[Building Sakai] Reaction: Sakai course on CD

Mark Norton markjnorton at earthlink.net
Tue Jun 9 10:15:28 PDT 2009


 > You mention that 'SCORM "could" be used to delivery encrypted tests

SCORM, as a specification, doesn't really state how tests are 
represented or delivered.  Rather, it defines a communication mechanism 
to report back results.  The problem comes when you are trying to 
determine if a user has answered a question correctly.  The low tech way 
is to include the answer in the learning object which is usually 
presented as HTML plus JavaScript.  As you point out, students are smart 
enough to know how to look at page sources and look for potential answers.

Well, suppose the answers were encrypted and had to be decrypted by 
JavaScript?  The answers are present in the web page, but are not 
visible to casual users.  This is made a bit more secure by moving the 
decrypt scripts into external JS files referenced by the page.  
Alternatively, there could be an "answer evaluation service" that the 
web page submits to.  Finally, assessment could be moved out of the 
SCORM package, but wrapped in SCORM SCOs (I think some commercial 
offerings are done this way).  None of these approaches violates the 
terms of the SCORM spec, to my knowledge.

Search is a different kettle of fish.  I don't see why Sakai search 
can't be made to work with content mounted on a CD drive, though you 
would have to handle the problem of the mount being ephemeral (not 
always on).  You'll note that computer games often have this same 
problem.  I occasionally play Sid Meir's Civilization, which offers me 
the chance to fully install all content onto my computer (consuming 
gigabytes of storage) or referencing content on the CD.  The later 
requires the CD to be in the drive when playing the came (which is a 
nuisance, IMO).  We could offer similar choices, I think.

