From pegasus.ATT.COM!hansen Tue Jul 31 08:15 EDT 1990
Received: by pyxis; Tue Jul 31 08:15 EDT 1990
FROM:       hansen@pegasus.ATT.COM (t.l.hansen)
TO:         coma!ehg
DATE:       31 Jul 1990   8:12 EDT
SUBJECT:    re: C++ answers

< As far as I can tell, nothing has arrived at research from you.

I shipped a cpio file to you last night. Please let me know when it's in
place and I'll try accessing the archive. Thanks.

					Tony

------m pegasus!hansen Tue Jul 31 18:26:25 EDT 1990
From-: Eric Grosse  ehg@research.att.com 908-582-5828
To-: pegasus!hansen
Date-: Tue Jul 31 18:22:46 EDT 1990
Subject: netlib

Your "answers" are installed here;  please give the system a try.  I decided
that all that subdirectory structure wasn't necessary in this application.
So to simplify my life and bring the structure into line with the rest of
netlib, I flattened out the tree.  To be more precise,
$ /bin/ls -C /netlib/c++
2.1    2.6    3.15   3.9    4.2    5.11   6.1    6.3    7.3    8.12   index
2.10   2.7    3.16   4.1    4.3    5.12   6.10   6.4    7.4    8.13   shape
2.11   2.8    3.17   4.10   4.4    5.2    6.12   6.5    7.5    8.2    tools
2.12   2.9    3.2    4.11   4.5    5.3    6.13   6.6    7.6    8.3
2.13   3.1    3.3    4.12   4.6    5.4    6.14   6.7    7.7    8.4
2.14   3.10   3.4    4.13   4.7    5.5    6.15   6.8    7.8    8.5
2.2    3.11   3.5    4.14   4.8    5.6    6.16   6.9    7.9    8.6
2.3    3.12   3.6    4.15   4.9    5.7    6.17   7.1    8.1    8.7
2.4    3.13   3.7    4.16   5.1    5.8    6.18   7.10   8.10   8.8
2.5    3.14   3.8    4.17   5.10   5.9    6.2    7.2    8.11   8.9

If you look closely, you'll see that 6.11 is missing for some reason.
Can you send that to me again?

So you can tell a user to mail netlib the message "send index for c++"
to get the general notes, or "send 2.1 from c++" to get an answer.

Best wishes
Eric

------m pegasus!hansen Tue Jul 31 18:31:04 EDT 1990
From-: Eric Grosse  ehg@research.att.com 908-582-5828
To-: pegasus!hansen
Date-: Tue Jul 31 18:28:59 EDT 1990
Subject: last minute change

I suddenly realize that there will be other C++ stuff eventually.
I'm going to put your files in a subdirectory "answers".  So, change
those commands to be  "send index from c++/answers" and
"send 2.1 from c++/answers".

From pegasus.ATT.COM!hansen Tue Jul 31 19:14 EDT 1990
Received: by pyxis; Tue Jul 31 19:14 EDT 1990
FROM:       hansen@pegasus.ATT.COM (t.l.hansen)
TO:         coma!ehg
DATE:       31 Jul 1990  18:54 EDT
SUBJECT:    re: netlib

< Your "answers" are installed here;  please give the system a try.  I
< decided that all that subdirectory structure wasn't necessary in this
< application. So to simplify my life and bring the structure into line with
< the rest of netlib, I flattened out the tree.  To be more precise,

Thanks. The flattened structure is just fine. I had linked the README file
into the top level; would that be appropriate? Also, I had left the README
file separate in each of the directories and using "send 2.x from c++" just
sends the shar file without the README. (I was trying to save space by using
a link for all the README contents.) Would it be better to repackage
everything with the README files enclosed within the shar files?

I just received the index and see that it's a reworking of my README file.
At the end of this message, I've put a slightly modified version. Would it
be possible to have a pointer to the index enclosed in the return mail that
netlib sends for each archive?

< If you look closely, you'll see that 6.11 is missing for some reason. Can
< you send that to me again?

I held back on it due to a reported portability buglet. After I fix it, I'll
send it to you. Okay?

< So you can tell a user to mail netlib the message "send index for c++" to
< get the general notes, or "send 2.1 from c++" to get an answer.

I'll give it a whirl. I'll eventually be posting a note to comp.lang.c++
after I verify that everything works okay.

< I suddenly realize that there will be other C++ stuff eventually. I'm
< going to put your files in a subdirectory "answers".  So, change those
< commands to be  "send index from c++/answers" and "send 2.1 from
< c++/answers".

