openwebwork-devel — WeBWorK development discussion

[Fwd: Re: [WWdevel] feedback e-mail subject line]

[John J. <jj...@as...>]

Oops, I meant to send this to the list originally.

John

Re: [WWdevel] feedback e-mail subject line

[Arnold P. <ap...@ma...>]

At 10:02 AM 9/16/2005, P Gavin LaRose wrote:

Hi,

This sounds like a good idea and a number of people have asked for 
something along these lines.

Another method of doing this would be to have 
$mail{feedbackRecipients}  be a ref to a user defined procedure which 
takes as parameters e.g. [CourseName] [SectionNum] [RecitationNum] 
[User Name]  [setNum] [probNum] and outputs a ref to an array of 
email addresses.  If we do this, the procedure should also allow the 
cc field to be filled in (e.g. send email to a TA with a cc to a 
prof).  Also the procedure could generate a subject line (say with a 
default subject line as we do now with the addition of [section] 
[recitation] if the generated subject line is null). The default 
procedure would do just what we do now except that I would add 
[section] [recitation] to the subject line.

This would be more flexible, but I'm not sure it's worth the trouble 
--- how many people would make use of it. If we do it, we should 
provide a few standard examples (e.g. email goes to prof a or b 
depending on course/section).  Just doing what Gavin suggests would 
satisfy most users.

Arnie

>Hi all,
>
>I'd like opinions on the desirability of the following, and, if it's 
>desirable, suggestions for a better way of doing it.
>
>Currently the WeBWorK system generates feedback e-mails with the subject line
>WeBWorK feedback from [coursename]: [User Name] set [setNum]/prob [probNum]
>
>We always have some big courses where there are multiple sections of 
>the course which have different instructors or TAs.  It would be 
>convenient for those courses to have the subject line to be something like
>
>WeBWorK feedback [CourseName] ([SectionNum]-[RecitationNum]): [User 
>Name] set [setNum]/prob [probNum]
>
>so that instructors could filter their incoming WeBWorK feedback 
>e-mail to isolate those from students in their section(s).
>
>In the past I've done this by just hacking the code in Feedback.pm 
>to accomplish this, but it seems it might be better to have a 
>setting in global.conf or course.conf that allows the customization 
>of the feedback subject line.  e.g., using some sort of format 
>characters---say,
>     %c = course name
>     %u = user name
>     %s = set number
>     %p = problem number
>     %x = section number
>     %r = recitation number
>then the default subject line could be
>     $mail{subjectLine} = 'WeBWorK feedback from %c: %u set %s/prob %p';
>and in course.conf it could be reset, if desirable, to something else (for me,
>     $mail{subjectLine} = 'WeBWorK feedback %c (%x-%r): %u set %s/prob %p';
>).
>
>This would require adding the $mail{subjectLine} value to 
>global.conf file, and doing some substitutions on that in Feedback.pm.
>
>Does this seem like a good idea?  If so, is there a better way of doing it?
>
>(The other option to resolve my problem would be to associate 
>instructors in a course with sections within the course, thereby 
>allowing e-mail to be directed to them, rather than to all 
>instructors of the course.  Would this be better or worse than the above?)
>
>Thanks,
>Gavin
>
>--
>P. Gavin LaRose, Ph.D.  Program Manager (Instructional Tech.)
>Math Dept., University of Michigan
>gl...@um...                        "There's no use in trying," [Alice]
>734.764.6454                             said. "One Can't believe impossible
>http://www.math.lsa.umich.edu/~glarose/  things." "I daresay you haven't had
>                                          much practice," said the Queen.
>                                                               - Lewis Carrol
>
>
>-------------------------------------------------------
>SF.Net email is sponsored by:
>Tame your development challenges with Apache's Geronimo App Server. 
>Download it for free - -and be entered to win a 42" plasma tv or your very
>own Sony(tm)PSP.  Click here to play: http://sourceforge.net/geronimo.php
>_______________________________________________
>OpenWeBWorK-Devel mailing list
>Ope...@li...
>https://lists.sourceforge.net/lists/listinfo/openwebwork-devel

Prof. Arnold K. Pizer
Dept. of Mathematics
University of Rochester
Rochester, NY 14627
(585) 275-7767

[WWdevel] feedback e-mail subject line

[P G. L. <gl...@um...>]

Hi all,

I'd like opinions on the desirability of the following, and, if it's 
desirable, suggestions for a better way of doing it.

Currently the WeBWorK system generates feedback e-mails with the subject 
line
WeBWorK feedback from [coursename]: [User Name] set [setNum]/prob [probNum]

We always have some big courses where there are multiple sections of the 
course which have different instructors or TAs.  It would be convenient 
for those courses to have the subject line to be something like

WeBWorK feedback [CourseName] ([SectionNum]-[RecitationNum]): [User Name] 
set [setNum]/prob [probNum]

so that instructors could filter their incoming WeBWorK feedback e-mail to 
isolate those from students in their section(s).

In the past I've done this by just hacking the code in Feedback.pm to 
accomplish this, but it seems it might be better to have a setting in 
global.conf or course.conf that allows the customization of the feedback 
subject line.  e.g., using some sort of format characters---say,
     %c = course name
     %u = user name
     %s = set number
     %p = problem number
     %x = section number
     %r = recitation number
then the default subject line could be
     $mail{subjectLine} = 'WeBWorK feedback from %c: %u set %s/prob %p';
and in course.conf it could be reset, if desirable, to something else (for 
me,
     $mail{subjectLine} = 'WeBWorK feedback %c (%x-%r): %u set %s/prob %p';
).

This would require adding the $mail{subjectLine} value to global.conf 
file, and doing some substitutions on that in Feedback.pm.

Does this seem like a good idea?  If so, is there a better way of doing 
it?

(The other option to resolve my problem would be to associate instructors 
in a course with sections within the course, thereby allowing e-mail to be 
directed to them, rather than to all instructors of the course.  Would 
this be better or worse than the above?)

Thanks,
Gavin

-- 
P. Gavin LaRose, Ph.D.  Program Manager (Instructional Tech.)
Math Dept., University of Michigan
gl...@um...                        "There's no use in trying," [Alice]
734.764.6454                             said. "One Can't believe impossible
http://www.math.lsa.umich.edu/~glarose/  things." "I daresay you haven't had
                                          much practice," said the Queen.
                                                               - Lewis Carrol

Re: [WWdevel] Stats.pm (logged as bug 841 in bugzilla)

[John J. <jj...@as...>]

It doesn't happen for me with the cvs version.  That function is defined 
in several places, so it would make sense to just use one of them (like 
the one in ContentGenerator.pm) and remove the rest.  That would 
probably also cure the problem.

