laurap says:
"hi, Daniel"
DanielP says:
Hi, I am just finishing lunch
laurap says:
and you didn't bring any for me?
DanielP says:
That would be a bit difficult :-)
(DanielP thinks: you just had breakfast no? )
laurap says:
yes, you're right...
DanielP says:
I wonder if Marcus is coming? From one of his post, I guess not
laurap says:
I think you're probably right. And PMR isn't coming either. Maybe we should think about rescheduling these meetings.
(DanielP thinks: I prefered the old schedule )
(laurap thinks: that wouldn't help with Marcus's situation, though )
DanielP says:
What is Marcus situation?
laurap says:
I think he's been going out of town a lot on the weekends.
(DanielP thinks: I know that he has "off loaded" many of his GNA's activities )
laurap says:
I think he's very busy IRL...
DanielP says:
but he is going to give a course for VOU
laurap says:
do you know what it's going to be about?
DanielP says:
some physics
laurap says:
not to change the subject, but... how many RL students are you planning to bring to this course?
(DanielP thinks: no idea if it is for beginners or advance ppl )
DanielP says:
don't know yet, we'll get the official number only in Septembe(probably from 0 to 25, with a realistic figure of 8 to 12 or 15)
laurap says:
So this will be all the students from one course that you'll be teaching?
DanielP says:
I am teaching two courses in the fall term, Data Struc and Algo and the course on Formal lang and automata (this is one will be mainly a compiler course)
At the moment I was planning to have their term project to be something for this course (don't on a volunteer basis or I "force" everyone to do this)
(DanielP thinks: don't know if on .... )
(laurap thinks: just typing something long... )
laurap says:
I was wondering basically what the RL students would be like, because I think it will affect the "diversity" of this course. I think that college students in general are a pretty diverse group, but students that have (1) enrolled in your particular college -- with whatever entrance requirements and focus it has, and (2) signed up for your class(es), with specific prerequisites and focus, would be a lot less so...
DanielP says:
At the moment the advantage that I see doing this, is that it will provide some "work power"
laurap says:
yes -- I didn't mean to imply it was necessarily a bad thing.
DanielP says:
we have to discuss the "procedure"
(laurap thinks: procedure? )
DanielP says:
For example, I like Jim's idea of interviews and "answer line
but we need people to do this. So far, what I have learn in this course, if you
propose something and you want to make it go forward, you essentially have to
do it yourself. but there is a limit to what a single person can do. We can ask
for "donation" but we need to have something more clearly define then what we
have at the moment"
laurap says:
oh, I see. BTW, I sent Jim mail -- GNA has a "reference librarian form" that's very similar to Jim's answer line idea.
DanielP says:
One good thing about ugrad is the unlimited enthousiam, if properly "geared"
laurap smiles
laurap says:
one thing we might want to do re interviews is set up special "office hours" for students to drop in and do interviews, rather than try to schedule individual ones.
DanielP says:
An the other hand, I worry about bringing in ugrad because they are not our typical audience. I got two volunteers for Nintendo session. I think that this Nintendo idea was targeted to a ugrad audience. So I have to conclude that we are servicing a different population
laurap says:
maybe, or maybe inertia has just set in...
Or maybe we should rethink what audience we're aiming at. If you think that your ugrad students will benefit from this course, maybe you'd like to always include some...
DanielP says:
I believe for the near future that the course will be open to everyone who register.
laurap says:
I didn't mean that it should have only people like your ugrad students, just that maybe we should expect to have some percentage of ugrads.
I don't think we can assume that the current population is typical. There are probably a number of people who signed up for this class because it was the only thing like it that they'd ever heard of. And when the class itself changes and becomes more organized (with lcs, etc.), I think that the way people react to the class will change.
