Hi everyone,

I'm trying to pick this conversation up again.  We can discuss on our next call.  Anyone have any other thoughts to add to Micah's?

Karen

From: "Dr. Micah Altman" <[log in to unmask]<mailto:[log in to unmask]>>
Date: Friday, September 20, 2013 4:14 PM
To: Karen Cariani <[log in to unmask]<mailto:[log in to unmask]>>
Cc: NDSA list serv <[log in to unmask]<mailto:[log in to unmask]>>
Subject: Re: [NDSA-INFRASTRUCTURE] Fixity check project


  2.  Fixity trade offs decision tool. Given different systems, different kinds of content, different scales of content etc. there should be different approaches to the frequency of fixity checking. There is interest in developing some kind of grid, or decision tree that could help orgs decide the best approach to how frequently to check the fixity of their content.

It seems to me that 2 is the logical foundation for 1. That is -- one should have a desired strategy for fixity checking in mind before developing guidance aimed at helping those new to the area to implement such a strtagey.

Anyone interested in number 1 specifically or number 2?  Should we start on both and see how it breaks out?
For number 2 were a number of ways we could break this out and look at it, or rather a number of parameters to frame it.  Here are the ones listed above:

System type
Type of content
Scale of content (not sure what this means)
Frequency based on those factors
Any others?

(I'm not sure what is meant by "system type". )

My hypothesis is that the factors that drive an appropriate choice of strategies are:
     - number of objects in collections
     - number of collections
     - size of objects
     - willingness to accept risk of loss ( or alternatively, budget --  which will indirectly bound risk acceptance)
     - content  type / structure
     - trust model

Based on these factors the strategy at the "leaf" nodes of the tree might involve recommended combinations of parameters :
      - compression algorithm, number of replication, fixity computation algorithm, verification algorithm, verification frequency, repair frequency

best,

Micah


Karen



############################

To unsubscribe from the NDSA-INFRASTRUCTURE list:
write to: mailto:[log in to unmask]<javascript:;>
or click the following link:
http://list.digitalpreservation.gov/scripts/wa-DIGITAL.exe?SUBED1=NDSA-INFRASTRUCTURE&A=1


--
------------------------------------------------------------------------------------------------------------------------------
Micah Altman, Ph.D. <http://micahaltman.com>           Twitter: @drmaltman
Director of Research -- MIT Libraries; Head/Scientist, Program on Information Science
Non-Resident Senior Fellow, Brookings Institution
"Entia non sunt multiplicanda sine necessitate" - Doctor Invincibilis (Corollary, "Ad indicia spectate.")


############################

To unsubscribe from the NDSA-INFRASTRUCTURE list:
write to: mailto:[log in to unmask]
or click the following link:
http://list.digitalpreservation.gov/scripts/wa-DIGITAL.exe?SUBED1=NDSA-INFRASTRUCTURE&A=1