Re: Looking for a good Ada 95 book

Re: Looking for a good Ada 95 book

[Ray Toal]

Richard A. O'Keefe wrote:
> 
> mfeldman@seas.gwu.edu (Michael Feldman) writes:

> >If the lexical style is really sufficient reason to adopt or reject a
> >text, I'd say you're not reading for content...
> 
> Unfair.  Not reading *JUST* for content, maybe.
> Students in this part of the world use the style they were first
> taught, and if a later lecturer wants something different, they regard
> the _lecturer_ as incompetent and pedantic, and don't do it. 

My experience has been totally the opposite.  Students can adapt
easily to different styles.  When they get enough experience with coding
you can even have very useful discussions about the different styles
in regards to the importance of readability and consistency.
Consistency is a big deal - you wouldn't wnat to sprinkle lots of
underscores in C code because the standard library avoids them,
but you like to in Ada because its standard library is full of them.

I've never seen a student complain about formatting styles, but then
I've only been teaching 11 years.

By the way, I've said before, the content in Feldman's book is
first-rate and it's the best book I've seen, and the non-standard
formatting doesn't have to be a big deal at all.

Ray

Re: Looking for a good Ada 95 book

[Michael Feldman]

In article <3290C33B.1772@cse.eng.lmu.edu>,
Ray Toal  <rtoal@cse.eng.lmu.edu> wrote:

>I've never seen a student complain about formatting styles, but then
>I've only been teaching 11 years.

I get complaints, but not about reserved words. Some of my students -
the ones who did BASIC in high school - always moan about how I make
them do all that (gasp) indenting! I've not had a single student
complaint that I can recall about the reserved words. 

I always tell students they are welcome to develop their own styles, 
as long as they are consistent. Naturally, most of the real beginners
imitate the style they see, then in later courses they start feeling
their oats and develop their own. There's lots more variety in my
CS2 class; that's exactly as it should be.

As a reviewer of AQ&S, I'm certainly happy with it. I even provided
the dining phils code example there (with lower-case reserved words!).
But my first-year students have quite enough to read, and I think their
text should be enough to get through the course.

I make AQ&S available to more advanced students.

>By the way, I've said before, the content in Feldman's book is
>first-rate and it's the best book I've seen, and the non-standard
>formatting doesn't have to be a big deal at all.

Thank you, Ray!

Mike Feldman

Re: Looking for a good Ada 95 book

[Richard A. O'Keefe]

Ray Toal <rtoal@cse.eng.lmu.edu> writes:

>My experience has been totally the opposite.  Students can adapt
>easily to different styles.

Could you send some of your students here, PLEASE?
I could do with students like that, really I could.

>Consistency is a big deal - you wouldn't wnat to sprinkle lots of
>underscores in C code because the standard library avoids them,

This is argument is riddled with flaws.
(1) Many people *DO* use lots of underscores in C code.
    Just about everyone who hasn't been brainwashed into using
    "Hungarian" notation, anyway.

(2) The standard library does NOT avoid underscores.
    Here is a partial list of identifiers in the ISO standard C library
    that contain underscores:
	FLT_ROUNDS FLT_RADIX FLT_DIG FLT_EPSILON FLT_MANT_DIG FLT_MAX
	FLT_MAX_10_EXP FLT_MAX_EXP FLT_MIN FLT_MIN_10_EXP FLT_MIN_EXP DBL_DIG
	DBL_EPSILON DBL_MANT_DIG DBL_MAX DBL_MAX_10_EXP DBL_MAX_EXP DBL_MIN
	DBL_MIN_10_EXP DBL_MIN_EXP LDBL_MANT_DIG LDBL_EPSILON LDBL_DIG
	LDBL_MIN_EXP LDBL_MIN LDBL_MIN_10_EXP LDBL_MAX_EXP LDBL_MAX
	LDBL_MAX_10_EXP CHAR_BIT MB_LEN_MAX UCHAR_MAX USHRT_MAX UINT_MAX
	ULONG_MAX CHAR_MAX SCHAR_MAX SHRT_MAX INT_MAX LONG_MAX CHAR_MIN
	SCHAR_MIN SHRT_MIN INT_MIN LONG_MIN LC_ALL LC_COLLATE LC_CTYPE
	LC_MONETARY LC_NUMERIC LC_TIME decimal_point thousands_sep
	int_curr_symbol currency_symbol mon_decimal_point mon_thousands_sep
	mon_grouping positive_sign negative_sign int_frac_digits frac_digits
	p_cs_precedes p_sep_by_space n_cs_precedes n_sep_by_space p_sign_posn
	n_sign_posn jmp_buf sig_atomic_t SIG_DFL SIG_ERR SIG_IGN va_list
	va_start va_end va_arg ptrdiff_t size_t wchar_t FILENAME_MAX FOPEN_MAX
	fpos_t L_tmpnam SEEK_CUR SEEK_END SEEK_SET EXIT_FAILURE EXIT_SUCCESS
	MB_CUR_MAX RAND_MAX div_t ldiv_t clock_t time_t tm_sec tm_min tm_hour
	tm_mday tm_mon tm_year tm_wday tm_yday tm_isdst
    That's 104 identifiers, which is less than a half of the total, but
    not a _lot_ less than a half.

(3) The C standard actually reserves identifiers beginning with "_" for
    use in the library.

(4) If your library uses one convention, and you do not have a decent
    package system, closely imitating the library conventions is the
    LAST thing you want to do because it increases the risk of accidental
    name collision.

>but you like to in Ada because its standard library is full of them.

As explained above, this is *not* a good reason, and it is *not* the
reason why I use underscores in Ada identifiers.  I use spelled out
words separated with underscores for the reasons given in the AQ&S
guidelines.

>I've never seen a student complain about formatting styles, but then
>I've only been teaching 11 years.

Well, I've been teaching 7 years, and the students complained every year.
They don't see why they should lay *their* code out to please *someone else*,
even if the someone else is the person who has to mark their ruddy code.
This year, marking about 40 2nd year students, I found that _none_ of them
followed the layout rules they were told they would be marked down for not
following, and only _one_ of them even came close.

