Content of this blog post:
- Welcome by Leo Appleton
- Keynote presentation by Karen Blakeman
- Lise Robinson OCLC
- Rapid fire sessions
- Bethan Ruddock MIMAS
- Panel questions and answers
Theme of this conference: Cloud busting. Given the number of
people we have here today, we think we’ve selected a relevant theme. The extent
of what cloud means to libraries is what we intend to get out of this
conference. There is a green slip in the welcome pack to write down questions
to submit to the panel for the closing Q&A sessions.
Welcome to Sheffield under the sun (we had snow one week
ago). I want to tell you why we did the move to the Cloud. The Cloud is
everywhere, we all have various devices, do shopping online etc., this brings a
lot of expectations. User interfaces are easy to use, we expect things to work
all the time. If it does break we expect support 24hrs a day. This expectation
is also on mobile devices. Most of the services we deliver is actually in the
hands of service providers and we have very little say. These are some of the
challenges that we face. We also have to do things overnight, over the weekend
etc. It’s so much part of our lives, we are so used to it.
We moved first to the
Cloud for emails. We were one of the 1st unis to go completely Google
for staff and students. We’ve been running like that for about 2 years. I think
we provide a much better service. When Google goes down for one hour it’s on
the news whereas, we once lost emails for 3 days, that would be unacceptable
now.
It's not a cost cutting exercise, we have saved money but
that wasn’t the main reason. It’s about providing better services, more
available and user friendly.
Delighted to be in sunny Sheffield compared to cloudy
Reading! I have my own company of information services. I don’t have access to
large database most of my research has to be done in the Cloud and I have to be creative to gain access to some
papers, mostly it’s contacting publishers an asking them.The issues of the Cloud, and benefits are very much the
same as a large organisation. We want to see where we’re coming from and in a
way it’s almost like going back.
The Cloud definitions: Stuff that is out there, it can be
all sorts, anything in electronic form available on the web. Another definition
is data & services that are serviced and accessible via the web so not just
publicly available sources but your own papers and documentation that you need
to run your organisation is hosted outside in a single interface (?).
If you remember when we had tape etc. you had to host them,
pick the one data source key to your business. The real progress then was having
only one device like the telephone, it gave you access to everything that was
available, you were no longer restricted to what you had in-house. There was much more data
that you could access. It’s the same now. The key is how you get to it. The
internet was a possible way of accessing these databases. Initially we were still
using dial-up methods etc. We had a pack of telecom software to access the
info. In 1994 netscape came, until then stuff sat on floppy disc, it was
telnetting, so point and click was key then. Now fast forwarding to my beauty,
my android. From this I can access everything, forget the laptop, with my
android I can quickly tweet people etc.
So Cloud computing is not really new, the main difference is
publicly available data. E.g. students working at a distant location, you can
access information outside of a physical location. Federated search engines, a
lot of you use them. There are different hosts, they go out identifying databases,
resources that are relevant and develop easy to use interfaces. Not only text,
but visualisation, images, clouds of data etc. There’s lots going on in the
presentation of data. Wonderful links to resources, in and outside of an
institution. Large amount of resources outside that you can link to your own
info.E.g. repository searches, open url, open access, etc., I won’t
go into details but the problem is that you may be able to find it but not
access it. There’s a large amount of different types of info.
My wonderful android is linked to my Google account so
everything is synchronised. It also means I’ve got backups. I have access to
all my various accounts, my apps, calendar, Googlemail, Evernote account, Mendeley etc. and it doesn’t matter what device I’m using. I can access my events etc. and don’t even
need to print my tickets because I can just show the screen. Mobile is the way
forward, it doesn’t matter what platform, it is it’s all out there.
My question is: where’s my stuff? This wasn’t a problem with
my desktop, I had a file manager, or file locater, that’s when you
really don’t remember where stuff is. But for the Cloud it doesn’t work. Google
tools etc. are not designed for real time stuff, but designed for indexing
stuff. There are other services designed for that, but there’re limited, e.g.
apps limit the amount of data you can pull back. E.g. topsy is a good one, or there’s the good
old fashioned rss but it’s not even really real time.
So how do I find my own stuff. CloudMagic does that but the downside is you allow
it access to your account. I did this for Facebook, twitter etc. then I
panicked, it’s a free service, what are they going to do with my data? I pulled
the plug and changed my password. So what’s the reputation, where is the data
stored, what security etc. are all important questions. If like me you travel a
lot, you might get wifi or not, battery is a problem etc. so Cloud is great but
when you’re mobile, you will have these kinds of problems. Stuff will be
unavailable. You don’t know what’s going behind the scenes. Privacy glitches
happen. Now security is tighter but we need to be aware they can occur with any
system. There will always be leaks, it’s a case of risk assessment.
