Future Use of SL

    Keywords: SL description

Table of contents

Sensei's Library hopes to become a central place of Go knowledge and discussion on the web - in true Internet spirit; open to anyone and made by everyone.

To this end, the open approach where anyone can edit and create everything is central in attracting contributions.

Discussions can be on any topic and with contributions from anyone.
Maybe SL can even be used similarly to a newsgroup:

It is easy to imagine the 'question' type discussion
someone starts a new discussion by starting a new page and with a question. Later users read the initial question and comment or answer it. Subsequent users will see all the previous text and can of course comment/correct/add information as they please. This way hopefully the original question will be answered in a comprehensive and complete way. Of course the originator can add more detailed questions ask for explanations on the way.
Discussion on current events
('Bill Gates plays Go!', 'Michael Redmond is 9 dan!') are again simply started by adding a new page and airing one's views.
Announcements
'Pair Go tournament in Antarctica' again, can easily be made. However, this type of typical newsgroup post would not benefit from the interactivity of the SL architecture.

Often, discussions of the first two types will after a while 'converge' and end. In that case, it is possible, and we hope it will be done, to reedit the discussion (called a WikiMasterEdit) to make it more informative and move it or copy it to the reference section. All this takes in theory is to edit the relevant page and add a link to it from the ReferenceSection page.
Of course, this doesn't mean that it is closed for good - just that it has been made easier to read and find for someone looking for just this information. Pieces in the ReferenceSection can be added/edited and corrected as well, remember.

Others may feel that there are clear holes in the ReferenceSection, and some pages may be created specifically for the ReferenceSection. (E.g. the SecretArtOfWar was inserted 'directly').

Hopefully, the combination of these two processes will mean that the ReferenceSection will fill up and stay updated.

Please add your thoughts here or comment on my thoughts above - use the 'Edit' link!


April 2013

I was struck by [ext] this article, although I'm usually not so hot for cutting-edge technology (because of [ext] software rot). Also, it would be interesting if it were possible to navigate a wiki truly back in time, not only navigate in the history of a page. On a tangentially related issue, the recent data loss convinces me (once more) that long-term stability and [ext] data survival is more important than flashy, cutting-edge features.

3 October 2011

Hyperpapeterie: Recently I've been thinking that SL needs categories in something like the way they work on Wikipedia, both for ease of navigation and for library work. Keywords are too coarse: what I am looking for is a tag like "players from the USA" that I attach to a page and then can see a list when I go to the category page.

axd: axd/Hypertagging

tapir: $0.02: Keywords are powerful when used in combination with the search (Advanced Find Page). Likely they outperform more specific categories in this task. Maintaining a category tree looks nightmarish to me given that we have trouble with the keywords already... but maybe it is just me.

axd: Same nightmare as maintaining Categories on Wikimedia (Wikipedia)!

(btw - this erasing of a block bug - I have had it before: it might happen when one is in diff view, and clicks "edit this block".)

tapir: Hope I got it right. I revived the old version and added this back.

axd: yes, thanks, I think this solves it.

Hyperpapeterie: My thought is that we already have a set of category pages. They are just not particularly well organized. See Go History or Outside The Kiins. It's true that many of these pages are quite old.

Towards 2010

Tapir: I feel like we have something to talk here... About the year 2009. All the new features, their success or failure. The overall development and the thesis of Arno's student. Some thoughts:

  • Article of the Week (success)
  • Big Question Mark (less and less questions) - there is a dan player question without answer or discussion at all!
  • Wu Ren Go / Captains Go whatever - the core activity right now. (success)
  • The final migration of chat and daily conversation to other pages
  • Lower quantity (without obviously rising quality) of content contributions
  • Dieter adds the full board position search, which should have been a breakthrough but does not seem to be wildly used.

And to repeat an old question:

  • What would it take to make the pool of (english-speaking) top amateurs view this as the obvious place for their discussions?

More personally I started several non-content pages this year which seemed lacking / to crowd activity to some topics. E.g. the EGD/Wishlist, which looks like a success, but owing to flatline and a number of interested players using it as a kind of obvious place for discussion. Similar, OGS/Status (despite Herman teasing me about it :) seems to be an useful place for dmg to announce maintenance status during OGS offtime. (On other days the OGS forum is surely the more obvious place). But those pages (similar to KGS wishlists etc.) doesn't necessarily create much activity outside themselves.

And pages like Latest promotions - which doesn't intend to inform about latest promotions most of all but to crowd activity to updating player data and adding new players - are far less impressive. Still more or less a one-man-show (by valerio) with some contributions by myself.

But the main problem is the absolute lack of activity on content pages. I did a WME of such a major page like the 3-3 invasion some months ago, but there is still not even a single comment about it. (And I am not even EGF 1d!) And the problem isn't the overall high quality... even a humble kyu player as I still am can notice flaws on quite a number of pages.

isd: I feel that there is a distinct lack of interest in completing the theory pages on the library. There is a lack of contributors with the appetite for this. History too is neglected, I decided to add some local history a topic which won't interest most I suppose. The article of the week is a good idea, it makes the library look busy.

