Metadata and Open Textbooks

Metadata and Open Textbooks


Okay, so today we are talking about the exciting
topic, one of the most hotly anticipated topics of the OER universe, which is metadata. Unfortunately, Laura Dawson, who’s a very
experienced person in metadata mainly from book publishing and database universe, who
was going to talk generally about things, unfortunately she’s not available. So, there will be three of us talking today. So, the topic is metadata. I’ll do a quick introduction in a minute,
but first of all welcome. As I think most of you guys know this is the
office hours, which is done in partnership between Rebus and the Open Textbook Network,
to talk about issues around open textbook publishing on a monthly basis. And just very quickly for those of you who
don’t know about Rebus, we are a nonprofit funded by the Hewlett Foundation. And we are developing a set of processes and
tools to help publish open textbooks and we work with universities and colleges in open
textbook projects of various kinds. And try to build out a really collaborative
approach so that we can scale up the publishing of open textbooks and help make it easier
for people who are starting in programs, or who already have programs, but to support
that process of publishing. So, that’s what we do, and Liz, maybe we’ll
put a link to the Rebus community in the chat, so you can find us if you want to help other
people get some open textbooks published. And I will turn over to Sarah to say hello
and talk a bit about the Open Textbook Network. And then I will do a very quick intro to the
topic we will chat about today. Hello everybody. My name is Sarah Cohen. I’m the Managing Director at the Open Textbook
Network. Really honored to get to do this work with
Rebus. And so glad that we’re hosting another Office
Hours together. Also, happy to see a number of familiar names
and faces here. The Open Textbook Network, just briefly, is
an alliance of close to 600 campuses across the United States, working to support access,
affordability and student academic success, through the use of open textbooks. We do that by supporting adoption, modification,
and creation. And also work to develop sustainable programming
for open education initiatives locally, at institutions, and with systems and consortia. And so I’m going to put our link in there,
and happy to talk to anyone after the call if anyone would like to learn more. The other thing that we do is maintain the
Open Textbook Library, which I’ve also added a link there. And we try to create a one-stop shop for open
textbooks for use anywhere, by anyone. And so, we accept books from all over, not
just OTN members, that are reviewed by faculty, for faculty. So, again, happy to answer questions about
that, too, if anyone has some. And happy to be here and happy to see you
all. Great. Thank you, Sarah. So, I’m going to do my best to do a quick
introduction to the topic, which is metadata. And I think that metadata is important. I think it becomes particularly important
as we imagine a universe where more and more open textbooks are produced. And particularly more and more textbooks available
on the web. And so, finding ways to discover that content,
to understand what’s in it, to understand maybe some other things about it, what license
it has. Right now, how many textbooks are in the Open
Textbook Library, Sarah? Sorry, I muted myself. Around 425. 425, so we’re still– this is in the scheme
of what a lot of you librarians might have in your collections broadly, it’s a pretty
small number. But we can imagine in a decade that that multiplies
by 10 or more, if we all do our jobs correctly. So, finding content, helping faculty find
content, knowing what’s in the content, knowing things about the content, this is the stuff
that metadata does. And we have a pretty long tradition in the
library world in the book publishing world for this. Open textbooks have come out of a slightly
different place and I think that it’s important that we think about these things as we think
about scaling this up, since we don’t just have a massive content that no one can figure
out what’s in it. So, that’s roughly it. I think it’s very important. It’s also very important, especially in the
web context, because there’s a lot of stuff on the web, and both in the library context,
in the web context, they’re slightly different, but it’s just critical that we get this right. It allows us to build a lot more on top of
open textbooks, if we do this right. And if we do it wrong, it causes everyone
a lot of headaches. But I think that’s about it, for the introduction
to the topic. I’m going to turn over to Naomi Eichenlaub,
who is a catalogue librarian, is that correct? At Ryerson University in Toronto. And Ryerson has been doing a bunch of projects
around open textbooks lately. And Naomi is part of that team, thinking particularly
about metadata. So, over to you, Naomi. Again, probably five or seven minutes to intro. And for those of you who don’t know, we’ll
try to keep what we’re talking about fairly tight, just around five minutes, and then
we’ll open it up and hopefully have a good discussion afterwards, to ask questions and
talk about things. So, over to you, Naomi. Thank you very much, Hugh. As you mentioned, I’m Naomi and I’m a librarian
at Ryerson University in Toronto, which is yeah, in Toronto. I recently worked on an OER project here at
Ryerson along with one of my colleagues, Trina Grover, who’s also in the audience today. And is also a cataloguer at Ryerson. So I’m happy to be here today to share a little
bit about our recent experiences. Because it was new to both Trina and I to
dive into the metadata world around OER, specifically. So, Ryerson became involved in an OER project