John


William Ziemer wrote:

> I just wiped my webwork clean (it was a mishmash of custom code and  
> webwork) and did a fresh 2.1.3 install via cvs.
>
> Clicking on "Statistics" brings up the error message that  
> underscore2nbsp is undefined.  Go back, click again and no problems.
> This happens the first time students stats are requested as well.   
> (multiple platforms, multiple browsers)
> I am thinking that this a mod_perl problem/feature.  Have you seen it?
>
> Error message:
>
> Warning messages
>
> Error messages
>
> Undefined subroutine  
> &WeBWorK::ContentGenerator::Instructor::Stats::underscore2nbsp called  
> at /var/www/webwork/webwork2/lib/WeBWorK/ContentGenerator/Instructor/ 
> Stats.pm line 115.
> Call stack
>
> The information below can help locate the source of the problem.
>
> in WeBWorK::ContentGenerator::Instructor::Stats::siblings called at  
> line 152 of /var/www/webwork/webwork2/lib/WeBWorK/Template.pm
> in WeBWorK::Template::template called at line 439 of /var/www/webwork/ 
> webwork2/lib/WeBWorK/ContentGenerator.pm
> in WeBWorK::ContentGenerator::content called at line 174 of /var/www/ 
> webwork/webwork2/lib/WeBWorK/ContentGenerator.pm
> in WeBWorK::ContentGenerator::go called at line 303 of /var/www/ 
> webwork/webwork2/lib/WeBWorK.pm
>
> ----------------------------------
> William Ziemer
> Dept. of Mathematics and Statistics
> CSULB
> wz...@cs...
>
>
>
>
> -------------------------------------------------------
> SF.Net email is sponsored by:
> Tame your development challenges with Apache's Geronimo App Server. 
> Download it for free - -and be entered to win a 42" plasma tv or your 
> very
> own Sony(tm)PSP.  Click here to play: http://sourceforge.net/geronimo.php
> _______________________________________________
> OpenWeBWorK-Devel mailing list
> Ope...@li...
> https://lists.sourceforge.net/lists/listinfo/openwebwork-devel

[WWdevel] Stats.pm (logged as bug 841 in bugzilla)

[William Z. <wz...@cs...>]

I just wiped my webwork clean (it was a mishmash of custom code and  
webwork) and did a fresh 2.1.3 install via cvs.

Clicking on "Statistics" brings up the error message that  
underscore2nbsp is undefined.  Go back, click again and no problems.
This happens the first time students stats are requested as well.   
(multiple platforms, multiple browsers)
I am thinking that this a mod_perl problem/feature.  Have you seen it?

Error message:

Warning messages

Error messages

Undefined subroutine  
&WeBWorK::ContentGenerator::Instructor::Stats::underscore2nbsp called  
at /var/www/webwork/webwork2/lib/WeBWorK/ContentGenerator/Instructor/ 
Stats.pm line 115.
Call stack

The information below can help locate the source of the problem.

in WeBWorK::ContentGenerator::Instructor::Stats::siblings called at  
line 152 of /var/www/webwork/webwork2/lib/WeBWorK/Template.pm
in WeBWorK::Template::template called at line 439 of /var/www/webwork/ 
webwork2/lib/WeBWorK/ContentGenerator.pm
in WeBWorK::ContentGenerator::content called at line 174 of /var/www/ 
webwork/webwork2/lib/WeBWorK/ContentGenerator.pm
in WeBWorK::ContentGenerator::go called at line 303 of /var/www/ 
webwork/webwork2/lib/WeBWorK.pm

----------------------------------
William Ziemer
Dept. of Mathematics and Statistics
CSULB
wz...@cs...

Re: [WWdevel] Re: Problem with num_cmp and strings

[Davide P. C. <dp...@un...>]

> I think it would be good idea to make a news announcement on the  
> bulletin board telling those experiencing problems with
> numerical compare with strings to either set useOldAnswerEvaluators  
> to 0 or to update the appropriate files. (which are?)

Just pg/macros/PGanswermacros.pl should do it.

> I'm still a little overwhelmed getting things back on an even keel  
> with hosted.webwork, so it may be a day or two before I can
> do this.  You can do in advance if you have time, or if you have  
> time to write up the explanation I'll post it.  Otherwise I can
> write up the explanation of how to make the fix and post it  
> sometime Thursday or Friday.

OK, I'll see if I have time tomorrow.

Davide

Re: [WWdevel] Re: Problem with num_cmp and strings

[Michael G. <ga...@ma...>]

I think it would be good idea to make a news announcement on the  
bulletin board telling those experiencing problems with
numerical compare with strings to either set useOldAnswerEvaluators  
to 0 or to update the appropriate files. (which are?)

I'm still a little overwhelmed getting things back on an even keel  
with hosted.webwork, so it may be a day or two before I can
do this.  You can do in advance if you have time, or if you have time  
to write up the explanation I'll post it.  Otherwise I can
write up the explanation of how to make the fix and post it sometime  
Thursday or Friday.

Thanks much.

Take care,

Mike

On Sep 13, 2005, at 9:03 PM, Davide P. Cervone wrote:

>> Thanks, Davide.  Sorry to have caused the trouble.
>>
>
> No problem; I didn't mean to accuse you of anything, just pointing  
> out where the trouble creeped in.
>
> Do you have a sense of how many sites are using the parser-based  
> versions of num_cmp versus the traditional versions?  This error  
> will affect any problem that uses strings and has a string-valued  
> answer, so could potentially cause problems for a lot of people if  
> they are using the old answer macros.  (This was in the 2.1.3  
> release, wasn't it?)  Do they need to be informed?
>
> Davide
>
>
>
> -------------------------------------------------------
> SF.Net email is sponsored by:
> Tame your development challenges with Apache's Geronimo App  
> Server.Download it for free - -and be entered to win a 42" plasma  
> tv or your very
> own Sony(tm)PSP.  Click here to play: http://sourceforge.net/ 
> geronimo.php
> _______________________________________________
> OpenWeBWorK-Devel mailing list
> Ope...@li...
> https://lists.sourceforge.net/lists/listinfo/openwebwork-devel
>

"Only dead fish swim with the stream."

