Erik wrote:
I feel that it is extremely tedious to have to click around many times and load many pages to get a complete picture of an issue, a person etc.
There is little difference between clicking on a TOC link in a huge article than clicking on a link to another article.
I think an article should have as much information related to its title as possible for that reason, and things should only be split off if a certain maximum size is reached (I tend towards 30-40K), or if they are not really related.
I really hate duplication of effort; If article A refers to event B and article C also refers to event B, it is MUCH better to simply have an article about B and short summaries in articles A and C. 30-40 KB is unreadably long for all but the most important topics (such as a major world conflict where simply providing short summaries of the major points would yield an article of that length). A max of 15-25 KB minus markup is more readable for most topics.
It is much better to chop things up into digestible bits. Then summaries of the spun-off bits should be left with links to full articles on those topics. That way the reader has a choice to read the summary or to skip right to the detail.
/That/ is far more useful for the reader and also minimizes duplication of effort by contributors. I also don't see a problem with this in a print version since on-topic summaries would always be left in parent articles (detail would be in daughters).
-- Daniel Mayer (aka mav)
From: Daniel Mayer on Thursday, December 18, 2003 12:58 PM
It is much better to chop things up into digestible bits. Then
summaries
of the spun-off bits should be left with links to full articles on
those
topics. That way the reader has a choice to read the summary or to
skip
right to the detail.
/That/ is far more useful for the reader and also minimizes
duplication of
effort by contributors. I also don't see a problem with this in a
version since on-topic summaries would always be left in parent
articles
(detail would be in daughters).
Moreover, keeping entries focused on distinctly individual topics makes successful collaborative editing possible. If it's not clear what the topic of the entry is (what the standards for inclusion are) then attempts to collaboratively edit the entry fall down.
Daniel-
I feel that it is extremely tedious to have to click around many times and load many pages to get a complete picture of an issue, a person etc.
There is little difference between clicking on a TOC link in a huge article than clicking on a link to another article.
When I read an article about a well defined subject, and I want to get all the information about it, I don't click around at all. I just read the article from top to bottom. I might skip a certain section, but I don't need to click anywhere. And when I do use the TOC, it is substantially faster in terms of latency than browsing of separate articles. Not only do I not need to wait for the page to load, I can also easily get context -- information about where I am within the structure of a topic -- simply by scrolling, rather than using back->click->back->click..
If I have an article like [[sports]], after Cunctator split it up into lots of tiny fragments, this possibility is no longer there. To some extent I agree with the split ups, but I find it questionable whether [[professional sports]], [[aesthetic appeal of sport]], [[nationalism and sport]], [[female sport]] etc. should really be separate entries. Even worse is that the article has a long list of "See also"s -- these are one of the worst ways to structure information.
In my experience, separating articles so much also leads to inconsistencies in style and neglect of articles about fringe subjects. For example, I predict that the newly created [[regulation of sport]] will be neglected, while I believe it would not have been if it had been kept as a section within the article. The reason is simply reduced exposure.
I really hate duplication of effort; If article A refers to event B and article C also refers to event B, it is MUCH better to simply have an article about B and short summaries in articles A and C.
I don't accept this as a general rule. It depends on the importance of the reference within each article. If the reference in article C is minor, it may well be completely acceptable to link to a larger article where event B is discussed. But there are of course many cases where for reasons of redundancy reduction this principle applies.
30-40 KB is unreadably long for all but the most important topics ... A max of 15-25 KB minus markup is more readable for most topics.
Well, these are mere statements of opinion. And of course the amount of information that people can stomach in one sitting varies greatly. I do notice, however, that these are not so far apart. If you take my minimum and your maximum and take the average, you arrive at 27.5K, which seems reasonable to me. The sports article is currently a mere 5K, which is IMHO a clear sign that it is far too fragmented. It also has no real summaries of the sub-articles it links to. Overall it gives a very unprofessional impression to me.
I would also probably be very pissed if Cunc had done this, without prior discussion as in this case, to one of "my" articles. One key reason is that the history gets completely lost in the process. Sure, you can still fish it out, but people will assume that Cunc wrote the individual pieces. I have invested many days of research in some articles I worked on, and getting fair credit in the page history is very little to ask for in return. Having a carefully planned out article structure messed up in this way would also make me quite angry. Be bold, but also be respectful toward other people's work.
/That/ is far more useful for the reader ..
