Service frameworks

"would like some easy to use tools, drawn to idea of modules or web services that can be joined together" (Ex 1, 1b-A)

"in terms of deliverable, beyond data collection: service frameworks; not so much specific architectures or products or sites, but rather agreeing upon service frameworks that will allow for data aggregation" (Ex 1, 1b-B)

"Framework in which many tools, archives, etc. can "play"" (Ex 1, 1b-C)

"basic humanities functions are: Find; Annotate (overlay tools/ personalize data tools); Re-use; Want digital tools for each stage and tools that allow both individual and community use/ sharing. Want Bamboo to create cross-systems tools that are function based, but can also be personalized. Lack common tools and architecture that allow scholars to share. Need inter-operability" (Ex 1, 1b-D)

"Fear that what we develop will be outdated by the time we're done; maybe by then everyone will be in Second Life; How to create tools that take advantage of changing environment that we can plug into?" (Ex 1, 1c-E)

"Bypass institutions, go right to various projects; In that model, Bamboo would consume those services; Would become points of integration and data sources; What do you think? Would it be disappointing if there's not institutional benefits? Is it okay to do it through national-scale infrastructure projects? Is there independent benefit to partnering w/ European campuses?" (Ex 1, 1c-E)

"Need a commitment from a resource provider to the interface (i.e. the API). That's the case with ARTStor. Or take the Museum of Fine Arts - there is no service level agreement. No guarantee that the service or the content will be there tomorrow." (Ex 1, 1d-A)

"common services to be a little more agile" (Ex 1, 1d-D)

"federated services. How do we make use of things." (Ex 1, 1d-H)

"Modularization of tools for exchange with others. Package, to make simpler for another user. Need wrappers into software & need to be extracted so don't have to reauthor again. Software extraction, to prevent unnecessary reauthoring" (Ex 2, 1a-E)

"Bamboo needs to build collaborative services & technologies that provides an environment that exemplifies the intellectual value of digital scholarship." (Ex 6b, 1b-D)

"Canonical Bad Humanities Proposal. A Humanities scholar should not worry about infrastructure only on scholarship." (Ex 6b, 1b-C)

"One interface that allows for the use of services all over the world. New services are added, new texts are added, so no need to learn something new. A service repository that is intelligent. Aggregation service. Actively filtered. Portal that mentions the tool you're working with, but with interoperability for different available services" (Ex 6b, 1c-A)

"Lowering cost and complexity of producing projects" (Ex 6b, 1d-F)

"APIs defined for various scholarly workflow components" (Ex 6b, 1d-F)

"Marketplace of services: digitization, project design, technical services. Apps and humans. A la Yahoo Pipes" (Ex 6b, 1d-F)

"It's an open question as to the right level of discipline specificity will emerge. Don't want projects that are very idiosyncratic & narrow. Also the question of taking idiosyncratic project & creating something more general. Is it possible a consortium can develop tools for multi-media authoring. It doesn't make any sense. Have to have a project with real needs and develop for that. But if you have projects with conform to certain standards, it opens up ways of collaborating. Don't want a top-down thing. Stanford has problems with Sukai because tools were developed at a high level. Let the tools develop at a small level and via adoption let the useful ones succeed." (Ex 7, 1a-E)

"Q about need for scalability consulting. varies with project: some have IT staff, others hack, and some have little idea. a question of maturity in some cases, when it becomes clear that a resource takes too much time to sustain, but is important enough to merit attention. Thinks Bamboo needs to provide scalable services from the beginning. yes, like Amazon DB that scales up in an easy manner. Would like Cloud Data store. - Cyber infrastructure and cloud storage, etc. would you have wanted a DB designer early on? yes, although did not know how to find someone when he started (not a Uni)." (Ex 7, 1a-H)

"He is often concerned with lots of data, and the problems of gathering and managing this data. He is worried about tools that are only demo tools that work on small data sets, but do not scale, and do not help gather really large data sets." (Ex 7, 1a-H)

"have translators to get data sets into their world. Generally have been helping collaboration on a project by project basis. Community contributes both primary data, and some info on schema mapping (they cannot do this themselves yet). Simple model in that just provide simple services with schema harmonization. Talked about Petra project and the emerging social models around that, as well as IP model (CC-BY use of data)." (Ex 7, 1a-H)

"Bamboo might have a place if could provide services that make it easier to digitize/transcribe materials." (Ex 7, 1a-H)

