Wiki/Report of Meeting 2023-12-28

Report of Meeting 2023-12-28

Present: Skip Cave, Ed Gottsman, Raul Miller, and Bob Therriault

Full transcripts of this meeting are now available below on the this wiki page.

1) Skip was congratulated on the huge number of Quora solutions that he has posted in J and the resulting promotion of the J language. Skip also thanked Ed for his work on the images that are working well. There were also issues with the spacing of text in Quora and the screen shots have prevented that spacing from being an issue. Skip is also finding the J Viewer very useful for exploring the J wiki. Not so much for finding new stuff, but more about finding out what is on the wiki. One consideration was the number of results that are displayed. Skip had not realized that only the first 200 were on display. The biggest challenge that Skip has found is knowing what search terms to use to get answers.

2) Ed showed J Viewer https://code.jsoftware.com/wiki/J_Viewer and the addition of Skip's Quora posts to the searchable space. Ed also showed the results of the forum posts. A suggestion was made to reverse the forum posts so that newest is shown at the top, since these are likely to be the most useful. Ed has also include the Max 200 notifications on the button to make users aware of the limitations. Skip suggested showing the number of results up to 200 which then would be a maximum. Clicking the button sends you to the J Viewer page that explains this limitation. https://code.jsoftware.com/wiki/J_Viewer#200.2B_.28Why.3F.29:_The_J_Viewer.27s_Search_Limit The problem is with changing the style requires us to know the night mode setting which is unknown. One solution is to change the height of the button.

3) Ed will need to crawl the new forum archive now that it has been changed to google groups. Skip said that he has had extensive experience with Google groups and offered his assistance if it would be useful. Ed is now allowing one character searches for code. Ed has a challenge in recognizing escape because certain sub-components swallow the escape and do not pass through. Bob wondered if it was possible to do a pass through of events. Raul mentioned that the Event Handling page in the wiki would be useful. https://code.jsoftware.com/wiki/Guides/Window_Driver/Event_Handling Ed said that he will look into this. Also Ed has included videos into the search for J Viewer so another area has been included. Ed was facing a challenge with having the external videos play inside a web view in JQt. Bob felt he had solved this with his video labs in the way that the video links were formatted. Bob will send the information to Ed.

4) Bob showed off work that he had done on the Community pages https://code.jsoftware.com/wiki/Category:Community_C. His plan is to aggregate pages under titles and make them available in the body of the page as well as in the category tree to the right. This will remove leaf nodes from the category tree until at the lowest sub-category where there will only be leaf nodes. There will be additional information on the body of the page to have extra information not available on the tree. Skip asked whether a page could be in more than one category. Bob felt that categories should be easily added, but may require group consensus to remove. Removing categories may break the wiki for some users who would no longer find the pages where they expected within a category. Skip wondered if you could see what categories the page was in. The challenge is that the category tree categories are hidden to minimize noise. The point was raised that the hidden categories could be very useful. Raul pointed out that the non-navigation categories do show up. Generally it was decided that it would be a good idea to expose the categories for the viewer as an aid to research. Raul also pointed out that there is a related change button in the side bar that may be useful. https://code.jsoftware.com/wiki/Special:RecentChangesLinked

For access to previous meeting reports https://code.jsoftware.com/wiki/Wiki_Development If you would like to participate in the development of the J wiki please contact us on the general forum and we will get you an invitation to the next J wiki meeting held on Thursdays at 23:59 (UTC) Next meeting is January 4, 2024.

Transcript

I guess I will start off by thanking Skip for all the work he's done with Quora.

That is amazing, what a resource.

We've had numerous discussions about it and Ed's done a yeoman job of going through things and clearing things up and figuring out ways to show your definitions and things like that.

But just, I mean, that many answers has to have a big effect on people joining the language.

- Well, Ed did a good job too.

A lot of the output from some of those J functions or J verbs are boxed and boxed stuff doesn't work if you just paste text into Quora, right?

It's a big mess.

So I've been doing screenshots, sticking them in there and Ed said, I don't think I can do a screenshot.

The screenshots are working great.

I can see it if you look at the Quora.

They're working great.

- Absolutely.

- That's great.

I mean, 'cause it's a huge mess.

If you got to paste, you know, J by itself is bad enough 'cause they don't do the, you know, the spacing's not right.

But particularly all the array outputs are all messed up.

If they're boxed, it's doubly bad.

- Yeah.

- Well, and skip the trick with some of the box stuff, the spacing, it's not all of it, but some of it, as long as you stay out of Unicode is to switch over to the ASCII view.

'Cause with the dashes and pluses and stuff, the spacing stays the same until you get to Unicode, which is completely messed up because Unicode is multiples of groups of four characters or whatever.

And depending on how you, I mean, the spacing is just completely out the window for J when you go to Unicode.

- Well, the other problem is that they actually only allow one space between characters.

So if you're getting an array and some of the arrays have three digit numbers and others are one digit number, it squishes all the arrays up.