- Mark

Philip Fedchin wrote:
> Mark, 
>
> thanks a lot for clarifications! It does help. Yes, I understand that it was wrong on my side to mess SCORM and IMS-CP in this case. And I agree that the whole idea to put the course on CD is technologically rather backward-looking. But in our context this is explained by the fact that it is the only way to get official registration of Sakai courses as e-publications (essential for tenure track and related faculty development issues).
>
> You mention that 'SCORM "could" be used to delivery encrypted tests - could you give any insights on what is the way for achieving the task? 
> And another concern I mentioned is related to the necessity to download full course content from CD in order to do the search/use the content. Do you consider this as a problem at all? 
>
> With best wishes,
> Philip
>
> ----- "Mark Norton" <markjnorton at earthlink.net> wrote:
>
>   
>> First off, lumping SCORM and IMS-CP (I assume you mean IMS Content 
>> Packaging) together does them a disservice, since SCORM is based on 
>> IMS-CP.  CP is an envelop specification.  As such, it DELIBERATELY 
>> doesn't place any  restrictions or requirements on what goes into it. 
>>
>> So saying that CP is flawed because it won't encrypt tests kinda
>> misses 
>> the point.
>>
>> SCORM (sharable content object resource model) provides a way to
>> define 
>> reusable learning objects and includes ways to deliver content that 
>> includes sequencing, navigation, tracking, communications (that
>> includes 
>> grade capture), and bookmarking (btw).  While SCORM "could" be used to
>>
>> delivery encrypted tests, in practice, most don't bother.  Why? 
>> Because 
>> learning objects are focused on learning.  Assessment, like tests, and
>>
>> high stakes tests, like finals and qualification exams require a more
>>
>> secure environment (as you have pointed out).  So the fact that SCORM
>>
>> doesn't encrypt tests isn't a flaw, it was a design decision.  That
>> may 
>> make it unsuitable for your needs, but it doesn't make it any less
>> useful.
>>
>> As for SQLite, while it may be the most widely deployed SQL database 
>> solution in the world, it uses a C/C++ interface.  Sakai is largely 
>> based on Java (there are a few exceptions).  As such, Java developers
>>
>> tend to prefer Java database solutions, like MySQL or Oracle.  Adding
>>
>> encryption to an existing Java database solution (even a generic,
>> cross 
>> platform solution) wouldn't be difficult to develop.  Finally, SQList
>>
>> doesn't include encryption out of the box.  A separate organization 
>> provides SQLiteCrypt, that does.
>>
>> All that aside, there is a need for good security in delivering tests.
>>  
>> Most of Sakai assessment applications, including Samigo and Melete,
>> rely 
>> on the fact that tests are delivered via the web.  The content
>> available 
>> to students at the time of the test DOESN'T include answers.  It would
>>
>> require direct access to the server and finding the particular XML
>> files 
>> in question to find answers to a question.  This is further
>> complicated 
>> by random question selection at delivery time.  So, used as designed,
>>
>> Sakai can deliver secure assessment.
>>
>> It therefore seems to me that your problems arise BECAUSE you want to
>>
>> deliver it on a CD.  It's no surprise that moving to an older delivery
>>
>> technology would re-introduce old security issues.
>>
>> - Mark Norton
>>
>> Philip Fedchin wrote:
>>     
>>> Dear Collegues!
>>>
>>> We do appreciate your help and constructive suggestions! And sorry
>>>       
>> for the delay in reaction. I would like to outline our position (and
>> reaction to expressed suggestions) in one message. Once again, I will
>> be really thankful for any additional
>> criticisms/clarification/comments etc. Sorry, if I missed any of your
>> arguments.
>>     
>>> 1) Yes, we did consider Solo initially as a possible option. But the
>>>       
>> problems with Solo, it seems, that offline search and test
>> functionality would not be possible to realize?
>>     
>>> 2) Concerning using SCORM (and IMS) packages as the background
>>>       
>> format for this offline instrument there are the following obstacles:
>>     
>>> - SCORM and IMS packages are zip archives as opposed to the solution
>>>       
>> we envision where data is stored in SQLite database;
>>     
>>> as a result of using these zip files the search functionality will
>>>       
>> be limited (all the data needs to be downloaded to active memory) - it
>> means the resources needed are rather large and it will be a rather
>> serious burden for end-user workstation; 
>>     
>>> - Am I right to suggest that both SCORM and IMS do not have any
>>>       
>> serious support of content encoding? On the opposite in the SQLite
>> there is AES password encrypting support and so even if you download
>> the database on other workstation, you would not be able to decipher
>> the data. This is essential if we want students to run tests offline
>> and then to synchronize the results. My understanding is that in SCORM
>> package case the student will be able simply to open the folder and
>> find the correct answers ;
>>     
>>> - Another functionality missing in case of SCORM/IMS would be
>>>       
>> bookmarking of the course content.
>>     