>By the way, I've said before, the content in Feldman's book is
>first-rate and it's the best book I've seen, and the non-standard
>formatting doesn't have to be a big deal at all.

Of *COURSE* it is not a big deal to someone who already fully agrees with it!

YOU have students who
 - don't mind following the style you give them
 - do have trouble knowing which are the keywords
WE have students who
 - DO mind very much following the style we give them
 - do NOT have trouble knowing which are the keywords
   They know because we TELL them, and give them lectures telling them
   what they mean.  If a student knows what "if then else end if" _means_
   s/he doesn't need to be reminded that "if", "then", "else", "end" are
   special words.  Is it really _possible_ for a student to remember how
   if-statements work without remembering that "if" is a special word?

And for die-hard "students who can remember if statements still need to
be reminded that if is a keyword" believers, why, putting "if" in lower
case when non-keywords don't begin with lower case letters is a perfectly
functional reminder.

-- 
Mixed Member Proportional---a *great* way to vote!
Richard A. O'Keefe; http://www.cs.rmit.edu.au/%7Eok; RMIT Comp.Sci.

Re: Looking for a good Ada 95 book

[Robert Dewar]

Mike says

"I always tell students they are welcome to develop their own styles,
as long as they are consistent. Naturally, most of the real beginners
imitate the style they see, then in later courses they start feeling
their oats and develop their own. There's lots more variety in my
CS2 class; that's exactly as it should be."

Ah! There is the fundamental difference. I strongly disagree with this.
Students should NOT be encouraged to think that figuring out your own
style is a normal and reasonable activity, on the contrary, they should
learn that in the real world, an external style is imposed and you
have to follow it. Encouraging this kind of individuality to me is
inconsistent with good engineering practice. It would be like telling
chemistry students in the lab to develop their own distinctive styles
of doing experiments.

So, it is not surprising that Mike doesn't see this as a big deal. He
notes that people tend to copy what they see at first, and this is why
I would like them to see something more standard.

By the way, I agree with others who like the *content* of Mike's book, but
there are some other very good books appearing, and, as I said earlier,
other things being equal, I would choose a book using standard style.
Perhaps I won't find one, perhaps I will!

Re: Looking for a good Ada 95 book

[Larry Kilgallen]

In article <dewar.848712351@merv>, dewar@merv.cs.nyu.edu (Robert Dewar) writes:

> Students should NOT be encouraged to think that figuring out your own
> style is a normal and reasonable activity, on the contrary, they should
> learn that in the real world, an external style is imposed and you
> have to follow it. Encouraging this kind of individuality to me is
> inconsistent with good engineering practice. It would be like telling
> chemistry students in the lab to develop their own distinctive styles
> of doing experiments.

If students do not experiment with styles, they will be ill-prepared to
contribute to the style discussion when a group such as that which wrote
GNAT chooses a style (as described many bytes ago in this newsgroup).

Larry Kilgallen

Re: Looking for a good Ada 95 book

[Michael Feldman]

In article <dewar.848712351@merv>, Robert Dewar <dewar@merv.cs.nyu.edu> wrote:
>Mike says

[snip]

>So, it is not surprising that Mike doesn't see this as a big deal. He
>notes that people tend to copy what they see at first, and this is why
>I would like them to see something more standard.

Sure they tend to copy what they see first, then they mature and start
developing their own styles. There's nothing wrong with that - it teaches
then to _think_ about developing a style (and NOT just lexical!) that is
courteous to the reader.

I am not training coders; I am educating college students. It is far more
important for them to 

- learn the fundamentals

- learn to _think_

- learn to express themselves articulately and clearly in both natural
  language and programming language

- learn to be ready for an industry in constant change and to view
  each new fad and "religion" with openmindedness and healthy skepticism, 
  not knee-jerk acceptance or rejection

That's why I'm dismayed that, of all the important educational
principles we could discuss here, we are wasting so much bandwidth on 
whether the keywords are uppercase or lowercase, for Heaven's sake!

It's a simple pedagogical device, not Holy Writ. It helps students
through their first year; if they are too stubborn to go with the flow
and change lexical style if a prof or a supervisor requires it, they
are definitely in the wrong major. Sheesh.

>By the way, I agree with others who like the *content* of Mike's book, but
>there are some other very good books appearing, and, as I said earlier,
>other things being equal, I would choose a book using standard style.

There are indeed a number of good books appearing; I welcome the
competition and have always subscribed to the notion that a rising
tide raises all ships. The more good books, the more schools doing Ada,
the more incentive for authors and publishers to revise books, and so
it goes, with each new book leapfrogging the older ones.

>Perhaps I won't find one, perhaps I will!

Naturally! Ice cream comes in lots of flavors, too.

Maybe our friends in industry don't realize the extent to which the 
"customer" for university textbooks is _teachers_, not students. The 
prof gets a free examination copy from the publisher, then adopts a 
text for the course. It's the students who get to _buy_ the books.:-)

I'm bowing out of this discussion of lexical stuff; I'll be happy
to discuss stuff that matters more!

Mike Feldman

Re: Looking for a good Ada 95 book

[jim hopper]

In article <1996Nov23.082018.1@eisner>
kilgallen@eisner.decus.org (Larry Kilgallen) writes:

> In article <dewar.848712351@merv>, dewar@merv.cs.nyu.edu (Robert Dewar) writes:
> 
> > Students should NOT be encouraged to think that figuring out your own
> > style is a normal and reasonable activity, on the contrary, they should
> > learn that in the real world, an external style is imposed and you
> > have to follow it. Encouraging this kind of individuality to me is
> > inconsistent with good engineering practice. It would be like telling
> > chemistry students in the lab to develop their own distinctive styles
> > of doing experiments.
> 
> If students do not experiment with styles, they will be ill-prepared to
> contribute to the style discussion when a group such as that which wrote
> GNAT chooses a style (as described many bytes ago in this newsgroup).
> 
> Larry Kilgallen


I have to agree with Larry here.  If the students are never taught to
think critically about an issue, only to blindly follow standards then
they will never be able to meaningfully participate in evolving these
standards!  Thats why students should be FORCED (in my not so humble
opinion) to use a variety of styles and to understand the good and bad
points between them.  Note, i am thinking not just in the limited case
of capitalization style but all elements of programming style
(including capitalization).

