wide...<pre> is a pretty ugly format, after all. That would be a project in its own right, though. Robin Hood  (talk) 18:04, 15 April 2017 (UTC)
In no particular order (and sorry this is dense, but a lot of topics were raised): It just occurred to me that a potential special use, on a game site, for a template with <samp> and CSS to make it look different than the default monospacing, would be to use it for marking up game-provided text, including NPC dialogue and book material. That technically is output of a computer game, but is not something we'd want to treat as code in a monospaced font. So, a special-styled version of this element (to be italic for Book material, for NPC dialogue, but not quoted signs, or for everything quoted, or this way for Oblivion and that way for Morrowind, yadda yadda – whatever rules are desired, for whichever namespaces, in however many templates), plus the <q> element as a wrapper for inline quotes and <blockquote> for block ones, and if a |source= parameter is used, the {{tag|cite|open)} element for that. Actually, thinking this all through more, one would apply most of this style stuff to the <q> or <blockquote>, and force the wrapped <samp> to inherit the font-related stuff from that.

One annoying WHATWG quirk to remove with CSS is the default forced italicization of everything marked up by <cite>. I've reported this to WHATWG as a bug, but they've not fixed it yet, so browsers are still doing it by default. They're going by the ancient HTML3–4 definition of the element, which was only for titles (not all of which should be italicized anyway), while it can be used for all citation information in HTML5 as long as it contains title and/or author info in it somewhere.

A simpler variant of the above CSS-on-samp stuff could also be done with a template to mark up in-game typos so they don't get "fixed" by well-meaning editors or bots later, though really just about any template, even an empty one, that looks something like {{in-game spelling|Bound_Helm}} will do the trick. It's less intrusive than {{sic}}, which probably should only be used on first ocurrence.

Greasemonkey: Sorry, I didn't mean UESP should have anything to do with that or anything like it, only that the most common means today of users applying local stylesheets in the browser is with such add-ons, since many browsers don't have an easy (or even any) means of creating and editing local stylesheets, despite that always being an intended use of CSS since version 1. I didn't have anything in mind with regard to Mediawiki's user JS, either, since it only affects logged-in users who opt to install a script.

"Isolates users from needing to worry about that sort of thing" – I think that was already addressed by "use templates, which are wiki" and "style guides are for gnomes and bots, not everyday editors".  :-)

Ugly <pre> – That'll be at least partially a matter of the site-wide stylesheets in the Mediawiki namespace, which is also where to fix things like forced italics in <cite>. Agreed that templates are the best approach for blocks of code markup. WP has a Mediawiki plug-in for syntax highlighting, but I'm not sure it would be useful on this site since it's language-specific and won't have anything special for TES-related scripting.

I am seeing italicized quotation blocks in Morrowind-related articles (haven't checked for even older TES games, nor have I checked to see how consistent it is in MW articles). It seems to be the prescribed style for block quotes and pull quotes here. Unfortunately, I'm also seeing italicization of all kinds of things that are inside quotation marks, inline in running text, and I remove that when I encounter it. An inevitable consequence of an "italicize big quotation blocks" rule (which I would advocate against were I ever to get involved in UESP style policy matters again, because the style is weird, unhelpful, amateur-bloggish, and visually excessive) is that people who have not memorized every detail of the style guide (i.e., almost everyone) will assume "italicize anything in quotation marks" is the house style. This compounding "unintended style spread" effect happens to other things, too, and needs to be monitored. Here, I've noticed excessive capitalization of nouns and of generic game-related adjectives and adverbs, as a side effect of the actually very specific capitalization rules here, which people mostly don't absorb. Almost every article (on MW, at least) has multiple errors of this sort, because a drive-by editor doesn't intuit that "Golden Saint Soulgem" should be capitalized, as a specific item/resource, while "soulgem" should not, as a generic item-type reference. "Major Skill" is capitalized as something special, "skills" is not. A specific skill is named "Enchant" and a specific NPC class is the "Enchanter", but magical (not "Magickal") items are "enchanted" with "enchantments". And so on. Sometimes a style rule is worth maintaining despite the creeping spread problem (I would argue that it is for caps on a site like this, because the distinction is meaningful), and sometimes it's not (as with italicized quotes, which do nothing useful and impede the effectiveness of intentional emphasis; see the Style Guide's own statement about avoiding unnecessary emphasis and remember that it also does not distinguish, presently, between semantic emphasis and typographic italics, thus these rules are in direct conflict, and it is not the only such conflict). I'm not likely to make either argument outside this user talk page, though, unless summoned [not Summoned!] to do so.

