PARTICIPANT NINE

- "Ok" through First Class Link
- "Ok" through Transclusion
- "Ok" through Generic Linking
- Showed Link documentation and quick reference
- Gave summary of syntax

4:25 VILLAINS
P: Strikes me immediately that...that the initial information is slightly different...in both. I'm guessing the transclusion isn't...quite the same as...I saw before.
P: Er, right, so...I need to add that his teeth are metallic.
- Looks at source of one of them for over thirty seconds
P: I'm trying to work out how I'd change it so that it updated both. It looks to me like I'd have to change it...in both.
- Hinted that can assume that have done any longwinded "real" processes, like cleaning pages to be same. More thinking, ~2x 7s.
P: So you mean I should do that, or I could do that?
I: Would you?
P: Erm. I'd probably try to, yeah. I guess in a Wikipedia sense I'd---there'd be a separate page on it and I'd link to a different---add to the main article, I suppose.
- Clarified that this is a real split from Wikipedia.
P: <to self> What I was interested in was that transclusion.
- Showed which node was used in demo
- Participant studies example transclusion node
- Helped with syntax, pointed out key part: element, and type=Transclusion, not id or aid. Showed example in help for regular links.
P: My thinking at the moment is that I'm not sure where is best to put the information to transclude it from, if you see what I mean. I'm not sure whether to have another page that I put that information in, and transclude it into both, or whether to have, have it in one...of these and then transclude it to the other.
P: So let's try it putting it in this one. I should be transcluding it.
- 10s thought
P: Where is this transcluding from...it's <I: It's the page Demo.4> the page Demo.4, and it translcudes the whole of page Demo.4. <I: Yes> So can I create new pages? <I: Yes.> Am I allowed to do that...
P: Presumably I need to put in a link to it first.
P: I'll put in trans...
P: I've put in a transclusion to the page that I wanted it to create...and possibly it will fill it in, if I understand the linking.
 - Creates a transclusion in one of the articles to a new page "Jaws" with text "Jaws has metal teeth"
P: So...that's correctly displayed the "Jaws has metal teeth". But...I just want to have a look at the...see if the link has updated in the source. Ok, so it's put in the IDs, which makes sense.
P: Now what I need to do to...also translcude that into the other one...which I'm already editing...and put; I'm going to try putting in [trans to=Jaws type=Transclusion...I'm going to try putting it in /without/ the text, to see if that transcludes it across. See if that works. Whether it then sets that as...empty, because I've not put anything in there.
P: Excellent, it's transcluded. Which is what I was /hoping/, but not necessarily /expecting/. I'm just going to have a look at the edit again to make sure---to see what it's put in there. I presume it's going to put in the---yeah, the text that I had in the other page. Excellent. Made some sense. Excellent.

11:55 MULTIPLE UNITS
P: OK <20s pause> To create a new page I'm going to put a link on the homepage, next to the two current links for Multiple Units and Elektrichka.
 - Puts link pointing to GenericMultipleUnits, type Link
P: What I was thinking there was, can I have names that have spaces in? Because that's how I sometimes...
 - Explain \ and "" notation
 - Checks link
P: Oh, I don't have any text. I'm going to see what happens if I don't put any text in; see if it reuses the name of the link.
 - Explain it won't while it's loading
P: So I now have an invisible link, so I'll edit that and add some text.
 - I: points out that it was in relations table
 - P: tries it, natch target doesn't exist yet
 - Creates text for link
 - Has a read of other pages to work out what to summarise
P: I'll probably just transclude the first paragraph or something. I think.
 - Edits MU of Ireland to put in transclusion link, for use into the generic page; saves
 - Does same with the Russian one
 - Notes HTML entities in source (bug)
 - Curious about what would happen if didn't escape space in to=...
 - Explain that [trans to=FirstBit is enough to make it a link, grammar would make the bit after the space content, participant grasps that the other attributes after this would also be interpreted as content.
 - Back to editing generic multiple units page; puts brief introduction because not sure what a generic MU is.
 - Wants put in heading per country; looks at markup quick reference
P: Just wondered why you used this markup for the headings, rather than <I: H2?> yeah, or the equals signs or something like that [wikimarkup]. Is it just for standard, um...formatting of all your---so your links are that format so you're going to do that with the headings and other formats?
I: Yeah, it's---it's---it's a consistent syntax, it's /all/ done with this tree structure of square brackets. <P: Yeah, ok>
 - Temporarily forgets what nodes called
P: If I put it in and it's wrong, will it then create that node, thinking that it's...
I: Er...no, not if you don't give it any content.

