[LON-CAPA-users] Forcing students to submit valid expressions for formula responses.

[Peter Dencker]

Great!

The feedback 'internal error' in the event of a maxima syntax error,
is not only not helpful but a real bother: Students are annoyed with
LON-CAPA and innocent authors as well as  course coordinators are
heavily bombed.

We will try the update after all current tests are closed.

- Peter


Am 31.10.2013 04:33, schrieb Stuart Raeburn:
> Hi,
> 
>> Or can someone finds similarities of the hosts 'mtal1' and 'scll1'
>> that may cause this bug :-/  ?
> 
> Yes, that is a good observation.
> 
> Both mtal1 and scl1 are running a Linux distro for which the distro's
> standard repos provide maxima rev. 5.23
> 
> The majority of other LON-CAPA servers in the production cluster use
> distros for which the maxima version is 5.20.
> 
> In the case of a syntax error, maxima 5.23 outputs "incorrect syntax"
> instead of "Incorrect syntax" as it did in previous versions.
> 
> The result of this change in case is that LON-CAPA's lonmaxima daemon
> rev. 1.40 included with LON-CAPA 2.10.1, and earlier, will not respond
> with a "BAD_FORMULA" response (which causes display of: "Unable to
> understand formula" as feedback to the student), but instead will
> timeout with (a much less helpful) display of: "An internal error
> occurred while processing your answer. Please try again later."
> 
> Anyway LON-CAPA servers with Linux distros with maxima 5.23 and later
> (i.e., CentOS/Scientific Linux/RHEL 6, SLES 10/11, Fedora 14 and later,
> SuSE 11.1 and later) should have someone with root access to the server
> update /home/httpd/perl/lonmaxima to rev. 1.41 as follows:
> 
> wget -O /home/httpd/perl/lonmaxima
> 'http://source.loncapa.org/cgi-bin/cvsweb.cgi/~checkout~/loncom/lonmaxima?rev=1.41;content-type=text%2Fplain'
> 
> 
> /etc/init.d/loncontrol restart
> 
> As restarting the LON-CAPA daemons impacts users, this action should be
> scheduled for a time of limited user activity.
> 
> Note: lonmaxima rev. 1.42 (which also includes the fix for bug 6653 to
> address occurrences of "code ran too long" a different origin -- this
> time seen only sporadically, and also on Linux distros besides
> CentOS/Scientific Linux/RHEL 6 -- which Justin had remarked on in an
> earlier post in this thread is included in the forthcoming LON-CAPA 2.11.0.
> 
> 
> Stuart Raeburn
> LON-CAPA Academic Consortium
> 
> 
> Quoting Peter Dencker <dencker at math.uni-luebeck.de>:
> 
>>
>> Hi,
>>
>> you wrote
>>
>>> Right now my LON-CAPA server times out on them,
>>> and asks them to try again later.
>>
>> I've tested it: You seem to suffer by the same strange maxima-bug
>> on host 'mtal1', see the 'catching maxima errors'-threads on
>> http://mail.lon-capa.org/pipermail/lon-capa-users/2012-November
>>
>> The problem is still unsolved.
>> Therefore you have to check *all* syntax errors.
>> This was the initial motivation for my library-project.
>>
>> Maybe I can offer a preliminary workaround-library ...
>>
>> Or can someone finds similarities of the hosts 'mtal1' and 'scll1'
>> that may cause this bug :-/  ?
>>
>> - Peter
>>
>>
>> Am 21.10.2013 16:23, schrieb Angus Findlay:
>>> When I give students formula response questions some of them don't 
>>> submit valid mathematic expressions. The most common problems are:
>>>
>>> 1) Not having balanced brackets. (i.e. number of left brackets do 
>>> not match number of right brackets)
>>> 2) Including invalid characters, such as quotation signs.
>>> 3) Not including brackets, when necessary. (eg. 2^-1/2 instead of
>>> 2^(-1/2))
>>>
>>> I want to tell them that the expression is invalid, tell them why, 
>>> and allow them to enter it again, without penalty. (Right now my 
>>> LON-CAPA server times out on them, and asks them to try again later.)
>>>
>>> Is there a piece of standard code I can use for this, or if not, 
>>> what it the easiest to implement a solution to this?
>>>
>>> Thanks in advance.
>>>
>>> Respectfully,
>>> ----------------------------------------------
>>> Mr. Angus Findlay
>>> Math Department, Mount Allison University
>>> Sackville, New Brunswick
>>>
>>> email: afindlay at mta.ca    tel: 506-364-2592
>>> ----------------------------------------------
>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> LON-CAPA-users mailing list
>>> LON-CAPA-users at mail.lon-capa.org
>>> http://mail.lon-capa.org/mailman/listinfo/lon-capa-users
>>>
>>
>>
>> -- 
>> Dr. Peter Dencker
>>     wissenschaftl. Mitarbeiter
>>
>> UNIVERSITÄT ZU LÜBECK
>>     INSTITUT FÜR MATHEMATIK
>>
>>     Ratzeburger Allee 160
>>     23562 Lübeck
>>
>>     Tel +49 451 500 4254
>>     Fax +49 451 500 3373
>>     dencker at math.uni-luebeck.de
>>
>>     www.math.uni-luebeck.de
>> _______________________________________________
>> LON-CAPA-users mailing list
>> LON-CAPA-users at mail.lon-capa.org
>> http://mail.lon-capa.org/mailman/listinfo/lon-capa-users
>>
> 
> 
> _______________________________________________
> LON-CAPA-users mailing list
> LON-CAPA-users at mail.lon-capa.org
> http://mail.lon-capa.org/mailman/listinfo/lon-capa-users
> 


-- 
Dr. Peter Dencker
    wissenschaftl. Mitarbeiter

UNIVERSITÄT ZU LÜBECK
    INSTITUT FÜR MATHEMATIK

    Ratzeburger Allee 160
    23562 Lübeck

    Tel +49 451 500 4254
    Fax +49 451 500 3373
    dencker at math.uni-luebeck.de

    www.math.uni-luebeck.de