[WWdevel] Re: Problem with num_cmp and strings

[Davide P. C. <dp...@un...>]

> Thanks, Davide.  Sorry to have caused the trouble.

No problem; I didn't mean to accuse you of anything, just pointing  
out where the trouble creeped in.

Do you have a sense of how many sites are using the parser-based  
versions of num_cmp versus the traditional versions?  This error will  
affect any problem that uses strings and has a string-valued answer,  
so could potentially cause problems for a lot of people if they are  
using the old answer macros.  (This was in the 2.1.3 release, wasn't  
it?)  Do they need to be informed?

Davide

[WWdevel] Re: Problem with num_cmp and strings

[Michael G. <ga...@ma...>]

Thanks, Davide.  Sorry to have caused the trouble.  When I get a  
chance I'll grep through the code and see if there are any other uses
of str_cmp which are affected by this change.

Take care,

Mike
On Sep 13, 2005, at 10:10 AM, Davide P.Cervone wrote:

>> I'm hoping you can help me out.  I'm getting a bug on the  
>> following problem below whenever 'DNE' is entered as an answer.
>>
>
> It turns out this is a result of your having updated the str_cmp  
> answer checker to return an AnswerEvaluator object rather than a  
> CODE reference.  The check_string filter uses str_cmp the check the  
> answer, but uses it as a CODE reference, so I changed that to  
> $ans_eval->evaluate($ans) and that took care of it.  I haven't  
> looked to see that there aren't other places where this might occur.
>
> Davide
>
>

"Only dead fish swim with the stream."

Re: [WWdevel] how should we report typos/errors in the new ProblemLibrary?

[John J. <jj...@as...>]

Michael Gage wrote:

> There is a typo in
>
> /ProblemLibrary/ma117DB/set1/srw1_1_57.pg
>
> There is an extra left brace in the first expression.  dvipng glides  
> right over it, but jsMath and TeX output catch it.

For the moment, the procedure is to e-mail me.  I just removed the 
extraneous brace and updated the tarball I have posted.

We had talked about setting up a cvs at Rochester.  At one time, we were 
thinking of multiple categories of files (untagged, tagged but not 
cleaned, tagged and cleaned) and then moving files around.   The last I 
recall, I said that we just needed a simple cvs, in part, because it 
looked like there was not going to be funds for extensive cleaning of 
problems.

Currently, I keep tagged problems away from untagged ones.  But, it 
would not be a big problem to have them together.  Untagged problems 
would not be visible through the problem library interface, but they 
would not be causing much harm either.

So, maybe we should go ahead and make a ProblemLibrary cvs at Rochester, 
and I can populate it with what I currently have.

John

[WWdevel] how should we report typos/errors in the new ProblemLibrary?

[Michael G. <ga...@ma...>]

There is a typo in

/ProblemLibrary/ma117DB/set1/srw1_1_57.pg

There is an extra left brace in the first expression.  dvipng glides  
right over it, but jsMath
and TeX output catch it.

Take care,

Mike

[WWdevel] Re: Problem with num_cmp and strings

[Davide P.C. <dp...@un...>]

> I'm hoping you can help me out.=A0 I'm getting a bug on the following=20=

> problem below whenever 'DNE' is entered as an answer.

It turns out this is a result of your having updated the str_cmp answer=20=

checker to return an AnswerEvaluator object rather than a CODE=20
reference.  The check_string filter uses str_cmp the check the answer,=20=

but uses it as a CODE reference, so I changed that to=20
$ans_eval->evaluate($ans) and that took care of it.  I haven't looked=20
to see that there aren't other places where this might occur.

Davide

Re: [WWdevel] Problem with num_cmp and strings

[Michael G. <ga...@ma...>]

Thanks very much John.  That was the problem.
I was pretty sure it was a flag set somewhere.
I'm not sure why the old macros couldn't handle this, but
I'm not going to worry too much about it.  Things seem
to be working well now.

Take care,

Mike
On Sep 12, 2005, at 11:30 PM, John Jones wrote:

> Hi Mike,
>
> For what its worth, this works fine if I use the parser based  
> version of the macros.  I get the errors only if I have  
> useOldAnswerMacros=1.
>
> John
>
>
> Michael Gage wrote:
>> Hi Davide,
>>
>> I'm hoping you can help me out.  I'm getting a bug on the  
>> following problem below whenever 'DNE' is entered as an answer.
>>
>> Do you recognize this as some configuration flag I'm missing?  I  
>> have the pg directory up to the latest CVS HEAD.
>> I'm also seeing this error on devel which wasn't affected by the  
>> latest crash.
>>
>> Rolling back the pg directory a month doesn't seem to affect the  
>> problem -- it is there in either case.
>>
>> I've listed the error message below the problem.  It would help if  
>> you could try this out
>> on your machine and let me know whether it works.  I'm puzzled as  
>> to why these problems
>> have stopped working on all three machine (devel, hosted and math)  
>> particularly since not much has been going on
>> on devel.
>>
>> Thanks much.
>>
>> Take care,
>>
>> Mike
>>
>>
>> DOCUMENT();        # This should be the first executable line in  
>> the problem.
>>
>> loadMacros(
>> "PG.pl",
>> "PGbasicmacros.pl",
>> "PGchoicemacros.pl",
>> "PGanswermacros.pl",
>> "PGgraphmacros.pl",
>> "PGauxiliaryFunctions.pl"
>> );
>>
>> TEXT(beginproblem());
>> $showPartialCorrectAnswers = 1;
>> BEGIN_TEXT
>> Enter DNE or a number \{ans_rule\}
>> $PAR
>> Enter DNE or a number \{ans_rule\}
>>
>> END_TEXT
>>
>> ANS(num_cmp('DNE', strings=>['DNE'])) ;
>> ANS(num_cmp(5, strings=>['DNE']))  ;
>>
>> ENDDOCUMENT();        # This should be the last executable line in  
>> the problem.
>>
>> #####
>> Warning messages
>>
>> Error in Translator.pm::process_answers: Answer AnSwEr1: |dne|
>> Not a CODE reference at line 4015 of (eval 366)
>> Died within main::check_strings called at line 545 of [PG]/lib/ 
>> AnswerHash.pm
>> from within AnswerEvaluator::evaluate called at line 1 of (eval 785)
>> Error in Translator.pm::process_answers: Answer AnSwEr1:
>> Answer evaluators must return a hash or an AnswerHash type, not  
>> type || at /home/gage/webwork/pg/lib/WeBWorK/PG/Translator.pm line  
>> 1158
>> Use of uninitialized value in numeric ge (>=) at /home/gage/ 
>> webwork/webwork-modperl/lib/WeBWorK/ContentGenerator/Problem.pm  
>> line 252.
>> Use of uninitialized value in pattern match (m//) at /home/gage/ 
>> webwork/webwork-modperl/lib/WeBWorK/ContentGenerator/Problem.pm  
>> line 253.
>> Use of uninitialized value in numeric ge (>=) at /home/gage/ 
>> webwork/webwork-modperl/lib/WeBWorK/ContentGenerator/Problem.pm  
>> line 254.
>> Use of uninitialized value in numeric gt (>) at /home/gage/webwork/ 
>> webwork-modperl/lib/WeBWorK/ContentGenerator/Problem.pm line 254.
>> Use of uninitialized value in numeric gt (>) at /home/gage/webwork/ 
>> webwork-modperl/lib/WeBWorK/ContentGenerator/Problem.pm line 257.
>>
>> "Only dead fish swim with the stream."
>>
>>
>