Dieter: Here are my thoughts

  • If we wonder about the lack of high dan contributors, the question "why would they" comes to mind. High dan players are busy establishing themselves through results. They became high dan players because they wanted to do whatever was needed to become one. And it is questionable whether "contributing to Sensei's Library" is a key activity to becoming a high dan. I remember Floris Barthel removed himself from SL and though he did not say why, I believe his teacher Guo Juan told him to, perhaps because she did not want him to spread what he learnt from her for free, but maybe also because he realized the energy was wasted in the light of his true purpose.
  • So we must ask ourselves: why do we? I believe the answer to this is: we want to establish a position in SL. Signed contributions serve to distinguish rumour and clutter from fact and contribution, but also to grow your SL personality.
  • The why do not more? The answer is manyfold:
    • Some people establish themselves elsewhere, at GoDiscussions, or GoProblems, or the GTL ...
    • The diagram rendering may still put off many. The lack of true sgf integration is definitely a minor, despite Arno's great efforts to facilitate the editing. GoDiscussions has the upper hand in this respect.
    • Many people have no need to establish themselves, and few people want to share knowledge for absolutely no reward.
    • Many people though want to get something for free, and I think SL fills a big need as an encyclopaedea of go terms and techniques.
  • The WuRenGo thing is currently what keeps most alive here but it's just an incrowd sort of thing. Comparable things are going on in many of the social rooms at KGS.
  • I believe - as I said before - the future of SL lies in merging or at least heavily interacting with aforementioned competing/friendly sites, combining the WIKI power with sgf power with flash power and getting all the communities together - though as with football clubs, only one community will take the upper hand and survive.

Bob McGuigan: Despite the talk pages and the big and quick question pages I think SL is not an easy place to have a discussion. GoDiscussions.com is more active in that way but it, too, suffers from having too few participants. The same old names are seen all the time, too few strong players participating. SL has an advantage over GD in that its archiving structure is better. Too, I think the wiki structure is an advantage for SL. Another problem for both sites is burn-out of active contributors. A person can only be active and make major contributions for so long before tiring out. The wiki structure means that active participants are constantly fixing questions posted in the middle of well-edited articles or on talk pages in irrelevant places. Sort of like dealing with a swarm of insects buzzing around you while you are climbing a mountain. The position search feature should be a big help in finding articles or information on special topics but I don't know how to make it be used more. So often I have seen people say they looked for something on SL and couldn't find it, but actually there is a lot of information on that topic on SL. The a question gets posted somewhere and someone has to deal with it even though the same question may have been answered many times before. There is also a difficulty in that even dan-level players are not necessarily sufficiently expert to be confident in responding to questions. If you don't think you have anything useful to say the tendency is to say nothing. This would apply to the theory pages in particular. It is no accident that the most active writing seems to be at the lower levels. Most of the questions come from lower level players, too. An improtant thing to do would be to make it easier for novices and weaker players to find things.

Tapir: "constantly fixing questions posted in the middle of well-edited articles" - I feel like this must be an impression from a time long ago. I neither see so much new well-edited articles these days. And despite checking recent changes daily I have not seen such questions for weeks or months.

Arno: I think SL has established itself as a repository for Go knowledge. On a typical week we have over 100000 page views from about 14000 unique visitors (according to my web statistics.) Like with many other social sites, people reading outweigh those writing many times. And of those writing, the top 5-10% create more than 50% of the content. I've read an [ext] interesting, recent article about Wikipedia: the same applies to SL more or less as well, details from my student's thesis are available early next year.

Apparently, the bar to contribute has become too high - either all the low hanging fruit is gone, or wikis are no longer as attractive to contributors as they once were. I don't know. I agree with Dieter that the lack of SGF integration may be partly to blame. I take some blame as well: anyone following my activity knows that apart from the position search I have not really been active for quite some time. I do maintenance work in the background, but the core of the site has not seen much improvement for a long time. I have a mindmap of interesting features lying around for far too long. Navigation, search and interaction with the site need improvement. The design doesn't look too fresh either ...

Reasons, why I have not integrated SGF yet (although I have a more or less finished SGF integration lying around for about 4 years now) are: lack of time, questions about legal ramifications (e.g. people reposting SGF files from commercial sites,) but most importantly: the correct user interface so that the SGF integrates with the wiki. Wiki pages and 10 move diagrams make it possible to read a game/position without stepping through the game, like you do with SGF files. You can also print the page. How would you discuss a position in an SGF file? Like you do on GTL (i.e. at the proper move)? Should discussion and diagrams be below? Should there only be a bare bones support for SGF (i.e. diagrams from SGF instead of ASCII?) And how would changes inside the SGF show up in RecentChanges, so that you can follow what is going on? Should the main branch be protected from changes? I think that SGF and the details of its interface with the rest of the wiki can radically change how you interact with the site. That is why I did not deploy my SGF prototype (it's like a regular SGF editor, embedded in the page, with full wiki markup in the comments.)

But I think, the real problem is that I lack the resources to give the site a technical boost. SL certainly could use it. If I have learned one thing during the last 10 years then it's this: maintenance of a code base for that period of time is hard. Adding features is easy, supporting them is not. Like with wiki pages, sooner or later maintenance eats all your resources.

Dieter: So the message seems to be:

  1. We're doing ok, since we attract a lot of readers.
  2. If we want to attract contributors we should increase the space to move around, meaning
    1. free up some unoccupied territory
    2. refrain from paternalist remarks, "help", guidance, let alone warnings, deletions or (at least premature or overly zealous) Wiki Master Edit

Today, 25/11/2009 I've made some attempts to show uncovered ground, see the front page edits of today.

Herman: I've tried a similar thing with [ext] this edit, but it doesn't seem to have helped much. And that edit uncovered easier open ground than the SL Wishlist or Master edit pages.

As for Floris Barthel, I've asked him about it already. He just lost interest. He was initially very into everything related to go (as were many Hikaru no Go inductees), but his interest has slacked off since then, and nowadays he hardly ever plays anymore (he did not try to qualify for the Dutch Championship this year, for example). There were no other reasons than that.

