Wiki/Report of Meeting 2023-07-20

Report of Meeting 2023-07-20

Present: Art Anger, Ed Gottsman, and Bob Therriault

Full transcripts of this meeting are now available on this wiki page. https://code.jsoftware.com/wiki/Wiki/Report_of_Meeting_2023-07-20

1) Ed led off with a demo of the current J wiki Browser https://www.youtube.com/watch?v=kJwj4Cz7yCk working in the JQt environment. We have entered the beta phase on this application and will be reaching out to add testers particularly for J'ers working on Unix systems as we have not had as much experience implementing on that platform. Contact me at bobtherriault AT mac DOT com if you would like to be added and we will put you on the list. We are planning to have a wiki page set up to provide support. A review with Stephen Therriault will be happening in two weeks time.

2) There was a discussion about wrapping up the prototype wiki in the coming weeks https://code2.jsoftware.com/wiki/Main_Page. Consider this an alert that if you have created information on the prototype you should really think about transferring pages across to the current wiki. This is most easily facilitated by the Prototype - Current wiki links in the sidebar and copying and pasting the mark up of the particular pages.

3) Bob gave a quick update that progress on the Primer is slow but moving along https://code.jsoftware.com/wiki/Help/Primer/001Start_here.

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 23:00 (UTC) Next meeting is July 27th, 2023.

Transcript

First item off is you were talking, Ed, about wanting to do some form of a demo to basically introduce, I sent an email back to you saying there's also a possibility of opening up some kind of a wiki page for documentation I think would be a good idea.

Yeah, and I started to reply and then I thought, I'll just tell the man.

Yeah.

So I think probably that whatever email, announcement email or recruiting email that we create would turn more trivially into a page for a page in the wiki on the viewer.

So I'm not too worried about that.

I don't think that'll require a significant additional effort.

I think we'll get that more or less for free just by doing other stuff that we know we have to do.

So I agree with you 100%.

I just don't think it's going to be a huge additional effort.

It'll be an incremental effort, if that makes sense.

It makes perfect sense.

The other thing I was thinking that it could be used for is as information comes in, it can expand.

So it becomes a resource for people doing stuff.

Absolutely.

And you alluded to the notion that there may be platform issues that we have to start documenting.

I hope not, but it's overwhelmingly likely that there will be.

And so I certainly agree with that.

Well, from what I've heard about Unix, each platform has its own little idiosyncrasies and the people that use them learn them and then they can make them work.

But if they don't know the specifics, then they have trouble with that.

They have problems, yeah.

And the people themselves, I'm sure, develop their own idiosyncrasies too, unlike us Macintosh folks, who are all perfectly normal.

Yeah.

So yeah, I guess what I could do, I actually developed a script, because I wanted to make sure I covered some important, what I regard as important points.

So I could just go ahead and share my screen and do that.

It's between four and five minutes as I timed it.

Perfect.

If that works for you.

Yeah.

All right.

Let me share my screen.

- And can you see my screen.

- It says you need to decline cookies.

Ah, there we go.

- It's more than I probably wanted to reveal.

- Well.

- Let's see here.

- I'm thinking maybe my waist tells me I should decline cookies too.

[LAUGHS] All right, I think I can get started.

OK, so I'm going to show a JQT add-on viewer for the JWiki and JForums that Roel Miller, Bob Terrio, and I have been working on.

In my JQT environment, it's set up to launch with a Control-Shift-H shortcut, so I'll do that.

It's divided into a left half for navigation, which is quite fast, and a right half for loading HTML, which is much slower.

The outline on the left shows the major categories in the wiki, as well as tags, search, bookmarks, the various forums.

And at the top of the outline, NuVoc To fit all of NuVoc onto the screen at once, the valence links are hidden until you hover over a glyph.

And when you see a link that you're interested in, and you can click on it, it will load up on the right.

What's interesting about this style of interface is that you never have to go back that ubiquitous, time-consuming, and disconcerting web operation.

You're always going forward to the next thing.