"Only dead fish swim with the stream."

Re: [WWdevel] Problem with num_cmp and strings

[John J. <jj...@as...>]

Hi Mike,

For what its worth, this works fine if I use the parser based version of 
the macros.  I get the errors only if I have useOldAnswerMacros=1.

John


Michael Gage wrote:

> Hi Davide,
>
> I'm hoping you can help me out.  I'm getting a bug on the following 
> problem below whenever 'DNE' is entered as an answer.
>
> Do you recognize this as some configuration flag I'm missing?  I have 
> the pg directory up to the latest CVS HEAD.
> I'm also seeing this error on devel which wasn't affected by the 
> latest crash.
>
> Rolling back the pg directory a month doesn't seem to affect the 
> problem -- it is there in either case.
>
> I've listed the error message below the problem.  It would help if you 
> could try this out 
> on your machine and let me know whether it works.  I'm puzzled as to 
> why these problems
> have stopped working on all three machine (devel, hosted and math) 
> particularly since not much has been going on 
> on devel.
>
> Thanks much.
>
> Take care,
>
> Mike
>
>
> DOCUMENT();        # This should be the first executable line in the 
> problem.
>
> loadMacros(
> "PG.pl",
> "PGbasicmacros.pl",
> "PGchoicemacros.pl",
> "PGanswermacros.pl",
> "PGgraphmacros.pl",
> "PGauxiliaryFunctions.pl"
> );
>
> TEXT(beginproblem());
> $showPartialCorrectAnswers = 1;
> BEGIN_TEXT
> Enter DNE or a number \{ans_rule\}
> $PAR
> Enter DNE or a number \{ans_rule\}
>
> END_TEXT
>
> ANS(num_cmp('DNE', strings=>['DNE'])) ;
> ANS(num_cmp(5, strings=>['DNE']))  ;
>
> ENDDOCUMENT();        # This should be the last executable line in the 
> problem.
>
> #####
>
> *Warning messages*
>
>     * Error in Translator.pm::process_answers: Answer AnSwEr1: |dne|
>     * Not a CODE reference at line 4015 of (eval 366)
>       Died within main::check_strings called at line 545 of
>       [PG]/lib/AnswerHash.pm
>       from within AnswerEvaluator::evaluate called at line 1 of (eval 785)
>     * Error in Translator.pm::process_answers: Answer AnSwEr1:
>     * Answer evaluators must return a hash or an AnswerHash type, not
>       type || at /home/gage/webwork/pg/lib/WeBWorK/PG/Translator.pm
>       line 1158
>     * Use of uninitialized value in numeric ge (>=) at
>       /home/gage/webwork/webwork-modperl/lib/WeBWorK/ContentGenerator/Problem.pm
>       line 252.
>     * Use of uninitialized value in pattern match (m//) at
>       /home/gage/webwork/webwork-modperl/lib/WeBWorK/ContentGenerator/Problem.pm
>       line 253.
>     * Use of uninitialized value in numeric ge (>=) at
>       /home/gage/webwork/webwork-modperl/lib/WeBWorK/ContentGenerator/Problem.pm
>       line 254.
>     * Use of uninitialized value in numeric gt (>) at
>       /home/gage/webwork/webwork-modperl/lib/WeBWorK/ContentGenerator/Problem.pm
>       line 254.
>     * Use of uninitialized value in numeric gt (>) at
>       /home/gage/webwork/webwork-modperl/lib/WeBWorK/ContentGenerator/Problem.pm
>       line 257.
>
>
> "Only dead fish swim with the stream."
>
>

[WWdevel] Problem with num_cmp and strings

[Michael G. <ga...@ma...>]

Hi Davide,

I'm hoping you can help me out.  I'm getting a bug on the following  
problem below whenever 'DNE' is entered as an answer.

Do you recognize this as some configuration flag I'm missing?  I have  
the pg directory up to the latest CVS HEAD.
I'm also seeing this error on devel which wasn't affected by the  
latest crash.

Rolling back the pg directory a month doesn't seem to affect the  
problem -- it is there in either case.

I've listed the error message below the problem.  It would help if  
you could try this out
on your machine and let me know whether it works.  I'm puzzled as to  
why these problems
have stopped working on all three machine (devel, hosted and math)  
particularly since not much has been going on
on devel.

Thanks much.

Take care,

Mike


DOCUMENT();        # This should be the first executable line in the  
problem.

loadMacros(
"PG.pl",
"PGbasicmacros.pl",
"PGchoicemacros.pl",
"PGanswermacros.pl",
"PGgraphmacros.pl",
"PGauxiliaryFunctions.pl"
);

TEXT(beginproblem());
$showPartialCorrectAnswers = 1;
BEGIN_TEXT
Enter DNE or a number \{ans_rule\}
$PAR
Enter DNE or a number \{ans_rule\}

END_TEXT

ANS(num_cmp('DNE', strings=>['DNE'])) ;
ANS(num_cmp(5, strings=>['DNE']))  ;

ENDDOCUMENT();        # This should be the last executable line in  
the problem.