The <code> styling: One approach would be to make the background color more subtle. I really wish Pickyweedia would do this, since so many other wikis take their style cues from it. Another would be to have an {{inline code}} template that removed the background with local CSS, for use in mid-sentence instead of code blocks.

"Only because you put it there" – Everything in every wiki was put there by some specific individual (given the nature of the work, often one who is still involved, and doing similar work at other projects). In this case, it naturally stands to reason that one of WP's main MoS editors put something in its MoS. And, yes, we did discuss it, and it was retained by consensus, because the wiki cares about the HTML specs, usability, accessibility (not a concern at UESP except for colorblindness issues, probably, but WP has a whole multi-page MoS section about accessibility and WCAG compliance), reuse, automated content processing, and user configurability. We started implementing all the semantic markup elements as templates way back in 2010–2011. That's six+ years of stability and incremental implementation by gnomes and bots. Yes, it is impossible to get random and often anon-IP editors to do it consistently, but this is true of every single facet of the MoS and WP's house style, and the same will hold on any publicly edited site with a style guide. Just the nature of the beast. The Template_talk:Em post was made shortly after creation of the template back in 2010, the bot was fixed, and there hasn't been a bots-and-emphasis problem since then. I'm not sure where you get the idea that it's evidence of dispute; it was a bug report that was quickly resolved, and is ancient history in WP time. Nothing added to en.WP's MoS goes unnoticed (it's highly watchlisted), and changes to it are quite difficult to get to stick, aside from clarity copyedits that don't substantively change any rule. Anything in it for years has general consensus to be in it, though virtually every line-item in it has its little camp of opponents. (It's another nature-of-the-beast issue, since almost every style point has divergent advice about it in different paper style guides, often with a nationalistic bent, and people are often irrationally convinced that one is "right" and all others "wrong". This prescriptivist and anti-linguistic viewpoint is reinforced since childhood by teachers who insisted on a simplistic set of "rules", often incomplete and decades behind actual practice, that students must follow exactly or be punished with poor grades.)

Now, I have a whole mess of daedra to go slaughter on an island off the coast of Tel Branora, and can't put it off any longer. >;-)
— Darklocq  ¢ 23:06, 15 April 2017 (UTC)

Capitalization cleanup[edit]

Refactored out as a sub-thread again. — Darklocq  ¢ 16:34, 15 August 2017 (UTC)

Just a couple of quick comments—I'll have to go through this in more detail later. First, I'm very familiar with changing the CSS styles in MediaWiki space. :Þ Second, I'm not surprised you're seeing a lot of miscapitalization in Morrowind's namespaces. The games are highly variable, but the older games, especially, often miscapitalized things so, naturally, we followed suit. I think it was only sometime in the Skyrim era that we decided to abandon that philosphy and we started switching things to lower-case as linguistically appropriate. The problem with a large wiki is that you can't readily retcon previous choices across thousands of pages, so by and large, the older spaces are still following the older philosophy. By all means, though, change them wherever you see things that need it. Robin Hood  (talk) 23:47, 15 April 2017 (UTC)

