User:Mstone/Commentaries/Sugar 2

From OLPC

< User:Mstone
Revision as of 14:15, 9 October 2008 by Mstone (Talk | contribs)
(diff) ← Older revision | Current revision (diff) | Newer revision → (diff)
Jump to: navigation, search

Contents

Introduction

     mtd> It strikes me that Eben's bulletin board designs just mentioned on
          IAEP seem to involve sharing things, and patches are things, and
          perhaps this is the answer to my patch-sharing itch. I should
          probably just spawn a tangent thread on the mailing list.
 m_stone> well, I found eben's answer quite non-satisfactory, first, because I
          think that bulletin boards are quite different from activities and
          second, because he didn't give any sense of _order_. narratives are
          _sequences_ of events. (and narratives which _will_ take place, with
          user input, are often tree or DAG-structured.) they are not simply
          sets of events.


  bemasc> I suggest a concrete example. I think neither I nor Eben has
          comprehended your concept of narrative correctly.
 m_stone> think of literary narratives. the point is that they have order. for
          example, the narrative-as-graph is like a choose-your-own-adventure
          book. [and yes, I'm actually thinking of hypertext and RPG dialogue
          code here...]
     mtd> narrative-as-equence? I understand the analogy but think that
          literary criticism has much to say about linear, one-way
          communication as "storytelling". I am going to have to cite the xkcd
          about literary criticism if I don't. There, I did. And I like your
          RPG-style thinking. But now I'll really stop, as, a bit like bemasc,
          I know nothing about education or psych.


 m_stone> did you understand walter's reply?
  bemasc> this discussion is not oriented in a way that works well for me. I
          think of Sugar as ... a virtual machine for untrusted code. I don't
          know anything about education or psychology, and I have little
          interest in speculating about either.
 m_stone> look, you've both undergone lots of education, same as me. we're all
          on equal footing here. I'm just pointing out the simple fact that a
          fairly large chunk of the education you underwent consisted of
          discrete experiences embedded in rhetoric.
          (http://en.wikipedia.org/wiki/Rhetorical_modes)
     mtd> rhetoric++. Ther'es a lot to be said for a clear, precise, compact
          explanation. Sometimes you need the answer to help you think about
          what you thought the answer was.


 m_stone> Bryan's remarks about the amount of effort and frustration stemming
          from the increased demand for lesson plans suggested to me that Sugar
          was making it hard for teachers to string together the experiences
          they wanted to share with their classes and their peers. Bryan's
          desire for Offline-Moodle clearly stems from the fact that he thinks
          that hypertext (and Moodle, which generates hypertext) is pretty good
          for stringing together experiences and he thinks that it's within his
          budget.
     mtd> ah, and lesson-plan-asnarrative-sequence, I get it. the best
          perormance improvement is going from a non-working stateto a working
          state. I suspect "sugar making it hard" was "sugar making it possible
          in a different way than just writing on a blackboard" in some cases.


Lessons

  bemasc> ok, lesson plans. that's something I know how to think about. I still
          don't really get it though. I don't see a big gulf between a teacher
          saying "now take out a paper and write a six-stanza poem with the
          rhyme scheme we just discussed" (an actual assignment from 11th grade
          or so) and "now open up Write and ..."
 m_stone> suppose you want the kids to work on it at home... it's true that one
          solution is to provide a write document containing instructions, but,
          imho, that's not a very good solution for a couple of reasons. in
          particular, because it seems to completely constrain you to working
          in Write. (or to giving instructions on how to flip between write and
          other stuff when needed)


 m_stone> another example I have in my head involves pippy examples. suppose I
          write up an awesome new Python trick and I want to explain it to you.
          I want my explanation to show you a buildup to the trick, a working
          example, and a collection of links to other tricks and tutorials.
          also, I want to prepare my explanation ahead of time, transfer it to
          you, and let you experience it without talking to you. how do I do
          that?
     mtd> 1) write is much less limiting than pre-printed paper with
          instructions and a blank space on the back for writing the poem; and
          2) literate programming gets you a long way.
 m_stone> well, one common technique used in many good CS textbooks is to write
          a series of examples which build up to the ultimate example. and
          people who do multimedia projects commonly say 'go make your collage
          over here, then glue it into place on the worksheet...'
  bemasc> so you're creating a category that includes textbooks and recorded
          lectures.


 m_stone> so what I think I'd want to try here would be a system that let me
          string together resumable activity instances and 'holes' for new
          instances to be put in. (maybe other things too later on, but those
          are the essential pieces. cscott tells me this is nothing more than
          ODF... however, I think that the UI is the important part). then I
          can just give you my explanation and you can pick whichever piece of
          it you like, resume it, and play with it.similarly, if there's
          something I want you to do with it (that you're going to give back to
          me), then I'll have made space for you. for example, maybe I want you
          to solve a puzzle or to take some photos. so maybe I put a blank
          record instance there for you?
  bemasc> why not just say "Now open Record"; "Now take some pictures with
          Record"? my textbooks are not interactive, and I managed to make it
          this far? They say "now solve the following problem" and so I take
          out my computer and write a program to solve the problem
 m_stone> I just think it's a better experience if you directly present people
          with the structure of what you want them to experience. it's like the
          difference between going on a hiking trip on a well-established trail
          vs. in the back-country... you can study the landscape from both, but
          the former is much more accessible. it requires less unnecessary
          decisionmaking, comprehension, and judgement.
  bemasc> I should say, I don't think what you describe is a bad idea, but it
          seems of minimal actual utility. I've seen many tutorials for various
          computer things that contain a picture of the icon that the author
          wants you to click