I'm glad that you are so confident that you are correct. For me the issue is still quite foggy. While I agree with you that summaries of subsections are useful, I feel they should not be created needlessly. That is, if we're below a certain size -- maybe the 27.5K from above -- we don't need to split the article up unless there is a clear logical separation. For example, in the [[Mother Teresa]] article we split away everything that is about the [[Missionaries of Charity]] (Teresa's order), which is fine with me because there's a clear distinction here. But stuff like "History of .." should really only be split off if we're approaching the size limit, in my opinion. Otherwise the disadvantages of splitting appear to outweigh the advantages.
Regards,
Erik
From: Erik Moeller on Thursday, December 18, 2003 4:57 PM Daniel-
I feel that it is extremely tedious to have to click around many times and load many pages to get a complete picture of an issue, a person etc.
There is little difference between clicking on a TOC link in a huge
article
than clicking on a link to another article.
When I read an article about a well defined subject, and I want to get
all
the information about it, I don't click around at all. I just read the article from top to bottom. I might skip a certain section, but I
don't
need to click anywhere. And when I do use the TOC, it is substantially faster in terms of latency than browsing of separate articles. Not
only do
I not need to wait for the page to load, I can also easily get context
--
information about where I am within the structure of a topic -- simply
by
scrolling, rather than using back->click->back->click..
Well, that's nice, but it's not fair or appropriate for you to assume that what you like should be mandated for all. Assuming that all readers would be happiest if they had the version you like the best--when there is evidence to the contrary--is not appropriate.
I, for one, when I read a sentence like
The [[asthetic appeal of sport]] is what sports art, including sport in film, attempts to capture.
..will follow the link if I don't feel as if I understand what the "asthetic appeal of sport" is; but if I feel that I do understand it, I am glad that the author didn't force me to waste time skimming through information I didn't need.
If I have an article like [[sports]], after Cunctator split it up into lots of tiny fragments, this possibility is no longer there. To some extent I agree with the split ups, but I find it questionable whether [[professional sports]], [[aesthetic appeal of sport]], [[nationalism
and
sport]], [[female sport]] etc. should really be separate entries. Even worse is that the article has a long list of "See also"s -- these are
one
of the worst ways to structure information.
Erm--I didn't break out [[female sport]]. Don't bame me for that one.
I *drastically shortened* the list of "See also"s. Again, don't blame me for that one.
The one complaint that is reasonable and honest is that my summarizations of the subtopics in [[sport]] were too drastic.
But if we take the case of [[professional sports]], the [[sport]] article discusses how professional sports rises from sport's appeal as passive entertainment, how professional sports have given rise to several related industries, and how professional sports comes into conflict with the concept of amateur sportsmanship.
And that is, in my opinion, a reasonable summary of the most important issues relevant to our basic understanding of sport that can be gleaned from professional sports.
I don't know if all my edits were exactly right, but the entry needed some drastic editing. By separating out some of the more digressive sections it's much more clear what information is needed and what is not.
In my experience, separating articles so much also leads to inconsistencies in style and neglect of articles about fringe
subjects.
For example, I predict that the newly created [[regulation of sport]]
will
be neglected, while I believe it would not have been if it had been
kept
as a section within the article. The reason is simply reduced
exposure.
Who knows? The longer the entry, the more difficult it becomes to properly edit its subsections--or to entirely reorganize it if necessary.
30-40 KB is unreadably long for all but the most important topics ... A max of 15-25 KB minus markup is more readable for most topics.
Well, these are mere statements of opinion. And of course the amount
of
information that people can stomach in one sitting varies greatly. I
do
notice, however, that these are not so far apart. If you take my
minimum
and your maximum and take the average, you arrive at 27.5K, which
seems
reasonable to me. The sports article is currently a mere 5K, which is
IMHO
a clear sign that it is far too fragmented. It also has no real
summaries
of the sub-articles it links to. Overall it gives a very
unprofessional
impression to me.
The version beforehand felt very unprofessional to me. I expect that it will achieve a happy medium at some point.
But I get the feeling that what feels "professional" to you is "what would look best on paper".
I would also probably be very pissed if Cunc had done this, without
prior
discussion as in this case, to one of "my" articles. One key reason is that the history gets completely lost in the process. Sure, you can
still
fish it out, but people will assume that Cunc wrote the individual
pieces.