"services around data sets: munging tools, versioning, citation, etc. there are many old data sets around, dead or half-dead. Need tools to resuscitate these. ..and a lawyer to figure out ownership and permissions. often just get a screen scrape. Need better tools. moving from punch cards to something modern. Bamboo as a virtualization space for data sets and assoc. services." (Ex 7, 1a-H)

"we're forced to work together because the digital is ubiquitous but must be highly structured. ubiquity does not mean independence. must be thought out in advance. "fascist" program of commonality, uniformity. potential objection: in doing this we are blinding ourselves, limiting ourselves... if we could all be google apps in the big google world... i don't think this precludes idiosyncracy" (Ex 7, 1d-E)

"one thing about the service-oriented approach, it makes your tools more sustainable; can switch out the technology easily and substitute it with another. need to make sure you can get your data out of the tools and so if you want to change tools you need to be able to get your data out. tools etc also need to be integrated in what people already do---this would answer the question of maintenance; find a way to fit with the research that people already do." (W2, Scholarly Networks, questions and concerns, group 12 notes)

"An open platform on which a variety of storage, preservation, access, and use services could be built upon and seamlessly interact. As the broader web moves towards cloud-level computing, one approach that would allow researchers and technologists supporting researchers to scale their efforts and provide next-generation services it to be able to build upon a common network-level platform that would provide open and consistent access to data, resources, and core services. If this platform existed, it would not only reduce the amount of recreating the wheel efforts, but it would also enable new and interesting research that could take advantage of data and resources in a much greater scale." (SN-0008 Network Platform For Open Data, Resources and Core Services, Jeremy Frumkin, 1/8/09)

"Need collaboration - a scale to collaborate w/ resources like JSTOR - faster than building from scratch. Put innovation into the platform. Need to take some risks. Need to change platform - open up access and utility to platform that does risk the way we think about things. Challenges how JSTOR has operated. Risky, but will help us move along. Learning from each other. Can all benefit from more communication along those lines. Collaboration has to be done carefully." (W3, Perspectives: Content, Timothy Babbitt, Chief Information Officer, JSTOR)

"A number of library dig repository systems in wide use among libraries. Would be helpful to work as community to build connectors so services become seamless, so we're not trying to re-engineer all the time. Could provide opportunities for new services for libraries to participate in creating/distributing content. Can we use Bamboo for round-trip to enhance collections w/ research products. How do we automatically get links for article written about set of images/audio files. Are there new additions being created? can we get the links back to link it up? we don't have to own it. If someone downloads images of 18th century letters, can transcriptions come back?" (W3, Perspectives: Content, Stacy Kowalczyk, Digital Library Program, Indiana University)

"When one talks about tech infrastructure, it's not about tech infrastructure. It's always more complex than just "computing". Then there's networks for things that aren't geographically together. You can have a wonderful processing thing that depends on data elsewhere, but if network fails, there's nothing you can do. Different phases in technology - true for everything. There's a period where you're installing (design & implementation), then "operate" (different kind of function), and "support" (also different). Leaving any of these off leaves an incomplete picture. Unless you worry about the other pieces, we end up with nothing. Complicated even further that each can occur locally, replicated, and shared. Replication not the same as sharing; "how do I get that so I can install it?" - that's not sharing. Replicating is more easy than sharing, local implementation are easier than replicating. If we're doing what we're talking about with shared services, must not make that confusion. There's no right answer, but the mistake is forgetting about dimensions in the cube." (W3, Prespectives: Information Technology, Greg Jackson)

"Q: I like how complicated you've made it- because it is. Is it even messier? We think about install/operate/support, but one of new models is perpetual beta - always changing. Would this be a model we should consider for Project Bamboo? Greg Jackson: People do perpetual beta for different reasons. Things are appropriately going to change, and don't want to freeze. Operating a system that's dynamically changing, part of the design. Not perpetual beta, but understanding things will change. 'Beta' is for limiting liability - this is ducking responsibility. Q: Constantly changing dynamic, and that's fine - but this kind of model is something that we assume a stable operating environment in which we have tasks and when tasks + technology + understanding of what we need are changing, then perhaps it's the first condition, a perpetual operating state that's in flux. Greg Jackson: There's no such thing- IT organization "architects" are there because there's no such thing as a stable operating environment; everything has to have clear principles and architecture behind it.We have to understand how evolving environment is going to affect applications." (W3, Prespectives: Information Technology, Q&A)

Bamboo tags: 

Add new comment