Wiki/Report of Meeting 2024-02-01
Report of Meeting 2024-02-01
Present: Art Anger, Skip Cave, Ed Gottsman, Raul Miller, and Bob Therriault
Full transcripts of this meeting are now available below on this wiki page.
1) Ed thanked Adám Brudzewsky for the pointers that he provided to accessing repositories including GitHub. Ed believes that he has now increased the accessible corpus of GitHub J code from 3000 to 7000. Ed is still considering the challenge of accessing the new J forums on google groups. The interface is through JavaScript and seems to be more suited to populating other google groups than retrieving information. Ed feels that he has gotten to the point where he will get in touch with Chris Burke to see if there are any tools that he has available that might provide the archived information. Ed has sent out the January update and is now in maintenance mode. Bob asked Skip if he had any suggestions regarding google groups since he works with it professionally. Skip is not as familiar with the google groups, but works with experts in these areas and will ask them if they have any suggestions.
2) Bob showed the further work that he had done on the Communities page https://code.jsoftware.com/wiki/Category:Community_C using expanding links. Bob had incorporated Devon's suggestion from last week and expanded the tables and relying on the table of contents for access to the different tables. He mentioned that he would like to put in a return to top on the tables. In the Reference section, Bob was working on Frameworks https://code.jsoftware.com/wiki/Category:Frameworks_R.2 which has a long list of sub-categories and because of this, he felt that it make more sense to have the add-ons heading link to the most useful page directly https://code.jsoftware.com/wiki/Vocabulary/Libraries. The other subcategories would remain available from the category tree for those that would like to explore the topics in a more serendipitous manner. The categories at the bottom of the page would also be a way to find linked information. With a smaller number of sub-categories, such as we find with databases then a table presentation could work as well. The final sub categories may remain as a page of links that over time may be curated into more accessible information. There are more add-ons that have had pages created to link to GitHub repositories, but have not yet been categorized.
3) Raul mentioned that there are many pages in the Systems areas that are out of date and some may never have existed. This is certainly an area that can be cleaned up. Raul found it difficult to find complete information on projects and the IDE's. Raul had included a comment on the Standard Library/Overview page.https://code.jsoftware.com/wiki/Standard_Library/Overview
4) Bob mentioned that he had exchanged emails with Eric and Eric feels similar to Chris regarding the landing page presentation. This means that aside from the Newcomers page the pages will adopt a more conservative approach. For Newcomers pages, Bob feels that there is a real advantage to having more colour to make the page more inviting. What remains most important in these pages is that the information be clear and correct and the Newcomers pages will be an ongoing development opportunity. Eric also mentioned that Videos and Labs have the most opportunity for community growth. Ed wondered if there was more direction provided and Bob suggested that that would probably come from a discussion at an upcoming meeting that has yet to be scheduled. Videos may give way to some form of interactive diagram. That is a direction that Marshall Lochbaum has gone with the BQN language https://mlochbaum.github.io/BQN/doc/functional.html and it has been well received. Another approach might be live coding videos that can last hours. Ed pointed out that those let you get into the mind of the programmer.
5) Ed wondered if there were categories that have multiple parents. Bob felt that there were pages that had multiple parents but categories generally have only one parent in the tree. Ed is exploring a flat category page by the W.W. Grainger company that presented their information in a flat, but very comprehensive and extensive catalogue most recently https://www.grainger.ca/en/category/c/1 older versions are less complete https://web.archive.org/web/19961225175510/http://grainger.com/ Raul mentioned that there are export pages on the media wiki https://code.jsoftware.com/wiki/Special:Export . Bob also showed that category pages can be investigated according to their trees through special pages. https://code.jsoftware.com/wiki/Special:CategoryTree
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 J forum and we will get you an invitation to the next J wiki meeting held on Thursdays at 23:59 (UTC) Next meeting is February 8, 2024.
Transcript
And what's going on in Ed's world?
Oh, gosh.
So, Adam showed me how to do a query in GitHub that would find all the repos that have J files.
So I did that.
And I think we're now up, we went from 3000 J files.
So financial institutions like to talk about assets under management, AUM, it's the standard measurement of their metric track.
And I like to think about documents under indexing.
And so we've gone from 3000 GitHub J documents under indexing to, I think, 7000, thanks to Adam.
So that has been achieved.
The bad news is, it looks like there isn't any clean way to get archived Google groups, posts out.
They have a very JavaScript interface when you go and look at documents, when you go and look at posts.
And unfortunately, there is an export mechanism.
There is an export mechanism that the admin can use, but it exports to a format that nobody seems to know how to read.
It's not entirely clear what its purpose is.
It's one of these write-only files.
And maybe you can use it to repopulate another Google group.
But I'm still sort of admiring that part of the problem.
So that is the major flaw currently in the whole operation, is that the archive on forum posts stops on the last day of 2023.
I wonder if you could enlist Chris Burk's help here, because, I mean, he has full control of administrative control of the system.
Maybe there's something he can do.
I think I may have gotten to the point where I actually need to bother somebody to whom I can offer nothing.