I think that SL is already very welcoming, and very open to newcomers. They are generally treated politely, their questions mostly get answered. Their contributions sometimes get edited quickly, but usually only to move them to the appropriate place, or wikify them, or merge them into the flow of the article. I think that is a good thing. It shows new users that their contributions are noted, are appreciated, and that we are willing to do some tedious work for them (like wikifying, or finding the right place for it).

tapir: Afaik, a number of club pages has been added since your edit and a number of new players joined the ongoing game. Tsumego just isn't popular. What puzzles me more is the inactivity of BQM.

I see the point, that contributing to SL isn't in itself part of becoming a strong player and sharing knowledge without reward isn't what people do so easily. (Or rather do in their local go club with new members.) But the initial question was not about finding some players to contribute all the high-level material lacking, but about SL becoming a place for them to discuss actual games, new ideas etc. The discussion approach would also encourage to ask questions, discuss own ideas etc. without feeling overwhelmed or intimidated and therefore preferring silence. (This is the reason I am involved mostly with meta pages for a long time, because I feel too weak.)

unkx80: I have been wondering for some time whether we are overzealous in editing newcomer questions and content, albeit misplaced. In other words, a newcomer inserts a question or comment at the "wrong" place, and then we quickly move it to the "correct" place. How many of these newcomers actually perceive such actions negatively, in the sense that established people on SL controls content tightly and/or are hostile to their attempts to join the community? How many of these newcomers know where to find their question or contribution after it has been moved? I suspect such questions are not easy to answer even with server logs of IP addresses. I wonder whether any such research has been performed in other wikis, such as Wikipedia, that helps to answer such questions.

As for tsumego problems, I am rather surprised to find that the godiscussions.com community is much more receptive to problems than SL. Of late, what I do is to cross-post new problems on both places. Each time, far more activity appears on godiscussions.com compared to SL for the problem. Another observation I have is that easier problems at the kyu level tend to attract more attempts than harder problems. One ponderable here: is SL flooded with so many problems to the point that people get overwhelmed and don't even know where to start?

More generally, I think SL is seen as a repository of information, and the mass of discussions and other chatter have moved over to godiscussions.com. As a result, most of the questions are being asked there as well, since they probably get answered faster. This may be a reason why we see a decrease in BQM activity on SL.

Dave: My contributions have fallen off greatly over the last two years (or more). There are several reasons, but two major ones:

  • The difficulty in finding out what is already written up here. This is particularly true for fuseki and joseki items. The content is spread across multiple pages for many of the major lines and we have been inconsistent in linking everything together. It is then difficult to determine whether something has already been covered somewhere, whether it conflicts with another page, etc. I find this frustrating. By the time I trace through what is already on SL, I have run out of time/energy to make the actual edit.
  • The diagram interface. This was so exciting when SL was created but, as Arno says above, is dated now. New material can be developed using Kombilo's editor and exported in SL format. However, editing existing diagrams or quickly adding a variation into an existing page now seems painfully slow.

tapir: Interestingly, technological datedness is referred to in recent discussion in Wikipedia (the german one) as well. How important is this factor? I guess with SL the factor is more about the place of SL regarding to other projects. I.e. if SL adds sgf support (and allows unrestricted upload) it would instantly develop into one of the biggest game databases. This would integrate well with the board search (which is much better here than kombilo imo), but I see some good reasons not to do this. Another question was the support effort Arno told us about. Isn't there a way to collectivise this. Projects like OGS e.g. have more than one developer as well, if someone is needed can't we put an ad on the frontpage?

Herman: Perhaps SL can become open source? According to GoWiki, the source is available on request, but not pretty. I'd be willing to develop on it, and I'm sure there's other go players out there that would as well.

Arno: Sure, I can open the source. I already have the infrastructure in place (tickets, repository, ...) I will not hold my breath though: over the life of SL close to a dozen people have expressed interest in the source or have been offering help. I did not receive a single patch in all that time.

Herman: Well, lets hope that the maxim "Past Performance Is No Indication Of Future Results" will hold. I think we should make it as easy as possible for people to develop. Have a live testing site that is always using the latest revision, regardless of stability, with a copy of the SL database, allowing people to have a very quick develop/commit/test/debug cycle. I think the main thing is to get some momentum. If there are visible results, people will become more enthousiastic, I hope :)

Dieter: I confess. A sudden desire to revive my programming skills was immediately followed by a recurring urge to hold a sabbatical at SL in favour of other interests. The only trace of my promise is a slight feeling of guilt whenever I see Arno's name tagged onto structural improvements.

Herman: Has anything happened with this yet? I'm still interested in contributing code :-)

Arno: I am willing to take applications from interested parties. I need to go through the registration process and setup process with some guinea pigs, before I make the repository public. Drop me an [email] email if you are interested. A healthy appetite for unhealthy code is required.

2007 Dec 8

UncleJam: I would really like to put the current tournament tables on the different pro tournament pages. This way people who follow certain pros could track their progress better rather than just someone popping up in the challengers seat suddenly. The problem is I can't find a good way of implementing tables. Other GoWikis? it is fairly simple, but I guess it hasn't been put in here yet. I tried hacking together a table together using sidebars but it was obviously not intended for that purpose. I could just use monospaced text but it would be a huge pain and look not as good.