Roberts comment that styles in the real world are imposed upon you is
an excellent point. But in fact, in the real world during a career,
MULTIPLE styles will be imposed upon a programmer. For example in the
last two years i have had at least 3 different styles imposed upon my
by the various prime contractors we have worked with, and i have
imposed my prefered style on our Ada folks at SAIC in dayton for our
internal work. :-) The best way to teach this (again IMNSHO) is to
train them by imposing various styles during their training to shake
them out of their religious opinions, and force them to understand the
issues that underly stylistic choices.

best jim  

Re: Looking for a good Ada 95 book

[Robert Dewar]

Larry says

"If students do not experiment with styles, they will be ill-prepared to
contribute to the style discussion when a group such as that which wrote
GNAT chooses a style (as described many bytes ago in this newsgroup)."

Well that's like thinking that students who have never been taught about
gotos know nothing about gotos! I doubt that requiring a consistent style
in the very first course guarantees a set of graduates who will never have
an opinion on style!

The point is that this early on, it is too early for students to make a
considered judgment anyway. After all in a first course, you typically do
NOT fully explore the programming language design space, that comes later.

You might as well say that they should be able to freely choose the syntax
and semantics they use for writing programs in the first course, on the 
grounds that otherwise they will be ill-equipped to contribute to 
discussions on language design!

Re: Looking for a good Ada 95 book

[Robert Dewar]

jim says

"Roberts comment that styles in the real world are imposed upon you is
an excellent point. But in fact, in the real world during a career,
MULTIPLE styles will be imposed upon a programmer. For example in the
last two years i have had at least 3 different styles imposed upon my
by the various prime contractors we have worked with, and i have
imposed my prefered style on our Ada folks at SAIC in dayton for our
internal work. :-) The best way to teach this (again IMNSHO) is to
train them by imposing various styles during their training to shake
them out of their religious opinions, and force them to understand the
issues that underly stylistic choices.
"


Best guess is that jim has not taught a CS1 course. I often find that
computer professionals GREATLY overestimate what is appropriate to teach
at this level. Teaching anything about programming and abstraction is
very hard -- I agree with everything jim says, but it is NOT an appropriate
excercise in a first semester course in programming, which is what we are
talking about here!

Re: Looking for a good Ada 95 book

[jim hopper]

In article <dewar.848785112@merv>
dewar@merv.cs.nyu.edu (Robert Dewar) writes:

> Best guess is that jim has not taught a CS1 course. I often find that
> computer professionals GREATLY overestimate what is appropriate to teach
> at this level. Teaching anything about programming and abstraction is
> very hard -- I agree with everything jim says, but it is NOT an appropriate
> excercise in a first semester course in programming, which is what we are
> talking about here!

No i taught physics to a lot of freshmen, but not cs-1. :-)
I agree my perspective is different as i have to deal with the products
of the CS courses :-)  

BUT i was not suggesting that first year students be taught a variety
of styles in the first year.  What i was suggesting is that if you am
going to teach, over a four year degree, a number of styles, with a
goal of breaking students of the habit of assuming their first method
is the one true way, does it REALLY matter which of the several styles
you want to address over the four years comes first??

best jim

Re: Looking for a good Ada 95 book

[S. McLain]

jim hopper wrote:
> 
> In article <dewar.848785112@merv>
> dewar@merv.cs.nyu.edu (Robert Dewar) writes:
> 
> > Best guess is that jim has not taught a CS1 course. I often find that
> > computer professionals GREATLY overestimate what is appropriate to teach
> > at this level. Teaching anything about programming and abstraction is
> > very hard -- I agree with everything jim says, but it is NOT an appropriate
> > excercise in a first semester course in programming, which is what we are
> > talking about here!
> 
> No i taught physics to a lot of freshmen, but not cs-1. :-)
> I agree my perspective is different as i have to deal with the products
> of the CS courses :-)
> 
> BUT i was not suggesting that first year students be taught a variety
> of styles in the first year.  What i was suggesting is that if you am
> going to teach, over a four year degree, a number of styles, with a
> goal of breaking students of the habit of assuming their first method
> is the one true way, does it REALLY matter which of the several styles
> you want to address over the four years comes first??
> 
> best jimGentlemen, 

As an earth science student, I forced myself to take computer science even though it was 
not required.  I rebelled as much as I could against styles imposed upon me by my 
instructors.  I would not (and will not) flow chart anything(or create diagrams).  Also, 
I only added comments to avoid failing the class.  I look back at the code I wrote to 
help me in the major and can't understand a thing.

--Back to the point, now that I am beginning to teach myself Ada, I follow the industry 
standard style guide.  My reserved words are always lower-case, and my identifiers have 
beginning caps.  I even write a few comments.  As far as the reserved word caps or lower 
argument is concerned, lower-case makes sense.  How many reserved words must we learn?  
Not many.  What gets confusing and what should stand out are the identifiers that we 
come up with ourselves.  They change with every new application written.  

Students should be taught the standard style for the particular language they are 
working with, but perhaps not penalized too heavily for their stylistic errors.  They 
should be rewarded for producing programs which work and which have source code that 
their PEERS can follow.
skip mclain

Re: Looking for a good Ada 95 book

[Robert Dewar]

Mike Feldman said

"I am NOT "stuck" with the intro coure, and (AFAIK) neither is Robert.
I teach this course at least once a year, with willingness and
(usually) enthusiasm. Ditto the second ("CS2") course.
"

I just filled out my teaching preference form for next year, and I put
CS1 and CS2 at the top of my preferences -- of course I don't always
get what I prefer :-)

Re: Looking for a good Ada 95 book

[Michael Feldman]

In article <dewar.848785112@merv>, Robert Dewar <dewar@merv.cs.nyu.edu> wrote:

>Best guess is that jim has not taught a CS1 course. I often find that
>computer professionals GREATLY overestimate what is appropriate to teach
>at this level. Teaching anything about programming and abstraction is
>very hard -- I agree with everything jim says, but it is NOT an appropriate
>excercise in a first semester course in programming, which is what we are
>talking about here!