DanielP says:
I think that there are (at least) two large groups of ppl interested in C++ at the moment, ppl in the private sector who would like to "test" the water and ppl in the university, who want to learn C++ or improve their knowledge. For example, a couple of weeks ago, I talked with a colleage and discuss to introduce C++ as the first programming (replacing Pascal), we have reached the conclusion that this will happen but for the moment we prefer to wait that there is a godd introductory book on this (btw I don't look at this course as a substitute for this, I think that this course server a more "useful" purpose)
(DanielP thinks: this course could serve ... )
laurap says:
yes, I think this course is more useful to people in the private sector (like me) who have fewer alternatives.
that's one reason why I think this course should be different things to different ppl -- if you compare this to a university settings, universities have
diffferent but similar courses where we might have just one.
DanielP says:
I was thinking to have a three level hypertextbook (beginner, intermediate, advance), we could have hyperlink to more advance section
laurap says:
yes -- I think we can do a lot with hypertext because the book won't have to be "linear".
did you read my (extremely long) "vision" message?
DanielP says:
and if it is a collaborative effort we can be more ambitious
DanielP [to laurap]:
I browse through it before connecting here, I will return
to read it more carefully later today
(DanielP thinks: I had to finish lunch :-) )
(laurap thinks: Daniel has his priorities straight. )
DanielP smiles
laurap says:
One of the reasons I was suggesting earlier that you might want to always bring in future students is that I'm concerned about being able to get CS professor-types to consult for this course when it's "finished".
I thought that maybe providing projects for RL students to work on would be an incentive...
DanielP says:
What do you mean by concerned exactly?
laurap says:
wait, I'm trying to find the right words for this...
I think we have a very good group of consultants now. I think a course like this will need to have a very good group of consultants forever if it's going to be successful. I don't think that this particular group will continue consulting forever (for example, I'll consult next "term" but probably not after that).
there's more...
So the question is how will you find people "like" the current consultants? I think it will be easy to find people like me -- there will always be people who take the class one term and consult for it the next term. But it will be harder, I think, to find people who start out as "experts" -- they won't for example, show up as students in the course.
Sunstone_Guest arrives from nowhere.
DanielP says:
I am not sure I agree with the last part.
laurap says:
about it being difficult to find experts?
DanielP says:
I think that, if we manage to keep a sufficiently good quality, then it will become a very interesting alternative, for a typical CS academic, to have his/her students taking the course and, we would ask as tuition that s/he particpates as consultant
laurap says:
yes -- that's why I think we might want to plan for a certain number of your RL students (or people like your RL students or RL students of people like you).
DanielP says:
Another aspect of the course is the projects, if we could set it up appropriatly, I would like to bring in some real life experiment being carried out. In Stroustrup lates't book he complains that there is not enough well documented software development effort so that we can learn form these experience, he is very keen on this. And I believe that this "course" could provide an interesting medium for doing this
laurap says:
that's a great idea. I don't think the idea of having students think up their own projects worked very well...
(laurap thinks: ( speaking of projects, I should write some documentation for mine)
one of these days. )
DanielP says:
After some trial and error, we will be in a better position to present students with a choice of option on how to follow the course (this is what I had in mind when i posted a message talking about badly applying information hiding in this course)
laurap says:
yes, and lcs will force ppl (students and consultants) to think about the alternatives up front.
DanielP says:
Marcus is keen on literate programming (he has proposed such a course). I definitely like the premise of litereate programming ie the documentation and the programming happen together (you would have to worry about your project documentation :-))
(DanielP thinks: wouldn't have ... )
DanielP says:
Should we record this session? I don't think anyone else will show up
laurap says:
I'm sure Marcus would appreciate your saying "I told you so" for him. :-)
DanielP says:
maybe we should not recorded the session :-)
laurap says:
I think you're right -- we've been here an hour...
I've been thinking about trying to make the transcript html-izer work with an emacs buffer of a session...
DanielP says:
As Marcus mentionned, it is better to use the recorder
laurap says:
I suppose we should turn the recorder on. (Then we can run out of things to say :-))
DanielP laughs
DanielP says:
The other thing that I wanted to know from Marcus, is that he wants to complete the whole tutorial. But there is officially two more week to go. We cannot really "dump" five chaps in two weeks
laurap says:
well, it would make the class more challenging :-)
which would you rather do, finish the class on schedule or work through the rest of the tutorial?