2007 March 17

floss: there is no accessibility policy. by "open to anyone" i think you currently mean "open to the default user". sl is voluntary and the maintainers can do what they want, but "true internet spirit" to me means open standards that exclude as few people as possible. i cannot use sl normally because of the buggy unnecessary top bar and the size of the board diagrams and marks. it is reasonable to ask that people consider these matters and possibly make an accessibility policy. i am not asking for special features. all i am asking is that people think about basic functionality like working with all font sizes and more than just one stone size. thanks.

update 2007 may -- this site is still not accessible. all that is required for the top bar is to move one of the links to the sidebar -- that's all.

2004 May 23

Dave: At the top of this page we see the statement:

Sensei's Library hopes to become a central place of Go knowledge and discussion on the web - in true Internet spirit; open to anyone and made by everyone.

To this end, the open approach where anyone can edit and create everything is central in attracting contributions.

Yet in the Fujitsu17YiWang/Discussion, the How To Use Wiki / Editing Freedom, and on this page (among quite a few others) we see quite different opinions about how to go about it. Much of the discussion (argument if you will) centers around the ability of everyone to edit everything - and whether this is a good thing. I believe that there is no absolute answer to this. Equally there is no one "right way" to structure SL. Go is too complex a subject with too long a history and appeals to too many different kinds of people for easy description. A vast array of people have something to contribute here. The Wiki approach makes it possible for all to contribute. The question for the future of SL that we need to come to terms with is:

What does it take to make it appealing for all to contribute?

I think this deserves serious consideration. I personally think it will require some form of compromise in the Wiki approach but I do not presume know what that should be. I think that considering what has gone into making SL already it should be discussed and hopefully a consensus reached. I like to believe it will be worth it. As we approach 9000 pages of content on SL we should admit that we have just scratched the surface of our beloved topic. How do we go beyond what has been done in the last few years. Specifically I would wonder:

  • What would it take to make SL a place where even Professional players would feel interested to contribute (short of actual cash to pay them :-)
  • What would it take to make SL a place where JF would feel comfortable to post his archives? (Personally I am dying for the answer to this one :-)
  • What would it take to make the pool of top amateurs view this as the obvious place for their discussions?

Any suggestions? :-)


2003 November 28

Charles I think anyone reading in bulk recent SL discussion (in particular the 2003-11-20 discussion that was at Metadiscussion) might detect some 'growing pains' here. When such debate reaches the stage of footnote, cross-reference and later interpolation in threads that it has, it is hard to see how to clarify things in place. Hence this fresh start by me on an appropriate page.

I'll try to get through points in summary.

1. Librarians

It may be that SL has too few/too many Librarians (vice-admins); but I'd say that the decisions on that are for Arno and Morten alone, or for offline discussion anyway; as is the question of how they are empowered to act and so on.

2. SL structure

We may not know what structure would improve things such as easy access to material here, and readers' chances of filtering out side-issues; but I feel it is clear that SL is in no sense too structured right now and the future must bring more structure, or else the site will collapse under the weight of the secondary material.

3. Master edits

Bill Spight puts great faith in these.[1] I believe there is a certain amount of the generals preparing to fight the previous war in that. Master edits here are quite rare (monthly?); I think few of the Most Popular pages are ever master edited, which is something to think about; and I see some drawbacks in the assumption that master edits create definitive pages (like the fact that they don't, that they can be assumed to block later edits, and that 'objecting' to a proposed master edit rests on a supposed right to comment on contributions which is detrimental at times).

4. Growth and quality

There are, fortunately, quite a number of routes by which interesting, good quality pages are created here. There are standard problems too: stub articles (like ribbon development of houses along main roads), weak articles on basic topics that are never reconsidered, left-over chat, reference pages that need much more work for completeness and good integration with the rest of the site. I would see these as fundamental things to work on; certainly in comparison with questions of terminology.

5. Conduct of contributors

This could be a painful topic. I think one can ask for some self-denying behaviour, anyway. Book authors, including me, should be reticent about mentioning their works; server admins should stick to factual postings about their servers; those with commercial products shouldn't use SL for more than small notices about them, allowing others to comment; those doing research in go shouldn't use SL directly to promote that or themselves, and so on.

I think that much might be common ground, based on previous discussion and custom here. Where it becomes harder is this: any site of this (wiki) nature will in effect reflect the views of the contributors, and in the long run in rough proportion to the effort they put into posting, and the interests of the site readers don't automatically match those of the contributors. Therefore there is always some need to ask the active self-consciously to consider the interests of the passive user, especially where those are somewhat different.

I also think that self-denial is a good idea in starting discussions that aren't 'matters arising'; there is no great virtue in posting here some view on a contentious topic that would be at home at rec.games.go.

6. Proposals (1)

There has been no recent debate, I think, on Page Type keywords. It might clarify matters to talk about them. Here are a number of ideas.

  • Discussion Page We could have a recognised page type, with actual conventions about threading, and the process of extracting material for document-mode pages (Advantage: deals with currently-perceived needs; Disadvantage: slow-moving, bureaucratic).
  • Menu Page We currently have both paths, and joseki pages inviting you to click on diagrams to get to variations. Adding menu pages more systematically could be a way to open up areas of the site, especially of reference material.
  • Game Page We have an Ongoing Game page type; a dedicated Game page type would cover pages posting game records of all kinds (at 10 moves per diagram). We could actually usefully introduce some format conventions, encouraging the posting of queries and comments on specified plays. (The future of the ongoing games here, a separate topic, could also bear discussion since rengo playing subsided over the summer.)
  • Variants and other games Page I'd like to see material on go variants, and the scattering of pages such as shogi's, clearly identified here.
  • Maintenance Page This type would be useful in relation to any voting procedures, going beyond WikiOrphans, and in organising Requested pages? etc. in a higher-profile way.

