User talk:Lee Carre

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

Hi, welcome to the wiki. We’re always glad to have experienced wikipedians start helping out. But there are a few things to watch out for, where we do things a bit differently than Wikipedia (Common Mistakes). Feel free to ask if you have any questions! —Nephele 13:56, 26 September 2006 (EDT)

Hi, thanks for the welcome and the links to some basic info, but I suspect that I’ll just end up being a lurker, rather than a contributor. Thanks all the same ☺ — Lee Carré 18:09, 26 September 2006 (EDT)


You asked: Beginner’s GuideTheRealLurlock Talk 10:17, 22 November 2006 (EST)

Thanks ☺ — Lee Carré 11:04, 22 November 2006 (EST)

Discussion: Markup Syntax for Talk-page Replies[edit]

From Template talk:ModInfo

Just to explain a bit better about the change from <p> to ::, the colons are the standard method of indenting on MediaWiki. The HTML definitions are just MediaWiki's chosen implementation of indents, as opposed to any kind of actual definition. They could well change the implementation in the future to use something more logical from an HTML perspective, but colons are still the preferred way of indenting text on a MediaWiki-based wiki. Robin Hood  (talk) 18:30, 5 May 2017 (UTC)

Please stop editing my comments; surely that has to be against some kind of sane rule, since they're rather different in nature to articles. Discussion, first, would be prudent+civilised, rather than attempting to impose one's preference regardless of merit. To that end:

I'm aware of the de-facto pseudo-‘standard’, but according to MediaWiki's syntax, starting lines with semi-colons and colons is for marking definition terms+descriptions. Thus, the semantics are wrong. The output is what matters (principal+correctness, accessibility, usability, non-human parsing, non-visual rendering, and whatever else others may see fit which I can't think of); internal syntax is an implementation detail. Elements should be used based on their semantic meaning, rather than how they might happen to be rendered by default in browsers (especially that not all browsers are graphical). Likewise, that default rendering may also change in future, since it, too, is an abstraction.

Even if wiki-syntax, itself, were perfect; if the resulting output is tag-soup garbage, then it's utterly irrelevant. Editing is a secondary function of the site; reading is the primary function+purpose, thus deserves more weight in considerations. There are surely vastly more readers (and reads) than editors; especially given this context (very much meta about a template).

I suspect that this habit came about via someone's personal hack, for convenience, and was simply adopted by others. This doesn't make it good, or any kind of meaningful standard.

As for wiki-conventions; one could make the same argument about other aspects of the UESP wiki, as compared to the wiki that is Wikipedia (and, somewhat, other WikiMedia projects). Namespaces are a prime example which come to mind. But, I've no objection because the reason(ing) for the difference makes a whole lot of (self-evident) sense. So, are we being selective about what's applied cross-wiki and what's not? If so, seems to be based on subjective preference.

Wiki-syntax is rather more ephemeral, and subject to change, than is HTML. One is an internal editing syntax, which is relatively young+immature, while the other is quite the opposite. Thus, my use of HTML was a more stable long-term option, in that regard. In fact, wiki-code typically focuses on presentation rather than semantics; for example, having no distinction between emphasis and a citation; it's all italics (which is merely the typical way of representing both of those (quite distinct) semantic cases). HTML doesn't have this problem. Thus, wiki-code is rather more a ‘presentation’ language, than is HTML.

An actual standard would be specific+dedicated wiki-markup for a talk-page reply; indentation is merely one possibly way to represent those semantics, but far from the only way. If there is such wiki-markup, then please do cite it; I'll gladly use it if it yields something sane. Using syntax which yields <dd> most certainly doesn't qualify, since I'm not specifying a definition. Relevant to this very reply; MediaWiki syntax lacks its own way to mark a block of text being quoted in order to reply to it. Seems that talk pages were not thought about, when designing the editing syntax. This would seem a bug (or several) with MediaWiki, for which there's no obviously-good work-around (which is acceptable to all).

What was the actual harm of indented-paragraph markup? It cleanly achieved the expected formatting while avoiding the harm(s) of a mislabelled definition-description. The first reversion seemed entirely pointless.