You never have to waste time reversing direction.

New Vokes ancillary pages are next in the outline.

Clicking on a page label will load up the corresponding page.

The search facility is next.

The way the search facility works is that it will hit both the wiki and the forums simultaneously.

So if you search for, for example, fret, what comes back is, in this case, 20 results from the wiki, including semi dot one, which is expected.

24 results from the J general forum, 60 results from J programming, some in J chat, J beta and J forum.

Results are saved until you explicitly clear them, so you can do multiple searches.

and that will all be retained until such time as you clear searches with the appropriate button.

All right, the forums.

Each forum has a selection menu for, we've got the J Programming Forum, General, Beta, Chat, and so on.

Each forum has a selection menu for year and month.

And if we go to at this recording the current year and month, we can see there are a number of topics.

The way it works is that all of the posts for a particular topic will show up on the right, even if they occur in some other month.

So a lot of threads cross months and we capture that appropriately here.

All right, the bookmarks facility behaves more or less the way you would expect.

I've got three bookmarks that I can load up.

I go back to JProgramming and I'm interested in Eric's post on cloud computing.

I can actually bookmark that.

And if I go back to my bookmarks, it appears and I can load it up.

I can unbookmark it, rebookmark it.

Tags are several hundred wiki categories that are not part of the formal wiki hierarchy Bob has created.

They're alphabetized and arbitrarily gathered into groups of 15.

This is a hit or miss section of the wiki, I would say, but there are a number of gems some interesting rabbit holes that you can go down.

The rest of the outline, starting with Home, shows the wiki category tree that Bob has created.

It is quite rich, quite deep.

The number next to each of the labels in the outline shows the number of pages associated with that node in the category.

And clicking on a page will load up-- clicking on a label will load up the corresponding page.

There were a couple of things I wanted to show.

This is an example of how we deal with very large subtrees.

So we actually break out another outline to the right of the main outline.

And clicking on an entry in that outline will show the corresponding pages.

The other thing we do-- let's see.

Yeah, it looks like I'm not going to be able to show it on this size screen.

Let's try system.

The other thing we've got is that columns have a lens mechanism, sort of a zoom mechanism.

So whichever column you're focused on is full width, which may be enough to show the full size of the label, the corresponding labels.

If it's not, then when you hover on a particular label, it will expand appropriately.

But the column lens mechanism lets us show much more, many more labels on the window than we normally would be able, otherwise would be able to do.

All right, with respect to ancillary controls, the shortcut button will show you how to add a shortcut to JQT.

So if we flip over to the terminal, that's the code that you would use to add the shortcut.

The two buttons immediately above the left and right parts of the interface will allow you to update the local database that gets used, as well as to update the add-on.

Currently, they're both up to date.

It's important to note that the viewer requires a local, roughly 10 megabyte database.

It's updated daily to reflect the wiki and the forum contents, and it's kept in J's temp directory.

It holds the structure of the wiki and the forums.

Essentially, it drives the left side of the interface.

So the database contains everything that you're seeing here on the left.

It also holds your bookmarks, your browsing history, your search results, and a debug log that you can turn on with the checkbox if you want to submit a bug report.

So that's this checkbox here.

On that topic, if you do experience a problem, it may make sense to turn on the debug log, reproduce the problem, and then send in this file, this database file sitting in your temp directory.

Because it contains a lot of application state as well as a trace log, that database can help us more quickly isolate problems that occur in the field.

So the viewer provides what I think is an unusually fluid and convenient mechanism for accessing thousands of pages of J-reference material, as well as countless interesting, I'll call them rabbit holes, that's what I find them to be.

Roel, Bob, and I are in what we believe is the final phase of testing, and we hope to deliver it soon.

That's all I got.

- Thank you.

- Well done.

That was good.

And I think you came in at probably just a bit over six minutes, but that's a good length for a demo to give somebody a sense of what they're looking at.

- All right, well, good.

I wanted to be thorough.

- Yeah.