I've long felt that we're all sort of carefully balanced towers of neuroses, and my base neurosis is not wanting to bother people.
So I have resisted getting in touch with Chris Burk.
But I think we may have gotten to the point where I have to do that.
As I say, there is an admin export function available to the Google Groups administrator, but it doesn't seem to export to anything that I would be able to parse.
But yeah, I think the time may have come.
So I will probably do that in the next couple of days.
And with the old system, I think Chris had told me that he was using Google Groups behind the scenes with it.
He was, apparently, but he was still populating, perhaps by forwarding, I'm not sure how, the old jsoftware.com forum archive.
And that's what I was going off of.
I was not going off of Google Groups.
And I don't think he's any longer doing that.
But I guess the question I've gotten, it's a question for Chris, is whether that could be done just to populate your records.
What, for me? Just alone?
Yeah, because he had been doing it in the past.
Yeah.
And he was taking it from Google Groups. Now he's just taking it from one source of Google Groups. He's not splitting them up.
Yeah.
So it might be fairly easy for him to channel it into something he can look at.
I'll present the problem to him and see what he says.
Yeah. But if that works, then that might be the easiest because it'll be something he's done before.
And it's something I've done before.
And it's something you've done before. So you're back working on what you're used to.
Yeah, that's possible.
Okay.
I'll compose something and send it to him in the next couple of days.
Yeah.
And then Adam had a couple of other suggestions, the Discord channel and a couple of other things that I haven't looked into in any depth yet, but I will.
Yeah.
I was going to say the Discord channel could be tricky because it's bridged to Matrix.
And I know a lot of times Zima is working to try and make that bridge better and more stable, but it sounds like it's a little bit creaky.
I did look at an archive and it is not parsable by me as far as I can tell. It's pretty involved on the inside.
I haven't given up. Adam had a couple of suggestions that I'll be looking into. I'm sure I can find something to work on. And as I said, we'll talk to Chris or send Chris a note.
But at this point I'm very much in maintenance mode. I sent out the January update, which had one great quote, "The J-viewer is a work of art." I'm going to carry that with me to my grave. That may be on my epitaph.
Yeah, I thought that was great. I thought it was also accurate.
Thank you. You're very kind. I didn't even make it up. It was an honest quote.
Yeah, it was good.
Yeah. So that's all I've got. Things are very quiet at this point.
Okay. And Skip has just come on and is connecting, I guess.
Oh, I see. Yeah.
And there's the man.
Hey, Skip.
And Arthur as well.
Skip, we cannot hear you.
You're muted, Skip.
On Zoom. There you go.
Okay.
So Skip, anything new for you at this point? Ed just finished talking about some of the stuff he's challenged with regarding J-viewer, and I think the biggest one is probably trying to get the new forums working.
But he may talk to Chris Burke about that. But I think, had you mentioned in the past that you'd had some stuff to do with Google Groups?
Yeah, I mean, my company is a Google partner, right?
Cave Consulting does Google consulting primarily businesses and lots of government cities and counties kind of things.
And we basically come in there and come into these cities and show them how to digitize all their documents.
That's probably a big step, convert all their paper documents to digital and then show them how to use Google Drive and Google Documents, store them in Google Documents, and how to set up a mechanism whereby you can do the retention schedule.
Because every city has to have a document retention schedule.
You have to know when to get rid of documents, how long to keep them, how to find them.
And all that gets complicated when you have, you know, 100 file cabinets full of paper documents, as opposed to everything is nicely filed in Google with the correct metadata for every document.
So you just look up the date of the document or the contents of the document with a search and you can find any document within 10 seconds.
So that's what about messaging systems, messages between different people?
Yeah, yeah, we have, you know, Google has, you know, a couple, you know, they have a Google Chat, which is a, you know, which is a secure messaging system.
We use that a lot.
You know, Google Chat is basically like, you know, like just a regular post, a phone, you know, a telephone text, but it's all, it could be web based.
In other words, you can do it all.
You know, you can do it from a browser, basically.
In fact, if you have a Chrome browser, you know, if you have a Google account, you click up in the corner, there's besides email, there's also a chat icon and you click the chat icon and you can set up a chat group, you know, and, you know, you can have all the people that are on the chat group can, you know, can see everybody else's chats and when they post, everybody gets the chats, that kind of thing.
Right. And is that same as Google Groups or is that sort of an offshoot of it?
That's different. Yeah, that's it.
Chat is very much like a text, you know, it's very much similar to text.
Go ahead, Arthur.
I think Arthur just got background stuff happening.
Oh, OK. OK. OK. So, yeah, so that's now Google Groups is what we recommend for, you know, for, you know, conversation, you know, involved conversations where you want everybody to see the conversations and be able to respond to certain comments and that kind of thing.
It's more organized than chat.
You know, chat just, you know, everybody sees everything you say. Right. That's chat. And so you can.
But in Google Groups, you can, you know, you can separate have certain side groups and, you know, and you can have certain people who are the moderators and some people can moderate a group and, you know, and kick people off or let them on that kind of thing.
Are those records that the councils and the municipalities have to track as well?