I have invested many days of research in some articles I worked on,
and
getting fair credit in the page history is very little to ask for in return. Having a carefully planned out article structure messed up in
this
way would also make me quite angry. Be bold, but also be respectful
toward
other people's work.
I'm sorry, but I don't think that it's appropriate to take editing changes like this personally. It's counterproductive to collaboration and egocentric. Taking edits for structure personally doesn't help anyone.
It's a lot better for everyone to get your high from the act of selfless contribution or the knowledge that you've insinuated your knowledge and perspective invisibly into the stream.
Once you contribute to Wikipedia, it's *not your work* any more. The original contribution is, but that's it. *You are explicitly agreeing to a merciless editing process*.
It's reasonable and right to say "I think your changes have hurt the coverage of this issue in these ways". It's unreasonable to say "I'm angry because you changed my entries, whether or not it's for the better."
If you're so worried about credit, then write code to allow people to rename sections to new pages so that the version history is remembered.
/That/ is far more useful for the reader ..
I'm glad that you are so confident that you are correct. For me the
issue
is still quite foggy. While I agree with you that summaries of
subsections
are useful, I feel they should not be created needlessly. That is, if we're below a certain size -- maybe the 27.5K from above -- we don't
need
to split the article up unless there is a clear logical separation.
For
example, in the [[Mother Teresa]] article we split away everything
that is
about the [[Missionaries of Charity]] (Teresa's order), which is fine
with
me because there's a clear distinction here. But stuff like "History
of
.." should really only be split off if we're approaching the size
limit,
in my opinion. Otherwise the disadvantages of splitting appear to
outweigh
the advantages.
Again, your idea of a reasonable size for an entry differs from mine, and from other people.
I think "History of X" entries are very different from "X" entries. The entry "X" should tell the person what X is--not what it was.
--tc
Cunc-
Well, that's nice, but it's not fair or appropriate for you to assume that what you like should be mandated for all.
That argument can be applied in both directions. So unless you come up with a very good scheme thad does not substantially affect performance, is easy to use and allows different "views" of an article, we'll have to go with what the majority prefers (and make some concessions to the minority). Given the experience of the recent poll on the Rick Santorum article, I find it likely that your view is a minority position.
I, for one, when I read a sentence like The [[asthetic appeal of sport]]
That should be "aestehtic", btw.
But if we take the case of [[professional sports]], the [[sport]] article discusses how professional sports rises from sport's appeal as passive entertainment, how professional sports have given rise to several related industries, and how professional sports comes into conflict with the concept of amateur sportsmanship.
The level of detail of this "discussion" is as high as the level of detail of the above paragraph. In other words, there is almost no information in it that goes beyond what virtually everyone knows. That is the kind of paragraph that makes articles look unprofessional unless you are fully aware when reading it that there is much more detail available. Print it and it looks really silly. "Participants are paid *by the audience* in professional sports? Wow, I didn't know that! Wikipedia is so informative!"
Who knows? The longer the entry, the more difficult it becomes to properly edit its subsections--or to entirely reorganize it if necessary.
Um, reorganizing won't exactly become easier with 20 or so separate articles where in many cases nobody will know where any specific piece of information is.
But I get the feeling that what feels "professional" to you is "what would look best on paper".
Not really, it's just that for me, the differences between what is good on paper and what is good on a computer screen are not that big. There's a certain hypertext myth that says that a "real" hypertext page needs to be small and full of links to other hypertext pages. In terms of usability studies, this has been found to be highly confusing and ineffective. Having links is good, but they should lead to reasonably sized chunks of information. The more fragmented an article is, the higher the likelihood that people will be frustrated by lots of tiny stubs.
I'm sorry, but I don't think that it's appropriate to take editing changes like this personally. It's counterproductive to collaboration and egocentric. Taking edits for structure personally doesn't help anyone.
It is not an irrational feeling. If you disagree with what someone has done to your article, then your own investments of time and energy into it are called into question, the other person in effect tries to *condition* you to behave in a certain way. It is therefore really a minimum requirement for collaboration to ask before making major changes, and to try to achieve consenus.
You know very well that your changes to [[sports]] probably would not have found consensus if you had announced them beforehand. And I believe that is the reason you did not do so. Please try to follow the defined Wikipedia decision making process. In case of minor changes, make them immediately. In case of medium changes, make them and then comment (in the edit summary or on the talk page). In case of major changes, announce them first and wait for objections before making them. In case of objections, seek consensus. In case of lack of consensus, you are free to start a vote.
