Meta Discussion
The future of SL [#142]
(moved from main page)
??? writes: The systematic joseki storm has calmed down, not that my metaphor says something about desirability. Many questions have been raised on the maintainability of joseki here at SL. Somewhere else on SL, someone questions the usage of WIKI for keeping player info up to date, suggesting database technology is far more useful for this end.
What has been going on here recently are the BQMs of Alex Weldon due to his writing a book, many people creating their homepage or go-blog or club/tournament info.
I've also had a look at rec.games.go and noticed that the days of game discussions (mostly thanks to Tim Brent) are now over, and what remains is SPAM, commercial advertisements related to Go and the usual flaming.
Several initiatives are going on in the spirit of SL: the new flavour of go4go, Benjamin Teuber's ultimate go server, the already existing gobase repository ...
My main question is: SL providing an engine for diagrams and discussion, does not seem to have yielded substantially more discussion of game positions, which I believe is the most interesting application. Is the problem still a matter of technology? I suspect that the link between SL-format and sgf is still too obscure for people. Or is the problem one of poor interest?
My secondary question: yes we may find Kogo, goproblems, GTL, gobase, GoGod valuable resources, perhaps each one better designed to do what it is aimed for than SL for all of them, but can we expect any kind of merge in the future? If not, duplicate effort is inevitable, since we do want to refer to players, games, db statistics, tsumego ...
(moved from main page)
Reuven: As for the sj stuff, personally I stopped it because I'm not going to have fluent net access for a long while. But even though the go4go author allowed linking to games in his archive, I belive the little people who did participate (Including me) already realised (on their flesh?) the size of the issue at hand... And that it'd be much more reasonable for a site with a local db such as go4go or gobase... go4go does support comments to most things in a forum like manner and in the future dev list there's a possible joseki dictionary maybe we could ask for a wiki like commentary page between the joseki and the forum like notes? Of course it's not as much freedom in actual vars editing which would probably only be given to relatively strong maintainers or something of that nature... Anyway - A merge may sound like a solution but it's not so simple to put lines between types of content (I've tried merging forums and wiki before:)... Also would like to note a "local project" here on sl, the ugs which according to the plans wants to contain all of the above one day....
macelee: I am currently reviewing exisiting wiki systems that can potentially be used for a joseki dictionary at go4go. My initial thought is that the dictionary should have 2 levels of access control. A few strong players can be given full right to edit everything just as if they are using a wiki; all other users can take advantage of the existing comment system to discuss related topics. The main contents are less likely to be flooded by unwanted noise using this approach. Just my initial thought.
Reuven: Sounds nice... :) I wonder about the location of discussion, if certain vars are discussed at different points, it might be hard for somebody to catch up... But it kinda does make sense to put them where relevant... :S And there's also the how-to-explain problem... Ohh yea, and you write that you aren't planning to make a pattern search just yet but it'd make sense to allow the editors at least to search for patterns and make lists linkable for examples and such... ^^*
(moved from main page)
Arno: Morten and I have been discussing similar issues for some time now. I think we should start a broad discussion about what we would like to achieve. Every now and then I hear that SL is nice for newbies, but does not cover serious stuff. Of course, I don't share this opinion. But it goes to show that the average visitor doesn't have an idea how large SL really is and how much wisdom and food for thought its pages contain. It also shows that we need to improve the navigation even further.
Furthermore, because of its ongoing growth SL also got more "noisy". What I mean by this is that not every edit is related to discussing games or positions. Some people view this as problem as well, I don't. SL is different things to different people.
We have been thinking about the possibility to create different types of pages: reference pages and community pages. Reference pages are written like an encyclopdia. Sober, neutral point of view, no personal opinion. Maybe even with some edit restrictions in place. Community pages is basically everything else: discussions, blogs, home pages, you name it. Although this split is in attitude rather than appearance, different types of pages could have different appearance as well. One could go even further with the split like having special types for problem pages or game discussions. The integration of SGF is in the works. I think it should be ready before summer.
The database (structured data vs. wiki text) issue: I have been giving this some serious thought. I think that a template system, for defining fields and searching in these fields etc. is the solution to this. People would be able to define new templates - basically (in database speak) this would create new tables in the database that could be searched etc. Think of it like Microsoft Access for the web (ok, with less functionality :o), with focus on integration into a wiki.