With broadband we made the assumption everyone has a fast
connection, but it’s not always the case. What’s the policy regarding backups?
What if everything goes wrong? Do we need to keep local backups? Don’t assume
that because your stuff is in the Cloud it will be well taken care of. I always
have my own backups just in case.
Obfuscation, fragmentation, isolation are important words.
Obfuscation if you read contracts (do you read them?), we don’t because it’s a
nightmare to try and read them but it’s concerning that there is a blur of
boundaries between public and private. Google is trying to combine your own
personal info with institutional info with search etc. If you search it will
give priority to content from your contacts in Google+. Or there are situations
where information only comes from the circles you belong to, this touches to
problems of confidentiality. You have to be very careful even when doing a
simple search, especially if you’re logged in.
Personal information is another tricky aspect. Again if
searching within your own circles all the information may not be public, but
it’s not always obvious. So we need to think how much we will incorporate in
our interfaces for our users. Another example is a facebook local group, there’s
a mixture of information and people are not always aware of the implications
and differences. Some parts of the pages are publicly available, others are
only for members.
There’s also an assumption that we are all well connected on
the internet. Having been online doesn’t mean that we are well connected. A big
chunk of users don’t have a good connection, in some parts there is no internet
at all. This applies to school children, students etc. So we should keep things
simple, not everyone has got megabus broadband with powerful downloading
capabilities.
Discussion:
Rapid fire sessions
“Something good is going to happen: an overview of developing a cloud based solution”
I am the product manager in UK and Ireland for OCLC WorldShare,
my approach for this workshop is quite personal, about my experiences of
working at OCLC, concerns about cloud based data management. So not a sales
pitch but a personal approach. I will have questions for you at the end because
I want to get feedback from users.
I am a librarian and have been working for LMS for a number
of years. I needed to integrate Dr Who in my talk, I’m very excited after Saturday’s
episode. Everyone is interested in the Cloud and the episode was about that, in
a scary way. It’s very sexy and everyone wants to know about this. But at OCLC
we’ve been doing it for 40 years. When OCLC was formed, the centre of what
they’ve done is WorldCat. It’s a universal library catalogue and has always
been hosted.
Why have OCLC gone into the idea of a hosted platform?
Today’s library environment looks like a model of interlinked services, data
etc. We all use similar systems with acquisitions, finances etc. and the
systems have become messy and libraries isolated. The idea of webscale is that
by bringing the libraries together, it’s giving us better services. I have a
positive look about what’s going on in the Cloud. The model we follow is one
that you see being successful with names you know such as e-bay, amazon, facebook
etc. It’s the idea that libraries come together to form a community because
they want to do something together at webscale, bringing together the data and
applications, the community gives you the power and the sharing abilities. It’s
made possible by Cloud computing. It’s helping the sharing and re-use.
When WorldShare Management started from scratch we looked at
these big names. We had to face questions such as: What you are going to store,
are you going to store my stuff on your server? It’s about saving costs.
Interesting what Leo said this morning, that it’s not about money. You have
to be careful talking about costs, you might be talking to IT people, library systems
managers etc. and we’re telling them we’ll take charge of the management of
your server. But we are taking away those mondaine tasks with the belief that
you can change from using 70% of your time to 30 % for the managing services to
developing innovation. It’s about freeing up people from management of hardware
and instead focusing on services to users.
We did research about library transactions (e.g. back office
transactions, OPAC transactions etc.) and look at how our software could
provide the right sort of infrastructure to provide efficiencies to libraries.
Another argument is that the Cloud is greener. It uses less servers and has a
lesser carbon footprint each library having their own servers. It’s maybe not
about internal efficiencies but it’s an interesting aside.
So we are pushing the idea of something new and maybe
better. It’s a struggle when you are talking about the Cloud because a lot of
people ask questions about traditional way of working based on a deployed LMS. We hear: What would happen if? And
most if the time the answer is well this won’t happen anymore. For example we
are asked about offline and backup, what if the server goes down but the answer
is it won’t go down. This is a positive thing of the Cloud, you shouldn’t have
to worry about that. We are not reliant on the one server. We have racks of
servers, we can invest in powerful systems. It won’t be the library server that
goes down, it will be the organisational network.