recently, that was an Ontario e-campus funded initiative. And the project was to build a prototype open
publishing infrastructure, that would integrate and extend the BC Open Textbook Library. And the content would be migrated from BC
Campus to e-campus Ontario. So, this project began in earnest in May of
this year. And the project has a completion date of tomorrow,
actually, August 31st. So, there were four deliverables. And this is just to give you a little bit
of context in our project, which were to create a working Pressbook prototype with enhanced
features to integrate an open source repository. And that’s where we came in, and I’ll talk
a bit more about that. Community building and mobilization around
OER and then a learning module authoring and distribution prototyping. So, Trina and I, along with our web services
librarian, Sally Wilson, were all part of the infrastructure workgroup, which was responsible
for the process of integrating the open source repository. That would be storing the objects, or that
would support the open textbook platform and other content. So, we were told that we would be storing
the objects in a D Space repository, that was the open source solution that was selected. And we were asked to advise on metadata issues
around access to the OERC collection. And there was a view in addition to this,
to expanding content beyond just books. So, again, this is a prototype project, that
is meant to provide access to OER, not just textbooks, but other OER as well. So, in terms of metadata we looked at trends
in learning object metadata and identified a couple of top, I would say, metadata schema
contenders that we would consider. And we wanted to keep in mind that we were
focused on, of course, on standardization, interoperability and openness, so those were
very important to us. And we set out with those priorities in mind. So, we did consider schema.org’s LRMI, Learning
Resource Metadata Initiative. And we looked also at OER Commons’ metadata
application profile. And we did start to do a little bit of work
mapping for example schema.org to Dublin Core. Dublin Core was the default metadata schema
that we were looking at because it was the default supported in D Space. And we also went as far as testing out registering
our LRMI and the OER Common’s metadata registering it in the D Space metadata registry. But at one point over the summer we had a
dedicated metadata meeting, and there were numerous stakeholders involved. Hugh was there with the development stakeholders
and the search and discovery layer of stakeholders of the project as well. And we had some really good discussion about
which schema should we move forward with. And after much discussion it was decided that
because we did have really, really tight time constraints on this project, just a few months,
and because this was just the first phase of a prototype project we decided that we
would move forward with Dublin Core for now. Even as it was the default metadata scheme
in D Space. So, we knew that we were going to– that meant
that we would not have some– we would not have the educational components of this schema
during this first phase. But we decided that in the interest of moving
forward and with our time constraints that was the best thing to do at the time. So, and also because we wanted to draw upon
the expertise of the broader learning metadata community, and work towards a common OER standard. So we wanted to make sure to put the time
and resources into having the conversations that needed to be had. And doing the work that needed to be worked,
to really build a shared, common, open OER metadata schema that we would hopefully work
towards in phase two. So, as I said, discussions, lots of discussions
and conversation, but we recommended that we start– we use Dublin Core as the single
metadata schema in the first phase of our project. And then, we proposed building out an OER
extension as a recommendation for phase two of this prototype project. So, down the road, so that we could include,
of course, OER specific elements that we definitely want to have in our repository, especially
for discoverability. So things like has this object been reviewed,
adopted, adapted, etcetera? And all of this, I just have to say, I’m just
keeping a quick glance on time, I think I have about one more minute, Hugh? All of this reminds me of that Randall Munro
comic, that you may or may not have seen about how standards proliferate. So, there are 14 competing standards, oh,
14 you say? Well, that’s ridiculous, but we do need to
develop one universal standard that covers everyone’s use cases. So, now that’s we how we end up with 15 competing
standards. But I think just to quickly summarize, some
of the challenges for us in addition to selecting and just selecting our metadata schema is
just thinking about the complexities of the different types of relationships between the
objects, in the OER repository, versions, adaptations, etcetera. So, we’ll definitely have to give that more
thought, going forward into our second phase. And then, we also had a hard time because
this was a prototype sort of imagining and visualizing the various workflows, metadata
workflows specifically. So, what would be automatically generated,
what metadata would be provided by authors, and what metadata may be mediated. So, that was another challenge for us. And I’ll just stop it there, but I’m happy,
of course, to pick up on any of these topics once we move further into the discussion. Oh, Hugh, you’re muted. Thank you, Naomi. So, again, we’ll– so, Sarah’s going to talk
now about the Open Textbook Library. I’m going to talk a little bit about what’s
going on at the W3C. ‘Cause I think that’s important as well, and