So where should we go from here?
(moved from main page)
kokiri: I for one would not be very keen on the idea of creating two types of pages on SL as I think it's unnecessary given the recent(ish) move to create parent/sub-pages - the need to differentiate between authoritative pages and opinions/discussion etc was why that move was made wasn't it?
A problem with destruction of content might necessitate 'locking' of pages, but I'm not aware of that at all. As it is, more vigourous WME should make it clearer what is content.
I can well imagine that some of the content (& there is a lot, right) might be hard to find or navigate around for the beginner, but that to me suggests that there is a need for more WME-ing & more linking between pages, especially from pages that are authoritative definitions, to pertinent examples and discussions from the BQMs etc. This is how I would envisage improving navigation.
I guess it's only natural that the editing of old material happens less that one would like - i for one have intended to do a fair amount of WME-ing, but life being what it is, have not got around to it.
I do think that it comes down to the question what is SL at the end of the day.
It's a library right, but does that mean it's a repository of knowledge, or a place where people seeking to learn come together, or both or something completely different?
For me, the points where the library really comes alive are the energetic, messy, in-need-of-a-WME discussions that arise from time to time. We seem to go through droughts and gluts of these, the most recent arising mainly from alex's questions, but it is reading and participating in them that I think that i would highlight to an newby as the best thing about SL, and that here is where the real lessons to be learnt are.
I think that in general, we don't make enough of the pages where interesting discussion has taken place by linking to them from definition pages to which they may offer some insight. This makes the most interesting parts of SL quite hard to find, i think.
Thus, what i think SL needs is, more questions, more BQMs, more heated debate, more WMEing and more links between pages!
(moved from main page)
Bob Myers: I agree with all of Arno's suggestions. Here are some ideas:
1. I do believe there is an important role for SL as the "Wikipedia" of go. In other words, a collection of "reference pages", to use Arno's term. I would not suggest editing restrictions; after all, Wikipedia has none. Rather, what is needed is simply a way to mark the pages as being reference pages and establishing guidelines and suggestions about editing them.
2. Again, I agree with a database solution to structured data. My only comment is that I think creations of templates should be limited to a small group of super-wiki-masters. I would not want to see a dozen different templates for pages about professional go players, for instance. Initially, I think we just need two templates (or new types of pages, if you will): professional players, and terminology. To make this really useful, it is not enough to just allow searching for all pages matching some criteria in terms of the new database fields. There need to be custom reports that present the results in useful form.
Much later, we can consider other applications for the database feature. Events would be one.
3. A lot of what goes on in SL is forum-like in nature. Which gives rise to the obvious question: why not have forum features available, supporting SL features like diagrams, and tied into the wiki format. The very simplest approach would be to have a forum associated with each page, which could probably end up taking the place of the Discussion subpage. Forums have lots of advantages. I can mark threads to follow, even set up specialized RSS feeds to track a particular discussion, search by author, etc .
3, At the heart of SL is its support of diagrams. Were it not for this, any wiki or CMS or blogging software would do almost as good a job. This leads to an obvious suggestion--improving diagram support. Supporting SGF, if I understand what is meant by that, is certainly a huge step forward. However, there are other killer features which could solidify SL's position as the leading repository of, and forum for exchanging, go information:
a. animated diagrams. b. in-place (on-board) editing of diagrams c. searching for positions
4. Blogs are great (although personally I might like an RSS feed that omitted them :-), or was otherwise selective). And SL is a good place for go blogs because of its go-specific functionality (diagrams). However, wiki technology is not that well suited to blogging. I'd rather see blogging-specific functionality in SL. By blogging-specific, I mean typical blog features like easy posting of daily entries, easy-to-read pages with the last few posts, categories, etc.
Just my two cents.
(moved from main page)
LukeNine45: Speaking of blogs: I adapted SL's diagram code into a plugin for wordpress. You can view it in action here: ![[ext]](images/extlink.gif)
http://blog.lukenine45.net/topics/go-game/go-problem/ I could make it available if there is any interest or if people would rather host their own blogs. I agree that more linking and WME-ing would improve SL for people that don't read the recent changes everyday... I love SL!