It's a lot better for everyone to get your high from the act of selfless contribution or the knowledge that you've insinuated your knowledge and perspective invisibly into the stream. Once you contribute to Wikipedia, it's *not your work* any more. The original contribution is, but that's it. *You are explicitly agreeing to a merciless editing process*.
That is all well under the assumption that you agree with what the other person has done. If you feel it makes the article worse, then this kind of rhetoric falls apart.
It's reasonable and right to say "I think your changes have hurt the coverage of this issue in these ways". It's unreasonable to say "I'm angry because you changed my entries, whether or not it's for the better."
Of course. I never said it would be reasonable to be angry about changes that are for the better. It is questionable whether we agree when this is the case, though.
If you're so worried about credit, then write code to allow people to rename sections to new pages so that the version history is remembered.
That's not possible. You *can* split up the edit history (with *very* much coding effort), but it will no longer make sense. Edit comments lose their meaning, the linear structure of the history is broken up as edits which were made to other sections between two edits of the same one are no longer part of the new history, etc. Aside from that, cut&pasting is typically not limited to clearly separable sections.
So losing proper credit is very much a given when cut&pasting out parts of an article. This is probably also in material breach of the FDL. That's one other major reason that it should not be done without consensus.
I feel that if you continue to split up articles in major ways without announcing this on the talk page first and giving authors a chance to react to your proposal, you are in serious violation of Wikiquette, and possibly other policies.
Regards,
Erik
From: Erik Moeller on Friday, December 19, 2003 5:01 AM To: wikipedia-l@Wikimedia.org Subject: Re: [Wikipedia-l] A quick thought about 1.0
Erik-
That should be "aestehtic", btw.
Ugh, "aesthetic" of course. There's got to be some natural law that
when
pointing out spelling errors in other people's messages, you will inevitably make at least one yourself.
It is a recognized phenomenon...not sure if it's been given a name.
On Fri, 2003-12-19 at 17:11, The Cunctator wrote:
> From: Erik Moeller on Friday, December 19, 2003 5:01 AM > To: wikipedia-l@Wikimedia.org > Subject: Re: [Wikipedia-l] A quick thought about 1.0 > > Erik- > > That should be "aestehtic", btw. > > Ugh, "aesthetic" of course. There's got to be some natural law that when > pointing out spelling errors in other people's messages, you will > inevitably make at least one yourself. > It is a recognized phenomenon...not sure if it's been given a name.
Typoietic conservation law? (just a suggestion)
-- Jussi-Ville Heiskanen
By the way, I apologize in advance if my tone in the last was too argumentative. I disagree with Erik, but I know his goals are to make Wikipedia as good as possible.
On Thu, Dec 18, 2003 at 09:58:13AM -0800, Daniel Mayer wrote:
Erik wrote:
I feel that it is extremely tedious to have to click around many times and load many pages to get a complete picture of an issue, a person etc.
There is little difference between clicking on a TOC link in a huge article than clicking on a link to another article.
For me there's a huge difference. My latency on wikipedia is usually between 5 and 10 seconds. OTOH I have high bandwidth. So I would greatly prefer to download a huge article at once.
I think an article should have as much information related to its title as possible for that reason, and things should only be split off if a certain maximum size is reached (I tend towards 30-40K), or if they are not really related.
I really hate duplication of effort; If article A refers to event B and
Why is this duplication of effort? We can simply copy-paste from one article to the other.
article C also refers to event B, it is MUCH better to simply have an article about B and short summaries in articles A and C. 30-40 KB is unreadably long for all but the most important topics (such as a major world conflict where simply providing short summaries of the major points would yield an article of that length). A max of 15-25 KB minus markup is more readable for most topics.
It is much better to chop things up into digestible bits. Then summaries of
I would characterize it as an information dump which forces the reader to make an effort to extract a coherent picture out of it. Further, it is often not possible to deduce what is the topic of the article in an outgoing link (the link text is usually a single word), so I click something and find its not what I wanted and get annoyed.
Arvind
On Fri, 2003-12-19 at 06:30, Arvind Narayanan wrote:
On Thu, Dec 18, 2003 at 09:58:13AM -0800, Daniel Mayer wrote:
Erik wrote:
I feel that it is extremely tedious to have to click around many times and load many pages to get a complete picture of an issue, a person etc.