then after that we can open up and have discussions. And I did mention earlier that Melinda is
here from OER Commons. So, she can– maybe we can invite Melinda
as one of the first commenters to give just an overview of their approach to metadata. I know you guys were involved in the LRMI. Okay, so let’s start with Sarah Cohen. Over to you. Thank you. And thanks so much, Naomi, that was super
interesting, thank you. So, I’m going to be in quite a bit of contrast
to Naomi, because I am not a cataloguer and so I’m going to give a brief overview of kind
of how the Open Textbook Network came to develop MARC records for the Open Textbook Library. A few things about the Open Textbook Library
that you should know is that we are actually a referatory, we are not a repository. So, the books in the Open Textbook Library
are actually from all of the different campuses, universities, institutions that create and
provide access to this material. So, we are not actually housing any of that
material ourselves. And so, that was a huge consideration when
we were trying to figure out how are we going to provide increased access to the content. We were very fortunate to start working with
Colorado State University. I want to say almost three years ago, now,
on trying to develop metadata for the Open Textbook Library. I do want to say that for us open is perhaps
the most important thing to everything about the Open Textbook Network. And so we were contacted by a number of vendors
who offered to create metadata for us. But we did not feel comfortable getting involved
with vendors to create any metadata around this content for fear that it would start
to become proprietary. And that’s really not what we’re trying to
do. So, that’s when we turned to the network and
asked for a network member to help us deal with this. We chose to make MARC records primarily because
we are an incredibly small team and we were using the benefit of a campus. We wanted anyone to be able to download this
content, or actually, sorry, records to this content from their library so that you could
use it in your discovery layer. But for even smaller or more marginalized
institutions, we wanted them to be able to have access to the content and every institution
now does have an OPAC. So, we did want to do something that was really
accessible for all types of institutions, including internationally. And so, we know the books from the Open Textbook
Library, our records are being downloaded heavily across the United States. And also through larger systems, like for
example, the Louisiana Library Network, Louis has downloaded it through their entire system. But we also know that it’s used internationally
throughout Asia, Africa and South America. So, we are trying to consider– that was a
big consideration for us. I did also just want to say that in terms
of being a referatory we also are facing those challenges of pointing to the right location
for content. And so, the Open Textbook Library does maintain
a dark archive, so that if the links that were being included in our metadata should
no longer be correct, that we’d be able to make that change. We do provide a cumulative file of our MARC
records that is updated. We try to update that monthly. It also does include a removal list to be
used. And we also rely on the community to let us
know when things are not working, and so we’ve been contacted just last week, with a record
that was in duplicate. And so, we’ve been working on trying to clean
up those files with Colorado State as well as OCLC. And so, I think that that’s actually all I’m
going to say, and wait for questions. Okay, awesome, Sarah, thank you. So, I’m going to talk briefly about what’s
going on at the W3C which I think has very important implications for open textbooks. So, W3C is a standards organization that sets
the standards for the web. And there’s a new initiative and Rebus is
a member of the working group developing this new standard called “Web Publications.” And the idea is to build a web native standard
for essentially a book on the web. And the reason we need that is the web has
the notion of single addresses, single URLs with a page attached to it. But there’s no fundamental protocol or standard
to say “Here is basically a unique item,” let’s say a book, “and here are several
different pages, or chapters, or resources associated with this book.” So, there’s no kind of sense of at the protocol
or standards level of a something that lives here, at a URL and here are associated URLs
with that. So, you basically can’t build a TOC into any
browser level applications in a way that’s standardized. So, at the W3C there have been a lot of discussions
about how to get this thing to work. And kind of the key elements for getting a
web publication, really the two key elements, there are a few others, but basically you
need to have a place where it lives, a unique place, so URL and then, what they’re calling
a manifest, which means what else is in this, what is at this URL? What are the other resources here? So, essentially a table of contents, there’s
some notion of defining a preferred reading order. So, mimicking sort of a linear route through
content but with the notion that applications could be built that would allow you to navigate
differently true content than perhaps a linear route. And then, the last– another piece is metadata
and how to represent metadata in these web publications. So, we actually somehow or other, despite
not having much expertise in this, but maybe I was inspired by my meeting with Naomi and
Trina in Toronto. Somehow we got saddled with the job of helping
to write this specification, or the draft specification for the metadata piece of web
publication. And the general approach we’ve taken is that