- So that's, yeah, that's the story.

And as I indicated in my email, and Art, I'll just repeat it.

My main concern right now is testing.

We don't have enough people and enough platforms who have used this thing on a regular basis.

I feel pretty good about the Mac.

I feel sort of okay about Linux, although from what you just said, Bob, a few minutes ago, it's clear that Linux is a bigger animal than I thought, and we've got no Windows experience.

I'm not at home right now.

I don't have access to any Windows boxes, so I can't test it out.

So I think the next thing is to recruit testers.

And I just don't have enough visibility into the J community to know who are good candidates.

I'm hoping that you do.

- That's a good question.

I guess the.

.

.

By a good candidate, you're looking for somebody who can provide feedback and is also interested in using it, I'm guessing, right.

Yeah.

I mean, one way to do it would be to send out a general call for help.

I'd rather not do that.

I'd rather keep it a little more controlled.

But I'm not even so concerned initially about whether they want to use it.

I'm just interested in whether they'd willing to be prevailed upon to use it.

You know, 10 minutes, install the thing, launch it, and see whether it blows up or not.

That would be a marvelous start.

And I can think of a couple of people that if I was going to tag people, I would could go to one would be Ian Clark.

Another would be Mike Day.

He's doing a lot of stuff on the iPad, but this would fit, I think, really nicely into what he's doing.

I'm just trying to think of whether it you need the JQT environment, don't you.

It's got to be JQT.

Yeah.

So that takes him off his iPad, but he might still have have some interest in going into it.

But there are a couple of others.

He's going to be a Mac guy, right.

I think, yeah, Mike will be a Mac for sure.

Yeah, that's that is less attractive.

I mean, it's interesting because it's another person and that's obviously a really good thing.

Yeah.

But what I'm hoping for is Windows people at this point.

Okay.

Well, I know Raoul has Windows in addition.

Oh, yeah.

I thought he was a Linux guy.

Oh, no.

Maybe you're right.

I thought he was Windows and Linux.

Oh, maybe he is.

I thought he only had one rather modest machine.

I sort of feel like I should send him something.

Well, I wonder whether he's got some form of a dual boot or something like that.

Oh, I see.

Yeah.

Because I've heard him talk about, he knows Windows quite well, whether I guess whether or not he's using it.

I think Chris Burke is using Windows.

I'm not sure, we can ask him, but he'd be a good one to put it through his paces for sure.

And I'm pretty sure Bill Lamb is, and well, actually Bill Lamb might be using Unix.

That's fine.

Yeah, yeah.

Would be good.

But I think we've got enough people who actually are off the Mac platform in development.

I can talk to Eric about it and see if he's got any that he really recommends.

I'm just thinking of the people I see do really good development, Norman Drinkwater, oh, Rick Sherlock, all those guys.

These are all names I recognize from the mailing lists and from the documentation.

Yeah.

But they're not people I know.

No and well, I've never met them face to face.

Well, of course, but you've had interactions.

I've had interactions with them and they've done stuff for me and I've done stuff for them.

Yeah, right.

Yeah.

This is why I'm hoping to leverage you.

Yeah, yeah.

No, I can absolutely do that.

And honestly, I think with that demo and seeing what it does, I think they'd be very interested in it too.

I don't think it's a hard push.

Okay, good.

Do you need anything more from me.

So I did a draft email.

I'll do another version of that because I've had a couple of additional thoughts That might be helpful Okay, you can just compose your own.

Well, no, no No, I think if you've had additional thoughts add them to it, but I think my next move is to talk to Eric All right, and and say what do you think and he can probably tell me exactly who he thinks would be the best target Oh, okay.

Good.

And then we may adjust the email to fit what they're, you know, make a personalized.

Well, then I'll hold off on sending you another version of the email and I'll let you talk to talk to Eric for which.

But don't don't lose your ideas, right.

No fear.

Yeah.

Okay.

Yeah.

Yeah.

So and the other thing I'll do is I'll put this up as a recording on, on, on my YouTube.