#####
Warning messages

Error in Translator.pm::process_answers: Answer AnSwEr1: |dne|
Not a CODE reference at line 4015 of (eval 366)
Died within main::check_strings called at line 545 of [PG]/lib/ 
AnswerHash.pm
from within AnswerEvaluator::evaluate called at line 1 of (eval 785)
Error in Translator.pm::process_answers: Answer AnSwEr1:
Answer evaluators must return a hash or an AnswerHash type, not type  
|| at /home/gage/webwork/pg/lib/WeBWorK/PG/Translator.pm line 1158
Use of uninitialized value in numeric ge (>=) at /home/gage/webwork/ 
webwork-modperl/lib/WeBWorK/ContentGenerator/Problem.pm line 252.
Use of uninitialized value in pattern match (m//) at /home/gage/ 
webwork/webwork-modperl/lib/WeBWorK/ContentGenerator/Problem.pm line  
253.
Use of uninitialized value in numeric ge (>=) at /home/gage/webwork/ 
webwork-modperl/lib/WeBWorK/ContentGenerator/Problem.pm line 254.
Use of uninitialized value in numeric gt (>) at /home/gage/webwork/ 
webwork-modperl/lib/WeBWorK/ContentGenerator/Problem.pm line 254.
Use of uninitialized value in numeric gt (>) at /home/gage/webwork/ 
webwork-modperl/lib/WeBWorK/ContentGenerator/Problem.pm line 257.

"Only dead fish swim with the stream."

Re: [WWdevel] Re: [WWcvs] webwork-modperl: Sam changed open(LOG,...) to open(my $log, ...), but

[Davide P. C. <dp...@un...>]

> here are the file-scoped lexicals that are hanging around,  
> potentially causing similar problems. (That is, if they get  
> modified. If they're just being used as constants, then in most  
> cases we can replace them with 'use constant', or maybe use a real  
> constant class that uses a tied scalar.) Fortunately, most of it is  
> in WebworkWebservice, which isn't production code yet, or in PG,  
> where I'm not sure if it's a problem, since every problem has its  
> own environment.

The ones in Parser and Value are almost all constants, plus a few  
that are computed and then retain their same value.  There is one  
exception, however, which is is $Value::context, that is a reference  
to the current Context hash.  (This is changed by the Context()  
function.)  It would take some doing to convert that to a thread-safe  
mechanism, I think.

Davide

Re: [WWdevel] Re: [WWcvs] webwork-modperl: Sam changed open(LOG,...) to open(my $log, ...), but

[Sam H. <sh...@ma...>]

On Sep 9, 2005, at 21:03, Davide P.Cervone wrote:

>>>>> (I'm not sure why it is necessary to make a local variable for  
>>>>> LOG, but at least it works now.)
>>>> We want WW to be thread safe, so that when we port it to Apache  
>>>> 2 (my next major project), people can run it under the threaded  
>>>> MPM.
>>> OK.  I'll keep that in mind as I work on other parts.  I guess  
>>> that means most of the other file access code will have to be  
>>> modified similarly.
>> Most of the file access code already uses lexical variables.
>
> A quick grep in webwork2/lib/WeBWorK turned up the following file  
> references that don't seem to be localized:
>
> [snip -- 27 examples of thread-unsafe filehandles]
>
> So it looks like there are still a few out there to be taken care of.

Yikes -- it appears so. In addition to those places, we should look  
at the places where the filehandle's typeglob has been localized  
before the open() call -- these are threadsafe, but dynamic scope  
isn't really what we want here.

Also, here are the file-scoped lexicals that are hanging around,  
potentially causing similar problems. (That is, if they get modified.  
If they're just being used as constants, then in most cases we can  
replace them with 'use constant', or maybe use a real constant class  
that uses a tied scalar.) Fortunately, most of it is in  
WebworkWebservice, which isn't production code yet, or in PG, where  
I'm not sure if it's a problem, since every problem has its own  
environment.
-sam