there will be some very limited– so there will be actually no required metadata at all
in a web publication. But there would be recommended metadata, which
would be– so there’s no must haves. There’s some recommended should haves which
would be: title, author, I think license information, there’s a handful of other things which are
escaping me right now. I can share the draft with you guys afterwards. And then, finally, for extended metadata the
idea would be linking to a file that would be to whatever specification metadata specification
that you care to choose. So, whether it be ONIX, which is what the
commercial publishers use, MARC records, which is what libraries mostly use, Dublin Core,
which actually now I get lost in the details of how these things work. I think MARC is kind of an implementation
of Dublin Core, if I’m not mistaken or vice versa. And then, schema.org, for instance, which
is the fundamental or becoming the standard for websites how they structure their metadata. So, the idea again, is that there should be
a very light specification that allows you to have, for instance, you could have four
different metadata files, one for each application. And I think in the world of open textbooks
where in a lot of cases they’re web first, or hopefully we’re going in that direction. Open textbooks are web first, but there maybe
e-pub files and PDF files, etcetera, associated with them as well. The general approach the W3C is taking is
that there be a very limited amount of required metadata for a web version, and then you could
have these other types of metadata files, which would be to a given specification. So, it was trying to avoid the exact problem
that Sarah mentioned of instead of saying, “Okay, what should be the web publication
metadata specification?” Just use what already exists and can sign
it off. So, again, what becomes exciting about this
is this notion of what we could start to do, if we had a really clear standard for what
open textbook looks like on the web and how the metadata is defined, how we expect the
table of contents to be defined. And if we do that, it means we can start building
new kinds of applications on top of books on the web, new reading experiences, new different
kinds of applications. And metadata obviously is a big part of that. Okay, so I think that’s it. I hope I didn’t go on too long, whoever is
in charge of time, I guess that’s me. So, I’m going to wrap that up there, just
say that the W3C work is kind of happening right now. I don’t expect anything real to happen sort
of I suspect if it was in six months I would be shocked. But the process is there and there’s lots
of good people and we seem to be moving in the right direction, in my opinion. And I think open textbooks are going to be
one of the first big use cases actually for this new specification. Partly, just ’cause I’m involved and I’m pushing
for that to happen, but also because the commercial publishers have constraints that open textbooks
don’t have on the web. And that’s exciting. Okay, so I think that’s it. Normally, we have taken questions in the chat,
but I think our group, we’re 27 people, so I think maybe we should allow people to ask
questions and have a more open discussion. I think you all have to un-mute yourself if
you want to ask a question, maybe type in the chat, I’m not sure. Anyway, un-mute yourselves if you’d like to
ask a question. But before we do that, I wonder if I can just
ask Melinda to talk a little bit about the OER Commons, and what they’re doing with metadata,
because I know you guys have done an awful lot of work including the LRMI stuff. I know you guys were involved in that. So, maybe you can just do it quick, just a
two minute overview maybe of what OER Commons is, and what your approach to metadata has
been. Melinda? Sure, I’d be happy to and invited our information
services manager to join, because she can speak better to our metadata strategy and
work than I can. (Laughs) But I am Mindy Boland, I’m A Senior
Product Manager of OER Commons. And in case you are not familiar with it,
OER Commons is a digital public library of open education resources. We both host and link to over 60,000 pieces
of OER, many of them are textbooks. And we have over 10,000 registered users. We do our best to make as much OER findable
in the world as we can. And Michelle, why don’t you talk a little
bit about where we have been, and where we are with our metadata standards? Sure. The OER Commons metadata schema used IEEE
LOM sort of as its first guiding profile and that was way back before I joined. It’s way back in 2007, I believe. And since then, we’ve sort of taken an approach
where we build different modules on top of that core that map to different metadata standards
in the field. So, for example, on top of that, we have built
extensions that map to LMRI metadata, that map to A11Y metadata. That map to I’m trying to think. We did a bunch of work with NSDL to map metadata
to NSDL_DC and to [?] for example. So, the approach that we take is to have the
a core sort of internal metadata standard that can support all of these modules and
then when we push content out onto the web, for example, or into the learning registry,
or what have you. Then we have metadata to whatever the application
that we’re pushing to needs. So, you’ll see if you look at OER Commons
resources on the web, that we’re using schema.org and LRMI together there, to provide metadata
markup inside the page or so it’s micro data markup. When we push resources to the learning registry
we’re using the “Go Open Node” now, so we’re standardized to the LRMI schema that
they’re using. And we also can map from OAI DC and NSDL_DC
and LAR. So, that’s kind of an overview, is that we