And I'll also send that to Stephen.

So he gets a chance to see it.

That was the next thing I was going to mention.

That's probably as good a presentation as I would give personally.

But of course, if he has questions, then it would make sense to.

.

.

Well, I tentatively asked him if he might be able to be available next Tuesday, maybe in the morning.

That was the original.

And I haven't heard back from him yet.

He is my son.

I believe my wife's going over to Vancouver for dinner with my mom and some of the cousins and I believe he's going to be there so I should be able to get some information from there.

Buttonhole him.

Yeah, yeah.

So anyway, I'll let you know as soon as I know for sure.

He's definitely back in town and the last time I talked to him he was interested in doing it.

But unfortunately Irish butter intervened.

It most certainly did.

Yeah.

Most certainly did.

Yeah.

I've been very careful since.

I haven't put groceries in my backpack.

I get a bag now.

I'm not being environmentally conscious anymore because it costs a lot of money.

It's a cautionary tale.

It is.

It is.

Yeah.

Not one you tell if you're trying to impress people at a cocktail party, let me tell you.

No, but I don't know whether you, I'm putting the transcripts up along with the notes on the meeting.

So it's a good story.

I saw that you mentioned it.

Yeah.

Yeah, yeah.

I was a little rueful when I saw that.

Oh no, I think it's a great story.

Yeah.

It is a good story.

It's four to the lower.

So that's, that's all I've got.

I think you wanted to talk about merging the wikis or cutting over or something.

or something.

- Yeah.

And I was sort of initially talking to Art about that just to give him the heads up.

Eric is wondering whether we need two prototype wikis or a prototype wiki and a running wiki.

And at this point, I don't think we do, but it does mean before they get rid of the prototype, anything that you wanna bring across, you should bring across.

And the easy way to do that is just to go in to edit a page in the prototype wiki.

Actually, I'll tell you what, if you stop sharing, I'll share my screen.

- Sneaker doodle.

- There we go.

And.

.

.

- But I didn't think of what's left on the blueberry.

- Well, you can do that.

Still there.

- I want to see what this is.

If you do that, (indistinct) - Share screen.

There we go.

And I'll share.

And Art, if you're listening here, the easiest way to get information is to go to the prototype Wiki if you've got the same, I've changed these pages, so there's no current page.

But if I get to a current page on the prototype wiki, I can edit it and copy it.

Maybe I should get to a page map or something like that.

There we go.

I'll edit this.

I can just click on it and copy everything.

And then if there's a current wiki, I can go to a current page.

There's none, but then you can create the page.

In this case, I'm not signed in apparently 'cause I don't have permission to create the page.

But if I was logged in, I'd be able to create the page, and then I just paste it into the edit and I'm done.

So you just pull the information across that way.

And that's the quickest way I think to do it.

And then from there, we can start to work out how we link into the tree and where it would hook in and what changes you would make.

We may wanna develop maybe under Wiki hints, I'm thinking possibly under development of the wiki, we could put pages in under here.

By doing that, by putting development of the jwiki as a category, you'd be able to store the pages off to the side if you were going to use them later.

Or if you're going to bring them into the main wiki, you can put them where you want to put them.

But I'm just thinking if they were something that really is in development, this might be a good place or something like this in the wiki area would be a place to save it.

And so in this case, you just add wiki.

w to the category of whatever page you put in and it'll show up in under that page.

- Hey, Bob, am I gonna be crawling the right thing once you're done with this exercise.

- Are you gonna be crawling the right thing.

you're gonna be crawling the current code wiki, not code two, right.

- I already crawl code, I don't crawl code two.

- Exactly, yeah.

So you're crawling the right thing.

- All right, good, thanks.

- Yeah, the prototype is really just there to have something we can test out and we weren't gonna affect the overall wiki, but you wanna be crawling code because that's actually what people wanna be looking at.

And when we get rid of the prototype, that's all they'll be looking at.

- Okay, got it, thank you, sorry.

- Yeah, no worries.