P: The strange thing is...you use the to attribute, but mentally I'm transcluding /from/ at the moment, because I'm not putting any content here. <I: Mmm.> So it's a little bit of a leap to...<I: ok> to sort of---to think about that. I'm saying to, but I'm actually pulling in content /from/ there. <I: That's a good point.>
P: It's not necessarily wrong, 'cause if that's your syntax throughout...
 - Explain that underlying hypertext structure would consider it from here to thing being pulled in, but can certainly see usability issue there.
P: Let's try saving that and seeing whether it transcludes the right Ireland (uncertaint about page name). ...no.
 - Looks up correct spelling
P: It's put that in the...have we got a way of cancelling? <I: Cancelling?> For if I /don't/ want to save...I guess if I've not changed anything I can just click save. So if I wanted to go back to /this/ view (view) without saving what I changed or without previewing it, can I get back?
I: Back button on the browser.
P: I thought I'd tried that sometimes and sometimes it skips past.
 - Browser bug
P: It might be worth having a link for that, just for simplicity.
P: So it's put in this transclusion even though it (the target) doesn't exist. It doesn't exist, it offers to create it for me (relation table?). That's good. Let me fix this.
P: Change the to= link of this transclusion, I'm saving it, and it should copy our summary of...?...which it didn't. Have I made a spelling error?
 - Has tried to pull SummaryMultipleUnitsOfIreland into summary, but transcluded 'out' the paragraph as SummaryGenericUnitsOfIreland by mistake.
P: Let's try changing it here (MUoI) [...] because that's what I actually wanted to call it.
P: So the Ireland page looks good, and if I refresh the generic multiple units page...it's correctly transcluded it. But there might be a floating page somewhere, called SummarGeneric(etc) <I: [Doesn't matter, can't delete pages, don't worry about making a mess]>

I: What if someone then needs to correct something in one of these summaries?
P: In one of /these/ summaries? So we can do it---I hope we can do it on this page because it should have copied the text in...
 - Tests that can make change (model number) in summary page, and then that this has been pushed through to appear on Ireland page; it has.

P: I quite like that. I like the fact that the text is in both, but if I change it in one, then it knows where it got that from so it goes back and changes it.

24:33 BELGIUM
P: Ok, so I'm going to have a look at the editing page to see how the existing attributes are there. They have a nice little editing area for that, but I don't know how to add one. [5 sec pause]
P: So there's an Other box at the bottom which has---has them in there. Attribute=Something format. I don't understand why they're there rather than in this...attributes list.
P: I'll have a look in the main text...body...but that's just the content, above the attributes. So... [5 sec pause]
P: Ok. I see that there's already an attribute that is EU accession...date.
 - Confirms that that's what was asked to add; decides to add; remind of date
P: I'll put it in as words---no I won't, I'm going to look at the rest of the list.
 - Spots another and notes format as year-month-day, uses that format
P: I'm just looking at the link next to it.
 - (Copies date in case lost in forward/back)
P: Ah, ok, so if I click on the link next to it it says that dates should be formatted numerically in the form that I said just now.
 - Back, checks present, saves
 - Wonders what happens if click on linked date, doesn't exist yet.
P: I think I've finished that, but what I'm intersted in is what the list of other things at the bottom is for. [pause] I wonder what---I'm just going to delete---can I do this? <I: Yeah!> I'm removing the date, I'm just going to try putting it in the other box to see whether that then fills that (the form field) in, or whether it just sits there.
 - Notes view puts them all in table
 - Notes that moving it to other, it is still in the view
 - Looks at edit
P: Ok, it's fill in the EUaccess date (in the table), and it's no longer in the other box. Hunh. Satiated my curiosity.
 - Suggest looking at an attribute
 - P spots that it is a node with domain and range. Is country defined?---oh, Country's just a description I guess...as opposed to...
 - Prompt to return to Belgium, point at type attribute, explain domain match.
P: Oh, I see.
 - Explain that Area is in Other as there is no Area node to specify domain
P: Got it. Interesting. I like that.