(DanielP thinks: or it would scare away the remaining students )
(laurap thinks: we could find out during Jim's exit interviews )
DanielP says:
I don't know why Marcus choose this deadline, but I find it, personnaly, very convenient since I will be offline for most of August
laurap says:
Okay, then we should probably just finish this chapter and templates. Going on vacation?
DanielP says:
The other thing about the tutorial is copyright, it is on a shaky ground, and I wouldn't want to problems with this after finishing up everything
laurap says:
I have a feeling that by the time we're finished with the tutorial (if we rework it for the next course) it won't be recognizable.
DanielP says:
Moving back to the university and getting my daughter used to this new env
I agree (about being not recognizable) but it doesn't change the copyright status
laurap says:
the alternative is to start from scratch (which I think is actually a good alternative, but labor-intensive) -- do you have a guess as to how many people will be working on it?
DanielP says:
not at all
laurap says:
has anyone been able to contact whoever holds the tutorial copyright?
DanielP says:
I was thinking that we should try to use a "template approach", ie we would extract the actual Corano part without disturbing the rest (this is a good app of templates)
I have asked marcus before the beginning of the course, and he said that he could reac anybody. But if you read the copyright carefully, what we have done is perfectly "illegal"
laurap laughs
(DanielP thinks: he couldn't reach )
laurap says:
I like the "template approach", but we might want to do a little more reorganizing than that.
(DanielP thinks: ( templates should be use when the "processing" is independent of)
the types (the hypertextbook should be independent of the underlying ascii
tutorial) )
DanielP says:
Like you said, it is labor-intensive, at least we have a start at the moment
laurap says:
Can you enlist any of your students to help? Or is the timing off?
DanielP says:
I will enroll some of my students to help the reorganization, but a whole rewrite would be asking too much (that's why this copyright issue bother me so much)
laurap says:
I think if we're going to do a whole rewrite, we'll need a lot of volunteers...
DanielP says:
that's the problem
laurap says:
Do you know any of those "typical CS academics" you mentioned earlier who might be willing to help at this point?
DanielP says:
I don't think many would be willing to undertake a whole rewrite of the tut, they would rather keep what they have for themselves and use a commercial publisher
laurap says:
too bad...
I wonder whether it would make sense to ask for help on the c++ newsgroup?
DanielP says:
There is many forums we can look for volunteers, and we have a "good looking" plan, it will work for sure. But we need an "infrastructure" to handle this "administrative" work. We don't have this at the moment from the GNA so we have to look in RL for help
laurap says:
GNA seems to be doing a lot of very different projects at once. Has anyone actually asked them for help on the text?
(DanielP thinks: ( lcs will hopefully work with students, but we have to nothing)
to handle a collaborative textbook effort except Marcus expertise )
laurap says:
we can automate some of the technical details like submitting chapter drafts, but we need someone with expertise for editing, etc.
DanielP says:
For example, going through everything that has been brought up on the mailing lists and integrate it in the hypertextbook would be a lot of very valuable work
laurap says:
yes -- actually I'd forgotten about all that material. I think there are actually two kinds of "advanced" information for the text -- advanced aspects of C++ itself and advanced programming concepts.
DanielP says:
in fact, just planning how to integrate this would be a good start, if we want the course material to "scale up" gracefully
laurap says:
yes -- even if the tutorial is completely rewritten, some of that work will transfer.
one problem I see with this tutorial is that it's basically a C++ tutorial, and includes almost nothing about design.
DanielP says:
This is not a problem for *this* tutorial because it is supposed to be an introduction to oop with C++, but it will become one, if the future versions evolve as I plan
laurap says:
I think it would be nice (but perhaps a little unrealistic) if we had something that was basically a text on oop that had hypertext links to text on "how to do this in C++".