In my experience, even _professors_ who do not teach first-term courses
underestimate the difficulty of doing so, and overestimate what can be
taught. First-year (and even first-term) courses are, these days, being
heaped with all kinds of stuff so that the students are "prepared" for
all successor courses.

In other words, the basic definition is now coming down to "teach
EVERYTHING in the first year, EXCEPT the special area I want to teach in
my own course."

As anyone hanging around the first-year teacher "network" knows, I am
only _slightly_ exaggerating.

Robert and I may have our differences about reserved word case and
"standard" vs. course-specific styles, but I think he and I are in
violent agreement about the politics and difficulties of intro
courses.

Luckily, Robert and I are both pretty senior faculty folks. We can
use both experience and seniority as ammunition to resist the pressure 
to load the freshmen up with everything except the kitchen sink.
I really feel sorry for the junior faculty people who - in many
institutions - get "stuck" with teaching the intro course and have
neither the expertise nor the seniority to resist.

I am NOT "stuck" with the intro coure, and (AFAIK) neither is Robert.
I teach this course at least once a year, with willingness and
(usually) enthusiasm. Ditto the second ("CS2") course.

We are doing this because we both think it's important to build the
right foundation under the students' education. We may have our
differences on some of the details, but (I think) we're quite in
sync on the underlying fundamentals.

Mike Feldman

Re: Looking for a good Ada 95 book

[Darel Cullen]

In article <329841F4.66DE@worldnet.att.net>, "S. McLain" <JSMCL-
endny@worldnet.att.net> spoke thus :-

>--Back to the point, now that I am beginning to teach myself Ada, I follow the 
>industry 
>standard style guide.  My reserved words are always lower-case, and my 
>identifiers have 
>beginning caps.  I even write a few comments.  As far as the reserved word caps 
>or lower 
>argument is concerned, lower-case makes sense.  How many reserved words must we 
>learn?  
>Not many.  What gets confusing and what should stand out are the identifiers 
>that we 
>come up with ourselves.  They change with every new application written.  
>
>Students should be taught the standard style for the particular language they 
>are 
>working with, but perhaps not penalized too heavily for their stylistic errors.  
>They 
>should be rewarded for producing programs which work and which have source code 
>that 
>their PEERS can follow.
>skip mclain

I agree, but i think forcing yourself to learn a strict format is
generally a waste of time, much better to use an interactive formatting
tool (such as 'FORMAT' in the rational environment, or the Ada-mode in
emacs) so that you arent using precious time messing about with format,
when you could be cutting code.

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Darel J. Cullen                    Software Engineer
Email: Darel@djcull.demon.co.uk                      
Url: http://www.djcull.demon.co.uk/                 
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Re: Looking for a good Ada 95 book

[Suzanne B. Zampella]

Michael Feldman wrote:
> 
> In article <dewar.848712351@merv>, Robert Dewar <dewar@merv.cs.nyu.edu> wrote:
> >Mike says
> 
> [snip]
>  
> Sure they tend to copy what they see first, then they mature and start
> developing their own styles. There's nothing wrong with that - it teaches
> then to _think_ about developing a style (and NOT just lexical!) that is
> courteous to the reader.

OK, so teach them the importance of style by making them read and <oh,
no> manually trace large programs in different styles.  That has much
more real world educational value than changing your style in creation. 
I remember having 3 inches worth of compiler dumped on my desk with only
one "isn't this neat" comment next to a bubble sort!  I valued comments
from that day.

> That's why I'm dismayed that, of all the important educational
> principles we could discuss here, we are wasting so much bandwidth on
> whether the keywords are uppercase or lowercase, for Heaven's sake!

Hear, Hear!  THat's what pretty printers are for!
 

Suzanne Zampella

Re: Looking for a good Ada 95 book

[Darel Cullen]

In article <329AB5C4.2E11@worldnet.att.net>, "S. McLain" <JSMCL-
endny@worldnet.att.net> spoke thus :-
>
>
>I don't want to sound like an idiot, but are you saying that something 
>will detect what I'm writing or read what I've written and properly 
>format it?  If so, will it it run with GNAT 3.xx on a 486 pc?  I'm still 
>trying to figure out what "object_oriented is.
>-Skip

Absolutely yes.. 
Im not sure if you are familiar with emacs. if not get it.. preferably
for X11 (Xwindows) , and make sure to download the ada-mode for emacs
(you can load it with M-x <whatever the adamode is called>), try reading
the FAQ that comes along with comp.lang.ada, I think it mentions some
sites you can ftp the ada mode (basically a elisp program) that will
format your code for you. What emacs does is have a record of all
reserved words, and templates of how each structure is formatted, the
case of your code will be taken care of you realtime, but the
indentation can be invoke with the TAB key on each line , which will
correctly (most of the time!) place each source line in its correct
position.
The Rational command 'FORMAT' requires a rational environment
(expensive) usually restricted to large commercial operations, I've
generally found the FORMAT command to be 100% successful in formatting
code where as the ada mode (and i think i have the latest) tends to get
a bit confused in less standard constructs.

Hope this helps.


-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Darel J. Cullen                    Software Engineer
Email: Darel@djcull.demon.co.uk                      
Url: http://www.djcull.demon.co.uk/                 
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Re: Looking for a good Ada 95 book

[S. McLain]

Darel Cullen wrote:

> I agree, but i think forcing yourself to learn a strict format is
> generally a waste of time, much better to use an interactive formatting
> tool (such as 'FORMAT' in the rational environment, or the Ada-mode in
> emacs) so that you arent using precious time messing about with format,
> when you could be cutting code.