Well, yeah, yeah. And the municipalities, I mean, all of their documents, that's that's totally different.
All their documents, you know, they may they may get public information requests, for example.
You know, somebody says, I want to know how much you guys spent on that last road road project. Right.
And we make a public information request and they're required to provide you provide the citizens that information.
Right. And so they have to they can go back, get that information, give the give the the citizen a link or attach the documents.
But normally just give them a link to the documents, the specific documents in Google Drive. Right.
And they go click on the link. They read the documents and then they can make their comments and stuff like that.
And if they make a comment, they can either email the council person or or the secretary of the city and say, well, I think this is bad.
Or they can come to the city council meeting and they can complain about what they're spending.
That's kind of thing that that but that's that's primarily not a group thing.
That's more of a, you know, of a people ask for information.
Yeah, yeah. No, the reason I keep coming back around to groups is is Ed's trying to figure out the way that groups might provide archived information so that if he's trying to grab the information coming off a group and pull it in.
Do you know anything about those formats.
Oh, so you want to save the information on the groups and put it somewhere for for.
Well, I mean, of course you can always you can always you can always highlight copy and paste to Google Doc and once you're in a Google Doc, you know, then you can.
Yeah, no, this is sort of, this is sort of more of a mass amount because what he's been doing in the past has been taking the J forums and putting them into his into the J viewer by putting it into an SQL SQL light I think database.
So what we're looking at is do you know anything about the format that Google groups would store their group, their, their huge amount of information, so that you could access that programmatically and pull it out and put it into an SQL like database.
I think I'm gonna have to talk to my one of my other tech experts that specialize I mean I, I use I don't use Google groups regularly and we have a couple folks who, that's all they do, you know, and they have, you know, they set them up for for our clients.
And I think there's a way.
I think there's a way to search Google if you're, if you're, well, if you're a member of the group. You can search for keywords in the group, right, I'm pretty sure you can do that.
Now, if you're not a member of the group but you want to, you want to put the group somewhere where anybody can search them. In other words, it's an open search, you know, then that's a different.
I'll have to, I'll have to check and see if that's a, you know, what's, what's available. I mean, if you if you want to be able to be if you want people to be able to see it but not be able to post to the group.
First of all, if you don't mind and post the group you just make them a member of the group then they can search anything they want in the group but you want to categorize stuff, you know, you know, take stuff out of group, put it someplace else, a nice categorized place.
You know, maybe a topic, all everything's about this particular topic, you know,
so I'm just I'll have to, I can find that out. Oh, I now I kind of see what you want I can make it. I can talk to the group, Google group folks and see exactly what they, you know what the option is but I'm pretty sure export.
If you get an if you get a chance to that that would be great. I think it'll be talking to Chris about it because Chris is the administrator for the group, but essentially, like when you're talking about being able to search and Google groups the advantage.
It has and putting it into an SQL like database is that that sits on your computer which is why it's so darn fast.
Yeah, that's true. That's true. And you don't have to worry about connectivity.
Yeah, I mean, I do an integrated search. So you do one search and you hit the J archive and GitHub and Rosetta code and all the other corpuses that I yeah yeah if you're looking you want to you want to span the search across all various categories of data, you know that are J data like you said all those list.
That's not a search within groups. That's a new group.
I need an I need a parsable export of a Google group to which I belong.
Yeah, okay, that's okay.
That's a reasonable, I think, I think it's a reasonable goal. And we're not scraping we have the permission of the administrators.
Google does not make it easy to scrape Google hides the data behind JavaScript.
Yeah, admin admin privileges.
Definitely.
Okay, so we got it. We got a good group admin privileges and we want to put the group data and SQL database basically, and we want to explore and be able to export it out and I can I can.
I don't think I've ever had anybody asked that question.
It's not something that most municipalities are going to need. It's really true.
That's about three levels above the expertise of most anybody in the in the municipalities that we do it.
So yeah, okay, I think I can find I'll find that out and I'll let you know I'll let you all know I'll skip. I very much appreciate that. Thank you.
What have you got in terms of community pages community pages. Yeah, let's see. I'll share a screen.
Not awfully, not an awful lot that's really new, other than the fact that I put into play.
Pretty much the suggestions that were done.
The last time.
So, what we were talking about before.
Notice that I always have this thing show up right in the wrong spot. There we go.
On this side.
We've got the regular tree of all the stuff that that goes into it with the different pages in this case.
This actual I'm just looking to see I don't think this actual page has any.
No, it doesn't. It's all pages it's not subcategories later we'll show us what it looks like was a subcategory. But in any case, what you end up with is this which is a clickable link that takes you to the same position as this forums link does.
And this information is essentially the same as what you get when you go to the forums page, slightly different format but essentially you're doubling up.
But it's more accessible right now.
In this level, and it may be in the future having the separate forums page allows it to be more searchable.
In terms of, there's actually a category that's forums and you're not just looking at community to try and figure out where forums is. So essentially I think it just aids search, and maybe narrows what you're looking for in that case.
But the this distribution.
I'm just going to look at this.
Okay.
This distribution.