30:30 SHELF
 - Familiar with disambiguation pages
 - Goes straight to relations table and spots the two pages (ignores Homepage)
 - Goes to albatross to see context
 - Searches in page for Shelf, and reads surrounding section
 - Something geographical; goes back to Shelf link page to see which option makes most sense; guesses if really making this fix would know anyway (yeah)
 - Looks at Continental shelf to check text
 - Back to Albatross; finds section and goes to edit
 - Finds section in editing, "lots of links", finds it fairly fast, changes to=, leaves content as-is
 - Explain the on-save tree generation, easier for manip, while waiting for slow save
 - Checks link correct to continental shelf
 - Uses relations table to get from ContinentalShelf to Shelf and checks that Albatross is no longer listed
 - Moves on to River Etherow; straightforward again. Idle curiosity as to how nodes selected for experiment while waiting.
 - Tests again
 - Had spotted ShelfWestYorkshire name in the relations table when it was cross-highlighted from Shelf body text (text does not use this name); no, wait, looked at browser status bar when hovering over link---relations table was scrolled offscreen at the time.
 - (Safari bug where relations table jamming highlighted after being clicked and going back)
 - Checks Shelf only has Homepage incoming; Shelf links to lots of others
 - What did think of Albatross when went to edit it? How easy to find the---
P: It was quite difficult, and there was lots of---there were a lot of links, and because they're quite big they take up a lot of space, so as you're scanning it it then interrupts where you are in the text, 'cause you have to scan through the---through the link, and then carry on with reading the text and then scan through the link.
P: So there was one bit; I'm not sure if it was in the Albatross one of it was in the other one, but there was a big section where there were lots of links, one after another, and it was quite difficult to read that.

I: Given that, can you think of another way you could have adjusted those links?
P: [4 secs] I guess I could have gone to the node page that...is there a node page that exists inbetween? So with the source and the target thing. So...how would I get to one of those? Is it in the relations table?
 - Knows that ContinetalShelf etc. in table takes to page, so "is it this via part?", tries
P: Ok, yeah, so there's the source, and the target. I could have changed it there. That makes sense.
P: To be honest, I had forgetten about---I remembered this, this page before in a previous one, but then completely forgot about it by the time I got to this one (task?)
P: It does make a bit more sense to---makes it simpler to /find/ what you were changing there, but...I dunno, I guess I now have to remember to come down and look at the relations table to find that link...whereas I'm looking at it /here/ (in the body text) and my---my instinct is to---to work on this bit, to make the change. <I: yeah> So if there were some other...I don't know if I could view like a, and /edit/ version of this page where this (link in the body text) gave me a link /to/ the page and a link to the link: to the node inbetween, that would let me edit it; it might keep me in---it might keep that in my mind to---to be able to go to that link inbetween. If you see what I mean. I don't how you'd---how you'd best interface that idea.
 - I: It's out of the flow of what you're doing, isn't it -> yes
 - P: Might just be that it's the first time, would remember in future
 - Pointed out tooltip; hadn't noticed it; don't normally get one (even though had been hovering over several times); hadn't really looked at it or read what it was.