I don't want to sound like an idiot, but are you saying that something 
will detect what I'm writing or read what I've written and properly 
format it?  If so, will it it run with GNAT 3.xx on a 486 pc?  I'm still 
trying to figure out what "object_oriented is.
-Skip

Re: Looking for a good Ada 95 book

[Jim Carr]

mfeldman@seas.gwu.edu (Michael Feldman) writes:
>
>I am not training coders; I am educating college students. It is far more
>important for them to 

>- learn to _think_

 http://www.unitedmedia.com/comics/dilbert/archive/dilbert961113-8774.gif

  ;-)

>- learn to express themselves articulately and clearly in both natural
>  language and programming language

 Do you require that projects be submitted with a clearly written 
 cover memo / executive summary as well as documentation? 

 Has anyone experimented with a series of projects where a student is 
 required to use a routine from project N-1 that was written by another 
 student, to emphasize the need for clarity and documentation? 

 Does anyone use teams / collaborative learning in their CS classes? 

-- 
 James A. Carr   <jac@scri.fsu.edu>     |  "The half of knowledge is knowing
    http://www.scri.fsu.edu/~jac/       |  where to find knowledge" - Anon. 
 Supercomputer Computations Res. Inst.  |  Motto over the entrance to Dodd 
 Florida State, Tallahassee FL 32306    |  Hall, former library at FSCW. 

Re: Looking for a good Ada 95 book

[Richard A. O'Keefe]

mfeldman@seas.gwu.edu (Michael Feldman) writes:
>In article <dewar.848785112@merv>, Robert Dewar <dewar@merv.cs.nyu.edu> wrote:

>In my experience, even _professors_ who do not teach first-term courses
>underestimate the difficulty of doing so, and overestimate what can be
>taught. First-year (and even first-term) courses are, these days, being
>heaped with all kinds of stuff so that the students are "prepared" for
>all successor courses.

>We are doing this because we both think it's important to build the
>right foundation under the students' education. We may have our
>differences on some of the details, but (I think) we're quite in
>sync on the underlying fundamentals.

I would greatly value the opinions of the two of you as to what _should_
be taught in CS1.  In this department, I'm on record as saying that I
would be perfectly happy if students didn't learn about pointers at all
in 1st year (or OOP), but wish that the students _were_ taught the
elements of Polya-style problem solving and more particularly of _testing_.

The reason why I'd dearly love to see "How-to-solve-it" approaches
and testing taught in first year is that these are skills the students
_need_ in first year as well as in all the other years that follow.

(Oh yes, and remedial English.  Did I need to say that?)

-- 
Mixed Member Proportional---a *great* way to vote!
Richard A. O'Keefe; http://www.cs.rmit.edu.au/%7Eok; RMIT Comp.Sci.

Re: Looking for a good Ada 95 book

[Michael Feldman]

In article <dewar.848961096@merv>, Robert Dewar <dewar@merv.cs.nyu.edu> wrote:

>I just filled out my teaching preference form for next year, and I put
>CS1 and CS2 at the top of my preferences -- of course I don't always
>get what I prefer :-)

And here I thought senior professors got perks like this...:-)

('Course maybe there are profs more senior than you getting in
line to teach the intro courses...:-))

Mike Feldman

Re: Looking for a good Ada 95 book

[Michael Feldman]

In article <329BD3A1.1CD7@accessus.net>,
Suzanne B. Zampella <szampell@accessus.net> wrote:

>OK, so teach them the importance of style by making them read and <oh,
>no> manually trace large programs in different styles.  

This is a good idea, but not for intro courses, which was the subject
of all this discussion. Real beginners have enough trouble following
_one_ style, let alone many.

I agree with your suggestion for more advanced courses.

>That has much
>more real world educational value than changing your style in creation. 
>I remember having 3 inches worth of compiler dumped on my desk with only
>one "isn't this neat" comment next to a bubble sort!  I valued comments
>from that day.

But why on earth did that programmer use bubble sort in a compiler?:-)

Mike Feldman

Re: Looking for a good Ada 95 book

[Michael Feldman]

In article <57gpd3$1ri$1@goanna.cs.rmit.edu.au>,
Richard A. O'Keefe <ok@goanna.cs.rmit.edu.au> wrote:

[snip]

Ah, now some _substance_ in this discussion!

>I would greatly value the opinions of the two of you as to what _should_
>be taught in CS1.  In this department, I'm on record as saying that I
>would be perfectly happy if students didn't learn about pointers at all
>in 1st year (or OOP), but wish that the students _were_ taught the
>elements of Polya-style problem solving and more particularly of _testing_.

It's getting a bit off-topic for CLA, but what the heck; we have lots
of off-topic stuff here.:-)

If I were able to design an entire curriculum from scratch, I'd go
along with you here. Unfortunately, many of us don't have that luxury.
Every time we think of deleting a topic from the first year, we have to
get our colleagues' agreement on where that topic _will_ be taught. In
my experience, that sort of coordination is very hard to achieve in a
faculty of more than a few members.

So I've tended to stick pretty close to "classical" CS1 and CS2, with
the addition of a bunch more "software engineering" stuff (well, I guess
you could call it SE...:-)) instead of traditional monolithic Pascal
style. Even this has some costs. I have a feeling that I'll catch hell
from a colleague this year, because I added some more OO-ish (inheritance)
content to my CS2 course, which made me skip my usual 2.5 hours on
hash tables.

You might want to check a couple of URLs for some discussion on this
issue. I was on a SIGCSE panel in March that discussed the first
year without reference to specific languages. I laid out my "vision"
of the first year in a set of viewgraphs at

http://www.seas.gwu.edu/faculty/mfeldman/papers/sigcse96.html

Among these viewgraphs are a couple that summarize "classical" CS1/CS2.

The overall panel, which gives others' positions and some other sets
of viewgraphs is at

http://www.cs.duke.edu/~ola/slides/lang96.html

>The reason why I'd dearly love to see "How-to-solve-it" approaches
>and testing taught in first year is that these are skills the students
>_need_ in first year as well as in all the other years that follow.