Essentially, it just gives you a quicker access to the pages that would be in your categories.
And I did go to previously. These links here were hidden in the table so they didn't show up as a table of contents, and I moved them out of the table on the headings and then they show up in a table of contents, so you can click and go directly to whatever
section you want. They do I should probably have a return to top.
That's something I'll think about is putting in a return to top next to these two, so you can pop back up to the top quickly, because some of them are pretty long.
So that's what I put into place with the community I like that format I like the way it works. I will put in a return to the top I think.
But then I started to look at it in terms of the next section which is reference that I was going to work on. I started looking at frameworks.
And this got interesting.
Because when I go to the add ons category for frameworks.
I have a well you can see I've got a huge number of pages attached to this category. So if I just did it by page.
When I go back to here.
I'd have an incredibly long open list now.
The, this will allow me to the table of contents will allow me to access through that list quite quickly.
But add ons for instance I think would probably be almost several pages on here, vertically of descriptions of the add ons, and essentially it's a duplicate of this page, which is much more useful, because this page has all that information already broken out.
And so I'd be duplicating what this page really has in a way that's going to cause a lot more work, because I think this is the place that it should be updated.
And so to go back to here, rather than put all those pages in a big file down below or a table down below. I decided just link this right to the add ons page, because I think that's the most direct way to do it.
If for some reason you're looking to do research on it.
And you really want to get into all the pages linked to the add ons category.
This will do it.
So you'll get access to all those links, but you don't get the access through the page itself. So that is a change from the way I was doing things that I've got a link on the page that will take you to the information you're probably interested in.
And then on the frameworks on the category tree. I've got pages that are more about exploring the categories and the, and the links that are have been put in.
So, any thoughts about that.
I think that makes sense. I'm a big fan of flat presentations but there does come a point when it gets a little ridiculous so.
Yeah, I and I just think it's it's, it gets to be so much work to duplicate which already there.
There's so much chance for error in in something gets updated one place and not another.
But I think it will take a little bit more explanation to people that what you're seeing in the category tree is a link to all the categories that have been put in all the curation.
And what you're seeing on the page is probably your most useful link.
So depending on what you want to look at.
So in the case when skip was talking before about having the categories, not hidden, which is a direction I'm going is I won't hide categories. So at the bottom of the page all the categories will be available.
The category tree is actually an even quicker way to get to all the categories you might be interested in.
And so that's maybe the way that I'll present it is it's, it's almost like the categories, the bottom of the page on steroids because you can see in this case, we do have these here are all categories, these all open up into their own links.
So they really do become an area that if you're looking across categories. This might be an area that you really would be wanting to look at as much as the bottom of a page that you've been looking at sort of different purposes but but there are different ways to the information.
So the word frameworks is really categories, what you're saying.
Framework is a category.
And it has subcategories, which are add ons and databases labs open plot publish all these have categories window driver has just pages.
But, like for instance publish. Yeah, pages.
Just trying to think of one.
So, some of them are pages databases isn't pages which is why it looks different on the page is I thought well it makes sense with this number of.
This is what the page looks like. But makes sense with this number of subcategories, just put them here. And then people can go straight to what this subcategory would be.
If that's a quick way to get to what there are again, it's information that might be interested in.
The page, showing all the subcategories.
But this would show you all the pages that are linked to it.
And I've been putting this placeholder in, and I'll clean these up so there's maybe a bit more informative placeholder, but it may be for especially in reference for a while that mostly what you're seeing are these links to pages because that might be the most
useful way to present it.
And go full on curation and try and describe what you're looking at, which is Bob, there are about 3000 pages right. Yeah, and that neighborhood.
Yeah.
A lot of information.
Yeah, and the challenge I found going into it and it's remained is a lot of information is out of date, but it's still good information. In other words, if you looked at the particular technique that might be used is very valuable.
But if you were looking at the particular application, it might be done with J 604 and completely out of date.
So, in terms of somebody trying to do research unless they're at a level where they're willing to overlook the implementation, and just look at the, the approach or the algorithm.
If they're not there it's probably just going to confuse them more than anything else.
We need a project to upgrade that stuff.
Either upgraded or what I did was a lot of it as I moved into archives.
And my sense was in the archive category.
It was put there.
Just to be a catch all so that if you go in there it's almost a there be dragon situation. And there's nothing wrong with having archive but it having a version which is more directly useful is probably a good place to be.
So if it's if it's valuable information just a matter of pulling it out and updating it. That would be the best.
But when I went go to archives.
This is the split I had on the archives, so I got subcategories here.
You know for these different, different pages.
As you can see some of them are quite large because there's a lot of stuff that is, you know, archival frameworks is one of them.
Having add on docs that's archive only is best. Well, in the case of the add on, say for instance that you know the, the art category.
It's no longer, we no longer have it for.
Well, what what what happened was john baker has moved it on to his, his into God. That's essentially what God has become here. Now, in terms of things like another place I was looking was API's but don't see them here.
There's been a whole bunch of API's that have been added to the
added to the add ons, which have not had pages done for them.
And so as a result, they're not categorized yet because there's no page to categorize.