There is little difference between clicking on a TOC link in a huge article than clicking on a link to another article.
For me there's a huge difference. My latency on wikipedia is usually between 5 and 10 seconds. OTOH I have high bandwidth.
This is an important point. There are people with 28kbps flaky bandwidth in the world. Lots of them.
I of course have a whopping 2 mbps download speed line but that is due to unfair global economics. Further my neighbours do not seem to need wikipedia as much as the places where there are no internet even.
So I would greatly prefer to download a huge article at once.
This helps to be saved on a floppy and be searched effectively (Ctrl+F for most browsers) Printing becomes more meaningful.
I think an article should have as much information related to its title as possible for that reason, and things should only be split off if a certain maximum size is reached (I tend towards 30-40K), or if they are not really related.
again the search utility becomes helpful. Anything very big will easily attract attention and then reorganisation can be thought about. I use tabbed browsing (a great recent trick to keep the number of open windows low) and find myself sometimes getting entangled in anything upto 50 tabs on a decent night.
Monolithic pages should not become a bad thing. Clear strucutre and named anchors can help monolithic pages be very useful.
To me 'standards' that seem to specify file size as a 'usability' issue based on average screen size a misled corporate PR feature.
But eventually I must admit, technology that interchange the data between monolithic and individual nodes an author wishes for will be a useful thing.
Ramanan
On Friday, December 19, 2003, at 01:30 AM, Arvind Narayanan wrote:
On Thu, Dec 18, 2003 at 09:58:13AM -0800, Daniel Mayer wrote:
Erik wrote:
I feel that it is extremely tedious to have to click around many times and load many pages to get a complete picture of an issue, a person etc.
There is little difference between clicking on a TOC link in a huge article than clicking on a link to another article.
For me there's a huge difference. My latency on wikipedia is usually between 5 and 10 seconds. OTOH I have high bandwidth. So I would greatly prefer to download a huge article at once.
Same here. At least. But for bigger articles, it's longer. I'd much rather read a short article and load expanded information in the background. I object to TOC entries, though, because the links have no context. The beauty of wiki links (and html in general) does not lie in TOCs or see-also's. It lies in the ability to connect information logically through context. If a single 'logical' entry is big enough to bring up the issue of using smaller 'physical' entries---that is, use a TOC entry with the equivalent of subpages containing sections thereof---I think the size of the entry needs to be addressed. I don't want to read that much in a single entry, and I certainly don't want to edit it. And editing *across* multiple pages? Forget it, it's not worth it, "Can't someone else do it?". That feeling hurts the wiki. A lot.
If an entry gets to that size and detail, great, let's use the information, but let's spread it out across a few entries, not just by breaking it into pieces, but by separating it into multiple topics. And if that means duplicating a little information, so be it. As you say, Arvind,
I really hate duplication of effort; If article A refers to event B and
Why is this duplication of effort? We can simply copy-paste from one article to the other.
Very true.
Also, when I look something up, I don't want a long article to read to find what I want to know. Remember, Wikipedia Is Not Paper. That means that rather than have to put all the information on one topic *and requisite info* in one place and the shorten it because there's not enough room, we can include just enough to understand the topic with a little requisite info---short enough to read quickly---and then link terms to further elaboration of each background concept. That way readers read only what they need to read. Having more and shorter articles also encourages them to grow longer, which means more information on the site---and eventually these can become multiple articles themselves. It's a wonderful circle of electronic life. Ok, the coffee can kick in now.
My apologies for rambling and twisting myself around. I'm running on very little sleep.
Peter
--- Funding for this program comes from Borders without Doctors: The Bookstore Chain That Sounds Like a Charity. --Harry Shearer, Le Show
On Friday, December 19, 2003, at 04:26 PM, Peter Jaros wrote:
On Friday, December 19, 2003, at 01:30 AM, Arvind Narayanan wrote:
For me there's a huge difference. My latency on wikipedia is usually between 5 and 10 seconds. OTOH I have high bandwidth. So I would greatly prefer to download a huge article at once.
Same here. At least. But for bigger articles, it's longer.
That was really unclear. 'Same here' should refer to the latency, not the opinion. Sorry about that.
Peter
--- Funding for this program comes from Borders without Doctors: The Bookstore Chain That Sounds Like a Charity. --Harry Shearer, Le Show
wikipedia-l@lists.wikimedia.org