This article was the "article of the day" today (deservingly). One gets a different perspective on the article reading it in unlinked plain text. In particular, this sentence stands out:
When these are linked, the sentence makes sense; when not, it looks confusing and even ungrammatical.
I would propose a rule: An article should be written so that it makes sense whether or not the links are there. A plain text, unlinked version of the article should be entirely helpful and nonpuzzling. Therefore, the expected following of links should not do the work of text; links should be for further information, not for getting essential information to the topic at hand.
What do you think? --LMS
Hmm. I am inclined to agree, though this may be harder in some cases than it first appears, and it may also make the text a bit more awkward than it might be otherwise for those who are reading it in original form. We have, after all, chosen the medium for our project, and a process that facilitates creation for that medium. The fact that the content may not work well in other media should not surprize us, nor should it bother us much. The hypertext concept isn't going away, so our information--and links are information too--will always be useful. Text such as the above can be reworked without much difficulty, but is it really worth the effort? What about articles with animations, or interactive content, or good use of fonts sizes and colors, or other aspects of this medium that can greatly increase understanding? Should we sacrifice those things to make articles mailable? Or should we add fallbacks at the expense of extra effort and possibly making the original uglier? I don't think it's necessarily a slam dunk, but should be left to the judgment of authors. Certainly articles that are primarily text with little need of extras should be written to fall back cleanly--articles on philosophy, history, and so on. But articles on math really need complex symbols to be understandable, and should use them. If that makes them non-mailable, sobeit. Likewise, my "hands" subpage here really needs the suit symbols for clarity. That makes them non-mailable too. I've tried to avoid those symbols on pages that didn't need them, but those that do really do. --LDC
I'm now totally convinced that we shouldn't aim at making all pages mailable. I think I did have a point, but I'm not sure best how to state it. --LMS
It might make sense for "accessibility" (in many forms, including the ability to survive email) to be point of quality, just not necessarily a reqired one. Math is the real toughie--there's just no way to do it without the symbols, and so we need to have specialized methods. But the change you suggest to the one sentence on the poker page does make sense, so I reworded it. It's now a better article, in that it expresses exactly what it did before just as effectively, and now it is also less confusing when read as email or on Lynx or a speech browser, etc. It's the same issue with things like illustrations: if you can make the article make sense without them with little work, why not? But if you can't, don't lose any sleep over it. --LDC