Having said that, the most recent or the thing that I was doing to get rid of the wanted pages a few weeks ago, I did create pages, I should have categorized them and I didn't.
But if I had categorized them those, those add ons would start to show up in the category page, but they would just be links back to the GitHub, which might be useful. It's better than that's a good starting point.
Yeah, I mean, that's what I remember doing on the add on pages we wanted to have a documentation page but when we didn't have that the GitHub pages.
The GitHub pages better than nothing.
So that was kind of the whole structure of the add-ons page
is to deal with that in between state of affairs.
Yeah.
Yeah, and I think that is a good approach to add-ons
because that's essentially where the live add-ons live.
And if you're looking for documentation,
whether or not the wiki is up to date,
the GitHub should be up to date with the person that's
creating it.
And it's even worse for the system.
We need a system page similar to-- there's an orphan system
page, but it's really horrible.
Yeah.
And most of those files don't exist.
One of them may never have been a file.
Yeah, which is a cautionary tale for those who
are looking to update wikis.
You go into areas that may never have existed.
And the one I was looking for isn't there.
So I was going back to looking at the source code
to try and figure out what the workflow was
on that system file.
Yeah.
Project.
If you do open project, it's part of J. It lives.
It's well supported.
I think Chris Burke uses it.
I think Bill Lamb uses it.
If he doesn't use it, he supports it.
There's no documentation on it as such.
Or there's a few hints and pieces for older versions.
But I think that's adequate or complete.
Ah.
And have you done the JViewer search on projects?
Because I thought there was a section in one of the areas
that Chris actually talks about projects.
I do remember seeing that.
And what I remember of that is that it
covers a tiny fraction of what's in projects.js.
Well, that's true.
It's not complete.
But there is a little bit there that gives you
hints about what you can do.
No, I haven't been able--
when I look for it in JViewer, I didn't--
I was looking for this stuff in JViewer.
And that one, I didn't stumble across the most recent time
I was on it, my most recent pass.
Yeah.
If I had to think about it--
The other thing is also the IDEs, QT, JKT, JHS,
those are system things.
Standard lib, I guess, is a system thing.
There's probably a few other system things.
I saw your comment on standard lib today.
Yeah.
No, that was on the system page.
OK.
Or I think it was on the system page.
Hopefully, it was on the system page.
It's an orphan--
I've made a comment on an orphan page today.
I hope.
Yeah.
It was standard library overview.
Really?
Yeah.
That's what I got it as.
Let me look, because that doesn't match my memory.
[TYPING]
Won't be the first time I made a mistake.
That's a fact of life for me.
[TYPING]
Yep, standard library overview.
Yeah.
See if that's what I thought I was working on.
Yeah.
OK, I guess it is.
Yeah.
See, this isn't the add-ons page, which is also
a standard library page.
Well, this is an orphan.
This is an orphaned page.
And all of the scripts here are system scripts.
It's near a second tail, except for unixlib.ijs,
which is a verb and not--
I don't know if it was ever a file.
Even standard lib, which does exist,
isn't found by that name.
Anyways, lots of oddball things here.
Yeah.
Now, the thing I was looking at for the reference page,
the reason I started working on the frameworks,
was I actually think what I'm going to do--
I've had an email exchange back and forth with Eric.
And I hadn't heard anything back from him
about the way things looked.
But he says he's pretty much in the same camp as Chris,
in terms of how many icons we use and how that goes.
So my sense is-- and we'll have a meeting about this.
I proposed having a Zoom meeting in the next couple of weeks
to talk about this, to figure out
what the best way forward is.
But I think probably what we'll do
is go to some format like this for reference pages
and everything but newcomers.
I really will make a strong argument that for newcomers,
we should have it look a bit more friendly.
It'll be two things to do with newcomers.
One, it has to be Balmer.
It has to be, as much as possible,
be able to present information clearly and accurately, which
may not be true with other parts of the wiki.
That's an aspirational goal in other parts of the wiki.
But I think in the newcomers area,
it's actually probably closer to a requirement.
One thing I would caution there is you're not actually
using icons, you're using emojis.
OK.
Because if you're using icons, they
can be replaced with other imagery.
And because you're using emojis, you're
limited to the set of emojis that currently exist,
unless you do a major up overhaul in how you're
presenting information.
And I think that's a key thing to be aware of when you're
talking about updating that.
Like a J icon is something that we have,
but it wouldn't fit there in its current form.
With an icon, would it not be able to be
presented as an image?
I'm just wondering whether icons are the preferable way to go.
Well, if you're talking about something that's--
you're talking about responsive design, varying fonts,
varying text sizes, stuff like that.
You've only got one version of the icon.
It's standardized.
With emojis, you have an entirely different situation.
Yeah.
And it's going to be--
especially if you've already tested it on multiple platforms,
it's a different testing experience.
Yeah.
Yeah, you're right.
But in any case, that's this sort
of a framework you see on the screen,
or this presentation on the screen,
is closer to what you would see on pages in the wiki.
And in the case of the newcomers,
they might be a bit more graphical.