Some months-later followup: I've been normalizing the case conventions, at least in Morrowind-related articles, to the extent I can. In-game "things" get the caps per the style guide, but I've been removing them otherwise. There are a lot of overcapitalizations remaining. E.g., I keep running into things like "The Adamantium Helm of Tohan is ... This Helm has a higher enchantability than other Adamantium Helms ...", where the stand-alone word "helm" should not be capitalized. I also often encounter stuff like "He is equipped with an ebony shortword and a daedric cuirass" where those item names should be capitalized. So, I fix 'em. — Darklocq  ¢ 21:06, 12 August 2017 (UTC)
Great! Sometimes, it can be a bit of a challenge to decide if something is a title or not. Is Shack for Sale a shack for sale, for instance? Or is Reeking Cave a reeking cave, or do people actually call it "Reeking Cave"? I remember there being some debate on one of the Oblivion caves as to whether it was a proper name or not, and the simple fact is that there's sometimes really no way to tell. I tend to err on the side of proper titles, personally, but I'm not going to fix something where it's a debatable call and someone picked the opposite choice than I would have. Glad to hear that they're getting fixed, though. Morrowind space, especially, could use a bit of attention. Robin Hood  (talk) 18:58, 14 August 2017 (UTC)
Same here. I generally see what the game itself does. E.g., "A Carved Ebony Dart" is done that way because that's what the game calls it; in the Journal, it's referred to with something like "I found a strange dart ...", so we wouldn't capitalize "Strange Dart", but the inventory name of thing literally is "A Carved Ebony Dart", capital-A and all. In running prose, the capital-A looks like a typo when it appears mid-sentence, so I created a Tribunal:Carved Ebony Dart redirect; most users would not think to include the leading "A " when looking for it anyway. This site is unusually redirect-averse (to the point of user-unfriendliness), but I hope that one sticks. — Darklocq  ¢ 16:32, 15 August 2017 (UTC)

Good Work[edit]

Hi Darklocq! I don't have anything to add concerning your edits to the style guide, so I thought it would be better to recognize some of the other edits you have made to the wiki thus far. Although I have not patrolled many of your edits (due to lack of Morrowind knowledge), I have reviewed most of them and you are doing a great job revising and adding content. The older games have received little attention in recent years, which makes your contributions all the more valuable. Keep up the good work! I also highly recommend joining UESP's Discord server, which is where a lot of wiki discussion happens. —Dillonn241 (talk) 05:35, 15 April 2017 (UTC)

Thanks much.  :-) For many OpenMW users on Macs and Linux like me, Morrowind is a new game. Which somehow arrived with thousands of mods already written for it, hundreds of walkthroughs, dozens of forums and wikis, and multiple code patching projects. LOL. — Darklocq  ¢ 09:59, 15 April 2017 (UTC)


Hi Darklocq. Thought I should let you know thay Tribunal and Bloodmoon are termed "expansions", rather than add-ons. —Legoless (talk) 14:44, 16 May 2017 (UTC)

Hokey-dokey! I've been running into "mod", "plugin", "plug-in", and various other terms, as well as "add-on", and had just picked the latter kind of at random, as something to normalize to, but will use "expansion" henceforth. — Darklocq  ¢ 15:46, 16 May 2017 (UTC)
PS: It might make sense to refer to DLCs with new quests and such as "expansions" (Tribunal, Bloodmoon, Siege at Firemoth, Adamantium Helm of Tohan, Master Index, maybe Entertainers – I've never played that one), and the ones that just add items or effects as "add-ons" (Adamantium Armor, Area Effect Arrows, Bitter Coast Sounds, LeFemm Armor). It's not really typical gamer terminology to label an add-on as an "expansion" if it doesn't actually expand the game in a meaningful way. I've been using this terminology split in the material I'm writing at the OpenMW wiki. — Darklocq  ¢ 13:53, 25 July 2017 (UTC)
Morrowind only has two official "Expansion Packs": Tribunal and Bloodmoon (source). There is no distinction to be made between add-ons based on whether they add a quest or not, and they are certainly not "expansions" in the sense used by Bethesda, or most video games for that matter. What you contribute to the OpenMW wiki is your business, but I would not recommend making such an arbitrary and incorrect distinction. If you're looking for official terminology, you should look at the old official site; the term used by Bethesda for the eight free downloads is actually "plugins", an entirely different thing from the expansion packs. —Legoless (talk) 20:25, 25 July 2017 (UTC)
Wasn't talking about "correct" according to Bethesda's marketing. I don't think we want to call them "plugins", either. No one seems to call mods that, whether official or third-party. It's kinda early-2000s talk, which fits for the era of the game but not present-day reader understanding. LOL. Anyway, I'll stick with "expansions" if that's an actual official decision here. I've not seen any consistency on the matter in the actual articles. — Darklocq  ¢ 01:12, 26 July 2017 (UTC)
We decided on "add-ons" for consistency across all three modern entries in the series. I believe this is the relevant discussion, if you're interested. —Legoless (talk) 01:35, 26 July 2017 (UTC)
Now I'm confused. You started this thread by telling me I should be using "expansions", which is what I've been doing since then. — Darklocq  ¢ 22:09, 27 July 2017 (UTC)