42:02 LONDON UNDERGROUND
P: Do I get to pick where to add it? <I: Yep> I recon I'm going to add it in the history, because it makes sense.
 - Adds new line with oldest in world fact
 - Checks that it's there
 - Uses relations table to get to TfL "cause I've used it a few times [and] now I know it's there"; was "only a guess" that it might be there; twigs that it's because TfL is linking to UG, not vica-versa. Avoids unnecessary Homepage.
 - Looks for suitable place; spots UG section. Compares paragraphs of text to see if they look the same (are already transcluded); concludes not.
 - Looks at source, not it's not transcluded
 - Uses fact that oldest in world left on TfL page (oops) for translcusion
P: What troubles me a little bit is that I'm transcluding like a /sentence/ of text, and that---that will create a node inbetween and I need to /name/ that node inbetween...and I'm not sure it's creat---having a... conceptually it seems weird having a whole node---a whole page essentially just for this one sentence of text. But I'm going to do it anyway, becuase it's only an experiment.
 - Calls it OldestMetro; wraps sentence Transclusion markup
 - Puts it into LUG page; checks not broken TfL first; uses main article link
 - Was going to take out previous sentence, but:
P: /Actually/, I'm going to see what happens if I transclude /this/ with some content [...] just to see that it /does/ overwrite the other one.
P: So this should now---it should stay /this/ in the LUG page, and it should also update the text in the TfL page to [text].
 - Checks same here
 - Looks at relations table, sees trans and visits node to see that it's the text P just changed it to, goes to TfL
P: Ok, so it has transcluded correctly, except the fact that because it was a sentence in---the start of the sentence in the other article, it's got a capital letter, so, I guess, whatever it was in this was the best one, I think it was [text about oldest].
 - Could fix, but convinced self that transclusional "links together", if that's OK

48:14 CAKE
 - Confirm that targets pages should be assumed to exist as per WP
 - Sugar, "keep forgetting to put the equals"
P: Do I need to tell it that it's of type link, or is it only type=Translcusion that I need to make specific?
 - Reads markup summary to self; clarifies that no, don't need to; ("link text" example slightly confusing, as "link" is also elem and type; also that everything proceeding is a single token)
 - Egg, butter, shortening (that probably needs a disambiguation page, but I don't know what it is), margarine, looks for some others, buttercream, marzipan, flour, crystalised fruit probably has its own page, and people may wonder what it is...
 - Links "piped borders" to "cake decoration"; must be a page on that
 - Been manually typing in links
I: So you, looked at one of the things in the relation table to see if it existed.
P: Yeah, mostly because they appear here (in the body text) as red if they don't exist yet---I presume that means they don't exist---but in the relations table they appear as blue whether they exist or not. <I: Yes> I think.
I: Well spotted. It's a strange little technical limitation thing. Or a 'bug', as it's better known.
 - Prompt for any other way to do that.
P: Nnnoooo...[thinks]...that was all I could think of to do
I: Would you perhaps have considered using generic links to do it?
P: Sorry, what do you mean by generic link? Is there an example?
 - Recovers demo; participant remembers on mention of CONTAINS at start; "didn't quite understand the CONTAINS syntax straight away"
P: Is CONTAINS the only thing that you can use, and can you only use text, you can't do regular expression matching or anything like that? Ok. Are you /planning/ to expand that?
 - Design allows for that, etc.
P: I'm just wondering if you could have a...a simplified version where it---you just rather than having a sourceQuery it's just contains=, and then you put your word in, and then tell it what to replace that with.
P: How do you...does a generic link work for all pages? <I: Yes> Ok, so it automatically replaces it for all pages.
P: Can I create one? <I: Yeah, go ahead> [Self] How /do/ I create one?
 - Goes back to Food to look for something to create generic link /for/; chooses word 'food'; looks back at 'syntax' from demo, editing.
P: Oh I see, it's just a node with some...attributes, so I /could/ edit the cake page and do that all inline I guess. Or could I?
 - Finding it faster to nav via relations table because alphabetically sorted and list, than body text (on Homepage)
 - Going to edit Cake page and put a generic link around 'food'; does that make sense? Lists reqs: "sourceQuery, target, type=Link, I /think/ some content---don't know how that...[doesn't?] make sense, let's try"
 - Tries it as a [link , does sourceQuery=CONTAINS('food'), target=Food,
P: I'm interested as to what targetQuery means, but I'm not going to investigate that now
 - Saves, suspects may not be right to create magical link
P: It's all gone wrong! [Errors, many, unknown attributes]
 - I: explains weird naming mismatch, e.g. target/to, sourceQuery/fromq
 - P: gets rid of all the errors in the source
 - Retypes link
P: The strange thing is that I'm putting the CONTAINS in, but I'm also putting some content in...I don't know whether to put that still within the---within the link, or outside of the link, or what.
 - Tries inside
 - Has created link
P: Thing is though...I don't know what my generic link is.
P: What's this? In my relations table I've not got subscript 1---it seems like a versioning.
P: So it's both of them. I don't understand why there's two versions of the link there, but nor does [investigator], so it's OK.
 - (it's the generic + anchor bug)
I: I think you've succeeded anyway.
 - Got two somehow.
P: Have they nested or recursed because I had food within the link and then it matched food within the link so then it replaced the food within the link, in the content of the link with a whole nother link?
 - I misjudge error as generic link versioning problem
 - Conclude that generic links tricky
P: They're a little bit trick because, if you can't add them in the page...they're a useful extra, administrative tool. But I'm not sure letting users...I guess you could let users have them but don't tell them much about it; it should be a very advanced feature that someone should sit there with a bunch---like with a generic link interface that they can add a bunch to, and then it'll go through and...
I: Adding them within a page is...I created the one in the example by creating a page which was a link.
P: Yeah
 - Agree we've demonstrated that creating them in the page is not easy

