I have started a conversation in [[Wikipedia talk:WikiProject U.S. Counties]] about the fact that the templates for country, state, county, and city all use the wrong markup, that is, sections headed with === instead of ==. I think this is important, but god knows there are going to be a lot of erroneous articles. I wish I'd noticed before there were so many counties done by the redoubtable Ram-Man.
Tom Parmenter Ortolan88
Tom Parmenter wrote:
I have started a conversation in [[Wikipedia talk:WikiProject U.S. Counties]] about the fact that the templates for country, state, county, and city all use the wrong markup, that is, sections headed with === instead of ==. I think this is important, but god knows there are going to be a lot of erroneous articles. I wish I'd noticed before there were so many counties done by the redoubtable Ram-Man.
Level-three headers (===) are well established as the standard header on Wikipedia. If you hate this so much, better to simply redefine it to pump out ideologically correct <H2> tags instead of <H3> rather than to prescribe the change of thousands of pages and demand a change in markup behavior.
-- brion vibber (brion @ pobox.com)
|From: Brion VIBBER brion@pobox.com |Date: Wed, 25 Sep 2002 13:35:41 -0800 | |Tom Parmenter wrote: |> I have started a conversation in [[Wikipedia talk:WikiProject |> U.S. Counties]] about the fact that the templates for country, state, |> county, and city all use the wrong markup, that is, sections headed |> with === instead of ==. I think this is important, but god knows |> there are going to be a lot of erroneous articles. I wish I'd noticed |> before there were so many counties done by the redoubtable Ram-Man. | |Level-three headers (===) are well established as the standard header on |Wikipedia. If you hate this so much, better to simply redefine it to |pump out ideologically correct <H2> tags instead of <H3> rather than to |prescribe the change of thousands of pages and demand a change in markup |behavior. | |-- brion vibber (brion @ pobox.com) |
It's not that I *hate* them. I don't. It's that you aren't supposed to start in the middle, with H3, but at the beginning with H1, title of article, H2, second level header, etc. That isn't "ideologically correct", that is standard markup, SGML, HTML, XML, SDML, you name it.
I wonder why I'm getting all this attitude from you when I am trying to deal with this in a polite manner, carefully explaining it on the talk page, announcing the issue here on the list.
If you read what I wrote on the talk page, you'd see I understand the issue, both sides, quite well.
The reason H3 has become the "standard" is that H2 is so godawful big that everyone is intimidated out of using it. There's nothing standard about bad markup.
Can the rhetoric and discuss the issue. Does it matter if article headers start in the middle? If numbering doesn't work right? If any number of existing or imaginable passes over Wikipedia articles don't work right? And, if not, then why do we have generic markup at all?
Tom Parmenter Ortolan88
Tom Parmenter wrote:
|From: Brion VIBBER brion@pobox.com |Level-three headers (===) are well established as the standard header on |Wikipedia. If you hate this so much, better to simply redefine it to |pump out ideologically correct <H2> tags instead of <H3> rather than to |prescribe the change of thousands of pages and demand a change in markup |behavior.
It's not that I *hate* them. I don't. It's that you aren't supposed to start in the middle, with H3, but at the beginning with H1, title of article, H2, second level header, etc. That isn't "ideologically correct", that is standard markup, SGML, HTML, XML, SDML, you name it.
The HTML standard neither prescribes nor proscribes such strict hierarchical use of the header tags, though it does note that "some people consider skipping heading levels to be bad practice". The levels are certainly *not* defined as hierarchical, but simply as a range "with H1 as the most important and H6 as the least".
See: http://www.w3.org/TR/1998/REC-html40-19980424/struct/global.html#h-7.5.5
So there's no justification for the level-two crusade based on the transformation of wiki to HTML markup. If you want to enforce this personal preference as part of wiki markup, then speak up and explicitly say so. The parser can be adjusted accordingly.
If numbering doesn't work right?
Automatic section numbering is dubious at best.
If any number of existing or imaginable passes over Wikipedia articles don't work right?
Passes of what?
And, if not, then why do we have generic markup at all?
Such as?
-- brion vibber (brion @ pobox.com)
--- Brion VIBBER brion@pobox.com wrote:
Level-three headers (===) are well established as the standard header on Wikipedia. If you hate this so much, better to simply redefine it to pump out ideologically correct <H2> tags instead of
<H3> rather than to prescribe the change of thousands of pages and demand a change in markup behavior.
I have to disagree. Level-three headers were certainly more common earlier on, but I see more level-two headers these days, and I've changed many from level three to level two myself. Redefining === would not help.