7. Proposals (2)

To go further: I'd like to see some page attributes such as Encyclopedia (only awarded to master edited pages), Example, Dictionary, to help separate out good material here according to the intention of the page.

Now, I think this addresses some of the most contentious points in recent discussion. Perhaps these attributes should be regarded as positive marks, only given to well-thought of material, and not mutually exclusive. Some of the issues between Bill and me may come down to a preference on my part to have them exclusive, for clarity: that it is helpful, in my view, to separate out encyclopedic writing from etymology.

The evolution of SL has been exactly that: evolution, rather than revolution. And so it should continue. I'm not advocating massive change in the rest of 2003. On the other hand, I think it is now challenging to say exactly how SL will continue with such success in 2004, given that a certain superficiality has been seen recently and threatens to become the norm. That is, mentally I award many pages a Chat attribute.


mgoetze: Hi Charles. I agree with you on most points, especially about "the site collapsing under the weight of secondary material". The thing about Master Edits is - they simply don't get done. I think the root of the problem is: many people have too much respect for the contributions of others. The original wiki spirit seems largely missing here - that is, if you disagree with it, just change it. Don't start up a long discussion about it at the bottom of the page.

Only if a change turns out to be controversial should it be discussed. Answers to questions should, if possible, delete the question and make the answer part of the main text. Example: at 4463 enclosure, a question diagram followed by an "I suppose..." answer is bad, whereas replacing the question with "If Black plays here then..." would be good. Similarily, at 4-4 Point, there should not be a note under the text "BTW, what's called a joseki in the text isn't really one"; rather, the text should be changed.

I would like to see a more Wikipedia-ish structure with each page having a talk page where people can discuss the contents of that page. The important thing then would be for people to actually implement the results of the discussion on the page itself. What we are doing now is kind of like building a house, and leaving the instructions for building it all over the place. I vote we tidy up the house and put the instructions in a filing cabinet.


Dieter: Well pondered thoughts deserve to be well pondered, so that's what I'm going to do now that I have returned to SL fields. Just one point I don't get is your position towards WME. Do you think the advantage of their being of higher quality outweighs the disadvantage of their disencouraging future edits ? Or do you think people misinterpret the status of something WME'd as final.

Charles I certainly think we should value good master edits; and even possibly give such pages a certain extra status here. I suspect that many pages we'd like to have master-edited (e.g. kikashi) will not be. After all, why work so hard to edit something, when you have to mention some wrong ideas, or deprecated usages (in your own opinion)? So much easier to start a new, sensible page. Well, it just doesn't happen, usually.

And I'm now upset that the WME idea, as the 'ideal', seems to put some people off editing (that is, people now may not edit if they feel they can't do a master edit). It really shouldn't be talked about in that way. There has been plenty of discussion about editing recently, but there is still some obstruction to it happening.

dnerra: I have another suggestion to add. I really think it would be better if we could, at least sort of, delete pages. There are various degrees of "deleted" that I could imagine:

  • A page type "obsolete". These pages would then, by default, not appear in searches, and links to them would be marked. e.g., with a different colour. But other than that, they would have no difference to other pages.
  • A state "deleted". This would mean that no more edits to that page can be made, and following a link to that page will lead you to the "Page Info & History" instead of the latest version.

The main difference to, say, marking all such pages as "Discussion" pages, would be that they are filtered out by default, instead of giving the user the option to filter them out (which most people will probably never bother to do).

Arno: I like the idea of 'obsolete' pages, not sure about 'delete'. I'd do it as a pagetype. Any other comments?

BobMcGuigan: I agree with Charles's suggestions above. As I've said elsewhere I am very much impressed with how the wiki nature of SL allows rambling discussions to be condensed into informative summaries which may be of a higher quality than any single contributor would write. However, I also notice a proliferation of unconsolidated material some of which, at least, is full of errors. Some of this comes from people writing about topics they think they understand but really don't. I'm reminded of the chat that goes on during live pro match games on IGS where we frequently see weak kyu players explaining and criticizing the moves of top pros. John Fairbairn has observed that there are many mistakes in Japanese term usage on SL, so many he despairs of correcting them all. Wiki master edits are very important, I think, as a way of sorting the wheat from the chaff. As for the rest, there is an understandable reluctance to make major changes in what someone else wrote; hence the many comments and footnotes. People should not be discouraged from contributing. And there is also the issue of who is qualified to evaluate the work of others: to make master edits, to delete pages, to brand someone's contribution as less valuable? Elsewhere there was reference to the "core" of SL. Pages in the core should be in some relatively final state, with the contents correct to the best of our knowledge. While I understand the intent of the term "core" I'm not sure what it means in practice, i.e. how one identifies a page as being in the core.


Older discussion

I am wondering what is the right way to think about and develop SL for future users. In particular I am interested in the collaborative nature of the Wiki Wiki approach. It is clear how we can use this in the initial stages of developing the content in SL. However, as stated above, there seems to be the idea that the contents will become 'mature' and be re-edited into a more-or-less final form. This does not seem necessary or necessarily desirable to me.

Perhaps we are partly trapping ourselves with the name Sensei's Library. A library has books. To use an ordinary library to study you go there, take down the books from the shelves, and read them. But SL is different. Here each and every user should not only study the books but also correct mistakes in them, modify and extend the contents, start whole new ones, ask other authors to clarify what they have said, etc.

