SL as an ageing Wiki

    Keywords: SL description

Tapir: Sensei's Library (as Wikipedia et al.) can be described as an ageing wiki. Aging wikis show some common features:

  • An increasing pages per editor ratio.
  • An end to exponential or even linear increase of editors.
  • Continued albeit slower increase in articles.
  • Increasing difficulty in keeping changing information up to date.
  • Decreasing churn rate.
  • Large parts fallen out of use altogether.
  • Lack of response - on requests, discussion in all library related issues.
  • Timid editing (because of lacking response).

Special for Sensei's Library in my opinion:

  • Still overall good mood. (Very good.)
  • Discussion left Sensei's Library towards other platforms. (Very bad.)
  • Dated technology. (Compare the easy .sgf integration nowadays in Lifein19x19.)

Some examples:

  • Deletion requests are usually handled by the librarian requesting deletion.
  • Article of the Week depending on availability of three persons. (And no one taking responsibility if those are not available by chance or accident.)
  • Joseki page redesign stopped midway - after a disagreement on 3-4 joseki page and a single comment on Systematic Joseki remove request. Calls for third opinion etc. unheared this either heads to no-change or an effort without consultation (which likely produces bad mood but no change). Not that change is necessarily a good thing, but it is sad, when it proves difficult to follow up on agreements altogether.
  • Pages to crowd ressources did not work out (Sensei's Wishlist, Library Work, various "change this"-templates).

Dieter: Here is my story with respect to SL. I joined in the very beginning, when I was at the peak of my go improvement rate. About 6 years ago I quit competitive go but was still lurking aound here and in my local club. SL was one usual suspect of my internet addiction and I made two attempts to quit. Then over the last two years I quit in a natural fashion.

On one hand I do not feel the need myself anymore to improve on SL, because I'm rather inactive in Go itself. Secondly, I believe (and foretold) there is and will be a general "community fatigue" especially when most relevant or easy to reap content is present. Look at Wikipedia, it's not rocketing like in its early days anymore either. Look at open street maps. Thirdly, indeed, if I would take up Go seriously again I would now put up my games for review in lifein19x19 instead of here, because sgf is integrated. As an early GTL reviewer, I notice and understand a decline in review requests. Even my former pupils put up their games in LI19x19.

Arno has made several requests for developers to jump in and improve on SL. If they don't, maintenance will remain laborious and sgf will not be integrated. There is no point in maintaining SL with the few human and architectural forces available, if other forces are moving faster. SL should probably merge with lifein19x19, or be accessory to it, or it will decay into a library with valuable but out-of-date material.

In the future, I believe we will see many efforts to integrate information that had been maintained in various sources and formats. Such efforts may be the most useful with respect to SL in its aging state.

This being said, I don't quite understand why people prefer a forum over a wiki in general. The sgf-integration seems key to me.

John F. I have a sense that the battle has already been lost. It struck me forcibly that there was a recent thread on L19 about starting a book review forum. Not once did anyone mention that SL already exists for this purpose. There have been several other topics there where enquirers have not even thought to look at SL.

The lower profile now of SL has several causes. Apart from what Dieter has mentioned, I think two factors are the increasing use of blogs and a general decline in the go population.

Quality control has always been a problem for SL, from Day One, as on all wikis. I believe that this is linked with the lack of attributions. Contributors do prefer to have their names attached, and so turn if they can to their own blogs instead. Readers prefer to know that what is said is by someone they can trust. Absent that on SL, I think they are inclined to turn to L19 and ask directly. They get a sense of who to trust from the discussion there.

There has also apparently been significant recent reduction in western go activity. On the basis of several figures, I'd estimate a fifth. But I get the sense that it is much more than that on SL. Also, turnover is a problem in that regard. It seems that many "big" names (e.g. Dieter's) disappear, and while it may be argued that they will naturally be replaced by fresh faces, two problems there seem to be: (1) the loss of a big name is not just loss of a body but of a wealth of experience, and that takes a very long time to replace; and (2) the fresh faces tend to be young and so come equipped with a rather different set of desires, expectations and skills. New internet technologies feature highly in that, and SL is probably seen as antediluvian by them. I can't see that their interests in the newest gizmos or programming languages would even tempt them to update SL.