Agreed; I'm going to think about whether I can do this without too
much diminution of the time I spend on other _expected_ topics.

>(Oh yes, and remedial English.  Did I need to say that?)

CSAB accreditation criteria over here require that CS programs focus
specifically on written and verbal expression skills. For details, see

http://www.acm.org/~csab

I guess in Australia you have similar problems to ours - often our 
foreign students speak and write better English (even if the syntax 
is not always right) than the native born.:-)

Mike Feldman

Re: Looking for a good Ada 95 book

[Larry Kilgallen]

In article <57i6n8$kvd@felix.seas.gwu.edu>, mfeldman@seas.gwu.edu (Michael Feldman) writes:

> But why on earth did that programmer use bubble sort in a compiler?:-)

To isolate the uppercase keywords, why else ? :-)

Larry Kilgallen

Re: Looking for a good Ada 95 book

[Robert Dewar]

Suzanne said

"Hear, Hear!  THat's what pretty printers are for!"

Fine, I know many people react that way, but if you review the thread,
you will see that I posted what I consider some serious objections to the
notion that pretty printers solve this problem, can you address these
objections?

Re: Looking for a good Ada 95 book

[Robert Dewar]

Mike says

"But why on earth did that programmer use bubble sort in a compiler?:-)"

Perhaps it was the most appropriate algorithm. I often find that students
go to the opposite extreme, they insist on using bazookas to execute
cockroaches in their code. In fact there are many instances where a simple
bubble sort is qiute appropriate if you never expect very large sets of
data to sort, and you think they will in general be pretty much in order
(an example would be the component clauses in a record rep clause, if part
of your processing involved sorting them into order).

Re: Looking for a good Ada 95 book

[Dale Stanbrough]

Jim Carr writes:

"Do you require that projects be submitted with a clearly written 
 cover memo / executive summary as well as documentation? 

 Has anyone experimented with a series of projects where a student is 
 required to use a routine from project N-1 that was written by another 
 student, to emphasize the need for clarity and documentation? 

 Does anyone use teams / collaborative learning in their CS classes? "

I suspect that almost all educators consider team work. However team work
is _really_ hard to get right. Often you end up with passengers in a team,
who pass having done nothing much (if they are hopeless, often other team
members take up the slack to ensure that _they_ pass).

Also on giving them work from other students... what if the work is
fundamentally flawed? How do you assess what they have done? Students 
don't like being penalised because of the failings of other students.

All of what you suggest is great - however the devil truly is in the
details.

Dale

Teaching Team/Maintenance Programming (was: Looking for a good book)

[Larry Kilgallen]

In article <57noj5$j8e$1@goanna.cs.rmit.edu.au>, Dale Stanbrough <dale@goanna.cs.rmit.EDU.AU> writes:

> I suspect that almost all educators consider team work. However team work
> is _really_ hard to get right. Often you end up with passengers in a team,
> who pass having done nothing much (if they are hopeless, often other team
> members take up the slack to ensure that _they_ pass).
> 
> Also on giving them work from other students... what if the work is
> fundamentally flawed? How do you assess what they have done? Students 
> don't like being penalised because of the failings of other students.

Although these concerns are real, they also exist in industry.
To the extent that team programming and maintenance programming
are a _given_ in the real world, devising methods to "teach" them
(or at least "experience" them) in an educational setting is quite
important.

Regarding the maintenance programming problem, would it be possible
to give each team some equally garbage code created by some team from
a previous year?  This actually provides an even better emulation of
industry, since one cannot go down the hall to get help from a "departed
programmer".

Larry Kilgallen

Re: Looking for a good Ada 95 book

[Frank Manning]

In article <57i65s$jkq@felix.seas.gwu.edu> mfeldman@seas.gwu.edu
(Michael Feldman) writes:

> In article <dewar.848961096@merv>, Robert Dewar <dewar@merv.cs.nyu.edu>
> wrote:
>
>> I just filled out my teaching preference form for next year, and I put
>> CS1 and CS2 at the top of my preferences -- of course I don't always
>> get what I prefer :-)
>
> And here I thought senior professors got perks like this...:-)
> 
> ('Course maybe there are profs more senior than you getting in
> line to teach the intro courses...:-))

Whoa, whoa, hold on a minute, time out.

Do I understand this right? Senior professors standing in line to teach
intro courses? What planet did you guys say you were from? :-)

I thought Standard Operating Procedure was to offload intro courses on
grad student TA's. And I thought academia's Holy Grail is to be promoted
out of teaching altogether...

-- Frank Manning

Re: Looking for a good Ada 95 book

[Michael Feldman]

In article <57q6g7$ako@news.ccit.arizona.edu>,
Frank Manning <frank@bigdog.engr.arizona.edu> wrote:

>Whoa, whoa, hold on a minute, time out.

>Do I understand this right? Senior professors standing in line to teach
>intro courses? What planet did you guys say you were from? :-)

>I thought Standard Operating Procedure was to offload intro courses on
>grad student TA's. And I thought academia's Holy Grail is to be promoted
>out of teaching altogether...

These questions merit semi-serious answers.

I talk to lots of intro-course teachers, on the net,
at conferences, etc. The only generalization here is that there are
no generalizations. It really varies with each school's "culture."

For example, the US is filled with colleges/universities that are
either 4-year colleges with no grad programs, or in some cases have
relatively small grad programs and so give a very high priority to their
undergrad programs. 

One thinks here of the many, many 4-year state campuses, but also of 
first-rate private schools like Williams, Dartmouth, Brown, Princeton, 
and others. In all these the intro courses are taught by regular faculty, 
and sometimes very senior regular faculty at that. I got my B.S.E. at
Princeton, for example, and (as I recall) every intro course teacher 
I had was established and well-known in his field. Grad students taught
labs and recitations. This was 30 years ago, but not much has changed 
in places like that.

My son is a freshman at Brown and reports a similar culture there.
The grad students are _assistants_, with senior folks doing the
lectures.