There might be a bit more verbiage explaining
what you're looking at, things like that.
Because I think it's necessary for newcomers.
And that's without talking about developers and that kind
of stuff, because that, again, is another part of the category
tree that, at this point, I've isolated,
because I wasn't really doing anything with it.
And I'd rather have it there for future development
than to have it in and sort of detract from the wiki.
I should add two comments on my last little spiel.
One is that, even if you are using icons,
you can do responsive design and have multiple sizes
and stuff if you do a different approach.
And also, if you went with SVG, that
could also give you a similar character,
but a little more resizable.
Yeah.
And you knew that, but I just didn't want
to leave it without saying.
No, that's right.
And with SVG, you retain the sharpness
across the different sizes, which is really nice.
You don't get in between sizes that way.
But it's still a different design approach
than what you currently have.
No, you're right.
Yeah.
Yeah.
Anyway, so that's where we're sitting now.
And then the other thing that Eric put forward
is he's really interested in developing further video
resources and labs.
He thinks those are the two areas that would most benefit
growing the community.
And so--
Bob, I wonder what he means by video resources.
Did he give you any detail about that?
I don't.
I have that level of detail, and it's an excellent question.
Because I've worked with video, and there's a lot to that.
There's a lot to unpack there.
But I can-- really, I think the most effective--
we're thinking in terms of getting information across.
The most effective way of thinking about video for me
is almost a just-in-time approach,
where your videos are limited to probably no more than 20
or 30 seconds.
They're very short, but they're very to the point.
And then you do end up with a lot of them.
But you basically have a way of searching through and finding
out one that addresses exactly what you're looking for.
You click on that.
In 20 seconds, you see how to do it,
and then you're back off it again.
That's a lot of work.
So you think about 20 seconds of video.
I mean, you've been in the business,
so you know that a speech, 30 seconds of text,
30 seconds of speech is not even a paragraph often.
There's just not that much content there.
And if you're out to deliver 30 seconds of talking
and a few J characters, it's not clear to me
that video is the most efficient way to deliver that.
No, it's a good point.
I think almost what I'm thinking of-- and this actually
goes back to a discussion I had on a podcast with Marshall.
He was pointing out that actually he
finds most effective diagrams.
And I really agree with him about that.
And for that kind of thing, it's possible
that what we're really looking at
is some type of an animated diagram
that will present the situation, make a few moves,
and complete the situation.
And so it won't have voiceover.
I mean, that's one thing that you're--
honestly, across languages, if you can avoid voiceover--
although English is a very common language.
It's--
And getting commoner.
Sorry?
And getting commoner.
And getting commoner.
But still, I think you're much stronger
if you can avoid having to rely on language.
True.
Certainly true.
OK.
I was just curious what Eric might have meant by video,
because it covers a lot of possibilities.
Absolutely.
Yeah.
And it depends on--
well, it depends so much, as always, on any of these media.
Depends so much on your audience.
If you haven't defined what you're
trying to communicate to your audience,
you don't really even know what your starting point is.
In terms of length, in terms of the way you present it,
in terms of where the challenges are.
So yeah, that'll be very much part of that discussion,
I think, to find out how he sees that being used.
And this point, I'd have to say--
and I guess to date myself, the MTV generation
is a reduced attention span.
And you're nuts if you do anything longer than two
minutes.
And if you can do it in 30 seconds, it's even better.
Right.
The only things I've seen that work long form
are the Twitch approaches, which are unedited.
And basically, you're able to sit in on a conversation.
And they're recorded--
Live coding.
Yeah, live coding, those kind of things.
Yeah.
And it may or may not be useful to you.
Those are great, because you get some insight
into the thought processes that go into the final product,
which I think is of value.
Yeah, it's tricky.
Yeah, and I would agree--
So does media-- oh, sorry, go ahead.
I was going to say, I would agree with you completely
about that.
But that's a case where if you were trying to solve a problem,
you probably don't want to drop in on a live coding session
unless it's interactive.
Yeah, right.
Well, yeah, if you have a question,
you want to ask somebody.
Yeah.
Yeah.
Yeah.
Yeah.
Sorry, you were saying?
So there are about 300 categories.
Is that right?
Going by my memory, but I think that's
probably in the neighborhood.
It might be slightly higher.
Like, I think it's between three and five.
Oh, really?
OK, and it's still the case that there
are some categories that have multiple parents, right?
Categories that have multiple parents.
No, I don't think so.
Pages that have multiple parents, but not categories.
Pages that have multiple parents, but not categories.
OK, OK.
And do you--
I could dig this up out of a SQLite database,
but I'm just curious if you know.
How many page occurrences are there?
In other words, if you count up all the dupes,
it's about 3,000 unique.
But how many, if you counted, if you navigated all the categories
down to the pages, how many occurrences of pages
would you reach?
I had thought that actually your 3,000 is probably
without dupes.
There's more than 500 categories here.
OK.
Yeah, I was thinking in terms of categories,
part of my category tree.
But yes, there's all the tag categories, right?
Those exist.
Oh, the tag categories, right.
Those are a different story.
