From: <bgardner@ATHENA.MIT.EDU>
Date: Tue, 23 May 89 21:33:37 EDT
Message-Id: <8905240133.AA12324@PINK.MIT.EDU>
To: bgardner@ATHENA.MIT.EDU
Subject: Screen design, etc.

 ~s Screen design, etc.
   Hmmm, well, I'm not sure what to say here.
^C
(Interrupt -- one more to kill letter)
^C<PINK:/mit/bgardner> mail mackay
   Hmmm, well, let me preface this by saying that I'm not upset
by these past few msgs, and I haven't been upset in my responses.
(I figure that needs saying, since e-mail communications are notoriously
bad at communicating moods.) So, this msg should be read as a fairly
unemotional, calm answers to questions you raised.

> It seems like we really need to talk.  Why is it being user
>tested before anyone in the group has seen it?  W

   I test every piece of code I write quite thoroughly. As soon as
enough of a section is written, it is tested. I test parsers
on dummy data-files, internal objects with debugging print calls,
and user interface screens on helpful (computer-illiterate) friends.
My resulting code is generally bug-free, because I put alot of effort into
making it that way. (In fact, being bug-free/well-designed/and seeing
the final results of a vision are what gives me enough pride in my work
to push myself through the writing the huge amount of boring internal
code associated with any project.)
   I'm sorry about doing stuff that Susanne was supposed to be doing.
(I just thought that testing things as you go was part of writing code
& UI well.) Honestly, I didn't know that was what Suzanne's job was,
she hasn't been around much (& at the last UI sub-group meeting,
that question came up (from Dan) and no-one seemed to know).
But, what you'd said to me, "You can't expect people to know what
you're working on if you don't tell them," perhaps this applies
to Suzzanne in this case? We certainly didn't know much about how she fit
into the big picture.
Now I know.
   As for people in the group not seeing it, ...
I have announced at the last two video-mail meetings that they can
(& should) try the current version of the program, and that
I consistently am keeping a working copy
in "/mit/bgardner/pygmalion/pygmalion_demo" for that purpose.
I have also sent two e-mail messages about it, one of which was soley
to announce that fact. I have brought in screen print-out of the
currently implimented screens consistently since version 2.
(Not doing that with version 1 was my mistake). For version 3,
I have even xeroxed copies and handed them out. I have personally
spoken to Brian & Evelyn about pygmalion_demo so they see it in action
and start working on the Help screens for it and suggest more page templates.
I even mentioned it to you one day when you were over by the hall
next to my cubical, as a response to your saying that you wish you
could demo it to someone. I have also put alot of effort into
writting up English descriptions of the internal workings and specs
for the screen interfaces and their implimentation/representation
details.
   Now, please don't take this the wrong way, but as a statement of fact
I have gotten *almost* no response to any of this -- and it has
taken me a fair amount of effort. I could probably put in more effort
and schedule the Demo Room (perhaps even on a regular basis) and
personally demo it to everyone. But this will have to wait until after
finals; I have already put alot of effort into Pygmalion and am
in danger of flunking my classes currently. Although I have gotten
next to no feedback so far, that might not be a bad thing. Perhaps
my efforts might not be helping me much, but hopefully all this dispersion
of information will be helpful to the people receiving it (even if they
don't let me know that). At least they know what I'm doing now.
(And my ass is covered because now I can point to a piece of paper
or e-mail and prove that I really did say the things I claim to have said.)
It does take away alot of my time from writting code, but I'm sold
that information dispersion is pretty important, too.

> Hi.  I probably shouldn't have sent the last email, since it
> probably sounded too curt.  Some things just don't get across
> via email.

  No problem. I understand (my outgoing e-mail often comes across
the wrong way, too.)

> I'm delighted that you're making so much progress.

Thank you.

> My concern
is that it is important that everyone know what is going on.

  I hopefully am putting enough effort into this.
And showing that this is a concern of mine, too.
I certainly am trying.
  I realize that I hadn't told anyone about my testing pieces as I go
(including with users), but i wasn't aware I should be saying that.
(I thought it was just part of debugging my piece of the project.)
I was expecting to go through a user-testing period after I was done
(as someone else's piece), but I thought that it would be to
mostly check for timing problems, potential scaling problems,
how long users spend actually using which screens, what the contents
of there messages are, and if there were any screen modifications
or re-designs needed based on the results of how either the system
actually responds as a whole or because of some initial design
flaws -- sort of testing the gestalt of the system. I didn't
think it would involve testing individually the sub-pieces or
debugging of individual screen implimentations. I had assumed
that as implimentor, that was part of my piece of the project.
(Especially, since its hard for me to imagine writting anything
without debugging it as I go.)