So that's, I think, I mean, I can tell Eric, we've got three week window, we've got a two week window or whatever to get the stuff across.

For me, most of my stuff is across because when I got what I wanted happening, I started to develop in the code wiki.

So my stuff's already there, even though it's not entirely exposed because I've put it into categories, which is hidden from a person coming in to look at the wiki right now.

But all I need to do to expose it is to switch the main page, forward it to my category page, and then they're gonna be looking at the category pages.

So I've kind of got a double structure going on, one that's not as apparent unless I send somebody to it, and the other is the current wiki.

And I think that's what Eric is most interested in is getting me to move off my ghost wiki and just bring it right on top.

And that's my plan for the summer.

So any feelings about the length of time that might take.

I'll try and get in touch with Raoul, at least by email to, 'cause I know he's got some stuff, whether he's got stuff he needs to do.

Last time I talked to him, he wasn't too concerned about it 'cause he'd done most of his porting early on in the process.

And I don't really know of anybody else other than you, Art, who's been working with that much.

Oh, there was Tom McGuire, but I already told him when he was categorizing to categorize on the blue wiki now.

So he shouldn't have anything else.

And I moved everything he'd done across.

I don't think anybody else has been doing very much in the yellow wiki.

And I'll probably put up a warning to everybody.

just letting them know if they've done things in the yellow wiki, when we have a timeframe that they've got that length of time to move the stuff over.

And then after that, it won't be available.

And my guess is it's probably a matter of weeks, maybe as much as a month, but I'll talk to Eric about his timeframe, what he's looking at.

So that's about all I've got in that category or in that topic.

Okay.

And so, is there anything else we want to talk about involving the wiki.

I don't think I have anything else, Bob.

Bob.

I got back in and started working on the primer again and found out how much I'd forgotten about how I was doing it.

So it took me probably an hour to get it back up to speed.

But once I did, it's it's yeah, I know what I'm doing again.

And I'm back to it.

Classic programmer experience.

Yeah, yeah.

It's like, well, yeah, classic programmer with undiagnosed ADD, I think.

Anyway, I think that's about it.

That's lots to work on.

I'll talk to Eric.

I'll find out about our timeline for the second wiki.

And we'll, how many, well, really Siri, I didn't ask you anything.

Have you got an idea how many beta testers you want, just so we can figure out.

10, 12.

No, I don't have a number.

I have a desire for coverage.

Okay.

part of my problem, which is more acute now than it was before this meeting started, is that I don't really know what coverage means.

I thought Linux was Linux, but Linux turns out to be, if I understand you correctly, a more varied animal than I thought.

Yeah, the Debian and the different.

.

.

Well, I haven't used it, but from what I've seen people talking about, if they get into.

.

.

And Bill Lam is probably the best guy to talk to about that, because he's usually talking people through issues with Android and stuff like that.

There you go.

Right.

But I'll see if I can get Bill on because that would be great.

That'd be terrific.

Yeah.

Yeah.

I don't have a number.

What I guess I have is a desire for a slightly broader conversation that will lead to a number.

Okay.

And you prefer probably the breadth with a smaller number.

Oh, obviously.

Yeah.

If I can get one person who has three different machines, three different platforms, obviously that's great.

Okay.

'Cause they'll, I mean, they can be much more efficient about it than three different platforms.

- And it keeps the noise down a bit when you're, yeah.

- Very much so.

And we're not bothering as many people.

We're not tapping as many people on the shoulder, which I think is probably good.

- I don't think people are gonna think this is a bother.

- Well, that's nice.

That's good to hear.

- I don't think that's gonna, I think they're gonna get on there and I think it's gonna become a bit of a game.

- Oh, well, good.

Oh, good.

Okay, that's nice to hear.

All right.

Well, Bob, thank you very much as always.

I'm gonna go to bed.

- Good stuff.

- Take care, goodbye.

- Bye Ed.

- Goodnight both.

- Goodbye Art.

[BLANK_AUDIO].