really try to be flexible and accommodate different standards and different contexts. So, I hope that answers the question. Awesome. So, I don’t know there’s probably different
levels of metadata expertise on here. So, for instance some of those acronyms, I
think was a note here, that there’s a lot of acronyms there in metadata standards. Sorry. I’m not sure if it makes that much difference
to know exactly what they are, I assume they’re different sets of standards of metadata schemas. Maybe you could just quickly, sort of broadly
respond to that. That kind of the character of the different
metadata standards and what’s different between them? Sure. Yeah. So, A11Y is accessibility metadata, and that
was– not sure if it– I think it was similar to LRMI. So, LRMI is specific to learning objects and
it was an extension that was recently accepted to schema.org which Hugh mentioned, is largely
what folks are using on the web. Because that’s what is used by the large search
engines such as Google, to sort of understand what’s in a resource. So, A11Y is similar to LRMI, in that it’s
just focused on accessibility. NSDL_DC was a metadata schema that was used
by the national science digital library. And that has elements of IEEE LOM which is
learning object metadata and I believe specifically for the web. And NSDL_DC does some interesting things,
where it combines some elements of the open archives initiative. And that’s the OAI DC standard I’m sure most
people on this call are familiar with that. Excellent. Okay. So I would sort of summarize that to say that
there are a lot of very specific metadata standards for very specific things, whether
it be accessibility or specifics around learning objects. And this is, again, just part of the ecosystem
that said those of us particularly involved in publishing need to be aware of and thinking
about. How to make it easy either to capture that
in the authoring process, or to support it later when OER ends up in repositories or
wherever. More questions. I think just maybe type if you want to have
a question in the chat, maybe, and we’ll– or maybe just jump right in, I don’t know. Maybe it’ll lead to chaos. Liz, looks like you have a comment, Liz. I was just going to say, it looks as though
Eric has a question about versioning. Yes, Eric, can you un-mute and ask your question? Hey everyone. Hi, Eric. Yeah, so one of the things I’ve been thinking
about for a while and starting to try to do some stuff about is agreeing how to deal with
versioning of open resources. One of the distinguishing characteristics
of open books, as opposed to traditional, non-open books is that a lot of them can be
modified by the user. So, if a Professor wants to replace a chapter
of a book with his own chapter, or if changing the order of a course, or something like that,
there can be many versions of a particular textbook. And we see that happening a lot for example
with some of the open access textbooks that are going around, and getting improved. And so, I wonder what people are– especially
the folks that have been working on cataloguing are doing about cases where books sort of
morph in small ways but ways that are important to users? Do we keep track of the formal tree of the
books, or do we pretend that they’re completely separate books with no relationship to each
other? I’m just curious if people have tried things
and run into problems and– or just closed their eyes to the problem and not worry about
it? Can I hop in about how we’re handling it in
the Open Textbook Library? Please. Yeah, please. So, it’s a really great question, Eric. And something that I think we ask ourselves
all the time. (Laughs) The way that we are at least addressing
in the Open Textbook Library is that we’re essentially saying that we’re maintaining
kind of the document of record. And we’re assuming and we try to provide resources
to support people who are going to– we’re assuming people are going to make changes. We want people to make those changes. But that does not mean that we’re going to
collect all of the different versions and all of the different changes that get made. And so, we’re actually taking the view that
that is up to people locally. However, if there is a significant enough
change to the content of a book, then we essentially will examine with you to say, “Does this
warrant it to be essentially a new book in and of itself?” But those kinds of changes you described,
in terms of moving– I mean, all the things we love about open, the ability to move things
around, the ability to insert your own research, the ability to take things out. Those are all things that we think people
can do locally, and therefore can also store locally, provide access to locally. I guess that doesn’t answer it from a metadata
perspective but it does say like we are maintaining that kind of the mother record. And the mother item. And allowing people to download from there
and then, whatever, again, because you have access to the MARC records, again, I’m taking
a very library centric perspective here, but the ability to then make those changes within
your own local metadata as need be, perhaps as a separate record, or another– with notes
in your 500 fields, or whatever that would look like. I hope that’s helpful. Yeah, that’s very helpful. I was just wondering whether you are maintaining
your own identifiers, to keep track of what your conception of the work is? What a great question, that I actually cannot
answer. Because I am again not a cataloguer, but a
question I’d be happy to pass onto the folks at Colorado State as well as our Director
of Publishing New Collections, who couldn’t be here today. They would probably better know that answer. So, I suspect there’s no– I mean, we may
have some strategy to deal with this now, but I don’t think other than sort of like