() Legoless means to say use the word "add-on" for the eight minor DLCs listed on Morrowind:Official Add-Ons, and the term "expansion" for Tribunal and Bloodmoon. —Dillonn241 (talk) 06:09, 28 July 2017 (UTC)

Ah! Ok. — Darklocq  ¢ 03:24, 29 July 2017 (UTC)

Creeper owning Ghorak Manor?[edit]

I have observed that you a month ago added on the Creeper page that Creeper claimed the Ghorak Manor, but the orcs were squatters who had recently invaded the house. I could not remember having found this in the game myself, but as it is a while since I played Morrowind I put a couple of verification needed tag on those. No-one seem to be able to verify it, so that is why I would like to ask where you found it? —MortenOSlash (talk) 08:29, 20 May 2017 (UTC)

Saw it in there somewhere, but that was months ago. It's not in Creeper's greeting 1 variations, so I dunno at this point. Was it in one of the book resources? Probably safe to remove it. The claim that Creeper is the Orcs' pet should not be reinstated; there's no evidence for that, either. — Darklocq  ¢ 22:36, 21 May 2017 (UTC)
Ok. Done! —MortenOSlash (talk) 18:23, 22 May 2017 (UTC)
PS: I tracked it down; it was in an NPC dialogue mod. While I was not running the mod at the time, that bit had stuck in my memory. I'm now CS-checking any dialogue-related edits I make here. — Darklocq  ¢ 15:49, 1 July 2017 (UTC)


Hi Darklocq, I stumbled across your comments about acefile on OpenMW's User:Darklocq/Mod_testing_notes page. Feel free to open new issue tickets for anything that does not work for you or is missing. For instance, I am not sure what you mean by acefile "does not like complicated paths". Note that you can instruct acefile to extract into a base directory instead of CWD using -d; the default behaviour matches that of other Unix command line archiving utilities. Thanks! Daniel Roethlisberger — Unsigned comment by (talk) at 20:06 on 1 August 2017 (UTC)

I don't recall at this point what the error was, and can't duplicate it (four revisions later), so I'll remove the note about that (which is at a different wiki. ;-) — Darklocq  ¢ 10:19, 4 August 2017 (UTC)
For any "talk-page stalkers": I can confirm that acefile is now working with every .ACE file I've obtained from Morrowind Modding History/Great House Fliggerty and from NexusMods. — Darklocq  ¢ 20:07, 12 August 2017 (UTC)

Dark Brotherhood Assassins[edit]

I've just spotted another 4 assassins that may have eluded you so far, Dark Brotherhood Journeyman (db_assassin1a), Dark Brotherhood Apprentice (db_assassin1c), Dark Brotherhood Operator (db_assassin2a), Dark Brotherhood Punisher (db_assassin3a). They can all go on the one page with redirects. The choice of page would be open given the number of names (ie, you could use Dark Brotherhood Assassin), though TR:Assassin would at least need to be a disambig from its current assumed status of being for Morrowind's class. The LEVC list db_assassins contains a list of levels that each of those NPCs should appear. dbattackScript contains the levels that the number of Assassins changes, as well as the scripting for their appearance. PS The faction template should not be applied to bugged factions, as it incorrectly places them in the categories; plain linking is preferred. Silence is GoldenBreak the Silence 20:01, 15 August 2017 (UTC)