> The message editor/sender/receiver (you need a name for this...)
> is unquestionably yours

   My design for the message editor/sender/receiver (screens & internals)
was rejected back in around February. It was thought to be too complicated
to write in time and too complicated to use. If you wish to check this out,
you can "diff" the files in "/mit/lens/src/bgardner/widget-builder" and
"/mit/lens/src/bgardner/video" against the files listed for "pygmalion"
in the Makefile for "/mit/bgardner/pygmalion"..you will find that
in fact, the only internals from my design of the message editor/sender/receiver
which are used in pygmalion v3, are two functions: "WBis_display_video_workstati
on"
and "WBisa_video_workstation" (which check whether a workstation has a
parallax board or not). (These funcxtions aren't part of Widget BUilder.
They just happen to be in that directory and start with WB, because
at the time Pygmalion didn't have a name or directory. Though,
version 1 was sometimes called MMM back then, and used alot of
ideas & pieces of code copied from Widget Builder.)
    Version 2 (called Media-Mail or Video-Mail back then)
was an effort on my part to try to salvage as much of version 1
code into the re-design of the message editor/sender/receiver
(co-designed, with the new UI group of Brian M., Evelyn, & I,
along with lots of input from the whole video-mail team, and
lots of friction, as you know.)
That friction stopped largely due to my giving up ownership of the
message editor/sender/receiver; instead, became a team-player
and began the re-designing the UI with Brian M. & Eveyln
from scratch with a clean slate and an open mind.
    I am currently implimenting version 3 (called Pygmalion).
It was designed as a team effort by Brian M., Eveyln, and myself
with constant review by the entire video-mail team.
The majority of the message editor/sender/receiver's functionality
is from Brian M.'s creative input. The idea of a pallette and optional
mouse-dragging of widgets between palette & page was mine. (Note that
a palette was introduced 1st in version 2, and not needed in my
initial version 1 design). And most of the consistancy between widgets
functionally was from Evelyn's input. All in all, version 3
seems to be a good design for SIGGRAPH.
    In short, while I have input into the design and implimentation
of the message editor/sender/receiver for Pygmalion, it is "equal"
input with the other UI people. To say that it is "unquestionably yours"
is a mis-statement in spades. If I were to be in control of
the message editor/sender/receiver piece of Pygmalion it would have a
*very* different functionality, look, feel, and internal representation.
The original functionality I'd designed for MMM (Pygmalion)
offered enough power in the UI to touch upon several issues
I was interested in and left room for some original work
that was entirely mine. In the current version 3 of Pygmalion,
there is no piece of my work which I can count towards my dissertation.
Pygmalion is just what I get paid to do now, it is
an interesting job, but no more than that. All of the
interesting ideas I'd wanted to research and express as original
work toward my dissertation are nolonger applicable to version 3 Pygmalion;
they were taken out of the project back in Feb. & March,
which is why I was so upset back then.


> I see Pygmalion as a group effort for SIGGRAPH composed of a
> collection of programs and projects that are created by individuals or
> small groups.  The goal is to make sure that everyone gets individual
>recognition for their parts, but to make the whole a team effort,
> greater than the sum of the parts.  Doing things by yourself is easier
> in many ways -- but it makes for communication problems and
> frustrations later.