P: I just wanted to know how I'd create my own link. So your generic link was clalled [demo name]
 - Create FooGenericLink in address bar, not sure of any other way of getting to "the generic link page" (a new page), I can put in the attributes
 - Presumes correctly that they have to go in attributes, not content box
 - Remembers that in page they get the "other" names (sourceQuery, etc.)
 - Looks for a repeated word in Cake to use, suggest the word "and"
P: Why are they [the attributes] different, [investigator]? It's just confusing.
 - Links to TfL for testing purposes
 - Point out that missed out one thing. Looks at example. Notes it's type Link.
P: Do I have to tell it it's of type link? Doesn't it know?
I: /Should/ it know? This is an interesting---should [explicit or duck]?
P: What would have a source and a target that isn't a link?
I: I think that's a good argument.
P: I think that---that makes sense. And that if you can't answer that, then the default should be that it---it's a link.
 - Looks at cake page to see how many "and"s replaced
P: It's only replaced one of the ands. Ah, I see, because it's following that---the Wikipedia sense of only defining it the first time. <I: Correct>
P: Which is sensible, otherwise you'd have ands all over the place.
 - Checks it links to TfL
 - Checks that that has many backlinks
P: So I've created a generic link.
P: Just going to delete the type, see what happens. See if it breaks.
 - Looks about, cache fun; while loading:
P: So does it go back and up---when I create the generic link, and save it, does it go through all the pages that exist? <I: No.> ...and change it...it's only when I first view this page that it...?
I: It's when you view it at all, it has to look and say which links match this.
I: So it's still taking effect.
 - Prompt to look at relations table. It is in the body, but not the relations table---fun bug.
 - Puts type back
P: It links to itself via the generic link. That makes sense. (TiL contains "and")
 - Checks about on some other pages for "and" links, and in rel table. "Cool"
P: That was good fun.
 - Final comments: "it was good; I think I've said everything that came to mind whilst I was working on it"

- Yes results
- Yes voucher