¡More assassins! At first I thought "am I losing my mind?", but now realize I didn't see them because I did a keyword search on "Assassin" in the text and got right to work when I found these guys, and didn't do a followup search on db_assassin in IDs. Good catch! Someone else added the Template:Faction around the faction ranks; I'd originally had it as just text. After seeing the change I had been meaning to ask if we really wanted to create new categories for DB faction ranks with nothing in them but redirects to this article [presently section]; I guess that answers that question. Anyway, I'll fire up OpenMW-CS and get the deets out. This actually highlights another reason to have a |notes= parameter for the NPC template; the level at which each appears could be added as a per-infobox note. It would be easier to parse the info on an assassinID-by-assassinID basis, rather than reading a bunch of prose, and then trying to relate it after the fact to the tabular data. — Darklocq  ¢ 14:21, 15 August 2017‎ (UTC)
Article done, at Tribunal:Dark Brotherhood Assassins. Redirected Tribunal:Assassins and Tribunal:Dark Brotherhood Assassin there. Not sure what to do about Tribunal:Assassin. Was thinking make it a disambig page. Not sure what to do about the categorization. Do we put that on the disambig page, or create a Tribunal:Assassin (class) redirect to hold the category link? PS: I note that we're wildly inconsistent with plural vs. singular for things like this, e.g. Morrowind:Guards and Morrowind:Followers versus Tribunal:High Ordinator and Bloodmoon:Reaver. I know this site doesn't like plural redirects as a general rule, but this inconsistency is a reason to use them when it comes to articles on groups of NPCs (as done with the Morrowind:Enchanters redir to Morrowind:Enchanter. I guess I can raise this at UESPWiki talk:Morrowind Overhaul Project. — Darklocq  ¢ 19:06, 1 September 2017 (UTC)
Because of its required use in the NPC template, a redirect will need to exist for the class. Class and faction redirects do not hold the same "weight" as other redirects, so it would not be beyond reasonable to create a redirect for the class and change the links, then hijack the Assassin redirect to point at the DBA page, and not create a disambiguation page, nor have a hatnote on the DBA page. This is further helped by the fact that the class is from the base game and not a new class introduced by Tribunal. In this case the "TR:Assassins" redirect would seem to be redundant. An easy difference in understanding what is (or has been) pluralised and what is not, is if a thing is in the game. For example, as every NPC gets a page, it is easy to use the High Ordinator page to accumulate knowledge on all High Ordinators without creating an additional page, whereas Followers is an artificial grouping where a page must be created anyway because there is no Follower page to use. In the case of a like Guards, because there are so many different names, there needed to be a single overarching grouping, so Guards was used despite the existence/use of the Guard page (for both named NPCs (who do not appear to be listed on the page) and the class (which has a redirect not noted on the Guards page, though the class is displayed there)). Changing/introducing a proper grouping structure is overdue, and though it will take time to fix this inconsistency site-wide, if there is acceptance for it it can begin, and if done properly and with care there should be zero problems. Silence is GoldenBreak the Silence 21:28, 1 September 2017 (UTC)
Okay, I'll proceed on that basis. "Dark Brotherhood Assassins" is also an artificial grouping ("Dark Brotherhood Assassin" is technically one NPC (db_assassin4a), and a faction rank, both covered at the same article. I guess redirs need to exist for the other named-but-generic DBAs, e.g. Dark Brotherhood Apprentice, etc., and properly categorized, with some categories being removed from the main DBA article. I ran out of time to do that earlier; had to go pick someone up from the airport. The second part of "create a redirect for the class and change the links" is going to be the tricky part; I can probably track down the extant usage easily enough with "What Links Here". Just hope none of the templates that use it are full-protected, or I'll have to ask for admin help to finish it up. — Darklocq  ¢ 00:00, 2 September 2017 (UTC)
Looks like this cannot be done until the character class is changed in mt_save_data (where ever that lives) to point to Tribunal:Assassin (class) instead of Tribunal:Assassin for all NPCs that need this done. That appears to be: Tribunal:Fedris Hler, Tribunal:Dravil Indrano, Tribunal:Ahnia, and Tribunal:Dandras Vules, and the redirects Tribunal:Dark Brotherhood Apprentice, Tribunal:Dark Brotherhood Journeyman, Tribunal:Dark Brotherhood Operator, Tribunal:Dark Brotherhood Punisher, and Tribunal:Dark Brotherhood Assassin, plus Tribunal:Assassin itself, about to become an NPC-name redirect when the database is updated. I would do this myself, but where mt_save_data is and how to update it don't seem to be documented anywhere. — Darklocq  ¢ 00:36, 2 September 2017 (UTC)
I'm not sure I followed the entire discussion, but I believe all you need to do to fix the Assassin links is change them to actual links, as I did here. I didn't do more than just the one, in case I'm misunderstanding what's needed.
I'm curious where you got the name mt_save_data from, though. That's the database table name, and shouldn't normally be visible anywhere (apart from the documentation). Robin Hood  (talk) 04:33, 2 September 2017 (UTC)
I did get it from the documentation (at UESPWiki:MetaTemplate#.23load, cross-referenced in the docs of Template:NPC Data). That template (unlike Template:NPC Summary) appears to be pulling the class directly from mt_save_data and auto-linking the class and other values, via {{#load:{{{pagename|}}}|titlename|race|class|gender|level|health|magicka|alarm|fight|faction|factionRank|loc|pageexists}} and subsequent code; it's receiving the class value as Assassin in a hard-coded way. I'll take a look at the page you just linked to and see if that elucidates. — Darklocq  ¢ 05:03, 2 September 2017 (UTC)