the world of Github which is a different universe where there’s tracking of every minute change,
talking about books and textbooks and larger works in this context is a very new universe
we’re about to enter into. I will say that Pressbooks, so we have a very
big, new feature that we’re pushing out, by the way, for those who don’t know, I do Rebus
community and I also do Pressbooks. And part of the work with Ryerson that Naomi’s
also been part of, was to develop, formalize the process of the ability to clone books
from one Pressbooks instance to another easily. So, this is supporting the five Rs of OER. And part of that was building in a formalized
very small, but a formalized un-editable, if you hack the software you can edit it,
but un-editable trace back to the original source content of the book. So, the idea is that there should be a history
of saying this book from here, from here, from here. And here is the original one. And that versioning we haven’t thought enough
about it, but the kernel of how we think we might approach it is there. And that’s something that we’re hoping to
be working on in the next number of months, to try to get something that makes a bit of
sense in that universe where it’s very easy to take content from one place, put it somewhere
else and then modify it. And have that be baked into that process of
things moving from one place to another. Now, that only would make sense if it was
going from one Pressbooks instance to another. It’s not the same if you’re pulling up the
e-pub and editing that, or whatever. There is a question here about– and maybe
someone can answer this. So, this is from C. Hollins, I’m not sure,
I don’t know if you want to ask that out loud? C. Hollins? Or I can read it out here, but the question
is: when does something and when does a work– how much modification happens until it becomes
a separate copyrightable work? And who makes that decision? Is it the author? So, that goes into licensing and I don’t know
if there’s anyone who’s got some open licensing chops who would like to tackle that kind of
complex question, on the call here? (Laughs) Anyone? As a copyright nerd, that’s– it depends a
lot on how you are approaching the intellectual property. So, you have a lot of examples in open source
software, where in order to change the license, people have to get consent of you know hundreds
of authors. And so, I don’t think there’s a really concept
of a separate copyrightable work, when we’re doing versioning. They are just derivative works. There’s no hard line as to when something
you know is so different from the work it’s been derived off of that it’s no longer you
know a derivative work, but a new work. But that’s really complicated and above my
level of nerd-ery. I will just say– Yeah. Go ahead, Sarah. Sorry. It took me a minute to scroll up to find the
question in the chat, so sorry there. I was just going to say that I think that
one of the benefits of– this is I think why there’s such advocacy for creating open textbooks
using CC by. And really trying to advocate for not only
technical openness but also legal openness. And I think that this is exactly one of the
struggles that we face, in terms of trying to take works that are– when there is more
to it than just attribution, that we run into these harder questions, that are not insurmountable. But I think raise a whole host of other issues. So, I just wanted to say, I think this is
a great reason for CC by, where it really is around, you know, here is the work that
I’m basing this off of. And I think that there’s great examples that–
and a lot of guidance that Creative Commons has offered, Lumen has also done quite a bit
of work in terms of and Open Stax has done quite a bit of work in terms of how they attribute
change, within their material. Especially when you go to sections or portions
of material, how do you actually attribute that? But I guess I would say that you know, as
long as you’re– the question that Claudia at least from how I’m looking at it, is that
the decision about you know, is this a whole new work? Or is this some– just a derivative work? I think is a great question, but I think in
CC by as long as you’re giving that attribution you’re moving– You’re continuing to create that new material
that you– as long as you’re saying that it’s coming from this other place, then you’ve
done your job. Would anyone dispute that? I know there’s some other– on this call– I think, Amy, I’m not sure if you have access
to a mic? To make your point there? But– Oh, I see Amy. There you are. Yeah. Can you hear me? Well, so, I think about this same question,
less from a licensing perspective because it’s what I’m thinking about is a derivative
of an openly licensed book, so it’s very clear that there is permission, etcetera. We know how to attribute it. But just from a collection standpoint, when
someone lands, for example, on Open Oregon’s Pressbooks page, is it useful for that person
to see the different versions? Or– sort of more along the lines– at a very
small scale of what Sarah was saying about the Open Textbook Library collection, you
know, and how different should it be? And how would a user know which version they
want to explore, etcetera? So, from that standpoint not a licensing collection,
but more like a management question. And one hopes this becomes a huge problem
in our world, right? Like this is a sign that we’re doing things
right? That things are getting adapted and used in
new contexts and improved for whatever particular context someone thinks they ought to be improved
for. And so, I don’t really have an answer, Amy,
except to say it’s like, it’s clearly an issue and a problem. But I think it’s the one we hope happens probably