Stephen
__________________________________________________ Do you Yahoo!? New DSL Internet Access from SBC & Yahoo! http://sbc.yahoo.com
--- Stephen Gilbert canuck_in_korea2002@yahoo.com wrote:
I have to disagree. Level-three headers were certainly more common earlier on, but I see more level-two headers these days, and I've changed many from level three to level two myself. Redefining === would not help.
I just saw your database querry, which shows level 3 as more common. So don't flame me. :)
Stephen
__________________________________________________ Do you Yahoo!? New DSL Internet Access from SBC & Yahoo! http://sbc.yahoo.com
Tom Parmenter wrote:
I have started a conversation in [[Wikipedia talk:WikiProject U.S. Counties]] about the fact that the templates for country, state, county, and city all use the wrong markup, that is, sections headed with === instead of ==. I think this is important, but god knows there are going to be a lot of erroneous articles. I wish I'd noticed before there were so many counties done by the redoubtable Ram-Man.
I think that having good HTML produced is moderately important. And I think that making editing easy and intuitive is very important. What is not important, however, is that the number of equal signs in the latter match up precisely with the number in the tag in the former.
When rendering a page, we should first measure all of the header markups and then render the shortest as <h2>, the next as <h3>, and so on. (Anybody that really needs a header of a specific size can still create this by putting the HTML tag in directly.) Then you can start with == or ===, or even = or =========, and it will still render as <h2> if it's the shortest one.
Best of all, the code to do this detection already exists in PediaWiki; it's being used to decide what style of automatic header numbering to use. (Not that I've ever looked at the code), but this should be an easy one.
Tom is happy, because good HTML is being produced. [I forget who the principal opponent is] is happy, because current articles don't have to be rewritten. At least I hope that y'all're happy; responses?
-- Toby
I'm happy too. Everyone wins! Can we do this easily?
Stephen G.
--- Toby Bartels toby+wikipedia@math.ucr.edu wrote:
Tom Parmenter wrote:
I have started a conversation in [[Wikipedia
talk:WikiProject
U.S. Counties]] about the fact that the templates
for country, state,
county, and city all use the wrong markup, that is,
sections headed
with === instead of ==. I think this is important,
but god knows
there are going to be a lot of erroneous articles.
I wish I'd noticed
before there were so many counties done by the
redoubtable Ram-Man.
I think that having good HTML produced is moderately important. And I think that making editing easy and intuitive is very important. What is not important, however, is that the number of equal signs in the latter match up precisely with the number in the tag in the former.
When rendering a page, we should first measure all of the header markups and then render the shortest as <h2>, the next as
<h3>, and so on. (Anybody that really needs a header of a specific size can still create this by putting the HTML tag in directly.) Then you can start with == or ===, or even = or =========, and it will still render as <h2> if it's the shortest one.
Best of all, the code to do this detection already exists in PediaWiki; it's being used to decide what style of automatic header numbering to use. (Not that I've ever looked at the code), but this should be an easy one.
Tom is happy, because good HTML is being produced. [I forget who the principal opponent is] is happy, because current articles don't have to be rewritten. At least I hope that y'all're happy; responses?
-- Toby [Wikipedia-l] To manage your subscription to this list, please go here: http://www.nupedia.com/mailman/listinfo/wikipedia-l
__________________________________________________ Do you Yahoo!? New DSL Internet Access from SBC & Yahoo! http://sbc.yahoo.com
Toby Bartels wrote:
I think that having good HTML produced is moderately important. And I think that making editing easy and intuitive is very important. What is not important, however, is that the number of equal signs in the latter match up precisely with the number in the tag in the former.
When rendering a page, we should first measure all of the header markups and then render the shortest as <h2>, the next as <h3>, and so on. (Anybody that really needs a header of a specific size can still create this by putting the HTML tag in directly.) Then you can start with == or ===, or even = or =========, and it will still render as <h2> if it's the shortest one.
That is certainly a very clever solution. However, I'm concerned that users may be confused by 3= producing different results in different circumstances.