Arno: could you detail how 3a and 3b would look or feel like? What do others think of having more features for thread like discussions? About 2: yes, creation would be limited to librarians or so. A sidenote on searching positions: yes, I agree that it is an important functionality, which I have been thinking about for a long time, but it is a very difficult problem. For server applications speed and resources is paramount. E.g. I like Kombilo very much and use it myself, but it would not be suited for SL if people started using the search seriously. So if you are an algorithm genius feel free to email me or take up the conversation on search algorithms.
Bob Myers: 3a: Each diagram would have a "play" button, which when pressed would, well, play the diagram. Under the covers, this could be implemented in different ways. There are Java applets which do this but probably a GIF-based script-driven approach is the most generic. 3b: Harder, but would make diagram authoring *much* easier and thus greatly foster the addition of useful content. Conceptually, it would be like having an SGF editor embedded in the edit page.
Migeru: 3b is actually *easier*. The process of opening an edit window and changing a single character of the diagram (say from . to O or X) could be easily automated. Inserting a full empty board at a given spot would always require editing the page manually. For the purposes of recent changes, the first of these automated edits could be a "major change", but all the others "minor changes".
(moved from main page)
P7A77: I know nobody knows me from Adam, but I agree that this site could use a major overhaul, and a handful of master editors. As an extremely anally-retentive person used to working out flexible, consistent, scalable systems (most notably my architectural AutoCAD layering systems), this is a task I would salivate at, then be terrified by the enormity of the task. And as a former graphic designer, I'd certainly like to see a slicker, deeper, more efficient layout. One that encourages more discussion while keeping things tidy, has better categorization and launching areas for beginners, and shows a face of which we can be proud.
Whatever hapless souls end up revamping the site, I hope they realize that a stitch in time saves a zillion. I think the idea of categories, heirarchy, ways of linking, article tone and layout, overall look and navigation, etc etc etc, should be hashed over to death with plenty of trial runs before any major work is done. We want this site to grow for generations. What's a few months (or years) spent working out the kinks? Usually such things are best left to a singular vision, but due to the nature of wiki and the like-mindedness of so many of us here, I think design-by-committee might work out.
...and of course I really want to be involved. :)
P7A77: I hope my last comment didn't seem too critical. I love SL and appreciate all the hard work everyone's put into it. It has growing pains, that's all.
Some thoughts on ways to approach this... first, I think this discussion needs a page (or group of pages) all its own. I feel it's best not to get too bogged down in details too soon. Start with general ideas first. For example, maybe have a brainstorming where everyone can give thoughts on the sorts of page categories they feel should be distinctive (joseki, blogs, game reviews, etc). Then, perhaps, we can all come in with ideas of how to take the scores of suggestions and fit them into scalable heirarchies. Once that's settled enough, then move onto navigational ideas, working out specifics of all of the great suggestions above, etc. When we're agreed on the skeletal structure, then get into things like graphical appearance. And keep in mind I'm coming from a layout/design perspective and know squat about coding. And this is just from my own experience, and what I've found has worked well for me in designing various things.
If I'm sticking my nose in where it doesn't belong, please thwap it.
(moved from main page)
Migeru: it seems to me that many pages are in dire need of the discussion being refactored somewhat. Sometimes there is a short definition of the term in the title, then a very long and informative discussion of how the definition is actually wrong, and no effort to condense the consesnsus (if any) into a more readable article and move the discussion that can't be refactored into a subpage. We have for a long time had a Discussion subpage of every page and a discuss-this-page link. It doesn't seem to be used much, and when it is the top page is not being refactored. Just my 2p.
Bob McGuigan: I just tried the advanced search functions for the first time. I didn't know, for example, that it is possible to search for pages with specific levels or keywords. That makes it a lot easier to find advanced material. When I started out at SL I just jumped in, without reading any introductory material where this advanced search function might have been explained. Now I think that people could get more out of SL if they would read the instructions for use. Any way tobe more emphatic about that on the entry page?
![[Welcome to Sensei's Library!]](images/stone-hello.png "Sensei's Library")