And so that's probably the biggest problem is that the spacing, for whatever reason, they tried to make it so that all spaces are one, one space big, which doesn't work for arrays that have different size numbers.

- Right, perhaps they're trying to conserve bandwidth.

- I guess, I don't know.

- Well, I think it gets even worse if you've got boxes of boxes and structures 'cause your boxes will be different sizes.

- Yeah, exactly right, the boxes are different.

- Yeah.

- It's all screwed up.

- Yeah.

- So the screenshots work great though.

I mean, I learned early on in Quora, I just take a screenshot, paste it into Quora, everything's good.

- Yeah.

Anyway, thank you so much for the work you've done in Quora 'cause I think it's exceptional and deserves recognition.

And I think Ed's device is gonna bring that level up a bit, like it'll make it more available to people.

I think we're at 291.

- I did that.

- Go ahead, Skip.

- I did that partially to promote Jake.

I mean, I started off, I'm like an engineer and I started off using Fortran and doing all this kind of stuff.

And then I heard about, I had to do some character recognition stuff and I learned about APL and I was like, oh, this is great.

So I was using APL for a while, but I had to use a Selectric terminal, Harris Selectric terminal with an IBM Type Ball to do APL.

And I was like, then I had to put stickers on the keys 'cause I couldn't remember where the things were.

So after a while I said, and then when Jake came, yes, that's what I need.

So I switched over to Jake and I've been ever since, I use the, 'cause I use that all the time in my engineering design work.

So it's, I use it just for my basic math calculator.

So, but, and so I just, I basically started off with Quora trying to learn, help me learn more J and I found it was a really handy way to learn it.

And as I just kept going and I got a lot of people who were interested in following me, started people started following me about J.

I said, okay, so I started promoting J and I ever once I'll put it in there, a link to the J, the J websites and I'll put in Arthur, what's his name?

The guy that said, civilization advances by the, by being able to do complex mathematics, without thinking about them.

I think something in that order.

And I put that little state saying in there once a while.

Go ahead.

- Yeah, well, I guess the only other thing I'll use this as an opportunity.

Have you had a chance to see some of the stuff we've been doing with the wiki?

- Yeah, I've set up the wiki, I messed it up once and I finally figured out what the problem was, got it all set up.

And I've been talking with Ed, a little bit about some of the things I stumbled across when I was starting using it and he's been very helpful in pointing out the issues and he's offered to fix a couple of issues I pointed out.

So, so far, I mean, I haven't gotten really into the using it to learn new stuff so much as just seeing what's out there so far.

As I just got it running a couple of days ago.

So, I'm just not getting that speed.

- So, your main window to the wiki right now is as J viewer, is that right?

- Yeah, well, I've been using the wiki up till, you know, till that J view thing.

I said, now I have J view up on my desktop.

I haven't got it running on my laptop.

I use a lot of stuff on my laptop too.

So, I got to get it running another place, but yeah, I'm on talking on my desktop now and I have it running on there so I can click control H and get into it and it works well.

I mean, like I said, I haven't had a chance to really start looking up stuff because I was confused about, I told it, I was confused about, there were only, you know, I couldn't find one of the things I was looking up and I didn't realize that there was only 200 listings in there and it was stopped at that point.

- Yeah.

- And the stuff I needed was there, some was further on down and it wasn't to me obvious that that was, I thought I had everything I had, so I thought it wasn't there.

But now I realized I just have to read, I have to narrow down my search criteria more till I get under 200, then I'm good.

It'll probably be in there.

- Yeah.

- That needs to be clarified basically, but that was just my- - I'll talk about that.

- Yeah.

Okay, good.

- It is, before I switch over to Ed, so he can talk about the stuff he's been doing over the last couple of weeks, is there anything when you were working with the Wiki, either with the viewer or before that, what were the pinch points?

What things about the Wiki were the most frustrating or?

- Well, it's just the semantics, I look up sorting, well, that's pretty easy, but some of the things like, how do I find the given a number, how do I find the index of that number in a big array?

That was something that came up just recently, and I didn't know how to, wasn't able to come up with the right set of questions or the right set of words that would give me that, get me to that area.

And I mean, that's generally a problem because some of the things you're gonna find are kind of hard to describe anyway.

I mean, like I say, what's the index of this number in a two dimensional array or three dimensional array?

- Right.

- Or three of them, what are the indices of that?

That for example, and I tried typing that and it was like, I mean, that wasn't, what's the indices that didn't work.

And that's, so you may need with some natural language understanding front end, AI front end or something.

- AI.

(laughing) - Well, I don't think that's off the table.

I don't know how soon it'll happen, but I think there's a lot of Wikis that it'll benefit for that sort of thing that you can sort of make your search more intelligent.

- Yeah, well, my background for years was speech recognition and natural language understanding.

Right, so my, basically I worked at a company called Intervoice and we built machines that would bank would buy them, put them in there and hopefully get rid of most of their customer service.

So you could say, call in your bank, push one for account balance, push two to transfer your money, push three to, you know, that was the start of it.