What about downtime and upgrades? Currently we are obsessed
with minimising downtime. But in the Cloud you probably won’t even notice when
an upgrade is made. We do it regularly, it’s a tricky moment we do about 4
times a year, and we need to be careful but libraries won’t notice.
It’s about efficiencies, so not cost savings but
efficiencies in the management of services especially where each library does
the same thing. It’s the idea of shared data and shared tasks. We are expert in
this field in the library world. There are other options for shared data to
create efficiencies. E.g. we may use the same vendors, check the serials in the
same way. The development process is community driven. The other one is ERM. If
one institution creates a licence data (or uses one that we’ve created), the
next library can simply re-use what’s already there.There is one version of WorldShare and you have an instance
of it. It’s not just about creating data, it’s also that analysis becomes an
interesting idea. Others might have specialities. You can analyse holdings on a
specific area and compare with other libraries. When working in silos, you
don’t have these options.
The big scary elephant is about security. What about my
data? Can I get it back? Is it safe? I don’t want to give you a law lesson. I’m
told (I did a bit of research) that a lot of people worry about the US Patriot
Act (this is my understanding), which is part of a number of security measures set
up after 9/11 and is about the ability to access data (terrorism etc.). In the
US we are horrified about the idea of keeping transactions data because we have
the idea that this Act means that data within the US is available, apparently a
lot of this is scare stories, what we do at OCLC is that we abide by local data
protection acts. We have to take consideration of customers’ concerns. We have
been thinking of these things. The Patriots’ Act gets all the publicity but in
the UK there is the Regulation of Investigatory Powers Act (RIPA). So we take
the security seriously. It’s about scripting the data. There is more information
on our website about this.
My questions to participants are (what I’m interested to know about you):
Why do you want to move to the Cloud?
How will your organisation benefit?
What would you move to the cloud? What’s stopping you?
Discussion:
Q: You were talking about offline backups. We have mobile
libraries going round. We are thinking of a cloud-based system but not sure how
that would work.
A: The default answer is that you can access the systems
from anywhere and that’s our default attitude. But in some cases we have to
acknowledge that an offline system is more appropriate and then
data/transactions can be uploaded to the Cloud. When we are so immersed in this
world, we tend to forget situations where it may not work so well. Most of the
users we work with are more static.
Q: Is there supply and demand for online video?
A: We provide a platform, one part of which is WorldShare Management
services so for video it would have to be streamed, it’s not a discreet service
we offer.
Q: Can we keep local copies?
A: I once heard of someone who couldn’t imagine that there
were libraries that had data they didn’t want to share. Our approach is that we
need to develop the ability to have records hidden from WorldCat. We have the
flexibility to catalogue all sorts of types of material, e.g. law material,
private papers etc. This is the positive advantage of silos! But it’s on our
roadmap. It will come, it isn’t there yet but it’s recognised as a high
priority. WorldShare offers the ability to catalogue in a shared environment. There
are more records in WorldCat that are not traditional books, there are other
types such as e-resources etc.
Q: We are going with Alma in July in the Cloud. I’m an IT
guy in the library. OCLC is out there already. Is there a way that both systems
are going to “speak” to each other?
A: Yes, you’ve got WorldCat and we’ve built around it an
alternative to Alma, and WorldCat still exists in its own right. So the access
of Alma to WorldCat should be possible. With WorldShare you don’t need to
download the records because it’s already there but with other systems we’ll
need to make sure it’s possible.
Q: What we are hoping from the Cloud is not only library
services but all the other stuff, api’s, finance systems, student record
systems etc.
A: That’s one of the ideas of WorldShare, it’s
the community data services and apps, the platform is built using open architecture.
Even what we describe as the LMS has services on the platform and users can
build other services on the platform. E.g. if you use EBSCO as discovery
service, you can build in on the platform because it’s an open system. Apps
that are written for WorldShare, we will quality check them and make sure the
rights are respected but it’s our idea of a community. That’s the real beauty
of the collaboration.Rapid fire sessions
Penny of Bailey Solutions: Tracking software
Outlook is not enough to monitor enquiries (there is an article
available about this). Why would you want to track enquiries? You want to
improve your service, it’s also about not missing deadlines, emails don’t give
you statistics so you have to enter them somewhere else. Our software provides
tools and it’s also about the quality of the information (statistics). Our
software is also especially useful if you respond to FOI requests. It’s about
sharing knowledge so that the enquiries staff can see how it’s been done
before. It allows to store information also for complex enquiries. It provides
a 24/7 service, as it keeps on recording your enquiries even out of hours.