The best I can see in a very murky crystal ball for the future of SL is for it to prune itself into a high-quality SL-Lite where the information remaining is attributed (and maybe has a "likes" counter to help show trustworthiness) and dated (a major lack at present), To my way of thinking, much of the discussion and stubs should be cut away in the pruning. Discussions belong on L19. If a subject has to be extensively discussed, it does not really belong on a fact-based wiki anyway. If someone disagrees with existing "facts" they should just post a rival page and let the readers decide who to believe. I dislike stubs because they have a negative tone. They stress what is not in the wiki rather than what is, and also convey the idea that whoever is posting the stub is not very knowledgeable anyway, otherwise they would fill out the page themselves.

However, I am not at all convinced that there are very many knowledgeable contributors available and willing to contribute, especially given that most of the data centres on source in "hard" oriental languages. It therefore occurs to me that SL might re-invent itself as much more western oriented. There is some European input, but I have often been surprised at how little SL has about the North American go scene.

Hyperpapeterie: I agree with a lot of what you said, John. However, you say the data is in difficult oriental languages. This is true, and an obstacle to a truly top-notch English language Go wiki, but I think that Sensei's even lags the English language sources. In reading your work, or Andrew Grant's Four Hundred Years of Japanese Go, I can easily take something from every single page that is not yet reflected on Sensei's. The same is true for material on blogs. That's even avoiding the issues of importing that content wholesale. Major improvements to Sensei's would be possible without even approaching the boundaries of fair use (just citing factual information). I see motiviation, organization and quality control as the biggest issues.

Andy: SL is regularly linked to in the kibitz of high-dan games on KGS, usually in the context of defining Japanese go terms for those unfamiliar with them. I've never seen Li19x19? or any other online reference used in this manner.

unkx80: Well, I don't think I have much useful things to add after Dieter and John. Like Dieter, I gradually lost interest in Go, and without playing the game for an extended amount of time, I found myself out of things to contribute on SL. Also, the dwindling activity on SL affected my enthusiasm to do work on it.

I pretty much agree with the views of John. However, it seems that a lot of so-called facts on SL is subject to challenge, often with respect to presentation. Also some areas such as opening theory is pretty much opinion rather than facts, but this is exactly the kind of thing that makes Go so hard. In any case, if users post rival pages to dispute what others said, then the system becomes more of a discussion forum rather than a wiki.

One thing I observed is that it seems to be unexciting to import information wholesale, especially if that information is already available online. I am not even talking about writing the information to avoid crossing the boundaries of fair use, which is more work. People would rather post their own questions, point of views, or other discussions. They seem content to extract whatever they wanted from the conversation, and then move on. People talk about the importance of preserving information, but few are actually interested in doing all that grunt work in sorting out and archiving the information. I guess that's the way it is.

I guess I'd resign to the fact that online communities will move every now and then. If something shiny new comes along with the right features and enough marketing is done, the community might just jump ship, especially when the old gathering place has some unresolved deficiencies. I'm the kind of person who's very reluctant to move once settled into a place, but I guess I must get used to this. The Go community is one example, and I have seen it when the majority of my contents moved from Friendster to Facebook, and also from ICQ to MSN.

That said, SL is massive enough not to go away very soon. Like Wikipedia, many people would point others to SL pages for definitions and other general information, but not contribute new material to SL. Assuming that SL continues to stay usable, I guess this will be the status quo for a long time to come.

Hyperpapeterie: Further thoughts. I think the reason that there are two intrinsic reasons to prefer a discussion forum to a wiki. The first is that for discussions that are by nature temporary, a forum is a better fit. You can look at unread posts and find what you're interested in. We have recent changes, but it's not the same at all. Less of it is interesting to an average reader. If I go to a professional player's page and update his dan ranking, or worse yet, delete three lines, that's not interesting (the first change might be a minor edit, the second won't be). Worse, for many changes, it's not at all clear what they are--whereas with L19, I know whether a forum is one that I follow, and the same for the topics. So there are some things that aren't our strengths.