Should SL become something like a collaborative set of notebooks on Go? One of the hardest things for me in studying Go on my own is to clearly identify my opinion/assessment of a position or concept. Normally it is more like, "OK, if Cho says black is slightly ahead, who am I to disagree..." Even when I make some special effort I virtually never get a second opinion to verify my assessment. I see the greatest potential in SL to provide a more or less permanent (right Arno, Morten? :-) record of work on problems, games, concepts, etc. Each person participating in SL would have the ability to build up a written record of her thinking on Go over a long period of time together with written criticism of that thinking from others. The use of user names together with SL's search capability will make the site much better for this than rec.games.go which is also collaborative but is more transitory in nature despite the existence of archives.

If this is a reasonable approach, then a question arises of how to structure the materials so that they do not mature, rather the opposite of the thread mode/document mode approach. I think that one of the fundamentals is separating the analysis/answers from the problems. If they are together and a significant amount of analysis has accumulated, then new readers will probably read the old materials before working on the original problem, etc. on their own. I have been thinking that a reasonable basic structure for a problem would be:

  • The problem statement
  • A list of analyses on separate pages (newcomers would add to the list, creating a new page with their own analysis and then read others' analysis and add comments where they feel like they have something to contribute)
  • A page with hints if any (separate so that newcomers can choose when to look at it - hopefully after working through the materials and writing a new assessment)
  • A page with a full solution for people to check with as a last step.

What do you think? DaveSigaty


Your comments about reediting (WikiMasterEdit) pages is valid. In many cases, it is not only the 'result' or 'conclusion' of the discussion which is interesting, but also the discussion itself. But the two are not necessarily mutually exclusive. If I were to WME a page, I would, if the discussion warrants it, keep the old page and put it in a ThisPageBeforeEditing place. This has the benefit that users can both read the abridged version and the discussion. Like you say, the benefit of SL over rec.games.go is that the information here is permanent - others can (and should) be able to retrace old arguments, and this is often more valuable than just being presented with 'the answer (TM)'.

Then again, with the current SL, there is only one page I can think of which has been WikiMasterEdited (Dame - and a Dame/Before Master Edit page was made) - with the current level of contributions, I do not think that this is (yet) a problem in practice.

Your suggestion on how to present and discuss problems is sound. It will be up to posters, when they set a new problem, to also create some dummy 'solution', 'hints' and 'analyses' pages to incite subsequent posters to use these. There could be an automatism in the Wiki to do this, but I am not sure whether we want to 'impose' restrictions on the use of the wiki :-)

You could make a 'PostingProblems?' page which gives these guidelines, and make the page part of the UsingAndContributingToSenseisLibrary path :-?

On a more general note: does anyone have any suggestions for how to attract more posters/contributors to SL?

-- MortenPahle


Dave, I agree with you on most accounts. We don't know how SL will turn out, and it's up to the visitors and contributers to make it a place they like to visit again. As SL is rather young, we all are still experimenting with different styles. My hopes are that eventually a set of informal rules and styles will emerge. People are quite happy to copy good ideas. Kind of servival of the fittest :o)

The idea of how to structure problem pages is nice.

--ArnoHollosi


On the issue of attracting more posters/contributors, I think there are two basic philosophies:

  • The 'build it and they will come approach' which says don't worry about the short run, people will naturally be attracted to SL as it fills up and has been around for a while. (On the other hand as John Maynard Keynes famously observed, 'In the long run we are all dead').
  • Go out and convince people that it is interesting and fun to participate.
    • I think the most likely starting point is RGG. It is interesting to compare the use of RGG for GTL and SL. There is a regular posting of the GTL FAQ. The FAQ starts with a short sales pitch, describes briefly how GTL can be used, and then gets into the details. In the case of SL there was the original announcement in January 2001 and a couple of brief mentions in other posts in the meantime. I think that it is time for a regular posting. It should include a sales pitch, how SL can be used, an update on number of contributors/pages, and whatever else we think is interesting. We might even go so far as to create an official SL FAQ already :-) Naturally in order to create the posting we can simply start a page (Sensei's Library Announcement for RGG?) and develop/maintain it collaboratively.
    • Improve the linking from other sites. A Google search shows that there are not many yet. The [ext] British Go Association link is so far quite uninformative. We might ask them to include us on [ext] The Teaching Page as well, right after the description of GTL. SL is not listed on Yahoo! in the Go page, Jan van der Steen's pages, Ken Warkentyne's go links page (no longer being updated), etc.

What do you think?

--DaveSigaty


I haven't contributed much yet, but I have some experience in web site promotion, so I'll chip in anyway...

As a veteran usenet contributor, I'm somewhat critical about regular announcements about specific web sites. It's usually considered bad style (though active services like IGS and the GTL are a somewhat different story). In fact, casual mentions in other posts can be much more effective, especially if many different people do it.

Yahoo is important, but not as important as it used to be. It also will take lots of time time until it really gets in.

I just submitted SL to the [ext] Open Directory Project, which is much more important. The resident editor "ywteh" appears to be fairly active, so that we can expect it to be listed within days (add: listed as of April. 17. 2001). Remember that the ODP data is used by most large internet portals (AOL, Netscape, Lycos, Google, etc.), and literally hundreds of small ones.

Apart from that, I think that the "build it and they will come" approach should work fairly well here. The SL already has enough critical mass to attract players from almost any level. It may also be good to remember that anchoring a web site in the public perception without a huge PR budget doesn't happen overnight or even within weeks. Expect something around 6 months from the official launch until it really catches on.

-- GeorgMischler


Charles Matthews Time I think for some more general discussion about SL.

This site already does some things well: problems and games; discursive themes and also bitty pieces that cumulate; terminology and anything Hikaru.