'Cause it was back when Bell started using the touchtone phone and the touchtone phone started working during the call, that people realized that they could make a really easy phone tree selections with numbers.

You know, problem was that customers weren't too happy about the 20 question problem.

You know, well, you want this one, this one, down there, the 20th one is the one that you wanted.

You have to wait, here's what number to push.

So we worked, I worked primarily the last several years on natural understanding.

So somebody said, I want my account balance, we can give them their account balance, right?

The problem with English language is people say, I wanna know how many coins are in my piggy bank and how much dough do I have?

You know, I mean, those are all essentially the same question.

This is the problem in natural language understanding.

Next, you know, the machine learning stuff was pretty good about converting speech into texts.

It got how much money do I have spelled out, it was right.

But to understand what that means is a whole nother problem.

You know, 'cause some people say, I don't wanna know how much money I have, I wanna know what check is cleared, what checks have cleared since yesterday or stuff like that, you know, and that was the kind of stuff 'cause once they thought they were talking to a human, that's what they got.

- Yeah.

- So that's my story on professional language.

- Anyway, thanks for attending.

Thanks for all the work you've done and I'll turn it over to Ed and feel free to speak up 'cause it's just free form at this point.

We'll just go work our way through whatever issues we see.

So go ahead, Ed.

And I think you just muted yourself.

Oh, sorry, I can mute you now.

Sorry, I saw a mute in the corner.

Go ahead.

- Let me see about sharing.

So I've done a few things.

So this is a search for slash co and perm.

Perm is a verb that shows up a lot in Skip's Quora posts.

And just to bring everyone up to date in case not everyone has seen it, when you select a Quora post now, you get-- this is leftover from earlier.

You get-- this is a good time to grow the browser.

You get Skip's verbs.

So in here, Skip uses perm.

And if you come down here, you can select perm.

And it shows you the definition of perm, which you can study.

- That's actually a J library.

There's a lot of J library stuff in there, of course.

- Right.

I didn't worry about mean, for example.

I just gave the definition of mean.

I didn't worry about the fact that it came from the stats library.

So the other thing I've done is I'm working my way up to getting rid of this menu.

Since there are only 200 results-- and we'll get to that in a second-- I'm not sure it's necessary.

And what I'd like to do instead is just sort the results, not by relevance, because there is really no notion of relevance with this particular full-text retrieval implementation, but basically by subject, subject line, title.

And what that works out to be is I added years to the forum post titles.

So this is a 2002 J forum post.

You might say, depending on the search you're doing, gosh, maybe it's relevant.

But if it's as far back as 2002, it's really not what I'm looking for.

So I think this might be useful.

And as you scroll down, you get to more and more recent years.

Maybe I should reverse this.

I don't know.

- Yeah, I was going to actually suggest that.

I think if you reversed it, you're going to get more current answers, and you can go back if you want.

- Yeah, well, yes, yes.

I guess that's true.

All right, reversing it is the right thing to do.

- The only thing I can see is if you reverse it, and it puts Quora and Wiki at the top, I guess that's not a problem.

- I don't think that's necessarily a problem.

- Yeah, OK.

- And then Rosetta, and then at the very bottom, Wiki.

So we reverse it, and Wiki will be at the top, which may, in fact, be a good solution.

- Yeah.

- So that's that.

I did add a button at Skip's urging, and it says max 200 results, question mark, exclamation mark.

And if you click that, it takes you down to-- essentially, it's the JViewer page, but I added a note at the bottom explaining the max 200 results result, which we don't need to go through.

Basically, it boils down to if I returned all 10,000 results, it would just be way too slow, and you'd have 10,000 results, which wouldn't really do you any good.

- Yeah.

- But is there-- where's that button?

Is that on there somewhere?

- The button is right between the two search input fields, so it's right here.

- Oh, I see.

OK, OK.

I got it.

All right, yeah.

- I did index all the-- yeah.

- I'm just going to-- is it-- would it be an idea to thicken the border when you have more than 200 results on that, so you see a visual change when that happens?

- Yeah, you need to know if there's more or not.

- I don't have any way of knowing whether there are more than 200 results.

- OK.

- I tell SQLite, give me no more than 200 results, and SQLite obliges.

- Well, if you could say something like, if there's less than 200, don't say anything.

It'll be-- you can say-- you can put the-- don't say anything if there's less than 200.

Just say there's 23 results.

You can do that.

But if it's 200 or more, say there are 200 results maximum-- 200 results, if you-- I don't know how you say this, but you may-- there may be more.

You may have to reduce it.

You may have to reduce it.

- What I was thinking was that I wouldn't even say how many results there were.

In other words, right now I say it in this menu, but I'm thinking about getting rid of this menu entirely.

I don't think I need it.

I think that real estate could be used for something else.

- And I guess what I'm saying is, you might be able to draw more attention to the max 200 results button.

It would still take you to the explanation, but it would only highlight when you had 200 results or more.

So in other words, if you-- - Yeah, just come up-- put a bright-- change the color or something.

Like you say, a bottom border that says, 200 results, maybe more, something like that.