Second, and more importantly, I think that discussion forums are easier. Blogs are easier than senseis, because you don't won't break conventions or step on toes, and can express whatever opinion you want, and discussion forums are easier than blogs. Twitter is easier than either, and it's starting to edge into the popularity of blogs and forums.

I think that ease of use has a cost. Old posts on a blog are often invisible, unless they're picked up by search engines. Ditto for a discussion forum (though there we often see collective memory helping). And building comprehensive information on any of these platforms is really hard. Btw: I blog sometimes, use Twitter a lot, and L19 way too much--I just think they can't do what SL is potentially capable of doing.

What SL can do best is provide a repository of information that is lasting, in the sense that it does not quickly become dated, and will continue to be of interest long after a blog post or forum thread would be buried. Book reviews and historical information are two good examples, but there are probably others.

The harder question is seeing how to keep people interested.

Tapir: I disagree with several points made by Dieter and John that tend to resurface in these kind of discussion (we still have some!). A library pruned of discussion or specialising as information repository would be a hollow shell. Still nice and educating to watch, but without life and development. Why should anybody care to edit it? Not that information and knowledge is not important, but in a lively place it is much easier to gather. There is one reason for this: collectivity. I have seen ample evidence that a bunch of amateurs consulting with each other can outperform the strongest among them over the board by far, why not in providing content?

And there is where my frustration lies. SL promises this collectivity (wiki was meant to be collective writing) but it does not happen but on rare occasions. I am probably now of about the same strength Dieter had during his most prolific time here (he started heavy editing as 1k here if i recall correctly). So I may add some pages with different misconceptions and biases to SL. A bunch of active writers working together - even if not stronger, even if weaker - may considerably improve most go knowledge we have here however. A few isolated people - even if considerably stronger - can't.

Also, I don't believe in a general decline of activity in Western Go, the numbers look different in every country. I looked at EGD data, there is no correlation whatsoever between e.g. Russia, Poland, Germany and Turkey. And no general downward trend at all. Just online activity is nowadays spread among much more platforms than earlier. (Why people bother to write a personal blog remains a mystery to me. More so, if it is about an endeavour like learning Go, really who cares? Who will read that? This should be actively discouraged. I once started one and just to delete it after two entries btw.) Maybe activity increases again when Starcraft players start to look for ways to sublimation (more refined addiction)...

John F. You have to be careful about confusing wishful thinking - in several respects - with reality. It may (or may not) be good to have a lively and developing library, but the point of the starting post was that SL appears to be showing its age and so is not fit for that purpose. In any case, there is an argument to be made that a library of unadulterated facts is better than one where opinion and debate obscure the facts, especially where L19 is available to host discussions.

It is also probably wishful to believe in collective editing as described. In office work or research, even when several people contribute ideas or research, the final paper is invariably written by one dominant person. There, the constraints of having a boss and a career make people cooperate. In voluntary cases such as SL, I don't think many people want to kowtow to a dominant writer. They blog instead, though I agree that most blogs are a complete waste of space.

By activity declining in western go, quoting Russia, Poland, Germany and Turkey seems a bit pointless for an English-language wiki, but in any case I meant activity as opposed to playing. It seems to me that association numbers are declining, land-based tournament activity is declining, book sales are declining, postings on SL and L19 are declining, blogs are drying up and other sites close or become inactive far more often than new sites open. Even if numbers playing online are going up (and I doubt even that), these are not the sort of people who will fill SL's pages.

Harsh reality may not be pleasant but it's a useful starting point for reform.

willemien I still think sensei's is the best internet resource for Go, I like the " no thrills" way it is, video, colourful background or sound it would only distract me,( and i think many others as well, there were not so long ago comments about the stripy background from sensei's. also the simple structure makes it easy to edit.

Discussions, for small discussions I think the best place to have them is at the end of the sensei's page itself, or on a discussion sub page. The talk pages, I find a bit of a nuicance. (especially if they are old and no longer of interest, can we not remove them after a year of the last comment or so? If the information is still relevant it should go to the main page) For longer (and easely going off tangent discusions life in 19 x 19 is better (why don't we have this discussion there?) The main goal for sensei's is I think to make available lasting information, to give links to other go websites ed.