It does function as an online community, though as in other cases the ratio of lurkers to workers seems to be high.

There's a shortfall of history, in proportionate terms. Since most of us could only copy material from elsewhere, perhaps that's something inevitable.

There's not a lot of Master Editing happening. Perhaps that's a tribute to the rate of accretion of new material. But it's leaving some of the early pages, which perhaps weren't intended as permanent, looking a bit exposed. Users are rightly respectful of existing discussion; but some housekeeping might be in order.

Format problems ... no, too boring to talk about format problems. I edit many pages I find for basics such as grammar and punctuation.

The areas that seem to need attention are: reference, indexing, systematics. I don't hear consensus about what to do. It is however already quite hard to get to know the whole site, without spending time here that one needs for the purpose of having a life. And we are hardly started on opening theory, or serious endgame theory, or rules ...

So, given that the announced role of SL is an online encyclopedia, anyone have structural ideas? Always bearing in mind that the hypertext can accommodate several interlocking systems?


Charles or some other English literate, what does "There's a shortfall of history, in proportionate terms." mean ? Before I start pondering your thoughts. --Dieter

Charles I only meant that while the site is growing, I don't think the number of pages with historical content is growing at the same rate.


Stefan: Part of the problem is that we tend to work in relative isolation instead of on common projects. A starting something and B, C and D adding a bit can lead to the kind of fragmented content that we have today. We could remedy the current situation by people volunteering as 'theme leader' for a topic (L+D, joseki...), and other people acting as 'subcontractors' to carry out part of the To Do list set by the leader. Could also be a way to produce some of the future content in a more systematic fashion.

Still doesn't say anything about the direction we want SL to take, of course.


Dieter:

There's not a lot of Master Editing happening. True. I think I am about the only WikiMasterEditor around and often I feel reluctant about it because I took part in the discussion, risking to let my opinions prevail. The last time I remember someone took the effort to perform a major WikiMasterEdit he got criticized for the way he structured the pages.

Stefan's idea of theme leaders and subcontractors pretty much goes against the Wiki spirit, in that anyone can contribute. We don't want people to ask permission to touch the page, at least I don't. The structure of SL should be so natural and self explanatory that people comply with it in a natural way and if they don't, many others know how to modify the new post according to SL practice.

Yet, there seems to be little consensus. Main reason for that is the small number of contributors. We know SAS doesn't like paths and that you do. Not really a basis for broad consensus.

I have created a PostingProblemsDiscussion page. First I was going to create a PostingProblems? page but then I realized that first we have the discussion then we distill it into a settled page. This looks like good practice to me but we can't impose this way of posting. SL will mostly settle by doing. Making rules and guidelines will not be very effective, and it certainly is not very Wiki.


Charles One reason for me to ask these questions is that I have material written and not published (I'm thinking of a book 'Fighting Fit', not my shape book 'Shape Up!, but also written with SJ Kim). The preferences Stefan has just given do seem to be related to what I was told by those who read the manuscript. It would be easy for me to divide it into 100 pages to post here, and many problems. But people wouldn't read it as intended, as a book.

The same material, organised into some other structures, might be interesting in a different way. As the title suggests, 'Fighting Fit' tried to be a new kind of manual for the middle game, based on analyses at 6 dan level of amateur games.

I also have seven tournament booklets I wrote, mostly with the same co-author.

Without well-considered structures in place here, what would happen is that a kind of tension in the writing would be relaxed in inconclusive discussion. Then master-editing it all would be like writing the books again, only about three times the length ...

So for me the structural question is important, really as an interface between older writing I have done and the possibilities of SL. It would be possible to experiment, and simply see what happens. But my interest in what others think is quite genuine - after all no permissions are required on a wiki, and the point is to learn from others.


Dieter: (slipping out of the library) You may have noticed that I modified a couple of pages describing techniques involving a small number of stones. The way I made them you can easily jump from a move to its follow up or predecessor. Page names become longer but eventually classification explodes anyway. At each point there are links to discussions or to a more thorough treatise. This is the way I like it but others maybe like a lengthy page with lots of comments rightaway instead.

I do not really understand that the reason you would not post your material here would be that the readers do not read it as intended. Time and finance look like more likely reasons. I have no idea how people read stuff at SL. I only know how I would like it to be when I read it. It would be nice if the lurkers would post more comments on how they like it here, what they read and so on. But that's what they are lurkers for ...

I see SL being mentioned more and more on other forums, such as Toriyama's or RGG. However, I know that ardent students of the game find more interesting stuff at Gobase. Truly you can make awesome analyses there, but few people seem willing to share that analysis with other people.


Charles I can state some things, more related to what I personally feel would be good to implement, speaking as a club-level sensei only (but as someone who constantly sees players of all levels), and as author.

I think people who come here expecting a kind of joseki dictionary service have it wrong. Sensei isn't a service industry - Sensei knows the difference between what players want, and what they need, anyway up to dan level. Sensei answers your question, but the answer isn't always in the terms you expect.

And as an author I feel that research is now not so hard - although to research any idea completely is a long task. An author should impose a considered view on a finite part of the research, to make a book. After all, if the book becomes obsolete just because pros start using some other fuseki or joseki, it is probably of low quality.

Stefan: I'm starting to warm to your idea of a discussion 'front end', with just enough reference 'back end' in place to enable efficient discussion. In which case a name change is indeed in order - instead of "Sensei's Library", we could call it "Sensei's Agora". Unless there is enormous goistic marketing value in calling it "Sensei's answer to Oprah's reading clubs".