DanielP says:
maybe I am bias but I prefer to look at oop from within C++ perspective
BillH has connected.
DanielP says:
Hi Bill
laurap says:
I'm actually speaking out of ignorance, having never actually taken an oop class...
laurap [to BillH]:
Hi.
BillH says:
Hi DanielP and LauraP
DanielP [to BillH]:
Do you have something that you would like to be discuss, we
have been talking about the course for the past two hours
BillH says:
Is there a consultant meeting?
(laurap thinks: this reminds me a lot of last week... )
DanielP [to BillH]:
I started two hours ago, but only Laura and I showed up
BillH reads a notice on Bulletin Board.
BillH reads a notice on Bulletin Board.
BillH says:
What time (GMT) is the consultant meeting should start?
laurap [to BillH]:
16:00
BillH says:
Oops. I thought it was for 18:00
laurap [to BillH]:
Where are you (geographically)?
BillH says:
What have you been talking about on the course?
BillH [to laurap]:
I'm in New Orleans
DanielP [to laurap]:
about oop vs c++, there is some language wars between the
Smalltalk camp and the C++ camp. I don't know any thing about Smalltalk but
this "contraversy" remind me a lot a similar situation that happened in
functional language community, where the Haskell language supported blasted
Standard ML for not being a pure functional language. I think this as to do
with "marketing"
laurap [to BillH]:
Mostly the future of the course and the text...
(DanielP thinks: I guess Im' getting tired I am making so many typing mistakes )
BillH says:
One thing about smalltalk is that everything is a class, like int, char, Bool
(BillH thinks: ( I like this: everyting being a class makes the language more)
uniform )
DanielP [to BillH]:
I have also heard that this is making it particularly
inefficient
laurap [to DanielP]:
I guess all I really meant was that the course could use a
little more "abstract" material on design issues, etc.
(DanielP thinks: ( there is not much material on design issues in the tut at the)
moment? )
BillH says:
Some students complained about that also. But shouldn't design be another course?
laurap says:
I think the focus of the text itself has been the mechanics of C++.
BillH agrees
laurap says:
I think that people should be able to take this course either as a design course or as a course on the C++ language.
DanielP says:
that's would be the ideal situation
laurap says:
some people (PMR in particular) have raised design issues on the listserv, but they're not an integral part of the text.
DanielP says:
I think that bringing these issues as an afterthought is not a very good pedagocical approach
laurap agrees
laurap says:
especially since they're often brought in "out of sync" with the "official" course...
DanielP says:
On the other hand, I guess PMR started mentionning about this because ppl ask for it in the survey
laurap says:
I think he also thinks it's possible that others in this course will have insights that he can use for RL issues and problems.
In a sense, PMR might be getting some of the same benefits from this course that he would if he were a student taking an advanced seminar...
DanielP says:
That's fine with me, but we should have isolate the "general public" from these discussion. PMR suggestion to have different mailing list might be an interesting approach to try
laurap says:
I think that some beginning students might benefit from being exposed to that kind of discussion. I know I did...
DanielP says:
PMR mentionned the isolation he was in when he first learned C++ and thought that this kind of course would help other ppl in a similar situation
laurap says:
but having various overlapping mailing lists might be a good idea (I know some people feel inundated by mail from this course).
(DanielP thinks: I was one of those )
laurap says:
Well, I certainly haven't felt isolated. :-)
BillH says:
What about using news groups instead of mailing lists. Do most people have access to newsgroups? Newsgroups better organize the mail into categories.
laurap says:
one thing this course has done for me is given me the motivation to actually learn C++.
Jacinth_Guest has arrived.
laurap [to BillH]:
You're right about the organization -- but anyone can post
to a newsgroup. I like the exclusivity of these mailing lists. And culturally
the mailing lists are different from most newsgroups (we want to encourage
people to post to the mailing lists questions that would be seen as annoyances
on most newsgroups).