[sh002i@devel] work$ grep -r '^my\b' webwork2/lib pg/lib
webwork2/lib/PSH.pm:my() variables.
webwork2/lib/WeBWorK/ContentGenerator/Instructor/FileManager.pm:my % 
uploadDir = (
webwork2/lib/WeBWorK/ContentGenerator/Instructor/SendMail.pm:my  
$UPDATE_SETTINGS_BUTTON = "Update settings and refresh page"; #  
handle submit value idiocy
webwork2/lib/WeBWorK/ContentGenerator/Instructor/SetMaker.pm:my % 
problib;      ## filled in in global.conf
webwork2/lib/WeBWorK/ContentGenerator/Instructor/SetMaker.pm:my % 
ignoredir = (
webwork2/lib/WebworkWebservice/MathTranslators.pm:my $debugOn    
=       1;
webwork2/lib/WebworkWebservice/MathTranslators.pm:my $PASSWORD   
=       $WebworkWebservice::PASSWORD;
webwork2/lib/WebworkWebservice/MathTranslators.pm:my  
$TEMPDIRECTORY            =        $ce->{webworkDirs}->{htdocs_temp};
webwork2/lib/WebworkWebservice/MathTranslators.pm:my  
$TEMP_BASE_URL            =        $ce->{webworkURLs}->{htdocs_temp};
webwork2/lib/WebworkWebservice/MathTranslators.pm:my  
$externalLatexPath        =        $ce->{externalPrograms}->{latex};
webwork2/lib/WebworkWebservice/MathTranslators.pm:my  
$externalpdflatexPath     =        $ce->{externalPrograms}->{pdflatex};;
webwork2/lib/WebworkWebservice/MathTranslators.pm:my  
$tmp_directory_permission =        $Global::tmp_directory_permission;
webwork2/lib/WebworkWebservice/MathTranslators.pm:my  
$numericalGroupID         =$Global::numericalGroupID;  # group ID for  
webadmin
webwork2/lib/WebworkWebservice/RenderProblem.pm:my $debugXmlCode=1;   
# turns on the filter for debugging XMLRPC and SOAP code
webwork2/lib/WebworkWebservice/RenderProblem.pm:my  
$COURSE_TEMP_DIRECTORY      =        $ce->{courseDirs}->{html_tmp};
webwork2/lib/WebworkWebservice/RenderProblem.pm:my  
$COURSE_TEMP_URL            =        $HOSTURL.$ce->{courseURLs}-> 
{html_tmp};
webwork2/lib/WebworkWebservice/RenderProblem.pm:my  
$pgMacrosDirectory          =        $ce->{pg_dir}.'/macros/';
webwork2/lib/WebworkWebservice/RenderProblem.pm:my  
$macroDirectory             =$ce->{courseDirs}->{macros}.'/';
webwork2/lib/WebworkWebservice/RenderProblem.pm:my  
$templateDirectory          =        $ce->{courseDirs}->{templates};
webwork2/lib/WebworkWebservice/RenderProblem.pm:my % 
PG_environment          =   $ce->{pg}->{specialPGEnvironmentVars};
pg/lib/AnswerHash.pm:my %fields =  
(             'score'                        =>       undef,
pg/lib/Circle.pm:my %fields =(
pg/lib/Complex1.pm:my ( $i, $ip2, %logn );
pg/lib/Complex1.pm:my @trig = qw(
pg/lib/Complex1.pm:my $package = 'Complex1';
pg/lib/Complex1.pm:my $display = 'cartesian';           # Default  
display format
pg/lib/Complex1.pm:my $eps     = 1e-14;                 # Epsilon
pg/lib/Complex1.pm:my %gcd;
pg/lib/Fraction.pm:my %fields = (
pg/lib/Fun.pm:my $GRAPH_REFERENCE = "WWPlot";
pg/lib/Fun.pm:my $FUNCTION_REFERENCE = "Fun";
pg/lib/Fun.pm:my %fields =(
pg/lib/Label.pm:my %fields =(
pg/lib/List.pm:my %fields = (
pg/lib/PGrandom.pm:my $multiplier = 69069;
pg/lib/PGrandom.pm:my $translate = 1;
pg/lib/PGrandom.pm:my $modulus = 2**32;
pg/lib/Parser.pm:my $pkg = "Parser";
pg/lib/Regression.pm:my $nan= "NaN";
pg/lib/Units.pm:my %fundamental_units = ('factor' => 1,
pg/lib/Units.pm:my $PI = 4*atan2(1,1);
pg/lib/Units.pm:my %known_units = ('m'  => {
pg/lib/Value.pm:my $pkg = 'Value';
pg/lib/VectorField.pm:my $GRAPH_REFERENCE = "WWPlot";
pg/lib/VectorField.pm:my $VECTORFIELD_REFERENCE = "VectorField";
pg/lib/VectorField.pm:my %fields =(
pg/lib/WWPlot.pm:my     $last_image_number=0;    #class variable.   
Keeps track of how many images have been made.
pg/lib/WWPlot.pm:my %fields = (  # initialization only!!!
pg/lib/ww_strict.pm:my %bitmask = (
pg/lib/ww_strict.pm:my $default_bits = bits(qw(refs subs vars));
pg/lib/Parser/Context.pm:my $pkg = "Parser::Context";
pg/lib/Parser/Context.pm:my $userContext;
pg/lib/Parser/Context/Default.pm:my $infinity = Value::Infinity->new();
pg/lib/Parser/Context/Functions.pm:my %Category = (
pg/lib/Parser/Legacy/LimitedNumeric.pm:my $context =  
$Parser::Context::Default::context{Numeric}->copy;
pg/lib/Parser/Legacy/Numeric.pm:my $context =  
$Parser::Context::Default::context{Numeric}->copy;
pg/lib/Parser/List/Vector.pm:my $ijk_string = ['i','j','k','0'];
pg/lib/Parser/List/Vector.pm:my $ijk_TeX = ['\boldsymbol 
{i}','\boldsymbol{j}','\boldsymbol{k}','\boldsymbol(0)'];
pg/lib/Value/AnswerChecker.pm:my %tth_delim = (
pg/lib/Value/AnswerChecker.pm:my @ans_defaults = (showCoodinateHints  
=> 0, checker => sub {0});
pg/lib/Value/Complex.pm:my $pkg = 'Value::Complex';
pg/lib/Value/Context.pm:my $pkg = "Value::Context";
pg/lib/Value/Formula.pm:my $pkg = 'Value::Formula';
pg/lib/Value/Formula.pm:my $UNDEF = bless {}, "UNDEF";
pg/lib/Value/Infinity.pm:my $pkg = 'Value::Infinity';
pg/lib/Value/Interval.pm:my $pkg = 'Value::Interval';
pg/lib/Value/List.pm:my $pkg = 'Value::List';
pg/lib/Value/Matrix.pm:my $pkg = 'Value::Matrix';
pg/lib/Value/Point.pm:my $pkg = 'Value::Point';
pg/lib/Value/Real.pm:my $pkg = 'Value::Real';
pg/lib/Value/Set.pm:my $pkg = 'Value::Set';
pg/lib/Value/String.pm:my $pkg = 'Value::String';
pg/lib/Value/Union.pm:my $pkg = 'Value::Union';
pg/lib/Value/Union.pm:my %subCall = (
pg/lib/Value/Vector.pm:my $pkg = 'Value::Vector';
pg/lib/Value/Vector.pm:my $ijk_string = ['i','j','k','0'];
pg/lib/Value/Vector.pm:my $ijk_TeX = ['\boldsymbol{i}','\boldsymbol 
{j}','\boldsymbol{k}','\boldsymbol{0}'];
pg/lib/Value/WeBWorK.pm:my @wwEvalFields = qw(
pg/lib/WeBWorK/PG/Translator.pm:my %shared_subroutine_hash = (

Re: [WWdevel] Re: [WWcvs] webwork-modperl: Sam changed open(LOG,...) to open(my $log, ...), but

[Davide P.C. <dp...@un...>]

>>>> (I'm not sure why it is necessary to make a local variable for  
>>>> LOG, but at least it works now.)
>>> We want WW to be thread safe, so that when we port it to Apache 2  
>>> (my next major project), people can run it under the threaded MPM.
>> OK.  I'll keep that in mind as I work on other parts.  I guess  
>> that means most of the other file access code will have to be  
>> modified similarly.
> Most of the file access code already uses lexical variables.