A light may be dawning: Is Template:NPC Summary actually auto-populating mt_save_data? If so, how does one set the class for NPC redirects, e.g. Tribunal:Dark Brotherhood Punisher? — Darklocq  ¢ 05:09, 2 September 2017 (UTC)

Maybe this is sufficient if the base page name of the redir and the name as given in the template agree? Now I wonder what happens in mt_save_data when two NPCs have the same name .... — Darklocq  ¢ 05:18, 2 September 2017 (UTC)

If there's a question here, the answer is that the category uses the displayed class. Robin showed the way we do it by making the class a manual link, rather than having the parameter in the template strip any "disambiguation" parts because some classes do use brackets. Silence is GoldenBreak the Silence 16:41, 2 September 2017 (UTC)
Yes, I kept answering my own questions with experimentation before getting a second-party answer. :-) — Darklocq  ¢ 21:17, 2 September 2017 (UTC)

NPC Data test[edit]

Dandras Vules Male Dunmer Assassin Dark Brotherhood Dark Brother(Dark Brother) 55 402 200 0 90
Dark Brotherhood Punisher Male Dunmer Assassin 30 Auto-calculated Auto-calculated 0 100
Assassin Male Dunmer Assassin 30 Auto-calculated Auto-calculated 0 100
Guard Male Dunmer Guard 20 210 122 100 30
Hlaalu Sharpshooter {{ }} [[Morrowind:|{{{race}}}]] [[Morrowind:|{{{class}}}]]
Ordinator Male Dunmer Guard Tribunal Temple Disciple(Disciple) 20 210 122 100 30

For now, the Tribunal:Assassin doesn't work at all because it still redirs to Morrowind:Assassin.

Others, like Tribunal:Dark Brotherhood Punisher are partially working, but the Faction (Rank) material isn't showing up, despite the Template:NPC Summary blocks in Tribunal:Dark Brotherhood Assassins being coded with this info. Note that this data doesn't show up in the Guard line from the Morrowind:Guards page, but it does show up for the Morrowind:Ordinator one which has its own article. This seems to be a bug in the templates. Some of the Guard NPC types also apparently don't have NPC-name redirects but should, e.g. Morrowind:Hlaalu Sharpshooter. From what I can gather, Template:NPC Data only works if a) Namespace:NPC_Name exists, and b) whatever it resolves to includes a Template:NPC_Summary with a matching NPC_Name in the |titlename parameter. It also won't correctly handle multiple NPCs with the same name (e.g. all Guard show up in Template:NPC_Data as Dunmer, when many are Imperials and at least one is a Nord; needs to be an array).

Also, I tried noincluding the * footnotes, and they are still showing up the NPC Data output anyway, so either that needs to change or we really do need a |notes parameter for Template:NPC_Summary, as I proposed last week.

Not sure whether to report the issues at Template talk:NPC Data or Template talk:NPC Summary; out of time for tonight, and will look into it tomorrow or so.
— Darklocq  ¢ 05:52, 2 September 2017 (UTC)