200 maximum-- - All right, let me spend some more time thinking about that, because I think you have a point.

And I will see if I can come up with a solution that is economical but still communicative.

- Yeah, I just think if it's a visual change, it draws your attention there.

And that's all you need.

You click on the button, it takes you and explains what's going on.

- Right.

- Yeah.

- I had the problem.

When I got on there, I pulled up what I thought was a-- I was looking for a particular problem.

So I pulled up and I started searching down through that.

And I realized there were 200 there.

And I said, it's not in there.

Why is it not in there?

I know I can-- yeah, that was a problem.

- Yeah, all right.

What I might just do is hide this, unless you get 200 results.

And then it pops up.

And that's-- - That's another option, absolutely.

Yeah, yeah.

- That's how I do it.

- I do it better.

- Perhaps it's a-- - You say something like, 200 results, maybe more, or something like that.

- The 200 result limit, there may be more or something.

I don't know how you say it.

It's-- I think we-- that's what you need to say, because it's really important when they know-- to know that what they're seeing is not enough.

- It's not everything.

No, it's true.

- Yeah, not everything.

That's where I-- I could spend a while looking, then.

I spent a lot of time looking for that, and couldn't find it.

- Yeah.

- So it turns out you can't just change the border width.

- Oh.

- You have to change the border width, the color, and the style.

So we talked about how you highlight the update buttons.

So right now-- - Yeah, highlight it, Reed.

- Well, I can't highlight them.

The problem is night mode.

- Yeah, yeah.

- So I don't know the colors that are being used.

So I can highlight it, but I can't unhighlight it later, because I don't know what to switch the colors back to.

What I can do is make the button tall.

- Oh, yeah, OK.

- It becomes a great big intrusive button in your face, which I think might work out pretty well.

- Yeah.

- And if I clicked it and we waited a while, the new data would come down, and the button would revert to being normal height.

- Yeah.

- I think-- - With a big button, you can put more text on it, too.

- Beg your pardon?

- With a big button, you can put more text on it and say, hey, there are 200 results shown, but there are likely more, or something like that.

- True, true, true.

Yeah.

What else?

Let's see.

Got notes on my phone, which keeps turning off.

I need to handle the new forum archive, I think.

So I crawl the old archive to get these results, for example.

But we're switching over to Google Groups, and I think what that means is that I'll have to crawl the new archive.

I seem to remember at some point long ago, Chris Burke saying that the old archive would be kept up to date, but I have to sync up with him and find out if that's true.

I'm pretty sure I'm going to need to do Google Groups, which I will do.

- I think you do.

- So I think his idea is that the old archive will remain available, but at a certain point, it won't be updated because nobody will be answering questions there anymore.

- Yeah.

- But I'm not sure whether he's going to do the Google Groups looking backwards, so he covers the last six months to cover most of the boundary area, or whether he's just looking forwards.

- I know at one point-- I do know at one point he told me that we were already on a form of Google Groups anyway, but it was just the front end is the difference.

But I don't know.

- I did notice that when you go to the Google Groups page, there is some traffic there from the last few days.

- Oh, OK.

- So it exists.

I had a restriction where you couldn't do a one character search, but now if you really must, you can search for pound sign.

It was a performance thing, but it's not necessary anymore, so I relaxed that restriction.

- Oh, nice.

- What else?

This is an escape close WD window.

- One comment about Google Groups.

My company, K Consulting, is also a Google partner, and so we have hundreds of companies that we teach them how to use Google Groups.

We teach them how to use-- I don't doubt that's a problem with you, because you've already figured out how to use Google Groups.

But in case you need some help, we do all kinds of Google stuff.

- Thank you very much.

So escape will work if certain fields have focus.

Like I could use it now, and escape would close the viewer.

But if other fields have focus, like an is a draw control, for example, like the left pane, escape does not work.

So I want to fix that, because I really like the idea of Control H to launch the viewer and then escape to dismiss the viewer.

Just make it very easy, very fast.

- And is that true when you define your overall parent?

Is that how you're having escape work?

- Yeah, certain controls swallow escape.

- OK.

- That's all.

- And it's at the level of the overall parent.

- You specify, I think, when you create the window or when you launch the window, escape closed.

- Yeah, that's right.

But it gets swallowed by the subcomponents.

- One of them, at least, yeah.

The input boxes, the text boxes are well behaved.

They don't swallow it.

So I could dismiss it right now with an escape key.

But the is a draw and the browser both swallow the escape.

So that needs to be fixed.

- Again, I haven't done this.

But is it possible to have the is a draw, because it's the name of a component, is it possible to have it do a pass through?

- I haven't checked that.

It's possible.

I don't relish having to do it for all of my components.

- OK.

- You see what I mean?

- Yeah.

But I'm just wondering if there aren't many.

Like, it may be the buttons will pass it through.

And it might just be is a draw that doesn't.

- That's possible.

All right, I should try that.

- Yeah, just to see.

Because I'm pretty sure when you do the explanation of how the session messages work, they play around with that a bit.