Yeah, yeah, yeah, yeah, yeah, yeah, yeah.
Right, right, right, right.
No, there are hundreds of tag categories.
Yeah.
OK.
All right, let me think about that.
And I guess MediaWiki doesn't have a generic export function
that just dumps everything onto a web page in some form, huh?
Everything--
Well, I'm thinking of the ultimate flat representation.
I'm just wondering what it would look like.
I think I mentioned in an earlier conversation
that the WW Granger catalog went online
and had hundreds of categories in just one home page.
And it was just a wall of words.
And people-- I don't know if it was--
I don't remember exactly if it was flat
or if there was categorization.
But these were people who had internalized the WW--
it was for users, customers who had internalized the WW Granger
catalog already.
And this was familiar terrain for them.
There were no monsters.
And it worked.
I mean, it shouldn't have worked.
By all the standard usability metrics, it was horrific.
But for that audience, it was a good answer.
I don't know if it was the best answer,
but it was a good answer.
And I guess--
I was going to say, I guess with Granger,
they are making changes to it.
It's not static, right?
I assume not.
But I assume the changes are slow.
I assume it's evolving rather than being revolutionized
from hour to hour, day to day, month to month.
You're familiar with the export pages feature of MediaWiki
right?
No, I'm not.
Let me just verify that it exists on our site.
It's not extension.
On any given page, you can do an export in a PDF form.
If you go--
I'll just paste the link into the chat.
There's chat.
[AUDIO OUT]
On the link I pasted into chat, there's
a option for all pages.
Right.
I guess what I was thinking of was some sort of--
maybe even like a tree map kind of a thing.
I'm not sure what it would look like.
But a scrollable-- I mean, something like the hierarchy,
Bob, that you've got in the right margin here,
but laid out in two dimensions, not just one.
And complete all the categories, all the page titles,
everything.
And I'm assuming MediaWiki can't do that.
That's kind of a hard problem and probably not
solvable in the general case.
[AUDIO OUT]
But I'm curious.
I mean, JViewer doesn't really give you a flat view.
It's still navigable.
I just wonder what a flat scrollable view would look like.
Well, I think it depends on how many levels
that you were able to open up.
But what I've done is--
Well, I think they would all be open.
Yeah, yeah.
But I'm just saying in this case, I can open them.
And then you can start to see.
Yeah, but I'd like to see page titles too.
Interestingly, this export feature
doesn't seem to be working.
[LAUGHTER]
But I-- or maybe what I've done is that's the problem.
OK, I'm going to change it to pages.
Here we go.
So again, this is based off of what I've categorized, right?
There you go.
But essentially, what you're asking for
is provided through the category tree.
If you give whatever heading you want of the tree,
the root of the tree, and then ask for all pages,
you're going to get something like this.
Right, and then my question would be,
what if you opened it all up at once
and laid it out not linearly as it is now,
but in two dimensions?
And you had formatting for category, subcategory,
sub-subcategory, and so on.
How big would it be?
How long would you have to scroll to get to the bottom?
And I just--
I find that an interesting question.
It would be pretty big, since that's just--
No, but keep-- yeah, but keep in mind that--
yeah, I know, it's 3,000 pages or more.
I get that.
But maybe you could get eight page titles per line.
So that's maybe only 400 lines, which doesn't seem too bad.
The thing-- when you're talking about--
this is a case where there would be pages
that have multiple categories.
And so there is--
Yeah, I get that.
--difficulty, for sure, yeah.
I get that.
Yeah, so yeah, the number of brides--
It would be more than 3,000.
Yeah, it would be more than 3,000, for sure.
Yeah.
For sure.
There is the special all pages.
But it paginates that, right?
I believe it does.
Yeah.
Let me check.
And when you're saying two dimensions,
this is actually technically two dimensions.
But what you're saying is, rather than showing it
as a line, you'd show it as a table.
So you're making better use of the horizontal dimension.
Number of columns.
I wouldn't go so far as to assume
that I know what I'm talking about when I say two dimensions.
But I do mean better use of the available page area.
Yeah, so in that case, for this add-ons, all these add-ons,
this might be four or five columns that would all
sit underneath here.
Yeah, right, right, right, right, right, right, right,
something like that, something like that.
And Granger wasn't even columnized.
It was just lines with circles separating different category
pages.
It was not laid out in tabular form.
It was just a dump.
Wow.
It was incredible.
And I just wonder, I just wonder.
And for all I know, they don't do it that way anymore.
That may not have survived the '90s.
But at one time, that is what they did.
Yeah, when stuff gets big, those kind of approaches
tend to blow up systems.
So they might--
Yep, yep, yep, yep, yep, yep.
And I'm also thinking somebody who
might be new to the system, like so many people are,
these days when they're looking at information,
they've never seen it before.
They would really be at sea with that.
But then you'd have some type of a search mechanism, I guess,
that would help you out there.
Search, you could also have lenses.
So you could say, all right, just
highlight the bits of this that are of interest to me
as a beginner, or of interest to me as a statistician,
or of interest to me as a, you name it,
graphics programmer, or whatever.