If following of (some) rules is beyond question (seriously? why?), then bad rules must be changed swiftly in order to minimise further harm. I'd welcome such an effort. Else, it's simply a matter of which rules are to be obeyed, and which ignored. Many more rules agree with my approach, than that of the arbitrary convention of wikis.

I'm quite willing to discuss actual merits and reasoning, sans authoritarianism. Questions welcome. Lee Carré (talk) 20:13, 5 May 2017 (UTC)

We're not trying to be authoritarian in any way. Formatting comments for things like indentation is not only allowed, Wikipedia considers users who don't follow the conventions as being mildly disruptive. As you point out, it is the de-facto standard. Perhaps more to the point, if the underlying code for colons is changed in some future version, everyone else's comments will be compliant with the new code, but yours won't. Even if I leave it alone myself, someone else is likely to come along in the future and format it correctly. Robin Hood  (talk) 21:38, 5 May 2017 (UTC)
Rule-following, regardless of merit else reasoning, seems to be the priority of several administrators (else editors-with-privileges). I accept that it may well not be intentional, but that's what appears to be happening, regardless. Reversions first, and then not even questions, only statements about arbitrary rules. Not to mention all the subjective bias, especially over what's positive and what's negative.
None of my points have been addressed, let alone adequately responded to. Certainly none of the questions.
The change you describe is merely some future hope; no evidence that it is, or will be, materialising. Especially since UESP is quite a few revisions of MediaWiki (v1·23·11) behind that used by WikiMedia projects (v1·29·0) (though I quite grasp why this might be, from a technical perspective); thus, even if such a patch were committed to MediaWiki's development source, that doesn't mean it'll be deployed here any time soon, or at all. Besides, and moreover, I very much doubt that any such change would be made, at all, since the relevant syntax is for specifying definitions, and thus would likely break existing+needed functionality. This is why ambiguous dual-use is inherently a bad idea, and separate syntax should've been specified in order that each may be changed orthogonally.
Even then; what would it be changed to? Something quite similar to the HTML I'd used, if the change were minor; else I expect that the changes would significantly redesign the way that talk replies are handled.
In any case, that doesn't address the matter+situation at hand. The point seems to be missed. I see no evidence that the problem is being reported, or worked on; the focus seems to be entirely on the resulting symptoms of it remaining unaddressed in MediaWiki's source-code. Is there even an entry in MediaWiki bug DB for this? Why isn't a wiki-admin on this (case|problem)? Particularly when several people clearly feel very strongly about following the conventions which lead to the problem results.
A de-facto isn't a standard, at all; it's simply an euphemism for what's commonly done, usually for historical consistency else tradition, and often including something non-good else it wouldn't be a de-facto ‘standard’. As I said; a standard would be something designed+fit for purpose, which didn't yield tag-soup. For example, the 4 tildes appended to talk-page comments, for generating one's signature. That syntax doesn't seem to be used for anything else, and thus there's no conflict. That isn't the case, here.
You state that formatting for indentation is allowed; that's precisely what I did originally, by making it explicit in a valid way. Replacing my HTML with syntax used for definitions, removed this; none is declared in the relevant CSS, it relies entirely on the assumed defaults of graphical user-agents. This is backwards. Thus, my original syntax is rather better adhering to the stated guidelines (or whatever one wishes to call them) than the replacement pseudo-markup. As said; elements aren't to be selected for their default rendering appearance (particularly since that may change), but their semantics. I wasn't adding a definition …
Re; attempting to improve on old habits as being mildly disruptive. To whom? Other editors? What about readers? Again, surely they matter infinitely more, in several ways. How's anything supposed to improve if nothing can be changed? Seems another iteration of ‘rules is rules’ sans substantive argument.
However, benefit of the doubt; talk pages are more-so for editors than those who only read articles like they would any other Web-site. That'd be a fair point. Yet, this discourages a large fraction of potential-editors from becoming so, due to precisely the accessibility+usability problems I've identified+described. Is editing only for people who need no adaptive-technologies, and who use graphical user-agents? Seems somewhat elitist or ignorant. Even worse, from the brief examination of the output I did at the time, it's unlikely that said output would pass an HTML validation, since (as I recall) only the dd element was added, sans the other elements of a definition-list.