Andy Tattersall, Uni of Sheffield
We’re in the process of moving from Talis to Alma. I want to
talk of what we think will happen in the future. My point of view is from a systems
manager so didn’t look at the legal point of view. Mine is about corporate
services. We’re on a good footing already because our IT people are wanting to
move, providing safeguards are in place. We need good relationships with the
technical people. Don’t forget to do that. Cloud stuff is good for user
interface, but you will have local systems needing to talk to your Cloud system,
e.g. firewalls so you need secure protocols, this is really important. Data
security is a biggy. We’ve all got worries about this, are we compliant legally
(e.g. EU laws)? What IT people asked me: what about backup data? What if it’s
stored in the US? The less sensitive data the better. It’s a good thing to
sprinkle you data down anyway and have disaster plans. I’m not talking of power
cuts etc., we all expect companies to respond to that, but what happens if
someone pulls the plug, if no money comes in, what plans are in place? What
about technical details: we have specialist people to manipulate the data, so
how much technical skills will I need in the future?
Jeff Newman, Kaltura
Kaltura is an online video platform for solving the problem
of who hosts when, for example what if youtube is not the solution? This is for
universities. We have a system to stream videos on all devices, it’s a SaaS
(Software as a Service). It works for all types of content. We created an app
called mediaspace, it’s a portal system allowing to create galleries of video
content using a logging so it’s not for the general public. There are different
levels of permissions/controls to secure videos on the internet. It’s an OS
platform. Its’ a matter of choosing the appropriate application. We are an open
and flexible platform so if you have your own system it could be integrated. We
have already a lot of customers, including in Denmark.
Dave Parkes, Uni of Staffordshire
I’m not selling anything but want to say how we create the
Horizon report, it comes out every 5 years and is a window on the education
sector. We usually get it about right, we said that tablets would be happening
etc. there’s 40 of us worldwide working in the Cloud, never see each other but we
argue and talk about stuff, we start to narrow things down, what are people
using in education, what are the challenges etc., we have a watchlist that we
look at. Anything that’s relevant to universities. It takes about 3 months for
things to take shape. It’s challenging when 40 people have a case to argue but
eventually we produce a little booklet. It’s then given away (my son was very
impressed with that). Our predictions for the next 2 to 3 years: increasingly
using games in technology, using analytics, and for the next 4 to 5 years,
freely printing and wearable technology.
Nicola Philpott, Ancoris
We are a partner company working with Google. Organisations
can buy directly from Google but Google doesn’t offer services around the
project itself, they provide tools and they have partners in the UK, we are one
of them. We work with organisations that for instance have the project to get
rid of emails, or companies looking at diff ways of storing docs so using
goggle drive. We look at the project from start to finish, help migrate the
data etc. We’ve just finished a project with Bristol University. We help with
the change of management, users changing platform and getting them learning to
doing that, so we provide training, face to face or webinars to help them use
your new service. We have an online e-learning platform, users can ask
questions, find answers etc. So helping you move your project on, including
your customisations. The Google Apps suite includes gmail, calendar, G+, vault,
docs& drive, sites (intranet?), talk (online conferences etc.). It allows collaborative team
working.
Robert Bley, Ex Libris
Random thoughts and musings about the Cloud. What is not the
Cloud? It’s not about just hosting legacy apps, it’s born Cloud, it’s not Cloud
server, it’s proper shared services, it’s supporting the application beyond
hosting. This means big change for libraries, for vendors, we need to re-engineer
our business around Cloud applications. E.g. it’s multi-language. How to choose
a Cloud provider? Do they have global facilities, the manpower to support you
and all other customers, full redundancy infrastructure, data centre certified
to relevant standards, etc.? Do they have EU hosting? If it has to be in the
States, are they part of the same agreements? Who else uses a service and are
they happy with it? Is there physical security? How open is the system, e.g. doing
mashups, taking the data elsewhere? Does the vendor support just the
application or the infrastructure as well, and the Code? We’ve realised very
soon that it’s very entwined. More evidence is needed, does it cost more? There
is a agreement that it’s Green, so energy savings. Implications for libraries
in terms of staff, not so much numbers but skills, and if tech skills, is it
the traditional ones? Or can we start doing mashups etc. Budgeting is also an
important question. We’ve done calculations about cost savings and we are happy
to refine those. It encourages people to be more creative.