I guess there are some tools around to make sensei's easier to edit.

  • for wikifying (make a link of the first occurences of xxx on every page)
  • page ranks, I sometimes have a look at pages by distance
  • age of pages, when was the page last updated?

Some pages (and adverts) are a bit old and should be removed/ updated.

Maybe we should promote sensei's more, (in discussions on L19

Also maybe some things are better done at other website's

and should we just abandom (and transfer them to there] (maybe only keeping an index here) although i do like some tsume go problem collections here.

I think we should sometimes import information wholesale (after permission of the author off course) and with link to the original information. The information here is lasting. some information on other sites is now lost. (see for example Nyobutsu)

And i do agree fully with tapir's last post.(where he says it is and should stay a community affair

isd: I find SL to be reasonably complete. Obviously it's not perfect, but with regards to basic material I find it satisfactory. There are not a great deal of pages which I would feel comfortable in attempting to improve. That is, what is left to be done is usually quite specialist. That's generally what stops me from contributing. AOTW isn't something that excites me a lot, so I have ignored that. :)

Tapir: While I believe that Sensei's Library is the single most important source of Go knowledge in English language by width, depth and accessability, I started to see its limits a while ago. (ref) Improving on that is not easy however and I do not feel too confident about my abilities myself. With most of the senior (by age or strength) editors with the exception of Herman being away from active play (and active editing), I and others around may have to grow into new responsibilities. As huge chunks are written by players not that much stronger, it may after careful thought even be possible to improve it here and there. Although that would not be the big quality jump I sometimes dream about (which could only happen if western players in larger numbers would actively discuss their ideas and concepts and approaches here [1]), but things will keep going.

[1] For starters: This may be Robert offering his thoughts on what else but vanity is at stake on the 9-7 point. By public scrutiny and discussion a lot of people will learn something about influence, large scale splitting etc., and if he improves himself I would consider this a major step for the western go scene. (E.g. by being less busy with rules discussion.)

isd: About your misconceptions formed from SL. I think some misconceptions are inevitable if the library is not kept up to date - new joseki lines will not be given preference. To take something standard like the 3-3 point invasion - there are stages of learning involved here. First everyone learns the standard gote joseki to answer it - then they learn the double hane, then maybe they learn the naughty double hane inside, then perhaps they learn the sente answers. This progression of knowledge (as normally aqcquired) usually reflects a progression in strength of play. I don't think it is accurate to say that SL misinforms here - it just doesn't totally inform us.

Dieter: (hoping this discussion is ordered top-bottom chronologically forward) Existentially, life is a waste of time. Whether a blog, or a forum or a wiki, it's all passtime but somehow we find gratification in it. I own a very active blog about songwriting in Dutch, which attracts about 20 visitors a day. I don't really care whether that is enought to consider it not to be a waste of time and space, because it is alive and I like it. I felt the same about SL 5 years ago except that it required a lot of collaboration and effort to maintain lingering content, posted by less careful people. I have always battled the sense of personal ownership, and teamed up with the side that wanted it to stand out for its content, not for the personalities behind it. Being in the music scene and blogging scene, I've come to understand that personality matters a great deal. It's true for art, for politics and probably also for go teaching and writing. It seems to matter a lot who's behind it, not only what is being said. The lines between fact and opinion are fuzzier than we may desire. All facts come to us through some kind of authority: there is frighteningly little we verify for ourselves.

If there is indeed an existential problem with SL, for its current contributing squad, I like the proposal by John F. to trim SL down to what it is used for and perform a large scale cleanup of all the noise. Cleaning old forums and actively referring to LifeIn19x19 for more discussion would be a step to come to a natural merge with that forum, but a key ingredient still seems to be a seamless integration of sgf. I repeat that Arno has been calling for programming help for such endeavours.