And I'm pretty sure there's a way to pass things through if you don't want to work with them.

- Well, you can trap the event at the component level, the control level, if you want to.

You can create a handler for the control rather than a handler at the window level.

- Right.

But it may be for the ones that are doing it properly, they're actually doing a pass through.

But for something like is a draw, maybe you do have to trap it and then send it.

- That's what I'm saying.

- Yeah, OK.

Yeah.

- Yeah.

And that's about it.

I've got a presentation to the British APL Association on January 4th.

They finally got back to me, which was nice.

So I'll just lay all of this out for them.

I'll just do the usual demo and see if they've got any suggestions for enhancement.

- Yeah.

- Let's see.

- Go ahead, Rolf.

- You can override WD handler if you want to do something special in event handling in how J handles dispatches events.

- Really?

- There's a whole page on-- - Sorry, go ahead.

- There's a whole page on overriding WD handler on-- even back to J602.

Just look up WD handler in your viewer and it should find that stuff.

- Yeah, event handling.

So where am I?

- If you look up that WD handler as a word, that should give you some more interesting stuff on this update.

- A lot of stuff.

[LAUGHTER] - And we don't have the old-- let me just paste something in the chat because it's not showing up in the viewer.

- Actually, that's-- it looks like the same thing you found there, so that's fine.

Never mind, chat.

- All right.

Yeah, yeah.

Right, I never paid any attention to WD handler.

I never needed it, so I just didn't know about it.

All right, that's worth looking into.

- Yeah, I think most of the time, none of us pay too much attention to it until suddenly you need that interaction.

And it's like-- - Yeah, right.

- --you don't know it's there.

And then you go-- you find it one day and you kind of store it in the back of your head.

- Yeah.

So that is the whole story.

Oh, the other thing I did get YouTube in.

So if you search for the Goldbach conjecture, for example, you do get a YouTube video.

- Oh, cool.

- The viewer, unfortunately, cannot display.

It can play some videos, but not others.

- I can give you a workaround for that, but I don't know whether that's what you want to do.

- Well, there is the browser button up here in the upper right.

So with that, it'll just launch the browser with whatever's in the web view at the moment, which we are not going to watch.

- Yeah, OK.

[LAUGHTER] My workaround-- and I do this for my labs because they're on YouTube.

I actually scrape the address and then reinsert it in a format that will display within a QT handler.

- Oh, really?

- You can do it, but if you look at some of my stuff on my labs, my video labs-- - Video labs, OK.

- I have to do that.

And the other thing it allowed me to do was I can actually do video labs even with console because all I do then is I launch a window because you can launch windows from console.

So you launch a web browser and then you populate it.

- Right.

- I think that's an old one.

Yeah, I think that's-- actually, that's just a list of the videos.

- Yeah.

- Try looking up Paddleland because that's where my-- I mean, I don't think I've done anything to explain it.

But it should show up in labs.

I don't know whether it maybe doesn't-- - It should.

- Oh, there's GitHub Labs Math Catalan.

- Oh, yeah.

- Yeah, OK.

So if you click on that, the first thing it'll show you is what I have to do-- oh, no, that's Roger's version.

Oh, maybe mine doesn't show up there.

Interesting.

- It should.

Oh, wait a minute.

MathMisc?

No.

Here's Catalan in MathMisc.

- Yeah, try that just to see.

- But it's an IJS file, not an IJT file.

- Yeah, that doesn't make sense to me, but-- no, that's not it.

- No, yeah.

- Yeah.

- But it is possible.

Is there anything you can send me?

- Yeah, I can.

- I appreciate that.

- Yeah, yeah.

No, I'll send you links to the different files.

- That would be great.

- And the biggest thing, I think, is actually in the IJT file.

I have to end up loading down three files.

Well, no.

Is it three?

Yeah, no, no.

It's two files.

One's an IJT file, and the other's-- I'll have to go back and look at it.

But essentially, what I do is I look at what you're running under, and then I can make adjustments so it shows up in a video.

And I use JQT to display it.

Yeah, so I'm sure it works.

- OK, interesting.

- Yeah.

- Thank you.

- You're welcome.

- So that's all I've got.

- Well, that's a ton.

- Yeah, that's great.

So all these are links to web locations.

There's nothing-- none of these-- [AUDIO OUT] --are stored on the machine itself, right?

- Everything on the right is coming off a server somewhere.

- Right, right, right.

- I actually did experiment-- [INTERPOSING VOICES] - Sorry, go ahead.

- So the only thing stored on there is the data-- only thing stored on your locally is your database, essentially.

- Exactly, so structure and index.

- Yeah, yeah, that makes perfect sense.

- Go ahead, I'm sorry.

I'm done.

[LAUGHTER] - All right.

- You're done.

- I'm done.

- Oh.

- That was awesome.

Awesome, absolutely.

- I'm going to look at share screen.

I should probably figure out what I'm doing with my Safari first to make sure I'm in the right area.

Yeah, that should work.

Let's see.