Charles One technical issue which seems to me to be related. The wiki format here doesn't have an equivalent to the HTML anchor point - which allows one to jump into a page at a particular section. Effective reference to a 'back end' would be helped by something like that - imagine one taisha variation out of thirty on a reference page. Of course one can use reference numbers or smaller pages, but both have disadvantages when one scales up.

Arno: is there need for such a feature? We could use footnotes. All that is needed is to allow that in links as well, e.g. TaishaJoseki#2. Of course, anchors may change or be removed that's why I didn't allow them in links. But if more people seem in need of such a feature, I can be convinced to implement it. On the other hand, it could be argued that such pages should be broken up into separate pages ...

Charles Yes - I'm not sure which choice I prefer. If there are many small pages on a topic the page names will either be long or look strange and technical. You could say that ugly page names can be hidden by the short|long construction from those looking from the 'front'. So maybe that is the correct solution. I haven't used footnotes here - I use anchors all the time in my Gobase articles.


Proposals (3) (May 2007) Groups and Templates

nachtrabe: Some thoughts:

Two things I think we are missing that wikipedia does (not to say we should emulate them in every way, but these two seem to be good ideas):

  • Groups
  • Templates

Groups would be, for instance, a "professiona players group" or a "jeongseok group" or a "korean language group"--whatever we need to define ("professional female women over the age of 40 who live in Korea"). These groups then let people who have a common interest form together to discuss more easily and helps them maintain a target list.

Templates are something that comes up frequently. I maintain a large number of Korean female professional players web pages (as about the only act I do on SL anymore) and its bloody hard to keep them all consistent with each other, even if I were the only one modifying them. The ability to do a template with a standardized sidebar would be invaluable here (so for korean players we would have various romanizations, hangul, hanja, date of birth, etc all tied up in a side bar which would have uniform information in a uniform location).

Arno: I have been thinking about templates myself lately. Should actually be quite easy to integrate, although I am short of time for the next weeks. Can you point me to a description of "groups"?

Steve: I think groups on Wikipedia are mainly represented by Wikiprojects, if I understand Nachtrabe correctly. An example is the Go wikiproject: [ext] http://en.wikipedia.org/wiki/WP:WPGO

Arno: Maybe I'm a little bit slow today, but what additional functionality is needed for groups? The page you reference looks like a normal wiki page. Granted, there is a todo list, but that can be emulated by a normal list in the page. Reading further into the project guidelines, it basically just looks like a social contract. Am I missing something?

Steve: No, I don't think you are. The only distinction on Wikipedia is that the projects aren't in the main namespace, so project pages aren't part of the encyclopedia. Kind of like the forums are distinct from the content on SL. However, I don't think that distinction is important on SL, since a group doing something on SL has relevance to the SL community who use the SL content. So I figure, it may be easier to address this by posting at the Coffee Machine looking for co-workers, and then starting "Pro players project", "Tournament project", "Go history project", "Systematic joseki project", "CGT project" pages. Of course, even a project of one can have a page, and it may be cool to have an index of such project pages so contributors can join existing projects - this kind of focus and organization may help make contributors more effective. To summarize, I think implementing groups may be good, but no new functionality is needed (as you said).

Velobici: Templates sound quite useful. Even just standardizing the position of the three elements: text, diagram and CJK term box would be huge step forward. Templates should help address both content (what data are required for a particular type of page) and display. An example might be that the "professional" template requires: certifying authority, date of birth, date of reaching shodan, current rank, name in native language, and active vs retired, date of death if deceased. Data could be divided into required, recommended, and optional catagories: required, etc. Particularly, if the template forced a consistent display of these elements.

As for Groups...one might speculate that groups exist as a formalization of what were ad hoc collections of people that policed certain pages at wikipedia to ensure that the page conformed to their vision/version. Groups formally enshrine these ad hoc collections of individuals into a recognized organizational body. Quite a change from the original tenets of wiki.

xela: I think templates could be quite useful. As for groups, I simply don't understand why they're necessary. People are already forming themselves into informal "groups" for things like the Advanced Study Room, so there's no need to change how SL works to make such a thing possible. Possibly I'm missing something here (I'm not a wikipedia groupie, I tend to just see SL on its own terms).

nachtrabe: As Steve indicates, I do not believe a group-like functionality would require any additional features. We just create a page called SLGroups and put everyone as a subpage of that.

The advantage of groups of this nature is that it forms an organizational structure for the discussion of things like templates and "metadiscussion" about a given topic. For example: where do we discuss what kind of romanization should be preferred? There are right now at least three separate discussions scattered around on this topic, and creating a new page for every discussion seems redundant considering the nice discussion system we already have in place. Following the programming principle of DRY (Don't Repeat Yourself), it would be nice if we could just automatically have a place to go for these kinds of discussions, knowing that this is the "right place" for it.

The further advantage of this is that it lets these groups indicate in their discussions and subpages what they feel needs working on. So someone who is interested in, say, Korean Professional Female Players (to use a personal example) could check to see which pages are listed as "needing work."

Steve: While we're on the point of useful ideas from Wikipedia, perhaps we can implement the concept of stubs. Again, this doesn't require any major work from the admins: all we'd need is a stub keyword. Anyone creating a new page which needs more attention, could give it the stub keyword, and anyone visiting a page they think needs more info/work that they can't/won't do, can just edit the page and add the stub keyword. People looking for something to do can then simply search for pages with the stub keyword?


See also:


[1]

Bill: I do? I have certainly seen good ones, and I think that most of them have been good. I have also seen not so good ones. <shrug>


Future Use of SL last edited by axd on April 9, 2013 - 10:00
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