The NPC Data template isn't able to display multi-NPC data properly. The DB Punisher is not working just like the Guard page because neither has the {{Non-Relevant NPC}} template holding data and both are trying to pull from multiple templates on the generic page. The NPC Data template wasn't designed for displaying generic NPC data, only specific data from one page, so you would need to create both "Guard (Imperial)" and "Guard (Dunmer)" pages with the Non-Relevant NPC template on each, and then display them separately in order to make it work, though that wouldn't be desirable. Silence is GoldenBreak the Silence 16:41, 2 September 2017 (UTC)
I don't understand yet why its not desirable; it seems to be what Template:Non-Relevant_NPC was made for (i.e. get the data inserted so Template:NPC_Data can use it, but only have the templates on redir pages. It's seeming to me that the redir commands could be noincluded, and that the actual templates could be inserted visibly into pages like Tribunal:Dark Brotherhood Assassins by transcluding them from the redirs. We're already using extensive transclusion to insert specific-something details in to more-general pages; this would just be another variant. — Darklocq  ¢ 21:09, 2 September 2017 (UTC)
They aren't desirable because they don't need to exist. Creating a page purely to transclude the information to another makes little or no sense whatever way you look at it. Redirects would need to be made for every variant, which makes a whole lot of clutter for no benefit whatsoever. For example, you would need to create 5 redirects for the Assassin due to the 5 different levels and faction levels. Those potential redirects aren't and won't be needed on any pages for linkage purposes either. The Non-Relevant template was designed to be used on uniquely named individual redirects, not generic multiple-entity redirects. Writing them out manually is much simpler, as well as allowing the opportunity to simply state "Assassins" multiple exist in a location, rather than label every generic variant that may or may not even be there. As below, a Generic NPCs template variant would allow the transclusion of any shared data (eg race and gender), simply filling in the missing parts manually. Silence is GoldenBreak the Silence 22:14, 2 September 2017 (UTC)
Got it (finally!) — Darklocq  ¢ 22:18, 2 September 2017 (UTC)

PS: I dig the approach taken at Skyrim:Bandit. Getting it template-compatible might be some work though; I'm guessing multiple values (race, gender, ID) would need to be put into arrays, or at least output templates like {{NPC Data}} changed to not auto-link anything in that format and just accept a literal value, which might be multi-line. — Darklocq  ¢ 22:22, 2 September 2017 (UTC)

Eventual content for Tribunal:Assassin[edit]

This will be the content of the eventual generic-NPC-name redir to Tribunal:Dark Brotherhood Assassins:

#REDIRECT [[Tribunal:Dark Brotherhood Assassins]]
[[Category:Tribunal-Level 1 NPCs]]
[[Category:Tribunal-Level 3 NPCs]]
[[Category:Tribunal-Level 10 NPCs]]
[[Category:Tribunal-Level 20 NPCs]]
[[Category:Tribunal-Level 30 NPCs]]
[[Category:Tribunal-Male NPCs]]
[[Category:Tribunal-Factions-Dark Brotherhood]]

All these categories except the last one can be removed from Tribunal:Dark Brotherhood Assassins as the final step. — Darklocq  ¢ 00:45, 2 September 2017 (UTC)

PS: All these generic-NPC-name redirs should also have Category:Tribunal-Leveled NPCs, if we want that category to exist. It will not contain anything but these NPC redirs. — Darklocq  ¢ 00:48, 2 September 2017 (UTC)

I was thinking we could make a version of the {{Non-Relevant NPC}} template for generic NPCs. This would be useful in all namespaces. Any of the more recent generic NPC pages and their redirected NPC pages can serve as examples for what parameters are needed elsewhere (eg SR:Bandit, SR:Vampire). Silence is GoldenBreak the Silence 16:41, 2 September 2017 (UTC)
That sounds like a good approach. — Darklocq  ¢ 13:15, 4 September 2017 (UTC)
The redir conversion and categorization above is done, aside from the Category:Tribunal-Leveled NPCs matter. — Darklocq  ¢ 22:26, 2 September 2017 (UTC)

Update articles[edit]

Also need to update articles to point to the new Tribunal:Dark Brotherhood Assassins article where appropriate. Starting list to check:

— Darklocq  ¢ 01:05, 2 September 2017 (UTC)

Update: Still need to double-check this, and also make sure it's consistent with DB lore in Oblivion, etc. I already fixed some issues in that regard. — Darklocq  ¢ 17:46, 18 May 2018 (UTC)

Bug Template Spacing Issues[edit]

The problem you were running into in the Bug template with needing to use &#32; instead of a space is caused by being inside <cleanspace> tags. That's a custom tag added by one of our extensions. The idea behind it is to allow you to format #locals, #defines, and other commands that don't emit any text. For example:

</cleanspace>Text goes here.