Is it not obvious how the points I'm raising are at odds with what would seem to be the entire mission of this site+wiki? Again; even if the internal situation were perfect; that's utterly irrelevant if the output is nonsense.
Perhaps back to basics; why is the convention for talk-page replies to use the syntax for definitions? Why is there not separate syntax for each (unrelated) use-case? Lee Carré (talk) 22:09, 6 May 2017 (UTC)
I'll make no bones about it, much of what you have typed here makes no sense to me. However, you seem to cite reader friendliness as something anti-mediawiki. How do you come to this conclusion? Mediawiki is user-friendly precisely because it simplifies the codes needed to make webpages. Keeping things to a de-facto standard is more user-friendly than changing our ways every few years, as pretty soon the oldest stuff becomes completely outdated, unreadable, and incompatible with the rest of the wiki. These standards mean that a reader has no compatibility or visibility problems moving from our newest page to our oldest page. Templates are the answer to changing standards, as with one edit (as if its ever that simple) you update every page using it, rather than requiring a bot to run through thousands of pages updating the now useless and outdated html codes that still exist on some of our older pages. The standard procedure for talk page indentation is much more logical for reader friendliness than some random html code that one user has, but another doesn't, and only encourages deviation away from the standard, which in a strawman example eventually leads to the page being unreadable. The rules on talk pages are lax for the precise reason that they are actually more used by readers than editors, not the other way around. However, the rules that exist are strict, with limitations on things like indentation and signatures (such as no vertical signatures, images, or being a template) to keep a page readable. If wikipedia has seen no reason to change its indentation coding after all these years with its much larger data sample to spot reasons for it, why should we seek to be different and therefore harder to read.
My last point is that you get absolutely nowhere coming to a wiki and lambasting all its users for being old-fashioned and new-phobic. When you come to a new place you learn their ways and then show them a new way of doing things if it is better. Trying to force new habits on people is as useless on the internet as it is in real-life. You must be patient and also accept that some things won't change, but some things can and will. Silence is GoldenBreak the Silence 23:01, 6 May 2017 (UTC)
Continuing on some of what Silencer said, if you dislike the standard, then take it up with MediaWiki and their people, not us. You can go on at length about the problems with using <dd> tags here, and it will do you no good whatsoever, since you're not addressing anyone in a position to make changes.
As to why you should be following the standard, regardless of whether it's a good standard, here's one very simple reason. When people open a page, especially newbies, they will learn what the standards are as long as a consistent standard is followed. When you break the standard, people get confused. As a bot-programmer, I can also tell you that bots get confused. While I haven't created such a program yet, it's not inconceivable that I will one day write a program to fill in missing signatures. Among other things, the bot would probably look at the number of colons to distinguish one reply from the next; it's certainly not going to pay any attention to raw HTML.
Lastly, even if we accepted your arguments, you're still being inconsistent with established formatting standards from a CSS perspective. You chose an indent level of 2% (and straight multiples thereof). For better or worse (mostly the latter, in my mind), MediaWiki chose a fixed indentation size of 1.6em (with multiples coming only via nested indentation, I believe).
And yes, as administrators and patrollers, it's our job to enforce the rules, so it's not surprising that we actually do that. Truthfully, I really don't care whether the HTML MediaWiki generates is good, bad, or ugly, at least not as long as it works. If I did, I'd be designing web pages, not writing wiki content. When in Rome on a wiki, do as the Wikipedians do. Robin Hood  (talk) 00:29, 7 May 2017 (UTC)

Continuation of Edit Filter discussion[edit]

Just to respond to your post about Bayesian filtering (which I've done here, in respect to Silencer's suggestion that the post is getting off-topic), I think that's probably a bit advanced for this size of wiki, and I'm not sure if any of us has the skill to program something like that. If anyone does, it would be the site owner, Daveh. There are several other programmers on the site of varying skill levels, but only Dave and I have any experience with PHP programming for MediaWiki, and mine is rudimentary at best. Robin Hood  (talk) 01:31, 9 May 2017 (UTC)