In big Ph.D.-granting universities, the culture still varies a lot from
place to place. I am aware of some in which the intro courses are
taught by senior faculty, others where grad teaching assistants
hold sway, and still others where the intro courses are taught by
non-tenured (and often non-tenurable) "lecturers" whose continuing
employment is at the mercy of the senior tenured faculty.

Robert Dewar is quite senior at NYU, and I am quite senior (22 years)
at GW, and we both enjoy teaching these foundation courses. I can't
speak for Robert, but I teach everyone from freshmen to doctoral 
students, sometimes both in the same semester.

One last point - the CSAB criteria for accreditation of US undergrad 
programs in computer science (http://www.acm.org/~csab) have something
to say on this issue:

   The number of full-time-equivalent (FTE) faculty required of the 
   program will be influenced by such factors as the number of students 
   in the program, the number of courses required by the program, the 
   demand for computer science courses by non-computer science majors, 
   the existence of other programs in which the faculty are involved, 
   and the teaching load of the faculty. 

   There must be sufficient FTE faculty with primary commitment to the 
   program to provide continuity and stability. Qualified instructors 
   other than full-time faculty may be used in a supplemental role, but 
   full-time faculty should oversee all course work and should cover at 
   least 70% of the total classroom instruction. 

   etc.

Of course, this does not say that _intro_ courses should be taught by
FT faculty, but in CSAB-accredited programs (see list on the web site)
they often are anyway.

Mike Feldman
------------------------------------------------------------------------
Michael B. Feldman -  chair, SIGAda Education Working Group
Professor, Dept. of Electrical Engineering and Computer Science
The George Washington University -  Washington, DC 20052 USA
202-994-5919 (voice) - 202-994-0227 (fax) 
http://www.seas.gwu.edu/faculty/mfeldman
------------------------------------------------------------------------
       Pork is all that money the government gives the other guys.
------------------------------------------------------------------------
WWW: http://www.adahome.com or http://info.acm.org/sigada/education
------------------------------------------------------------------------

Re: Looking for a good Ada 95 book

[Robert Dewar]

iFrank Manning says

"Do I understand this right? Senior professors standing in line to teach
intro courses? What planet did you guys say you were from? :-)"

Well I know that some residents of this country consider New York to be
another planet, but in fact we have a number of senior professors here
at NYU who are actively interested in teaching the beginning course,
including me :-)

Re: Looking for a good Ada 95 book

[Richard A. O'Keefe]

Mike Feldman wrote
[that he _chooses_ to teach CS1 and CS2]
Robert Dewar wrote
[that he put CS1 and CS2 at the top of his preferences]

I have never put CS1 at the top of _my_ preferences for one very
simple reason:  I am terrified of the responsibility.  I think a
lot of education "status" is backwards:

	early parenting is more important than
	kindergarten is more important than
	primary school is more important than
	secondary school is more important than
	first year of tertiary education is more important than
	later year undergraduate education is more important than
	graduate/postgraduate education.

By the time the students are in _my_ hands, they _ought_ to be mature
enough to take responsibility for their own education, using me as a
resource.  It doesn't actually _work_ that way, because resources have
not been applied appropriately at lower more important levels.

Last Saturday I heard that most (school) science teachers "wish they
were doing something else", which has me as a new parent _horrified_.
I suppose it won't matter much; the price of tertiary undergraduate
science education is about to double, and physics departments around
the country are downsizing and/or merging with engineering. I dunno
who bought this government, but it's sure not the government _I_ paid for.

Any student who gets Feldman or Dewar in CS1 is fortunate indeed.

-- 
Govt saves money by cutting legal aid, guilty plea rates soar;
poverty is a crime!  (See also recent Sci.Am.)
Richard A. O'Keefe; http://www.cs.rmit.edu.au/%7Eok; RMIT Comp.Sci.

Re: Looking for a good Ada 95 book

[Frank Manning]

Robert says

> Well I know that some residents of this country consider New York to be
> another planet, but in fact we have a number of senior professors here
> at NYU who are actively interested in teaching the beginning course,
> including me :-)

I hope it was clear from my article that I think this is terrific.
Although it's hard to judge by Usenet articles alone, certainly by
that standard I think students of both Robert and Mike are fortunate
indeed to have them as teachers.

-- Frank Manning

Re: Looking for a good Ada 95 book

[Laurent Gasser]

In article <57noj5$j8e$1@goanna.cs.rmit.edu.au>, Dale Stanbrough <dale@goanna.cs.rmit.EDU.AU> writes:
> Jim Carr writes:
>  
>  "Has anyone experimented with a series of projects where a student is 
>  required to use a routine from project N-1 that was written by another 
>  student, to emphasize the need for clarity and documentation?"
>
... 
> Also on giving them work from other students... what if the work is
> fundamentally flawed? How do you assess what they have done? Students 
> don't like being penalised because of the failings of other students.
> 
> Dale

Is this possible to distribute only the works with a certain level of
quality.  Something like not every copy returned to the teacher, but
works which were rated A-B, no less.  Or for each student, a set of
three works from which to choose, with motivated rationals for the
selected version.

-- 
Laurent Gasser (lga@sma.ch)
Computers do not solve problems, they execute solutions.

Re: Looking for a good Ada 95 book

[John English]

Jim Carr writes:
>  Has anyone experimented with a series of projects where a student is 
>  required to use a routine from project N-1 that was written by another 
>  student, to emphasize the need for clarity and documentation?"

At Brighton we used to run an exercise for our 2nd year students
which we called a "Software Hut" (i.e a small software house)
where at the end of phase 1 each group had to market their
product to the other groups for use in phase 2.  Worked quite
well as I recall, although I was never directly involved; it
got them doing presentations, sales pitches and advertising,
and also gave them an insight into product evaluation. There
was an article in the Computer Bulletin about it a few years
back (sorry, but I don't have the details).