As for a decline in activity, I confirm partly what John says. My feeling is that physical activity is downgraded to local clubs and people do not move as much to do stuff together, since the Internet has resolved availability of competition and mental activity. National bodies decline. I would expect however that such localized physical activity is complemented by an increased online activity. Why this is not so much the case, may be due to the younger generations being not as eager to do efforts for a large, anonymous organization, a higher goal. Let's face it: youngsters have been spoiled in the West. They take more than they give. If it is fun to do, like playing online, they'll do it. If it brings some financial reward, or immediate personal reward, they'll do it too. In the spur of programming new stuff, they'll do it to, out of a desire to experiment and learn. But keeping a website alive for a community, like Arno has done, let alone maintaining a diffuse content, like we've been doing, may not be a calling for the newer generations anymore. But maybe that's the perspective of an aging wiki contributor.

willemien: Started a parallel discussion at [ext] http://www.lifein19x19.com/forum/viewtopic.php?f=10&t=2870 to see what users/ people at life in 19x19 think about Sensei's library / us

Tapir: I must admit that I was hyperventilating when I read those "pruning" parts. A gardener is pruning some branches, to give the other branches more space, light and perspective. What is proposed here sounds to me like the proposal to cut off all remaining branches and sell the rest as dead wood. This looks even stranger when this proposal of John is seconded by senior librarians who did not prune anything at all. (As far as I can tell: I started playing in 2006, editing here some month later.) And I feel both angry and sad about that.

The reactions on Lifein19x19 are until now the predictable ones. Navigation, technical datedness, comparing wikipedia to sl, complaint without contributions, because the minimum level for contribution looks pretty high (I felt the same way, I thought I need to be 5 dan before being able to work on contents, so I ended up working on technicalities, people pages, library stuff) etc. etc. For tsumego I feel the presentation on SL is the second best after plain paper (without solution). Clicking through a tsumego is not solving them. Also this merits a quote: "I've always used it and still do. What's frustrated me of late is a "corrosion of content": topics are not presented locally anymore but referred to with "there's a very interesting discussion on this topic in this thread at godiscussions." Those threads are long gone and nobody distilled whatever their content was into Sensei's pages."

Dieter: I never liked Systematic Joseki pages and regularly joined the side that wanted to merge them back into the more organic treatment, but at the time SL was prospering and growing and MortArno has always favoured allowing any content and appreciate SL for its imperfection. I was ok with that and followed suit as a librarian. That does not withstand that if we now discuss the future of SL in different conditions, prompted by your perception of an aging wiki, I can favour a more radical approach. You know, you come and go, and as you and your environment evolve, you can change opinion. At the time I had lots of arguments with Charles Matthews and Bill Spight about personalized content. I shifted towards their opinion, due to other experiences and maturing, which doesn't say I am more "right" now.

Tapir: I am in favour of allowing any content and give it a chance to prosper and grow and strictly against any self-limitation of Sensei's Library myself. And I am not interested in a Wikipedia of Go, where every sentence has a reference but the whole article is useless nonetheless. (Pretty much like the worst film ever "300" uses classical quotations but still manages to misrepresent about everything.) But being open to all kind of contribution and never cleaning up attempts that went nowhere are two unrelated areas in my opinion. See the Go news page, which was practically dead in 2007 (and still filled with its old content, nobody bothered to remove), it got revivified mostly thanks to Hyperpapeterie and Valerio. It surely isn't the world's most important outlet for reporting and notifying about current Go events - but having this page allows SL to keep up to date all those entries on tournaments, title holders etc. An agreement to refer to outside SL for go reporting would not do that. Latest Promotions tries to achieve the same for the other end of the professional scene, keeping track of entrants and promotions. And if you look into Wikipedia deep enough you will notice that their "facts only" approach ironically undermines their endeavour to keep track even of the most basic facts at many a place.

Arno: although the number of active participants has been declining, SL attracts a lot of traffic: currently about 700,000 page views per month (search engines and crawlers add another 600,000 views.) As far as I can tell (e.g. according to alexa.com,) no other English go site attracts this much traffic. So all is not lost :o) That being said: I believe that the days of fast growth are over: as several people have stated, the low hanging fruit has been written and added already, the entry barrier has risen. Heck, I even get emails asking me to make changes for other people - apparently people no longer know what a wiki is. What is left then is either tedious library work or "hard stuff" (i.e. you have to be 4d+ or do a lot of research.) In a world, where instant gratification becomes ever more important, this is bad news. Maybe we should make a more detailed survey and prompt SL's visitors to complete it. I think it would be easy to get 200+ responses, so that we can see what is really keeping back contributions.