Go back to Zoom share screen.

This is an example of what I did.

I haven't done very much over Christmas.

But what I did do is-- actually, I started today doing it-- is working on the community category.

And the way things display, of course, we've got the category tree down the side.

And that's expandable.

And of course, when you don't see the arrows, then you've got pages.

And I'd like to reduce the number of pages so it cleans this up a little bit.

And the way I've done that-- and you can see the gray arrows here-- I created a category for conferences, promoting Jay and Ken Iverson, because all these pages actually fit within those categories.

So what I can do that way is I can go in, change the category that this is attached to, so it's no longer just generally attached to the community category.

It'll now be attached to conferences.

And that means that you'll see these arrow categories here.

You'll also have the community category.

You'll also have these lines down here that have the general title and then just a one-line description of what you're going to see in there.

And that'll be your access into those groups of pages.

Now, I think probably what will happen is when I get a line leveled down-- like, for instance, if I click on this, you can see the pages that are in here.

And that'll still remain true, except that when I click on conferences, you'll see these conference pages show up.

So you still have a way of going directly from this page.

But if you were more interested in looking at the category as a whole, you would click on conferences here, and it would take you to the conference page, and you'd probably just see a list of the pages.

So it's a little less-- Sorry.

I really didn't understand most of that.

I'm sorry.

OK.

Can we start with the outline on the right?

Yeah.

OK.

So starting with the outline on the right-- oh, where did that go?

[CHUCKLES] Is that kind of going to-- Oh, right.

Yeah, I know.

What happens when-- interesting.

I ended up covering it up with my visuals, because when I click this, you see it moves horizontally as well.

Yeah.

So the page we're on right now is the community page.

Right.

And it's a community category.

So I click on here, and now I've got these subcategories.

And the gray ones are new subcategories.

New in what sense?

I just created them.

Is that new?

There's nothing in them yet.

OK.

My plan is to take, for instance, these community conference pages.

I'll go into them and change the category they're attached to, so they will now be attached to the conferences category.

OK, right.

And so they'll show up under-- like for now, we've got Journal of J Journal and JOJ, one article.

I should also have a vector in there as well as a link, but I haven't done that yet.

So there's some work to be done there.

So ultimately, there will be no leaf nodes in this outline.

It will just be parrots.

Is that right?

That is my hope.

And then as you get down towards the lower levels of the tree, you'll see possibly only leaf nodes.

I see.

But if we get down to that level, before you get to that one level up, say in forums, I can access those this way.

So I don't actually have to go to the forums page to access the leaf nodes.

Because I can access-- And why would I go to the forums page?

Well, if I go to the forums page, then I do see the leaf nodes, but I don't see anything else.

But that's the same information that I can see in the outline on the original page, right?

It is, yes.

It means that you have a way-- you have a way, like you can see the big picture here, and you can go at least one level down to see lower level pictures.

You don't have to go to the page to see it.

So it is a bit of a shortcut.

I don't have to go to forums to see what's in forums.

What's the incremental value of having the forums page, I guess, is my question.

I think the incremental value of having the forums page is it's possible with all of these, there may be another level below them.

That's not true in community, but in things like essays in reference, you can go down three levels before you get to-- Those three levels will all be visible in this outline, right?

Yeah, I suppose as you click on.

But then you need a page to-- you need a category page in order for it to show up on this tree.

I recognize that there's an implementation issue.

Yeah, yeah.

My question is the user experience.

Does the user need to know anything about these implementation pages that you're creating?

What's the incremental value of them to the user experience?

I guess when you get down to the lowest levels, there'll be more explanations like this of what you're actually seeing.

There's a bit of context that would never show up in here.

I see, OK.

Yeah.

But you're right, there's not-- I have a question.

Yeah.

No, I'm sorry.

I had a question.

If I have a document and it happens to fall under two different top level categories, can you do that?

In other words-- Yes.

--this document fits in this category and this category both, or three.

Yep, absolutely it can, yep.

Yeah, it's-- what is it, Ed?

It's a cyclical digraph?

Is that what you referred to?

It's a digraph, a directed acyclic graph.

OK.

And yeah, the MediaWiki supports that.

And what that means is I can have a number of pages attached to a category.

So the fact that you might fit in one category doesn't mean you can't be in another category.

The challenge there is there's a fair amount of work in doing the cataloging and figuring out what things fit in which categories.

And I think that's going to be ongoing as the wiki develops.

People will find pages that probably shouldn't be under some category.

And they'll go in and change the category to another category.

It'll move to a different category.

But as long as they're doing it responsibly, it should improve things.

And I would probably tell people the best thing to do is not to take categories away, but to add categories.

Because if you take a category away, it will disappear from someplace somebody might have expected it.

Whereas if you add categories, it's just going to show up in a new place as well.

Well, if you have a document and it's under a particular category and somebody decides it needs to be another category, that may be their opinion.

But the guy that put there in the first place disagrees.

The best solution is to have that under two categories rather than letting it move.

Yes.

You're going to make a choice.