Bethan Ruddock, MIMAS
LAMP: your data uses lots of fabulous info so how to put it
in a Cloud service to allow you to benchmark with other libraries? This project
is a partnership between JISC, Huddersfield and Mimas. We had a meeting the
other day, and we talked about: do you think you might want this so we’ve created
a dashboard with our discussions. The idea is that you’d click on a single
thing and then it takes you down to further details. It’s early days. Potential
use cases of things we think people might want to do with this data, we’d like
to hear from the community.
Liz Robins, OCLC
Why web scale makes sense for libraries? OCLC has a good
background in delivering library services. WorldCat is essentially web-based
and we have been doing this for 40 years. Sharable data is not just for bib
data, but vendors, licences, knowledge base etc. It’s innovative, WorldShare is
the name of the platform, started from scratch, built on open platform to allow
for other technologies as well. The situation now is there are lots of silos
not speaking to each other, so what if we brought all these services together.
We are all over the world. 100 sites , the first in the UK was on 4th
June.
Bethan Ruddock from MIMAS
Opening up - bibliographic data-sharing and interoperability (view slides of the presentation)
Opening up - bibliographic data-sharing and interoperability (view slides of the presentation)
Why would you want to share data? And how would you do it?
Your data could be enhanced from being shared. We have developed a lot of
services that are about shared data (e.g COPAC, Archives Hub etc.). COPAC
combines your record with other peoples’ records. With Archives it’s different
because they are unique but hey are stored in the same database. With LAMP we’ve only just started so we don’t really know yet but we would also like to
combine your data.
So you need to think of the format. If it’s just a library
service, it’s probably only up to you, it’s whatever works for you. But with
the general caveats of trying to do something that’s stable, so if you can
store things in fairly compact and sustainable ways, it will help because
combination services will want data in certain formats. They may not always do
the transformations for you.
Have a look at what formats you can get data in and how you
can get it out? If the data is in a locked down system it will be hard to get
it out. Which one is most appropriate for what you want to do? Which version
should you choose? Xml is one of the best formats, it’s text based so it’s
sustainable, it’s compact and transformable. If you don’t have the expertise
in-house, there are services that can help you. We’ve developed a programme to
transform data in EAD for Archives.
Consistently bad data is better than inconsistently good
data, that’s “my rule”. It’s easy to fix something in the wrong place
everywhere, whereas it’s hard to correct something that’s inconsistently wrong
in 5% of your records.
Barriers and risks are about licensing and data ownership.
For example there are legal barriers to sharing. Think also of the risk of not
sharing your data not only the risks of sharing it. What will happen to your
data if you don’t share it?
Things to think when choosing a Cloud solution, think of why
you want to share your data? Does the service meet your needs? Is it something
your users have asked for e.g. they want your data as part of a collaborative
service? Is it for resource management? Once you’ve decided where and why you
want to send you data, think of the right format for the purpose. And will you
be able to get it out of the Cloud in an appropriate format if you wanted to
send it elsewhere? Will it be kept in a standardised format?
A note on Linked Data:
it’s the idea that you describe something with the use of triples and they are
reversible. So they are built to describe a relationship. This allows you to
build a picture and this can then be extended: example: Ruth works at MIMAS,
and reverse this info and then link it; so for instance who else works
at MIMAS, what else does this person do etc. everything has a unique identifier
so things can’t conflict even if the name is the same, as an example. The idea
is that it can go in the internet because of uri’s and url’s. We use
standardised vocabulary. There are definitions of RDF types and you can look up
what it means.
Panel questions and answers
Panel questions and answers
Should we teach Cloud literacy (especially for students)
We (Staffordshire) provide literacy about the Cloud and it’s
all about staying safe online, so it’s about raising awareness, including of
the “digital health”. Understanding the consequences of an online self, so as
not to be limited in the future. It’s about letting go of a level of control.
It’s also having the resources there and educating about the risks. It’s
extraordinary to have tools allowing to collaborate. There is often an idea
that teachers don’t understand technology. But we have to explain the different
layers of media. Recommended reading Marc must die
140 characters to convince us not to go to the Cloud
They’re all watching you. Where’s my data. They are going to
sell it back to you. Data security issues. Buy in isolation, be off grid. Some
people in Microsoft world send their kids to schools where they are no
computers. Don’t be part of the technological singularity, where software
becomes self-aware. Don’t put all of the stock. Be careful to give away too
much and not being able to get it back. How much will you outsource? If you
exit the Cloud and you have no more infrastructure to fall back onto.
Other presentation slides:
Other presentation slides:
- Staffordshire University: Koha the cloud and the crowd
- Collaborating to create eBooks in the Cloud
- Welcome to the University of Google