Yeah, which is kind of the direction I was going
with the category trees, is that newcomer
says it's on subcategory.
And if you look at essays, there's
a subcategory for statistics and combinatorics.
And--
Yeah, that's great.
But it's not flat.
Right.
Yeah, yeah.
Yeah, something to think about there.
Yeah.
Anyway, I guess what I'm saying is
I'm looking towards becoming more
pragmatic about the approaches that I take as I look
to basically bridge over to the new look of the wiki,
not using as much emojis and looking towards more text
in terms of the reference, communities,
and things like that.
And probably transitioning the work from the presentation
more to the curation, which is a substantial amount of work.
But I think that's where you start building in the greater
value of the wiki.
And curation may include upgrading old information
to make it useful.
But that'll be a question to--
in this meeting that I have with the J software group,
to say where they want to go and what kind of resources
they want to put into this.
Because it's a substantial amount of work.
And it would have its uses.
But it's whether or not that's where the best use of the energy
is being put.
Having said that, this whole wiki exploration and stuff,
I think, has been hugely successful
and has spawned a lot of really great things
that may not have come to pass in the community.
But in terms of the wiki, then it just
becomes a question is, where does the wiki fit into that?
And how much resource do we want to put into that?
Well, I will be very interested to hear
your report from that meeting, which I suspect will not
be a single meeting, if I had to guess.
Yeah, I don't know.
We'll have to see.
I don't know how clear their vision is, to be quite honest.
If they have a very clear vision,
it could be done in one meeting.
But again, your question about the video,
if they're not clear about what video can do
or where it could be applied, that
is multiple meetings, for sure.
Yeah, right.
Cool, anything else?
Really nothing that I've got.
Good.
I have a note into our Google Groups guy.
I just sent him a note.
So we'll see what he comes back with.
Tell him we need to extract the data of the Google Groups
into SQL, see what he says.
We'll see.
I'll let you all know when I find out.
Super.
Thanks very much.
Skip, thank you very much.
Yeah, that's very kind of you.
Yeah, and we'll work with Chris on that
as well to see what--
as the administrator, he might have some ideas as well.
But whatever you find that would be useful for us
that would be useful because it's a different viewpoint.
And there may be new solutions that he might not be aware of.
Chris is also new to this game, right?
So I'd be very interested to hear
what Skip's expert has to say.
Roel, do you have anything?
Are you still working on simulation?
No.
I actually am.
Some of the rabbit holes I was talking about today
are things that I went down looking
at documentation of things that I was
trying to use in that context.
Cool.
I'm glad that you have not abandoned it yet.
Not yet.
That makes me happy.
Excellent.
All right.
I was able to put JIG into JHS.
I like to do it smoother so that I'm using a function key,
but I haven't been able to crack that one yet.
But because I think it involves some JavaScript,
and I'm not--
I'll have to bring my JavaScript up and figure out
how JHS's interaction goes.
I'm learning that.
But I have built it so that if I--
I don't know if I could show it.
Let's see.
I don't know whether I've got a working--
basically, what I do is if I put a command in,
I change the whole interface to JIG.
So whatever information you put in
comes back as an SVG with its interactive.
And then if you type in off and hit return,
it takes it back out to the regular text version.
So you do get access to it that way.
But what I really would like is to be
able to highlight a line with a cursor,
hit a function key, and just have it pop up.
Right.
Right.
Right.
And I think JHS actually would provide,
in terms of what we were talking about in terms
of visualization, I think it's actually a really good platform
to do that because there's a lot of stuff that's
built in that's interacting with JavaScript,
interacting with SVG, and interacting with J
through different components of JavaScript and HTML
that we may be able to do some stuff with that.
But again, that's another step beyond where I've gotten so far.
But I haven't hit dead ends yet.
And--
It's always good when there are more--
sorry.
It's more flexible than JQT, which
I remember in our last meeting you were talking about,
don't get bound by your tools.
Well, JHS is more flexible than JQT is.
Yeah.
I'm glad we have more ideas than we
have resources to pursue them.
I think that's a good sign.
What is it?
Your reach should exceed your grasp.
Is that what it is?
Yeah.
I've never understood that statement.
It seems to me it could go either way.
Your grasp exceeds your reach would be equally sensible.
I don't know what it means.
If you grasp it, you've got it.
Yeah.
So you should be able to reach for further than you
can actually pull back in.
I think it's-- is it macaques, the monkeys?
One of the-- how you trap a macaque
is you take a glass jar with a narrow neck
and you put a betel nut in it.
And the macaque sticks his hand in, grabs the betel nut,
and cannot remove.
And his grasp, which equals his reach,
he has no control over as long as there's
a betel nut in his hand.
He is hardwired genetically to hang on to that betel nut
at all costs, even though the hunters are coming for him.
I don't know where that fits into grasp exceeding reach,
but I always think of it when somebody mentions that.
It's more like your ideas exceed the number of resources
available to implement them.
No, that makes sense to me.
That makes sense to me.
All right, gentlemen, I will see you all next week.
Skip, once again, thank you for sending off that question.
And I hope you all have a good week.
All right.