Don't get rid of a category.

Add a category.

And what I kind of foresee happening is if we get to that stage where we're really discussing categories, that might be something that we spend time talking about in these meetings so that it's not just one person's decision.

If you want to take a category out, you present the idea to the group.

And then people make a decision about whether that's a good way to do it.

You could notify the owner of the page and say we're looking at moving this category.

Could you do a redirect?

No, probably wouldn't do a redirect because you're bouncing around categories there.

You don't want to do that.

But there are things you can do.

There is a redirect category template.

There is a redirect category template?

Yeah, it's a complicated process.

But it does need to happen so they do support it.

But you see the challenge there is you'd be redirecting every page that goes through that category.

This is migrating from one category to a different category is what this is about.

Yeah, we're thinking about taking one page and moving it out of one category into another, which you can do by just editing a category out.

Redirecting category has a different effect.

Another thing that you can do-- And what happens if you have-- Go ahead.

What happens if you have one category-- I don't know, one document in three different categories?

How do you tell the user that?

He goes down and sees that document and he needs to be able to see that it's under three different categories because he may want to go up and look at those other categories because they may be irrelevant to what he's talking about.

So you need to be able to go back up as well as go in.

Well, yeah.

Actually, I think that's where the JViewer comes in because this is a category tree that you can work your way down the tree, but the fact that you could be in more than one category-- I suppose the answer is if you really want to know that information, you look at the source of the page and you can see what categories it's in.

It will tell you what categories it's in.

But that's a little bit more advanced than most users are going to need.

And I think the more likely is that you'll go down the tree in a different way to find a different-- like you might say, "Well, I want to look at something for tensor fields.

" And then partway down you say, "You know, that might also show up as some form of a-- I'm just trying to think.

There's a guy out of Australia that's doing a lot of nuclear physics and isotopes and stuff.

" And you might say, "Well, maybe there's stuff in that there.

" And you might see the same page in both areas, but you're looking in the areas where you're expecting to see them.

Yeah, no, I think that Skip raises a good point.

It may well, for research purposes, be nice to know that what you're looking at shows up in a category you didn't expect.

Well, there's two things that are kind of going on there.

I mean, we've got hidden categories here for these trees because they're conceived of as navigation rather than information.

But regular categories, the normal categories that we've already been using in JViewer, those show up at the bottom of the page.

You can just click through to them.

But the other thing is content on the page can also be used to find related content, and that's where the JViewer comes in, and this is kind of halfway in between.

Actually, that's a really good point because it's right here.

What links here will show you all the categories.

If you're looking at a document, it would be nice to see something that tells you all the categories this document's under.

So if I go to this page and I go to what links here, well, it's linked to community conference, and it's Eric's site map, which was a web crawl.

But I guess it's not showing me the other categories because they're hidden.

Is that right?

Well, because they're not pages.

Oh, pages that link.

But this is namespace all.

If I do category, what happens?

Look at your filters.

I think this is the filter, isn't it?

No, down below.

You've got three hides.

Hide translutions.

What happens if you click those hides?

Okay, well, I'm going to hide translutions.

It doesn't make a difference.

Okay, so the hide means it's future, not present.

Okay, never mind.

Yeah, yeah.

But I'm just thinking if I go to category now, as usual, we're exploring stuff, which is good.

And I go go.

It says no -- but you see, I think that's the issue of hidden categories.

Right, right.

You need to have -- if this document has three categories, you need to have all of them easily spotted.

Well, the other option is to not hide the categories, which I can do.

And I guess those would just mean they show up at the bottom of the page, right?

Yep.

You know what?

I don't think there's a reason.

Yeah, I think originally I was hiding the categories because they were messing up the titles of the pages.

But I don't think I have to do that.

I can change the titles of my pages anyway.

You just need a separate field or something that just lists the categories in this document.

Yeah, well, and I'll show you an example.

Let's go back to something that's -- oh.

Let's go back to home.

Let's go to reference.

Let's go to something that's probably got a bunch of categories.

Let's do a book.

Let's see where the primer shows up.

So I go to the primer and I say what links here.

Oh, there it shows the category.

Huh, interesting, eh?

Oh, it's because that's -- I've actually made that a category page.

That's why.

Sorry.

Primer is probably not a good example.

That primer should show up under -- should be under learning J or, you know, some more broad category that covers all of the broad instruction material, not narrow instruction materials on a particular topic, but it was a broad instruction somehow.

It should be a, you know, instructional J category or something like that.

Well, in fact, that's actually what it is, Skip.

That is what the books category is.

What you were looking at a minute ago is exactly that.

But in this case -- okay, so now this is a page, not a category.

But it does have a category there that it does show up on that one.

And that's a hidden category.

So I'm not sure what's going on there.

And it also shows up under newcomers.

Is that the only category?

No, it shows up under all these -- it shows under the category books and the category newcomers.

It's in two categories.

Okay.

Bob, I think perhaps this can usefully be taken offline.

The to-do is pretty clear.

It's just a question of figuring out how to do it.