Usability

 m_stone> does your opinion change if the people you're teaching can't read?
  bemasc> no, except that for people who can't read, none of this will work
     mtd> I think people are going to have to read to get lots of benefits out
          of narrative.


 m_stone> put differently -- why are screencasts a valuable instructional tool
          when almost all screencasts arise from written instructions?
  bemasc> I don't know; I've never used a screencast
     mtd> s/read/use language/
 m_stone> as the study of semiotics demonstrates, that difference matters. I'm
          inquiring about the utility of a fairly small visual language for
          'here are a collection of things you should try doing'. (probably in
          some fairly specific order)
     mtd> hmm. I don't disagree there could be power in that, but I think the
          difficulty of creating even a small symbolic language that
          presupposes no existing symbolic language should not be
          underestimated.


  bemasc> if you're simply proposing a way to allow one Activity to expose a
          widget that launches another activity, then I have nothing to argue
          about. I think that's a fine idea. I just don't think it represents a
          major leap in functionality, or enables any new behaviors that
          weren't easy already. it removes a single step of "find this icon on
          the home view" and removes zero clicks.
     mtd> I think the lesson plan UI is just going to help teachers build (or
          just perceive) more easily the plans for using the XO in the
          classroom. Fostering that perception is quite useful. I think it's
          actually a big leap in usability - it's moved the number of switches
          away from the activities involved to zero. From one per activity
          involved. To a computer user it's of course a "chasm" we've all
          crossed long ago. but perhaps not for first-time users (obviously I'm
          not able to imagine that situation myself, and I am unfamiliar with
          reasearch in that area, so i'm pulling this out of my ass)
  bemasc> for users who are truly uncomfortable with the interface, I see how
          having to switch to the home view and launch an instance could be a
          barrier


 m_stone> it also avoids the need to invoke object choosers everywhere
          (P_DOCUMENT). if you have several activities linked together inside a
          single narrative container, then they can all interact happily w/out
          crossing any security boundaries.
     mtd> yes, for pre-existing documents...actually that's a lot of clicks
          you've done away with. I have noticed - completely anecdotally - the
          high cognitive load non-developers seem to experience keeping track
          of open / planned-to-open applications. (whereas I think most
          developers run several WMs on the wetware)


   nteon> where in the current interface do you invision "now take some
          pictures with record" fit in? from your original email I had been
          thinking more about allowing kids to pull activities they had done
          into narratives, but I think that could fold into this broader
          concept you're talking about, although it is still somewhat nebulous
          to me. also, how would you link them together in the first place,
          through the journal?
 m_stone> I'm not sure yet.
     mtd> I did indeed think you were distributing an activity that allowed
          launching existing (and potential) instances of activities together.
 m_stone> that's one potential implementation but I don't have strong feelings
          about either the implementation or the graphics; instead, I'm
          thinking about the interaction design -- screens, input, waits, and
          so on
     mtd> the UI is definitely the innovation/interesting thing here
 m_stone> my other concern is how it should be displayed in the journal.


Cost/Benefit Ratio

     mtd> as bemasc and others have pointedout, the workflow/usage pattern is
          poosible now
 m_stone> 'possible' != 'efficient, smooth, polished, pleasant, ...'


 m_stone> my current complaint about the journal is that the journal doesn't
          have any way of showing you things you _might_ do in the future. (or
          that someone wants you to do)
   nteon> I think it would really help the list thread to get your ideas down
          digitially, even if it is scanned in scribbles (like cscott's
          journal). I myself am having a hard time imaging things other than
          ebens bullitin board or a ppt style tool


  bemasc> what you're describing sound to me like EPGY and other "distance
          learning" programs. more specifically, I would call this a "self-
          guided interactive lesson"
 m_stone> okay.... how does that conceptualization affect your judgement of its
          relevancy to encoding educational experiences for consumption by or
          through XOs?
  bemasc> I think this is very valuable but I maintain that it's already very
          much possible and that the technical barriers are negligible compared
          to the social barriers of finding people to write and translate these
          lessons. An HTML content bundle is better than close enough for me.
 m_stone> do you also assert that it's any of 'efficient, smooth, polished,
          pleasant, ...' ? also, how do I make one of those on an XO?


   nteon> what about students wishing to string together a thread of activities
          telling a story of their experiences or of something they've learned?
  bemasc> they should write up a lesson. certainly if it's not easy to create
          content bundles on an XO... well, it should be.
 m_stone> (there are other problems with content bundles too, such as the
          copying-into-journal problem, the invisibility-of-bundle-in-journal
          problem, the what-the-hell-does-Bitfrost-say-about-these? problem,
          ...) but I take your point. as for the 'well, it should be'.... I'll
          reply: 'well, it certainly isn't yet, and no one is working on the
          problem either'. this is, in some sense, is the cause of my email --
          I think that more people would work on this use case if the sugar
          vision said that this use case was important.


 m_stone> in conclusion, what I'm saying is that it is currently all but
          impossible to write and share slick lessons on an XO, or to keep
          track of one's progress through a curriculum of lessons in the
          journal.
  bemasc> making content bundles work better sounds very valuable. We certainly
          don't provide nice content creation tools. I heartily agree that this
          is an area in which improvements are worth pursuing.
 m_stone> lovely. now if only you weren't in engaged in pursuit of further
          education... :)
  bemasc> right.
Personal tools
  • Log in
  • Login with OpenID
About OLPC
About the laptop
About the tablet
Projects
OLPC wiki
Toolbox