Yes, SL is like a zoo sometimes, and not everything is easy to find. Granted. Also, apart from templates and position search, there was not a major technical overhaul for more than 5 years. Like many of you, my interests have moved away from Go. Yet, I'm proud to say, I have been able to keep this site running for over 10 years now (this is not to diminish the hard work the most active contributors (typically librarians) have put in.)

I think, finally I have a good idea of how to implement SGF in SL (my first prototype of this is now well over 5 years old). For me, the ability to read a page online as if it were printed on paper (like in a book) is key - I really don't like the way it is integrated in L19x19. The discussion of variations has to be easy to follow - I don't want to jump around in the SGF applet to see the position that is being talked about.

Also, looking at the top search terms (joseki, fuseki, kgs, tesuji,) I think an easy improvement would be to offer a position search, when you search for one of those words. Also, with regards to auto-linking content I have some ideas. However, in the end, 20000+ pages - with varying degree of detail and different intents - will never be a homogeneous body to search.

But: don't hold your breath. These are only ideas and without help, I may not find the time to implement anything soon.

In the end, I think SL is far from dead :o) We still have a lot of potential. I've not met a single English speaking Go player who has not heard of and used SL. Maybe all we need to do, is finding new incentives and ways to communicate with the viewers & lurkers out there.

Dieter: Thanks Arno for your good humoured thoughts which remind us of the great asset SL really is to the go world. Nothing better than some hard numbers to defy existential remorse.

I also agree that the sgf inclusion is not necessarily more helpful because it is easier.

As you may notice I've done some library work again. I took up go again albeit leisurely. We'll see where it's going.

(later)

Some analysis

Random search

I did 25 random pages. Here's what I found:

In depth search

  • About Ko: search for ko, ko is the third page, the page fills the need but is a bit hard to read due to lay out. From there, more advanced learning is confusing because there are so many takes at it.
  • About Go Seigen: first page in search is the man's page. Splendid page. Digging deeper is also rewarding. Seems we're well organized about this subject.
  • Which go book to read as a beginner? Search yields: recommended first books which is opaque discussion. I'm discouraged here and probably won't give SL another chance. If I am really determined, I do an advanced search. As Dieter, I can readily see the advanced search is made for editors, not for readers. I look for "beginner" difficulty with keyword "books & publications". I get 7 results which look much like a google search result and not some man made intelligence.

Browsing from the reference section

This time I'm a newbie interested in life and death. From the reference section, I choose Introduction to Life and Death. A great body of work, but I'm discouraged by the heavy browsing needed. As Dieter, therefore I turn this page into a path, making it easier to navigate.

Conclusion

Maybe we deshis should focus on user stories and not so much on fragmented Library Work. The more user stories ("I want to learn about ko", "I want to find which book to read as a beginner", "I want to understand the basics of L&D") we can emulate satisfaction for, the better SL will become. It is also more rewarding to work on one topic at a time than to tackle the thousands of stubs, WME-able stuff, obsolete stuff, which may be superfluous effort because who cares what else is in the library if all the books I'm looking for are easy to find?

Bob McGuigan: I like Dieter's suggestion of "user stories" as a way into SL. Maybe construct a sort of FAQ with links into the library. Same thing as Dieter's idea I guess. I've noticed that a lot of these user stories appear as questions on L19x19 when the answers are (easily) found on SL.

Bob: More ... Dieter pointed out there is a lot of stuff on SL that isn't really useful, some of which is the remains of incomplete or failed projects. I'm wondering whether there could be some sort of "Failed Projects" page that gives brief descriptions of failed projects like Systematic Joseki or Systematic Fuseki and why they failed. This might prevent people from recreating the failed project in a different form.

tapir: There is the abandoned template. Btw. the random pages can be customized for more meaningful results. (My personal random page link looks like this [RandomPage|phpwiki:?random=keywords&term=-HomePage+-Alias+-Humour]) Though the homepage part does not work :) How can I exclude some pagetypes and some keywords?