more than most other problems, right? Like it’s a sign that something’s happening. I mean, that to me is a signal of your success
here in that book moving to somewhere else and getting used, but used in a way that meets
someone else’s requirements. And I think that’s exactly what we want. Hi, Hugh, this is Anita Walz from Virginia
Tech. Just wanted to add that there was a Twitter
chat about this, a little while back, with the hashtag #askcc. In which some of the Creative Commons US people
responded to a lot of similar questions that might be a good resource to look at. Yeah. In general, just reflecting that back, Creative
Commons is very deeply engaged in this, these sets of questions and are helpful. But I think (laughs) it does get very complex
very quickly and I think that’s an important thing that those of us thinking about this
from a system point of view. How do we try to reduce the complexity for
most people? And these questions about– there was a comment
here by Jonathan, about GIT Hub. And maybe I’ll turn over to you in a second,
Jonathan. But GIT Hub solves things at a technical level,
in a way that we could do here. But in a lot of sense, the question we’re
asking here isn’t so much technical as kind of cultural and day to day use. And if you have 50 different versions of the
same book, and you search for that book and you find all 50 versions, what do you do with
that? Is that useful for anyone? And how do we start thinking about handling
that? But these questions do get very complex and
I think we’re just going to have find sort of things that work well enough. But maybe, Jonathan, I don’t know if you wanted
to make– if you can come on the audio and make some comments about GIT Hub and Forks
and Commits and Repositories? And you might be muted. Hi, can you hear me? Yes. Yeah, I just agree that there’s a technical
issue, it strikes me that– if you want to keep a complete sort of versioning tree, this
is a problem that has been well discussed in the open source I use the acronym FLOSS,
Free Libre Open Source Software community for a long time. And things like GIT Hub or other kind of versioning
systems are– I don’t know, I feel like this is a problem,
and I agree, it’s a highly technical solution but it seems like in terms of textbooks, and
the kind of that the differences between the versions maybe significant for a particular
user and if a user cares enough to really explore. You want to make it discoverable if you do
enough work, but if it’s– if you’re just want to get a survey of what is available
in the general area, then you don’t need to dive deeply. So, I think that the issue is somehow how
to make it easier you know, to make easy things easy and hard things possible to steal the
line from the definition of Pearl. But you know, that a lot of this has been
done in places like GIT Hub and other FLOSS-y communities and so I wonder if we can just
steal some of that stuff? I don’t know. That’s a little bit off track from the metadata. I think the metadata idea I think helps then
people who are crawling these kinds of resources, to make them more easily searchable, findable. So, I think good metadata and well organized
is a key to help making easy things easy. Anyway. Yeah. I think– so what we found is the metaphor
for– like GIT Hub is the obvious metaphor to go to when thinking about open textbooks
when you first think about it. Translating to reality is a little bit more
difficult, it seems. Not sure why. So, there’s– sorry. Go ahead, Jonathan. You know, if you want to make another metaphor,
of course, there’s always the Wikipedia page view of things, right? When Wikipedia was a– if you look at the
revision history of Wikipedia pages and the talk pages, it’s– you know, and of course
the common line about Wikipedia is that it’s a complete failure in theory and it works
wonderfully in practice. And I don’t think we know why. If you want our textbooks– it seems like
what we were saying about the Open Textbook Library, or maybe other people who are trying
to have the most current version should be some sort of canonical version should be recognized
by the community. That’s sort of what we put– the Wikipedia
talk pages are about, making people make changes to some common resource and then there is
the canonical, current version. Yeah. You want to keep all possible versions. But, I as an instructor, would like to see
maybe I don’t like the canonical version so much, I wish chapter two and three were switched
in order, just to be able to see all of the forks and I would like to be able to drill
down to that kind of precision. Yeah. So there’s a comment here from C. Holland,
saying getting back to metadata, which is how do we know when to incorporate these authors,
multitudes, there may be multitudes contributing to a work. How do we know when to incorporate them into
the metadata of the work? I wonder if anyone can take that one? And if no one jumps in, I’ll give my general
opinion from a technical point of view. Again, this is the conflict that we’re dealing
with here. There’s a technical problem, and then sort
of the real life problem. So, from a technical point of view, you could
have a million authors attached to a book. And there’s no problem having that metadata
from a digital– in a digital book. Already been in the ten pages of the back
of the list of people who’ve contributed. What becomes more problematic is when we surface
that. And when do we want to see it and how do we
want to see it? I don’t know if anyone else has any other
comments on that question? Someone had a question about Open Stax and
how they track versioning. I don’t know if– I guess it was Jeremy had
the question. And maybe Anita has an answer. So, Jeremy’s question is how does Open Stax