...instead of...

{{#local:var1|abc}}{{#local:var2|def}}Text goes here.


-->{{#local:var2|def}}Text goes here.

Obviously, with only two commands in the example, it's not quite as obvious what the benefit is, but when it gets to be dozens, you start really appreciating it. Indenting is also possible.

The customary solutions to the issue you ran into are the one you found, temporarily closing the <cleanspace>, and I believe either <nowiki/> or the paired tags will work as well, but I'm not 100% sure of that last. Robin Hood  (talk) 18:08, 2 March 2018 (UTC)

Ah so! Aside from minor "gotchas" like that, I really, really like the additional functionality that stuff provides, though I'm still examining how it all works. I badly wish Wikipedia itself had these extensions, though I'm not sure which ones they are. The ability to define variables, in particular, is priceless. — Darklocq  ¢ 18:17, 2 March 2018 (UTC)
There are two custom extensions we have here that were both designed by a now-retired bureaucrat: UespCustomCode and MetaTemplate. I know I've seen other extensions that allow variables, though I don't know if any are in use on any Wikimedia wikis at this point. Robin Hood  (talk) 21:48, 2 March 2018 (UTC)
Too bad that didn't get pushed back up into the MediaWiki pipe. All wikis could benefit from that stuff. — Darklocq  ¢ 22:57, 2 March 2018 (UTC)

Vertical Alignment[edit]

Just a quick tip: for the table where you have vertical-align:top all over the place, try using class=vtop at the table level. Then you can remove all the alignment statements. Robin Hood  (talk) 07:02, 31 May 2018 (UTC)

Which article? I did this per-cell thing recently in a user page, but I've probably done it in an article somewhere as well. — Darklocq  ¢ 07:30, 31 May 2018 (UTC)
Sorry, I meant to say on your user page—the "Pages and Major Sections Overhauled" section. Robin Hood  (talk) 07:37, 31 May 2018 (UTC)
Coolio, and thanks for the tip. Just implemented that. I may need to do a code search; I'm pretty sure I've done it the long-winded way in one or more articles. — Darklocq  ¢ 07:51, 31 May 2018 (UTC)
Jeebus, there's the long version in exactly 1.37 gazillion pages. Also, the class only appears to work at the {| level. It failed at the |- and |- levels, though the longer vertical-align:top syntax worked at the |- level, at least. So the replacement will not work in all circumstances, e.g. where particular rows or cells need particular treatment. — Darklocq  ¢ 01:07, 28 July 2018 (UTC)
Yeah, it's currently tagged as being a table-level-only class in MediaWiki:Common.css. That's easy enough to change if desired, though. Robin Hood  (talk) 04:53, 28 July 2018 (UTC)
Should probably apply to any table content element like td and th and col and, wrappers for several like tr, colgroup, thead, etc. Many of these use middle alignment by default. MIght even be non-table elements for which it would be useful; not sure. — Darklocq  ¢ 10:58, 28 July 2018 (UTC)
Last I heard, col/colgroup isn't supported in any way on MediaWiki, though I don't really keep up to date on stuff like that. I'll add it for the other table elements later today. Robin Hood  (talk) 21:00, 28 July 2018 (UTC)
And least as far as my CSS knowledge extends. :) I just added .vtop and .vmid options, so they can be applied to anything that supports vertical alignment. Robin Hood  (talk) 03:24, 29 July 2018 (UTC)
Schweet. — Darklocq  ¢ 22:05, 12 August 2018 (UTC)


Hi. In case it's fallen off your watch list, I got interested in Tribunal's Dark Brotherhood attack script today, found your post, and replied to it. Tribunal_talk:Dark_Brotherhood_Attacks#.2F.2A_Max_number_of_attacks_.2A.2F Shmoot (talk) 00:42, 31 October 2018 (UTC)

Great work! PS: I wouldn't consider it necroposting at all; it's an open issue from just last year. If it was some "I think I found a bug, cuz a rock moves when you go into this dungeon while wielding a two-handed sword" post from 2012 and no one able to confirm it since then, then probably too old to bother responding to. — Darklocq  ¢ 18:39, 31 October 2018 (UTC)