---------------------------------------------------------------
 John English              | mailto:je@brighton.ac.uk
 Senior Lecturer           | http://www.comp.it.bton.ac.uk/je
 Dept. of Computing        | fax: (+44) 1273 642405
 University of Brighton    |
---------------------------------------------------------------

Re: Looking for a good Ada 95 book

[Michael Feldman]

In article <57tlca$5to$1@goanna.cs.rmit.edu.au>,
Richard A. O'Keefe <ok@goanna.cs.rmit.edu.au> wrote:
>Mike Feldman wrote
>[that he _chooses_ to teach CS1 and CS2]
>Robert Dewar wrote
>[that he put CS1 and CS2 at the top of his preferences]

>I have never put CS1 at the top of _my_ preferences for one very
>simple reason:  I am terrified of the responsibility.  

I agree with you. But the older I get, the more I realize "it's a
rotten job, but somebody's got to do it, so better they try to do it well."

[snip]

>By the time the students are in _my_ hands, they _ought_ to be mature
>enough to take responsibility for their own education, using me as a
>resource.  It doesn't actually _work_ that way, because resources have
>not been applied appropriately at lower more important levels.

Hmmm - I'm not sure if I should be comforted by the fact that it's not
just in the US that we have these problems.:-)

>Last Saturday I heard that most (school) science teachers "wish they
>were doing something else", which has me as a new parent _horrified_.

Yep. One of my more interesting and creative doctoral students was a 
burned-out high-school math teacher who gave up and joined the military.
(The military had the good sense to send this individual back to GW
for a Ph.D...:-))

>I suppose it won't matter much; the price of tertiary undergraduate
>science education is about to double, and physics departments around
>the country are downsizing and/or merging with engineering. I dunno
>who bought this government, but it's sure not the government _I_ paid for.

This sounds _awfully_ familiar.

Are you talking about the Australian government, or the American one?
Maybe your PM and our President had some interesting discussions last
week.:-)

>Any student who gets Feldman or Dewar in CS1 is fortunate indeed.

Wow! I sure hope you're right. I teach CS1 enthusiastically; I hope I do
it _well_. 'Course I'm ruining my students' minds with those *&^%%$#
uppercase reserved words!:-)

Mike Feldman

Re: Looking for a good Ada 95 book

[Dale Stanbrough]

Larry Kilgallen writes:

"Although these concerns are real, they also exist in industry.
 To the extent that team programming and maintenance programming
 are a _given_ in the real world, devising methods to "teach" them
 (or at least "experience" them) in an educational setting is quite
 important."

Yes I understand this. However when someone leaves a company and
moves on to another, their lack of ability doesn't reflect on the
company. However when someone graduates from a University/College/
whatever, they are given a piece of paper which says "we think this
person _is_ good enough". Often the _only_ thing a University has
of worth to employers is a reputation. If you sully that with students
who get a free ride, you are in trouble.


"Regarding the maintenance programming problem, would it be possible
 to give each team some equally garbage code created by some team from
 a previous year?  This actually provides an even better emulation of
 industry, since one cannot go down the hall to get help from a "departed
 programmer"."

Great idea. I even did it this semester in my Concurrent programming 
subject where students had to make old code concurrent. Students quickly
appreciate the value of comments, identifiers, and working, _well tested_
code :-).


Dale

Re: Looking for a good Ada 95 book

[Michael Feldman]

In article <584rno$294$1@goanna.cs.rmit.edu.au>,
Dale Stanbrough  <dale@goanna.cs.rmit.EDU.AU> wrote:

>Yes I understand this. However when someone leaves a company and
>moves on to another, their lack of ability doesn't reflect on the
>company. However when someone graduates from a University/College/
>whatever, they are given a piece of paper which says "we think this
>person _is_ good enough". Often the _only_ thing a University has
>of worth to employers is a reputation. If you sully that with students
>who get a free ride, you are in trouble.

Well, of course in the US at least, we give students grades, and based
on their overall grade average, they graduate. On the standard US
scale of 0.0 (fail) to 4.0 (A), in most schools students will graduate
if their average is 2.0 or slightly above (at GW it's 2.0 overall,
2.2 in CS and other technical courses). It's reasonable to assume
that grades are reasonably representative of ability; over the years
they've gotten a bit inflated, but still, a C student is a C student.

Sure, universities have reputations, and I'd definitely like to know
about it if one of my good students is disappointing an employer. On the
other hand, if you hire a C student, you get what you deserve.:-)

Mike Feldman

Re: Looking for a good Ada 95 book

[Debora Weber-Wulff]

: Jim Carr writes:
X-Newsreader: TIN [version 1.2 PL2]

: "Do you require that projects be submitted with a clearly written 
:  cover memo / executive summary as well as documentation? 

:  Has anyone experimented with a series of projects where a student is 
:  required to use a routine from project N-1 that was written by another 
:  student, to emphasize the need for clarity and documentation? 

:  Does anyone use teams / collaborative learning in their CS classes? "

I wrote about such projects (among other things) in 
"Software-Oriented Exercises for Learning Teamwork", 
In: Softwaretechnik-Trends, Vol. 16,
Nr. 1, Febr. 1996, S. 70-74
This issue of STT contains the proceedings of the ISEEW'96
(International Software Engineering Education Workshop). I've
tried to recreate the .ps files for you but the blasted LaTeX
implementation here has been changed and some MetaFont file can't
be found :-( 
If anyone is interested, bug me about it, and I'll bring a copy
from home where my LaTeX works as I need it to.... 
--
Debora Weber-Wulff (Professorin fuer Softwaretechnik und Programmiersprachen)
Technische Fachhochschule Berlin, FB Informatik, Luxemburger Str. 10, 
13353 Berlin, Germany        email: weberwu@tfh-berlin.de   
<http://www.tfh-berlin.de/~weberwu/> 
--
Debora Weber-Wulff (Professorin fuer Softwaretechnik und Programmiersprachen)
Technische Fachhochschule Berlin, FB Informatik, Luxemburger Str. 10, 
13353 Berlin, Germany        email: weberwu@tfh-berlin.de   
<http://www.tfh-berlin.de/~weberwu/>