track or link to derivatives of their works? And maybe Anita, looks like you’ve got an
answer for that. Hi, Jeremy. I’m not sure that they do. There are various derivatives in their CNX.org
site, where you can find lots of different ways that faculty have adapted these kinds
of– the Open Stax books. But then, there are other places where they’re
simply hosted and they’re the same, so BC Campus has a lot of their books. The e-pub version, and all the other versions
available as well as, I believe, that they have some of those in Pressbooks. So, I don’t know that they track derivatives. We could ask. Eric has a comment here. Talking about pushing metadata downstream
with web hooks. So, that sounds intriguing to me. Eric, do you want to jump in and talk a bit
about that? Well, it also relates to what Jonathan was
talking about, using GIT Hub. I don’t know if people are aware of what we’ve
done with the [GIT-enburg project 00:52:23] where we put all of project Gutenberg into
GIT Hub one book per repo. And the idea was to make things open so that
people could make corrections and revisions. And so with that, and we’re working with Project
Gutenberg to get them up to speed on moving all their stuff to GIT Hub. So, that should be fun when it happens. But one thing we’ve done is we have made it
so that when you make a change in one of these GIT Hub repos, that there is a Travis process
that goes and builds the e-books. And when that happens, it tickles a web hook
on our on [?] and then on GLUIT will go and read the metadata file on GIT Hub that has
been created as part of putting it on GIT Hub. And see okay, well, there’s a new version
number, or there’s a– we have to re-download the new files. And so it creates an updated metadata record
and on GLUIT and then when New York Public Library, which has been using our feed, to
feed the e-books into simply e they will read our modification date and re-download the
books, whenever there’s a new modification date. So, it’s sort of– it’s not very extensive,
right now. But it at least in this one chain, we’ve got
it to work. Yeah. I think again, one of the challenges is that
if you sort of control the production place, then it becomes easier to manage, thinking
about this. If you don’t, if there’s lots of things happening
in different kinds of ecosystems it makes it different, more difficult, which is why
having standards around this makes it a bit easier. Yeah. I agree. There was a comment here from Jeremy, which
I think is actually maybe not such a crazy– at first I disagreed. But anyway, the comment is how is this different
than new editions in books? You know, when a library has 16 editions of
books, of a book, each one gets a new metadata record and it’s up to the user to determine
which is the best for them. And I think the concern would be well, there’s
going to be a thousand different versions of one textbook, but maybe that’s not a real–
maybe that’s just an assumption that isn’t borne out by reality. Yeah, go ahead, Sarah. No. Again, I’m not a cataloguer, but I will say
like I think that what’s the challenge, and I think you’re right, Hugh, that you know,
there could be a thousand versions in part, because there is no longer that authoritative
voice of who makes the decision of what is a new edition. Right? So, I mean, that’s where I think your point
earlier Hugh about this is still a new area. And we are trying to figure out kind of what
does it take for us to live with all of the potential of openness and to be able to provide
as much access as possible. And I think a question or a tension and again,
I am not a cataloguer. But a tension there is, is access to every
single possible version really in the best– like is that going to facilitate, improve
discovery and access? So, I think that’s a challenge of working
on the Commons in that way. And that these kinds of conversations bore
that out and allowed this kind of community to form to answer them. And that’s really, I think where both Rebus
and the OTN love having these conversations and hearing these questions and bringing us
altogether around them. Yeah. So, we’re at 3:00 now. So, it seems like we could probably talk for
a lot longer. Metadata feels to me like one of the most–
so kind of combining metadata and versioning, it’s sort of a very fundamental building block
to what we’re doing here. But very complex, so I think as Sarah says,
this is a conversation that we’re going to keep having. And I think what our hope is that we can have
these together, so that we can try to, as we’re making decisions in the various projects
we all work on, that we can try to have those converge, rather than splitting off into 15
and 16 and 2,000 new metadata formats or standards. So, I think we’ll wrap it up, the video will
be available. Thank you to Sarah and OTN. Thanks to Liz, who’s kind of the force of
nature behind these Office Hours. And maybe Sarah, I’ll let you do the final
sign off. I– Oh, you’re mute. Nothing else to say, except thank you, thank
you Liz and Hugh, thank you to all of you for coming. And onwards we go together! Excellent. And Mindy from Open– having a total blank. OER Commons? OER Commons, thank you just ’cause your email
address is different and my brain just went rnk. So, OER Commons has probably done more thinking
about big different collections of metadata and trying to standardize it than anyone else
here. I think. So, definitely a good resource. And I think as we think about getting these
conversations maybe moving to more concrete directions that we’re certainly going to want
to have you guys on board. Or driving the conversation, or however it
ends up. But okay. Thank you everyone.

Daniel Ostrander

Related Posts

Leave a Reply

Your email address will not be published. Required fields are marked *