That's definitely more appropriate. Might I suggest the directory name
"answerbook" instead of "answers"?

Thank you!

					Tony

------------------------------------------------------------------------
Notes on the C++ code for "The C++ Answer Book" by Tony L. Hansen,
Addison-Wesley, 1990, ISBN 0-302-11497-6.

...

------m pegasus!hansen Tue Jul 31 19:30:52 EDT 1990
From-: Eric Grosse  ehg@research.att.com 908-582-5828
To-: pegasus!hansen
Date-: Tue Jul 31 19:28:16 EDT 1990
Subject: changes

Ok, the new index is in place.  The style in netlib so far has mostly been
to put this sort of information in the index file, which everyone is
expected to read before requesting any of the specific items.  This way
someone who asks first for one exercise and then the next day another,
doesn't have to download and read the same instructions.

The name has been changed to answerbook, and I left in answers as an alias.

From pegasus.ATT.COM!hansen Sat Aug 25 20:43 EDT 1990
Received: by pyxis; Sat Aug 25 20:43 EDT 1990
FROM:       hansen@pegasus.ATT.COM (t.l.hansen)
TO:         coma!ehg
DATE:       25 Aug 1990  20:19 EDT
SUBJECT:    re: answerbook

< Subject: minmax.h

< David Forslund at LANL asks where to get minmax.h.  I could be snide and
< say Appendix B, but obviously it should be online.  Are the two of us
< overlooking something?  Sorry if this should be obvious, I haven't pawed
< around in your library much yet.

Others have also asked the same thing and I just told them to look in
Appendix B. The header files were so small that I didn't think anyone would
have problems retyping them. Anyway, I've shipped you appendixB.shar to add
to the collection.

< I've tried to clarify the index in response to your complaint.  The
< original idea was that you would have the book in front of you and would
< request a single exercise solution, not try to download a couple
< megabytes.

It looks good. I've updated it to include info on appendixB, plus a couple
of other notes.

How much traffic have you seen for my sources?

					Tony

From pegasus.ATT.COM!hansen Tue Sep  4 08:44 EDT 1990
Received: by pyxis; Tue Sep  4 08:44 EDT 1990
FROM:       hansen@pegasus.ATT.COM (t.l.hansen)
TO:         att!relay.EU.net!eurtrx!hansk
CC:         research!ehg
DATE:        4 Sep 1990   8:29 EDT
SUBJECT:    re: send index from c++/answerbook FAILS all the time
REFERENCES:  <9009041143.AA21484@cs.eur.nl>

< I try to send the msg: SEND INDEX FROM C++/ANSWERBOOK to
< netlib@research.att.com but my colleague who tried the same thing and me
< failed : no answer.

I just verified that the netlib server is working just fine. One possibility
is that the server is having problems parsing your return address. I got
"relay.EU.net!eurtrx!hansk" as the return address. (I hope this mail gets
through.)

Eric, could you check your logs? Can the server syntax be extended to accept
a return address, as in:

	send index from c++/answerbook to relay.EU.net!eurtrx!hansk

< I am interested in the code on pg 395 .. 432  discrete event simulation in
< the DEMOS-style. I would like to try to translate the SIMULA-based
< DEMOS-package into C++ with the help of your code in your well-written
< answerbook. But I like to try out your code first in ZORTECH ZTC and I do
< not succeed getting it. What do I do wrong ?            hans kraamer

I'm glad you like the book. When I wrote it, I was using a much earlier
version of ztc; current versions of ztc should compile the code with only a
few minor adjustments to add "const" to some of the function definitions.

					Tony Hansen
				att!pegasus!hansen, attmail!tony
				    hansen@pegasus.att.com

!/bin/mail pegasus.ATT.COM!hansen
Subject: return address

People can put a line "Really-From: valid-return-address" near the bottom
of the mail header and netlib will use that.  Here's what the logfile
shows for eurtrx
1990  8 21 0540 [relay.EU.net!eurtrx!hansk] 4826 c++/answerbook / index
1990  8 21 0540 [relay.EU.net!eurtrx!hansk] 2550 c++/answerbook / tools
1990  8 21 0540 [relay.EU.net!eurtrx!hansk] 23940 INDEX
1990  8 28 0948 [relay.EU.net!eurtrx!evas] 6751 c++/answerbook / index
1990  8 28 0949 [relay.EU.net!eurtrx!hansk] 6760 c++/answerbook / index
As best I can tell, nothing bounced.

Eric
!D