Yeah.

And the effect it will have if we unhide categories.

Right.

There must have been a reason why you chose to hide categories.

It would be good to remember what that was.

I think it was because it messed up this titling up here.

I see.

And I can clean that up a different way.

Back on the earlier topic of incremental value from the intermediate category pages.

Yeah.

There is something that I stumbled over while looking up some of these other things.

If you go to the recent changes pages of a category page, it will show you recent changes of all pages in that they have that category.

Oh.

That's nice.

So this is a book -- so if I go to recent changes, and so what am I seeing here is all -- I'm seeing recent changes, but do I have to filter for the categories?

Maybe it was -- maybe I've got them wrong.

I need to go back to the previous slide.

I'm not sure.

I need to go back to what I was reading before because maybe it's a different recent changes page.

I'm wondering if it's in special pages.

I'll do this offline.

I'll do this offline and email.

Okay.

It may be in this special page and categories group somewhere in here, like tracking categories.

Yeah.

When you're looking for a solution to a problem, when you're looking for a solution to a problem, you know, being able to, you know, if what you're seeing in this document isn't right, the next step would be to look at categories that are in the same area.

You see what I'm saying?

And that's the perfect way to do a search for, I mean, to me, the hardest part of the whole J experience is, you know, I know what I want to do.

Yeah.

And how do I, where do I find out how to do it?

That's the basic problem.

Yeah.

Right.

And so, you know, if I can explain it in some simple terms, it gets close and then keep looking around in that area until I find that area.

That's just, I find the one I want.

And that's, that's, that'd be, that'd be a marvelous learning tool.

I think the solution is going to end up is at the bottom of the page will be displayed.

All the categories of that page shows up in is that, would that solve your problem?

Yeah.

Okay.

I think that's, I would certainly, you know, I would certainly go down there, you know, if I found pumpkin for something and I get something that's close, but not no cigar.

Right.

Yeah.

So the next step is to click on the categories and say, okay, this is all playing to do the same thing or close to it.

So I ought to keep looking and there's a good place.

And that list of categories at the bottom of the page is at a place.

You'd look through the whole page.

You'd see at the bottom, there are other categories.

You click on one of those.

It would take you somewhere else.

Yeah.

Okay.

Perfect.

Yeah.

I think that could be done.

And then I'll, we'll take it offline from now so that we have a, a few minutes to talk about anything else.

Anybody wants to talk about that's what I've been working on.

Yeah.

Okay.

Well, I've, I've, I've definitely gonna.

If you get that run and I'll test it out because I have, I have several issues that I'm trying to solve right now.

This'll be perfect.

Yeah, 'cause these Quora problems, occasionally they get the same old, same old, but about every five or 10, and it's like, oh, how are you gonna do that?

You know?

- Yeah.

(laughing) - Well, and in the short term, I'd encourage you to use the JViewer 'cause I think that's probably gonna give you the best narrowing down options right now.

- Yeah, yeah, yeah.

I'm still gonna use it.

I'm gonna use that.

I'll start using that on a regular basis, I think, as soon as I get my, I'll start working on some more problems here next few days.

Hit me in, Sharp.

Yeah.

- And it was related changes, not recent changes.

That was basically pointing out.

- Okay.

- Ah, okay.

- That's down like three lines or four lines from recent changes.

- Okay.

So related changes and categories filters back out through other things that have changed.

Ah, good.

Good to know.

Thank you.

Is there anything anybody else wants to bring up?

(laughing) Ed and I have our work cut out for us.

(laughing) I haven't made contact yet to talk to Eric or Chris about where they see the wiki fitting into J software in general, but I'll do that to sort of get clarification on that.

And in the meantime, I'm just gonna be working on the categorization stuff 'cause that doesn't really affect the concerns that they have.

- It's still a little bit complicated to get it running.

I mean, I figured out, you know, after the first problem was, it wasn't sure when, it wasn't very clear when the database finished loading.

It took a long time.

And I finally just walked away and came back later and it looked like it finished, but I wasn't sure.

And then I, you know, and so, you know, the installation- - I gotta give that some love then.

Thank you.

You're not the first person to comment on that and not the first person to think, oh, it's broken.

- I didn't know what I was doing and it stopped.

It was just sitting there, sitting there and it wouldn't do anything.

- Yeah.

- What's going on?

Maybe it's still loading, but it doesn't say it's still loading.

- Yeah, sorry, sorry, sorry.

I'll add that to the list.

Thank you.

All right, gentlemen, it was a pleasure.

Take care.

I'll see you in a week.

- Great, thank you so much again.

- Thank you.

- Thanks, Dave.

- I learned a lot.

- Do you want me to put you on the regular list for these meetings?

I can make sure they get set up.

- Yeah, you can put me on there.

- Okay.

- Yeah, I'll, I love to listen in when I can.

I've got a lot of stuff going on, but- - Yeah, yeah, no, but every week you'll get an invitation, your choice if you show up.

Yeah, I'd love to have you though.

Thanks for the input.

Really important.