Regarding Dieter's last post: It is the first time someone classified BQM as not helpful. People stubs are useful, they usually tell me there was pro with this name in a certain period - and I have an idea where to look for more.

Arno, we look forward to sgf-editor beta-testing!

Dieter: BQM are absolutely helpful, while they're active. They're not helpful as a reference. I was making an assessment on SL as a reference. Clearly every discussion page serves the purpose of discussion. Reflecting on my analysis, it perhaps doesn't really matter how many pages are not useful. What matters is how many useful pages there are, how easy they are to reach and to which extent they're obscured by less useful pages.

  • (Sebastian:) I agree with this statement. I think the main reason these are not helpful is because they are out of context. I therefore have a radical suggestion: Abandon the numbered pages, and instead keep questions where they fit best for any other future readers. I would start by simply asking new posters to do that. (We can keep the page layout and the numbers for those who like that format and want to keep count.) Next, we'd have a nice community project: Go through the 550+ BQMs so far, first just categorize them and link them from places where they are helpful in context. Then, if there is a good place for the page, move it or its content there. (If not, we might consider deleting or mothballing those pages that aren't useful anymore, but that depends on a separate discussion; it doesn't look like there's consensus that useless pages really harm us.) The beauty of this project would be that it doesn't manifest itself as an ugly construction site but will be useful during any stage, even if it is abandoned before completion.

Andy: The goban pattern searching is very useful in this regard.

Bob McGuigan: I agree with both Dieter's and Andy's points about BQM. The questions, when answered, are immediately useful at least to the person who posed the question. Some are of limited usefulness, such as some of the questions about "how should I have played here?" and the like. Some questions produce valuable information of importance beyond the BQM context. Such BQM content probably should be transferred or copied to a page on a specific topic. The problem then becomes which page and answering it and copying the information requires knowledge of what is in the Library and doing some editing. That process takes place slowly if at all. Meanwhile, position search can allow people to find something in BQM as part of a general search. In thinking about how to improve SL, though, we can only go so far. There will always be people too lazy to make even the simplest search after all.

tapir: The idea to evaluate the value of SL as reference by looking at random pages is somewhat ironic. I just imagine myself opening random files in the National Archives or ordering random books in the British Library - of course most of them won't be very helpful to me. The random pages is not a search tool after all, but an exploratory toy.

There are however many small and medium improvements to the search tools that would increase the overall functionality.

Some ideas

  1. Path functionality - as far as I know Arno is working on improvements here
  2. Awareness of available search tools - overall pretty low, although every diagram and the side bar features a link to the board search, a guide to the advanced find page and to customizing (recentchanges, randompage and similar functions) might be useful here
  3. Path/index quality - I often see odd humour pages linked from "serious" pathes. I do not mind humour pages at all, even if I do not laugh at them, but linking the "finger tesuji" from the technique path feels wrong. Also there are plenty of non-maintained or duplicate pathes.

Dieter: I was only trying to provide an analytic angle to what otherwise would be my gut feeling. There was no intention to upset anyone. I also tried the other angle, by diving into the material with an intention. If you find these investigations funny, I'm sorry for that. What would be a serious analysis instead?

tapir: I am not upset. That people have trouble finding some of the interesting materials is a well established fact, but random page is obviously not the guide to the most relevant pages. This guide may be missing, but random page will never turn to this guide (although you can customize it for better results).

Dieter: The only way to have an unbiased view on what is present in the library is looking at random pages. If the ones that pop up are not the ones you'd want to see pop up, then why are they there? Of course a random search is not the best metric for the usefulness, which is why I also made three functional analyses, from three different use cases. I'd say, if the random component of the analysis is so disturbing, let's just add more functional analyses, instead of focusing on the ironic angle.

tapir: SL has more than 2000 home pages, we will never see most of them, unless we are hitting the random page button hardly. They are not a problem at all. It is however a problem if you are looking for something valuable (directed search) but do not find that, but something else instead. Two proposals come to my mind instantly: how about removing all other keywords from humour pages and taking most humour pages off from "knowledge" pathes.

axd: some observations.