DanielP [to BillH]:
I think that hyper-mailing list as the archive created by
Diego is superior to newgroups
laurap says:
on the other hand, I don't think the "use the first word of the subject as a keyword" idea works very well.
Jacinth_Guest reads a notice on Projects Board.
DanielP [to laurap]:
Many different hyper-mailing lists is probably the best
way to go (for example one would be an anonymous one for question)
(laurap thinks: having one anonymous list is a good idea )
Jacinth_Guest reads a notice on Projects Board.
(DanielP thinks: anonymous but restricted to the students population )
Jacinth_Guest reads a notice on Projects Board.
(DanielP thinks: we should have recorded this session! )
Jacinth_Guest reads a notice on Projects Board.
laurap says:
I have this all in an emacs buffer. I can make hypertext out of it.
Jacinth_Guest reads a notice on Projects Board.
Jacinth_Guest reads a notice on Projects Board.
Jacinth_Guest reads a notice on Projects Board.
DanielP [to laurap]:
let me do something

Jacinth_Guest reads a notice on Projects Board.
DanielP drops Dan's_tape.
Jacinth_Guest reads a notice on Projects Board.
Jacinth_Guest reads a notice on Projects Board.
Jacinth_Guest reads a notice on Projects Board.
Jacinth_Guest reads a notice on Projects Board.
Jacinth_Guest has puzzled look.
DanielP pastes:


this is a trial paste


DanielP says:
I haven't succeded in pasting my whole emacs buffer (I get a "ran out of ticks" error message, anyone know about this?)

(BillH thinks: I don't know. Can you try smaller chunks to paste? )
laurap says:
no, but don't worry about it -- I'll just html-ize mine.
(BillH thinks: laurap method might be simpler )
DanielP [to BillH]:
The small paste worked just fine
laurap says:
if you paste your whole buffer, it will look ugly anyway (you'll have both '>say hello' and 'You say "hello"' for everything you say)
DanielP thinks that Laura is right again
laurap says:
probably whatever @paste uses has some buffer that's overflowing...
laurap smiles
laurap says:
do we have anything regarding the current course to discuss?
(BillH thinks: I don't )
DanielP thinks that things are very quiet at the moment (ie summer vacation
time)
laurap says:
BTW, g++ 2.6.0 was announced Friday. Their template implentation is completely different from their 2.5.x template implementation. Everything I know is wrong...
DanielP says:
Should we call it a day?
DanielP [to laurap]:
That's the nice thing about the computer field, you have
to keep up to date!
laurap [to DanielP]:
Oh, I don't know -- in another ten minutes, the people who
think the meeting is still at 19:00 should start so show up. :-)
(laurap thinks: interesting definition of "nice" :-) )
Aquamarine_Guest joins you.
Aquamarine_Guest arrives from nowhere.
laurap says:
Seriously, though, I think I'll only be able to come up with "template" compiler notes for the 2.5.x g++ by the time the template chapter is announced.
BillH [to laurap]:
Maybe 2.6 will overcome all the template problems?
DanielP [to laurap]:
No one will blame you if you "only" do this
Aquamarine_Guest reads a notice on Bulletin Board.

laurap [to BillH]:
I haven't fetched 2.6 yet -- from the writeup it does look
easier to use, although I think the templates still need to live in .h files.
Aquamarine_Guest reads a notice on Bulletin Board.
Aquamarine_Guest reads a notice on Bulletin Board.
Aquamarine_Guest reads a notice on Bulletin Board.
(DanielP thinks: I have to go shorthly )
Aquamarine_Guest has disconnected.
Assistants of the local psychology institute arrive to cart Aquamarine_Guest
off to their dream-research labs.
(BillH thinks: I'm ready to go too )
DanielP waves Laura and Bill
laurap says:
let's call it a day, then..
laurap waves
BillH [to DanielP]:
Bye
DanielP says:
Goodby
DanielP has disconnected.
BillH [to laurap]:
bye