There are thousands of pages with spelling mistakes, thousands of "it's" that we can't even do a search for, and thousands of pages that use bolded text instead of headings -- these all need to be changed, and WikiWeeders like myself are gradually working through them. I don't think that adding the changing of top-level 3= to the list is a big deal. (there are also thousands of page titles to be bolded, names of novels and films italicized, links to be fixed, <BR> to be replaced with * & so on)
I would also like to add my vote to the "please change the stylesheet to H2 is smaller" issue. :) -- I agree with whoever said that the current H2 font size is the reason people don't use 2=
|From: Toby Bartels toby+wikipedia@math.ucr.edu |Content-Disposition: inline |Sender: wikipedia-l-admin@nupedia.com |Reply-To: wikipedia-l@nupedia.com |Date: Wed, 25 Sep 2002 21:36:09 -0700 | |Tom Parmenter wrote: | |>I have started a conversation in [[Wikipedia talk:WikiProject |>U.S. Counties]] about the fact that the templates for country, state, |>county, and city all use the wrong markup, that is, sections headed |>with === instead of ==. I think this is important, but god knows |>there are going to be a lot of erroneous articles. I wish I'd noticed |>before there were so many counties done by the redoubtable Ram-Man. | |I think that having good HTML produced is moderately important. |And I think that making editing easy and intuitive is very important. |What is not important, however, is that the number of equal signs |in the latter match up precisely with the number in the tag in the former. | |When rendering a page, we should first measure all of the header markups |and then render the shortest as <h2>, the next as <h3>, and so on. |(Anybody that really needs a header of a specific size |can still create this by putting the HTML tag in directly.) |Then you can start with == or ===, or even = or =========, |and it will still render as <h2> if it's the shortest one. | |Best of all, the code to do this detection already exists in PediaWiki; |it's being used to decide what style of automatic header numbering to use. |(Not that I've ever looked at the code), but this should be an easy one. | |Tom is happy, because good HTML is being produced. |[I forget who the principal opponent is] is happy, |because current articles don't have to be rewritten. |At least I hope that y'all're happy; responses? |
Tom is unhappy because Tom doesn't care about good HTML; Tom wants good generic markup in the Wikipedia that will permit various processors, HTML and otherwise, including future processors that we know nothing about, to produce output that is organized and presented as it was originally designed to be presented in the Wikipedia.
Spaghetti code is spaghetti code, even when it's text.
Put garbage in, and garbage comes out, even when it's text.
Each article in the Wikipedia is an entry in a database. Entries in databases ought to be organized so that they will be interpreted correctly by whatever application is looking at them. That means putting things in the right slots, which is why we use templates for common forms of articles, and why we attempt to follow a regular style, and what is the real purpose of generic markup.
In the specific case of the == and === and ====, that means that not only will the headlines be presented in appropriately sized faces, but that the organization of the article can be "understood" by an application that is, say, looking only for External Links, or, say, creating an outline from a Wikipedia article, or, who knows what some future application might be able to do with properly organized material that it can parse according to some well established rules.
Of course, that is all an ideal and there are thousands of articles that don't conform, but if new articles conform and old articles are fixed by people who understand that the present Wikipedia is not the only way the articles in the Wikipedia will or can be used, then the Wikipedia will be better, worth more, used more, and last longer.
The genial and hardworking Ram-Man, by the way, has already fixed all the city, county, state, and whatever templates to work correctly in the future without any fuss.
That naive and cranky guy,
Tom Parmenter Ortolan88
Ortolan88 wrote in part:
Toby Bartels wrote:
When rendering a page, we should first measure all of the header markups and then render the shortest as <h2>, the next as <h3>, and so on. (Anybody that really needs a header of a specific size can still create this by putting the HTML tag in directly.) Then you can start with == or ===, or even = or =========, and it will still render as <h2> if it's the shortest one.
Tom [Ortolan88] is happy, because good HTML is being produced. [I forget who the principal opponent is] is happy, because current articles don't have to be rewritten. At least I hope that y'all're happy; responses?
Tom is unhappy because Tom doesn't care about good HTML; Tom wants good generic markup in the Wikipedia that will permit various processors, HTML and otherwise, including future processors that we know nothing about, to produce output that is organized and presented as it was originally designed to be presented in the Wikipedia.
Ah, so it was the *wiki* code that you wanted to be robust, not so much the resulting HTML; that's a good attitude.
In the specific case of the == and === and ====, that means that not only will the headlines be presented in appropriately sized faces, but that the organization of the article can be "understood" by an application that is, say, looking only for External Links, or, say, creating an outline from a Wikipedia article, or, who knows what some future application might be able to do with properly organized material that it can parse according to some well established rules.
Why can't these other parsers (which I agree will exist in the future) parse the header markup according to the same rule that I just suggested?
Of course, that is all an ideal and there are thousands of articles that don't conform, but if new articles conform and old articles are fixed by people who understand that the present Wikipedia is not the only way the articles in the Wikipedia will or can be used, then the Wikipedia will be better, worth more, used more, and last longer.
And in the meantime, these thousands of articles are rendered badly. If we change the HTML rendering (thereby changing the standard for any future parsers) to make the primary header always <h2>, then these thousands of articles will be rendered correctly. And then you and Ram-Man -- and me for that matter, for I do this too -- can put the wiki code in specific pages in a consistent style at our leisure.
-- Toby
wikipedia-l@lists.wikimedia.org