   Originally, the project was rather loosely defined as to who
does what exactly or what the exact programsa and projects would be.
I was promised that the user interfacing was mine, including video
user interfaces. It was to be my individual project and I won't have
to give up any part of it and that no-one else would be placed
on it without consenting me first. You urged me to cut this piece of
the whole off as my own and should use it towards my dissertation
research. I clearly understood that there was a larger collection
of programs and projects greater than the sum of the parts.
You also made it clear that, being one of the founding people on
the larger project, it was neat because I got to carve off which
ever piece I could turn into "my own" research.
   When Brian & Evelyn signed onto the project, I was not
consulted first. I don't know how they joined on, but when I found out that
they were being listed as User Interface Designers I was worried
about a conflict with "my piece". I came to you to voice my concern.
You assured me that I was still in charge of the screen design,
user interface stuff, and widget coding. You said it that I was right
to come to you, and I should feel free to express my concern and
insure that I can keep that piece of it for myself & my research.
You said that you didn't know how Brian & Evelyn would fit in yet,
but that you needed help with things like the kios design.
This set my mind to ease because the kios *is* a user interface
but doesn't overlap with my project. So, I was conforted.
   Well, I won't go into this further. Surfice it to say, that i have
come to you every time that someone has signed on to the project
to express my consern about loosing my piece. I have always
been open about this.
   the user interface is a sexy area, especially in a video mail
project, and everyone wants to sign on to help with it.
But as Don Davis says, "we have far too many cooks, especially
on the user interface, already". If you walk to my booth now,
you'll see me working hard on the internal data structures
and parser implimentation (which "my piece" has been expanded
to include (as much as I *hate* parsers), because no-one else
is doing them). If you walk into a video-mail meeting,
you'll see Brian M., Evelyn, Suzanne W., yourself, Debbie, Rich,
and ..oh, ya,...me all involved in some part of the UI design.
That's 7 people for doing the UI. There is only
one person doing the scheduler, one doing the database retriever,
one for the mailer interface (the same one doing the database retr.),
one doing the video server, and one doing the bulk of the main internals...
but 7 people doing UI design. Now, this lop-sidedness seems
to be obvious enough that Don Davis boldly states it in meetings
(out of fear that the project is suffering from it.) And it is
less boldly stated as an unfortunate fact in external conversations.
    In any case, I guess the purpose of this message wasn't to
to ramble (sorry about that).
I sent this message simply to open communications.
I wanted to know afew things, if they weren't already obvious:

       1) I don't own the message editor/sender/receiver.
          Its design is part of a team effort.

       2) I have put in great effort (and have tried especially
          hard this past month or so) to appease and inform
          the concerns of the other members of the video-mail project
          and to become as good a team-player as possible.

       3) Simply and honestly put, you broke your promise to
          me. I have not been allowed to cut of the UI piece
          of the project I wanted, and consequently, there
          is no part of Pygmalion relevant to my dissertation research.

       4) I am spending most of my time on Pygmalion currently,
          doing work which is thoroughly dis-interesting to me.


   Unfortunately, this is very similar to Athena Lens.
I joined on Athena Lens because writing I was interested in
the rule firing code and in the neat user interface stuff.
Unfortunately, Jeff and Rich also greatly wanted to do that;
there were two interesting sub-projects, but three programmers.
You may recall, you were in abit of a dilema as to how to resolve
that. You held off for awhile hoping someone would change there
mind or an answer would come somehow. It was obvious that it was
a sticky spot, and to resolve it, I backed down and didn't do
any of the stuff I was interested in. Jeff did the rule firing code,
and Rich did all the user interface stuff. I ended up writing
tyhe larger internal objects, the parsers, a trivial main() for
runrules (to call Jeff's do_rules), and (when you were in a tight spot)
I did a 6-hr temporary "rsedit", because Rich was behind on it
and wouldn't make the 1st release date.
   I ended-up writting internals and parsers(which I hate),
and giving up doing the stuff i was interested in...but I figured,
I'll do the neat stuff next time.
   Unfortunately, the same thing happened on the next project, Pygmalion.
I got involved because there was neat artsy Ui design stuff
and interesting interaction issues and some neat possibilities for video
spiffy interactions. Again there was too many people wanting to
work on the interesting stuff, and too few isolated pieces;
only this time I was unvoluntarily displaced. And I am
unable to do any of the interesting-issues stuff (mostly
that stuff has been voted out of the project totally).
   But worst of all...yup, I'm sitting here pouring out massive overtime
hours doing...most internal coding and parsers (which I *hate*).
And I'm trying hard to grin-and-bare it. But the fact that
(on the good side) I do have a small say in what I'm doing
(the parts that were approved by the team),
and that as implimentor I do get to put in tiny artsy variations
on other people's designs, is a consolation. (   :-) actually,
those will probably get removed in the upcoming screen design
meeting and user testing sessions.   :-)
Sorry, I always joke at the most ackward spots.
   So, after this long message...I have a question.
(This is an honest question. I'm not being sarcastic or curt
or however this might come across...just worried.)
This is my second project under you where I've ended up
doing all the parser writing because no-one else is.
Are you planning for me to be around in Sept.?
If so, is there a real chance that I'll be able to ever
be in charge of the user interface design for a project?
And if I'm continued,
am I going to be assigned the writing of parsers again ?
    In some sense sending this message to you is good,
because I should have a way of telling you what up on
my end. But, in other ways its bad, because because
I'm probably easier to deal with when my mouth is shut.
Besides, i'm sure you have enough problems as is
without hearing from me. But alot has happened, unfortunately,
with "my piece" of Pygmalion.
   Hope this letter wasn't a mistake.
        Yours,
          Brian