  1. SL is by design not made for discussion (nor for game play). LI19 and RGG serve that purpose. Forums are not for permanent knowledge because (1) the text cannot be edited (by others) and just keeps growing (2) forums get increasingly noisy because they invite visitors to ask the same questions over and over rather than look up the information in the forum - where would that be? all the data in a forum is discussion by definition, searches will highlight discussions rather than information. Attempts to crystallize information (eg sticky threads, [ext] summarizing posts) are just as clumsy as SL trying to set up discussions. But - discussions might become more interesting if true time navigation? of SL is possible (p:8360). Also, inline discussions should be more frequently replaced by answers.
  2. knowledge is becoming increasingly fragmented across the Web because of an accelerating Web technology pace: it becomes more and more easy to create yet another (go) site that will live for a few years and disappear. see Replacing Go Discussions, [ext] rgg. IMO, future-proofness increases when choosing for simplicity over flashy features.
  3. SL misses a more powerful categorization tool such as (1) the [ext] WP Category tool (2) see also my hypertagging proposal. Both tools might help improve "indexing", the tag concept allowing to reach pages from different angles. the current structuring tools are primitive (single-level subpaging is odd and has its obvious limitations, organizing links in pages is a manual activity, paths are linear only, keywords are limited and hardcoded, Undefined Pages is huge, hyperlinks in diagrams try to emulate the navigation of an SGF editor).
  4. what should be fought is the tendency to turn content into lists of links to external (ie volatile) content, as this lowers the quality of SL.
  5. all living systems eventually face death. therefore, a system must prepare itself to be resurrected as from its conception. one of the design choices must be that the content of a wiki must easily be transferable to a radically different system that doesn't exist yet. The zoo aspect of SL might make it less attractive to new visitors and might carry the cancer of SL, but I think this is such a fundamental concept of wikis that we therefore must embrace rather than fight.
  6. I'm surprised and worried about "performance pressure": why would SL need to have to attract a sufficient amount of changes, of editors? The real danger is not the absence of editors, but SL owners giving up the site. Arno's traffic stats say enough! When LI19 introduces book reviewing and there is no reference to SL, I'm tempted to think that this is a conscientious effort to move SL aside.

isd: SL isn't a bad place for game play. I put up a request to play Tibetan Go here ages ago, but nobody has responded yet. :(

[2] It is interesting that while Josekipedia is an excellent source of joseki information, it frequently links to Sensei's for more information about its sources. -Chaz

RobertJasiek: For a comment on "I did 25 random pages. Here's what I found: 20% are confusing pages, forum talk or personal projects in disguise (Bob and JF on readability, External and Internal Ko, BQM ...)", please see [ext] http://senseis.xmp.net/?KoOverhaulOfMaterial%2FJasiek

Update?

(Sebastian:) I just stumbled across this thanks to recent changes, and I'm sorry that I didn't see this sooner. There are a number of good ideas and suggestions in here, above all Dieter's idea of user stories, and Arno's SGF plans. I would like to help with both of these. What's their current status?

Dieter: Well I just picked ko as a user story, imagining I were a relative beginner and wanted to understand more about ko. I found myself very soon to resist the thought it's only an exercise: if it weren't I would have run off in confusion. SL does not give a good overview of ko. Signed comments insisting on non-intuitive terms like ko master or ko monster on introductory pages, paths where the second page is already a rant of ages ago ... Everything the go community says to be repelled by is there. Hence the ko overhaul of material project.

(Sebastian:) Hmm, that page seems to use an entirely different approach: I don't see any focus on user stories; it never even mentions the user. When I read your post, I thought of a [ext] persona approach, where you begin with the user in mind. As for me, I'm scared of that discussion because most of the articles in question seem pretty esoteric to me.


SL as an ageing Wiki last edited by Sebastian on December 7, 2011 - 04:26
RecentChanges · StartingPoints · About
Edit page ·Search · Related · Page info · Latest diff
[Welcome to Sensei's Library!]
RecentChanges
StartingPoints
About
RandomPage
Search position
Page history
Latest page diff
Partner sites:
Go Teaching Ladder
Goproblems.com
Login / Prefs
Tools
Sensei's Library