A quick grep in webwork2/lib/WeBWorK turned up the following file  
references that don't seem to be localized:

Utils.pm:606:   if (open LOG, ">>", $logFile) {
Utils.pm:623:   if (open LOG, ">>", $logFile) {
ContentGenerator/Grades.pm:98:    } elsif ( open(FILE,  
"$filePath")  )   {
ContentGenerator/Grades.pm:141:         open FILE, "$filePath" ||  
return("Can't open $filePath");
ContentGenerator/Hardcopy.pm:495:               open TEX, ">",  
$hardcopyFilePath or die "Failed to open $hardcopyFilePath: $! 
\n".CGI::br();
ContentGenerator/Hardcopy.pm:553:       open TEX, ">", $texFile or  
die "Failed to open $texFile: $!\n".CGI::br();
ContentGenerator/Hardcopy.pm:636:    open(LOGFILE,"<$filePath") or  
die "Can't read $filePath";
ContentGenerator/Hardcopy.pm:648:    open (TEXFILE, "$texFilePath")
DBv3/Utils.pm:70:       sysopen LOCK, $lockfile, O_RDONLY|O_CREAT
Utils/CourseManagement.pm:252:      open(INSTRUCTOR,">> $courseDir/ 
instructor.txt") || warn "Can't create 'instructor.txt': $!";
Utils/CourseManagement.pm:256:    open(TITLE,">> $courseDir/ 
coursetitle.txt") || warn "Can't create 'coursetitle.txt': $!";
ContentGenerator/Instructor/FileManager.pm:474:         if (open 
(OUTFILE,">$file")) {
ContentGenerator/Instructor/FileManager.pm:716:                 if  
(open(NEWFILE,">$file")) {
ContentGenerator/Instructor/FileManager.pm:818:         open(UPLOAD,"> 
$file") || $self->addbadmessage("Can't create file '$name'");
ContentGenerator/Instructor/PGProblemEditor.pm: 
1024:                    open OUTPUTFILE, ">", $outputFilePath
ContentGenerator/Instructor/ProblemSetEditor.pm:618:    open  
(PROBLEM, ">$probFileName") ||
ContentGenerator/Instructor/ProblemSetList.pm:1477:     if ( open  
(SETFILENAME, "$filePath") )    {
ContentGenerator/Instructor/ProblemSetList.pm: 
1673:                     open SETDEF, ">$filePath" or die "Failed to  
open $fileP
ContentGenerator/Instructor/SendMail.pm:758:    open (PROBLEM, "> 
$probFileName") ||
ContentGenerator/Instructor/SendMail.pm:775:            open FILE,  
"$filePath" || do { $self->addbadmessage(CGI::p("Can't open  
$filePath")); return};
ContentGenerator/Instructor/SetMaker.pm:204:    if ( open  
(SETFILENAME, "$filePath") )    {

and in pg/lib, we have

AnswerIO.pm:52:             open(LOG, ">>$accessLog") or $error.=  
"Can't open course access log $accessLog";
WeBWorK/EquationCache.pm:93:        sysopen(DB, $db, O_RDWR|O_CREAT)
WeBWorK/PG/IO.pm:114:       open(INPUT, "<$filePath") || die "$0:  
readWholeProblemFile subroutine: <BR>Can't read file $filePath";
WeBWorK/PG/IO.pm:173:       open(TEMPCREATEFILE, ">$fileName")
WeBWorK/PG/ImageGenerator.pm:75:    open FILEH,  "<$filename" or die  
"Unable to read $filename";
WeBWorK/PG/Translator.pm:508:       if ( open(SOURCEFILE, "< 
$filePath") ) {

So it looks like there are still a few out there to be taken care of.

Davide

Re: [WWdevel] strange log messages

[Davide P. C. <dp...@un...>]

> Anyone seen this before?

Yes.

> Ambiguous use of -i resolved as -&i() at (eval 8172) line 1.

This is from cplx_cmp when a complex number of the form 1-i appears  
in the answer.  The -i could be treated as some kind of file test  
(don't know what it is off hand) and perl is telling you it has  
decided to use the (implicit) function call to i() instead so is  
treating it as 1-&i().  If the Complex1 package had stringified 1-i  
as "1 - i" or "1-i" you would not see this problem.

It was this error that prompted me to Paser-ize the cplx_cmp macros.

> Unquoted string "n" may clash with future reserved word at (eval  
> 8172) line 1.
> Operator or semicolon missing before *n at (eval 8172) line 1.
> Ambiguous use of * resolved as operator * at (eval 8172) line 1.

I don't know specifically where these ones are coming from, but it  
looks like something is calling eval on an expression that has not  
been parsed by AlgParser.  I know the cplx_cmp does that with the  
professor's answer, so check to see if someone is passing strings  
that contain expressions with n, f, t or x to cplx_cmp.  The cplx_cmp  
macro calls PG_answer_eval with the (unprocessed) professor's answer,  
so errors like this could easily be produced in the log, but are  
trapped by the PG_answer_eval; cplx_cmp simply ignores the errors, so  
the professor simply doesn't know they are occurring.  I don't know  
if other answer checkers call PG_answer_eval or PG_restricted_eval  
under similar circumstances.

Davide

[WWdevel] strange log messages

[Sam H. <sa...@uo...>]

Anyone seen this before? We're getting this in our logs on  
math.webwork.rochester.edu occasionally, since we rebuilt the server.
-sam