>>> With best wishes,
>>> Philip
>>>
>>>
>>> Interesting idea, John, but there are some pretty serious
>>>       
>> restrictions
>>     
>>> in IMS-CC.  There is no sequencing defined in the spec, for example.
>>>       
>>  
>>     
>>> Melete modules would need special treatment, though SCORM is
>>>       
>> supported
>>     
>>> as a data type in CC.  Glossary should be simple enough - burn it
>>>       
>> as
>>     
>>> HTMLpages.  CC has no support for Search.  Also tests would have to
>>>       
>> be
>>     
>>> in the restricted form of QTI supported by CC.
>>>
>>> Zach and I worked on getting IMS-CC import to work in Sakai some
>>>       
>> time
>>     
>>> ago.  The spec definition changed out from under us (in part because
>>>       
>> IMS
>>     
>>> took waaay too long to validate the spec and make it "public".  Zach
>>>       
>> and
>>     
>>> I both moved on to other (paying) projects, so there is some code,
>>>       
>> but
>>     
>>> it's broken.  It's a good place to start from, however.
>>>
>>> - Mark Norton
>>>
>>> John Norman wrote:
>>>   
>>>       
>>>> In addition to other comments, this sounds like a slightly unusual 
>>>>         
>>>> (from Sakai perspective), but very reasonable scenario for IMS
>>>>         
>> Course  
>>     
>>>> Cartridge support. If you could find or create a lightweight  
>>>> distributable IMS Course Cartridge player and distribute/develop
>>>>         
>> that,  
>>     
>>>> then Sakai will be wanting to support Course Cartridge export (if
>>>>         
>> we  
>>     
>>>> don't already) as part of its archiving strategy. The reason I
>>>>         
>> mention  
>>     
>>>> this specification is that the scenario you outline seems very
>>>>         
>> similar  
>>     
>>>> to the Publisher use-cases that informed the development of the
>>>>         
>> spec,  
>>     
>>>> and distributing materials on CD that include content and
>>>>         
>> assessment  
>>     
>>>> is very similar to some of the online active textbook projects that
>>>>         
>>  
>>     
>>>> publishers are developing and distributing using the spec.
>>>>
>>>> A key issue might be around expectations of communication back from
>>>>         
>>  
>>     
>>>> the system running the CD, about the activities and scores of the 
>>>>         
>>>> student.
>>>>
>>>> John
>>>>
>>>> PS there are commercial Course Cartridge players.
>>>>
>>>>     
>>>>         
>>>   
>>>       
>>>> On 5 Jun 2009, at 11:31, Philip Fedchin wrote:
>>>>
>>>>  
>>>>     
>>>>         
>>>>> Dear Collegues!
>>>>>
>>>>> We have a rather urgent need at our university to create
>>>>>           
>> CD-versions  
>>     
>>>>> of our Sakai courses.
>>>>> In relation to this need we are considering a development of  
>>>>> specific tool on the basis of OCW. We would be really grateful for
>>>>>           
>>  
>>     
>>>>> your comments/suggestions/expressed interest to participate.  
>>>>> Although we are very much looking forward to Sakai 3, but the task
>>>>>           
>>  
>>     
>>>>> is much more urgent and we are forced to base the process on
>>>>>           
>> current  
>>     
>>>>> Sakai 2.5.4 version with the addition of couple Contrib tools. May
>>>>>           
>>  
>>     
>>>>> be you have some ideas concerning more productive way of realizing
>>>>>           
>>  
>>     
>>>>> our short term goals in terms of Sakai future.
>>>>>
>>>>> So, we have in mind the following list of tools/functions
>>>>>           
>> available  
>>     
>>>>> from the disk/local workstation:
>>>>> 1) modules content (Melete and Resources, perhaps Conditional  
>>>>> Release);
>>>>> 2) tests (Test&Quizzes);
>>>>> 3) glossary (glossary);
>>>>> 4) search (new tool for the search on CD).
>>>>>
>>>>> The preliminary plan is to develop the tool on the basis of OCW.
>>>>>           
>> The  
>>     
>>>>> tool would consist of server and client parts. The server part
>>>>>           
>> would  
>>     
>>>>> be fullfilled by Sakai through the creation of standard course
>>>>>           
>> site  
>>     
>>>>> with modules created using Resources, Melete, Test&Quizzes. The  
>>>>> client part is a player (interface), which allows to present on
>>>>>           
>> the  
>>     
>>>>> CD the course downloaded from Sakai.
>>>>> And we plan to use Adobe Air (Flex) as the technology behind the 
>>>>>           
>>>>> client part.
>>>>>
>>>>> The standard scenario is the following:
>>>>> 1) Professor creates a course in Sakai using Melete, Resources,  
>>>>> Test&Quizzes and Glossary
>>>>> 2) Using the tool the course materials files are downloaded
>>>>>           
>> together  
>>     
>>>>> with the XML file reflecting the structure.
>>>>> 3) The files are copied to CD
>>>>> 4) The course is played from CD using Adobe Air based player.
>>>>> There will be an opportunity if needed to create client interface 
>>>>>           
>>>>> using SWT (The Standard Widget Toolkit) as we are planning to  
>>>>> document and make available the code.
>>>>>
>>>>> Looking forward to your reaction.
>>>>> Best,
>>>>> Philip Fedchin
>>>>> Associate Director for Information Resources
>>>>> Smolny College of Liberal Arts and Sciences
>>>>> (St.Petersburg State University & Bard College)
>>>>>       
>>>>>           
>>>       
>
>   



More information about the sakai-dev mailing list