Ambiguous use of -i resolved as -&i() at (eval 8172) line 1.
Unquoted string "n" may clash with future reserved word at (eval  
8172) line 1.
Operator or semicolon missing before *n at (eval 8172) line 1.
Ambiguous use of * resolved as operator * at (eval 8172) line 1.
Unquoted string "f" may clash with future reserved word at (eval  
8172) line 1.
Operator or semicolon missing before *f at (eval 8172) line 1.
Ambiguous use of * resolved as operator * at (eval 8172) line 1.
Unquoted string "n" may clash with future reserved word at (eval  
8172) line 1.
Operator or semicolon missing before *n at (eval 8172) line 1.
Ambiguous use of * resolved as operator * at (eval 8172) line 1.
Unquoted string "t" may clash with future reserved word at (eval  
8172) line 1.
Operator or semicolon missing before *t at (eval 8172) line 1.
Ambiguous use of * resolved as operator * at (eval 8172) line 1.
Unquoted string "n" may clash with future reserved word at (eval  
8175) line 1.
Operator or semicolon missing before *n at (eval 8175) line 1.
Ambiguous use of * resolved as operator * at (eval 8175) line 1.
Unquoted string "f" may clash with future reserved word at (eval  
8175) line 1.
Operator or semicolon missing before *f at (eval 8175) line 1.
Ambiguous use of * resolved as operator * at (eval 8175) line 1.
Unquoted string "n" may clash with future reserved word at (eval  
8175) line 1.
Operator or semicolon missing before *n at (eval 8175) line 1.
Ambiguous use of * resolved as operator * at (eval 8175) line 1.
Unquoted string "t" may clash with future reserved word at (eval  
8175) line 1.
Operator or semicolon missing before *t at (eval 8175) line 1.
Ambiguous use of * resolved as operator * at (eval 8175) line 1.
Unquoted string "x" may clash with future reserved word at (eval  
8850) line 1.
Operator or semicolon missing before *x at (eval 8850) line 1.
Ambiguous use of * resolved as operator * at (eval 8850) line 1.
Unquoted string "x" may clash with future reserved word at (eval  
8850) line 1.
Operator or semicolon missing before *x at (eval 8850) line 1.
Ambiguous use of * resolved as operator * at (eval 8850) line 1.
Unquoted string "x" may clash with future reserved word at (eval  
8850) line 1.
Operator or semicolon missing before *x at (eval 8850) line 1.
Ambiguous use of * resolved as operator * at (eval 8850) line 1.
Argument "x" isn't numeric in multiplication (*) at (eval 8850) line 1.
Argument "x" isn't numeric in multiplication (*) at (eval 8850) line 1.
Argument "x" isn't numeric in multiplication (*) at (eval 8850) line 1.

Re: [WWdevel] Re: [WWcvs] webwork-modperl: Sam changed open(LOG,...) to open(my $log, ...), but

[Sam H. <sh...@ma...>]

On Sep 8, 2005, at 11:33, Davide P. Cervone wrote:

>>> (I'm not sure why it is necessary to make a local variable for  
>>> LOG, but at least it works now.)
>>
>> We want WW to be thread safe, so that when we port it to Apache 2  
>> (my next major project), people can run it under the threaded MPM.
>
> OK.  I'll keep that in mind as I work on other parts.  I guess that  
> means most of the other file access code will have to be modified  
> similarly.

Most of the file access code already uses lexical variables.
-sam

[WWdevel] Re: [WWcvs] webwork-modperl: Sam changed open(LOG,...) to open(my $log, ...), but

[Davide P. C. <dp...@un...>]

>> (I'm not sure why it is necessary to make a local variable for  
>> LOG, but at least it works now.)
>
> We want WW to be thread safe, so that when we port it to Apache 2  
> (my next major project), people can run it under the threaded MPM.

OK.  I'll keep that in mind as I work on other parts.  I guess that  
means most of the other file access code will have to be modified  
similarly.

> Thanks for fixing this btw.

No problem.  One of the faculty here wondered why she couldn't get  
past answers.

Davide

[WWdevel] Re: [WWcvs] webwork-modperl: Sam changed open(LOG,...) to open(my $log, ...), but

[Sam H. <sh...@ma...>]

On Sep 7, 2005, at 22:54, dpvc via activitymail wrote:

> (I'm not sure why it is necessary to make a local variable for LOG,  
> but at least it works now.)

We want WW to be thread safe, so that when we port it to Apache 2 (my  
next major project), people can run it under the threaded MPM.

Thanks for fixing this btw.
-sam

[WWdevel] Re: hosted.webwork disaster

[Sam H. <sh...@ma...>]

On Sep 8, 2005, at 9:15, Michael Gage wrote:

> There is one source of recovery that we haven't explored yet.  The  
> transaction logs in each course have the student responses to
> problems.  The answer_log also has similar data.
>
>  In some cases if we can recover the exact structure of
> the problem sets we might be able  determine if the student  
> answered the quesition correctly.
>
>  Sam, could you think about ways of automating the recovery of this  
> data?

Mike,

We can parse the transaction log pretty easily, although we have to  
watch out for multiline base64-encoded answers. ;) It looks like the  
transaction log actually contains most of the data we would need to  
reconstruct sets and problems:

We should look at the most recent line in the log for each <user_id,  
set_id, problem_id> tuple.

We can use user_id, set_id, problem_id, source_file, value,  
max_attempts, and problem_seed to recreate the "skeletons" of  
problems. However, for source_file, value, and max_attempts, we won't  
know if a user has an overridden value for their UserProblem,  
although we can do some analysis to determine the most popular  
values, and assume that these are the global ones.

We can create skeletal User, Set, and UserSet objects from the  
user_ids and set_ids mentioned. We can't get any details about the  
user (name, email, password, permission level, etc.) out of this, nor  
any about the set (header files, dates, published).

We can set the status, attempted, last_answer, num_correct, and  
num_incorrect fields in the UserProblem as well.

We know nothing about problems that no one has attempted, or users  
who have attempted no problems.

This definitely seems worth looking into though. I can take a look at  
it today if there's nothing else that's more urgent.
-sam

[WWdevel] Another Legacy answer checker

[Davide P. C. <dp...@un...>]

Folks:

I have added a new Parser-based answer checker to replace cplx_cmp  
ass part of the Legacy package.  Upate your pg/lib/Parser/Legacy.pm  
and pg/lib/Parser/Legacy directory to get it, then move pg/lib/Parser/ 
Legacy/PGcomplexmacros.pl to pg/macros and restart the server.  You  
can use $useOldAnswerMacros to control which version of the answer  
checker is used (as with num_cmp and fun_cmp).

This is a bit more delicate than the num_cmp or fun_cmp replacements,  
as the strict modes need to allow operations in some circumstances  
and not others, but I think I have it worked out.  Let me know if you  
see any problems.

A question arose when I was working on the strict-numeric modes.   
Both the original and the new versions accept e^(pi i), since there  
are no operations performed within the real numbers involved, but e^ 
(pi/2 i) is not allowed because of the division by 2.  Would it be  
reasonable to allow expressions of the form a*pi/b within the  
exponent even in strict_numeric mode?  Should there be new modes (or  
a new flag) to allow/disallow this?  The old checker would SHOW  
answers in this form, even though the student couldn't type them in  
that way (but the new one won't).

In any case, try out the new checker and see if it works for you.

Davide