although I don't have a detailed DSpace vs Sharepoint comparison, I wanted to share some general findings, and a recent feature analysis we did for one of our prospects.

@mire is a company specialized in consultancy for DSpace, covering all activities in setting up a repository, customization, support & maintenance.

The most general observation, when comparing DSpace and Sharepoint would be that DSpace is specifically designed as a digital document repository, with great care for the ingest process (different workflows with quality control) and accessibility through search, browse (including indexing of uploaded full text so that it becomes searchable in your repository, but also in search engines like Google).

I suppose it would be possible to customize Sharepoint to offer repository-alike functionality, but the fact that DSpace supports a range of these features out of the box, makes it very easy to start with and elaborate on.

Below you find our recent feature analysis, based on the Mental Map.Repository Platform comparison - mental map.pdf Hopefully it contains information that is useful in your context, and enables you to explore the possibilities of DSpace faster.

Please contact me with any questions or feedback & welcome to the DSpace community !

Technical Aspects

++ Technical Requirements

- Operating system: this can be both a windows or linux based platform, but choosing a linux platform has the advantage that most instances are run on Linux, and you can get more support from the community, documentation.

- Programming Language: JAVA for the DSpace core code, XML/XSLT (based on Cocoon) for the user interface

- Database: by default, this is the open source database system Postgres, but some instances run on Oracle. We advise to run Postgres, but even some of our clients run Oracle.

++ Staff Knowledge Requirement

- Of course, it depends on which tasks you want the staff to perform. The DSpace web user interface is intuitive, in the sense that people experienced with web interfaces can easily get into the role of "repository administrator" (creating & administering collections, access rights, web-content in the system such as collection homepages, ...). Same goes for staff, responsible for submitting new content.

If you want to train internal IT staff to perform installation or customization of DSpace, the XML user interface has a steep learning curve for people who haven't developed in the Cocoon framework before. ( http://cocoon.apache.org/ )

+++ Adoption factors

- Comparing to other repository platforms, DSpace has by far the highest installed base and most lively community. The non profit DSpace foundation, who's being merged into the Duraspace organization on the 1st of July, organizes a global outreach committee and has a technical director who manages the roadmap.

- The number of support providers is larger than for any other open source repository platform and continues to grow: http://www.dspace.org/index.php?/Service-Providers/Service-Providers.html

++ Documentation

- There's a very extensive set of powerpoints called "the DSpace course": http://www.dspace.org/index.php?/New-User/New-User-Training.html

End User Interface

+++ Surrogate/Splash

- Browse features are very elaborated and configurable, as is the case with comprehensive metadata.

- The basic DSpace distribution contains a limited package for statistics, but @mire has a very comprehensive add-on module: http://atmire.com/USB/resources/reporting_suite.html . Furthermore, Statistics is one of the main themes that will be heaviliy improved, even before DSpace 2.

- Exporting metadata to different file formats from the web interface is not built in, but will be very related to export citation (mentioned below), and will require a limited amount of customization.

+ Interaction/ Web 2.0 Tools

- automatic emails to authors & RSS feeds are included

- tagging, comments, export citations, ... are not present in the basic DSpace, but are features that will require a limited amount of customization.

+++ Additional features

- File conversion is not built in in DSpace, but we have a module that does the trick: http://atmire.com/infocon.php

Requirements

Metadata export

- Exporting metadata to different file formats from the web interface is not built in, but will be very related to export citation (mentioned below), and will require a limited amount of customization.

+++ Search

- Full text search, Browse/clustered and indexed by search engines: all natively present and supported in DSpace

Version Control

- One of the more difficult customizations in the current versions of DSpace. This is typically a usecase//requirement that can be implemented much better in DSpace 2.

+ Complex Object Model

- Which use case is actually meant here ? Currently, DSpace has items, that can contain multiple bitstreams (attached files). Metadata per bitstream is limited, as the metadata is managed on the level of the item. But we have done pretty complex metadata structures, within these limitations

+++ Usability

- easy to deposit - including customizable workflow of who needs to review//edit

- Easy to retrieve/find by human users

- Easy to find/retrieve by search engines and harvesters: especially true because DSpace is OAI-PMH compliant out of the box, so registering with initiatives like OAIster is only a matter of configuration.

Taxonomy

- Depending on how you want to manage these controlled vocabularies, this might require some customization instead of configuration. The DSpace inputforms allow you to construct your own selection lists//vocabularies, and make them work as a list where users need to pick from. If you want to connect to a thesaurus or taxonomy, offered by a webservice (such as the FAO Agrovoc ( http://www.fao.org/aims/ag_webservices.jsp ), you will need some customization.

Standards

Bibliographic citation formats

- Not implemented in DSpace out of the box, but we have experience with all of these at the moment except for RefWorks

+++ Metadata

- DSpace is configured to support Dublin Core out of the box, but has a configurable interface to support other metadata schemes.

+++ Interoperatbility

- Z3.50 is not natively supported by DSpace, but we have experience in this type of customization

- ReDIF: open source scripts are available to make this work: http://ideas.repec.org/c/rpc/script/dspace2redif.html

- OAI-PMH - Out of the box supported

++ Syndication

- Atom is not (yet) out of the box supported

- RSS is present

++ Resource Identification and Localization

- DOI can be present as a metadata field

- integration with HDL supported out of the box

- I know we recently did an OpenURL customization, but I'm not aware of the details. Let me know if you want some info on that.

Adaptability

++ Document & Metadata Migration

- the file format registry allows administrators to stay in control which filetypes are being uploaded so these file types can be managed, and migrated if necessary. If you need further information on this, I should get more detail from our technical folks.

- our metadata quality module, allowing batch edit for metadata, is relevant here: http://atmire.com/quality.php

+ Platform Independence

- In theory, DSpace can run on any Java application server and Postgres database. However, I don't know many people who run it on for example Solaris or other platform.

Programming language independence

- DSpace is pretty much tied to Java.

++ Modular Architecture

- thanks to the Maven build system, add-on modules and isolating functionality has become a lot easier over the last versions and is still being improved further. (as you can imagine, because we sell add-on modules, we have a big stake in modularity as well)

+++ Metadata

- see earlier defined metadata advantages

System Management

+++ Security

- not sure which of these is currently out of the box supported, but I know we have experience with Active Directory//LDAP, Domain specific restrictions and Access limitations on file access.

++ Submission

- Approval process // workflow is elaborated in DSpace

- Once approved, there is no real infrastructure for versioning present, as indicated before, this is probably the most challenging customization you'll be looking at

++ Document Upload Page

- Dropdown metadata fields are possible

- the inputforms are elaborated and easily customized

- SWORD compliance is included in DSpace 1.5.2

++ System generated usage statistics and reports

- as mentioned earlier, not so elaborated in DSpace out of the box, but very elaborated in our statistics module.

with kindest regards,

Bram Luyten

@mire - http://www.atmire.com

 Hopefully it contains information that is useful in your context, and enables you to explore the possibilities of DSpace faster.

Please contact me with any questions or feedback & welcome to the DSpace community !

Technical Aspects

++ Technical Requirements

- Operating system: this can be both a windows or linux based platform, but choosing a linux platform has the advantage that most instances are run on Linux, and you can get more support from the community, documentation.

- Programming Language: JAVA for the DSpace core code, XML/XSLT (based on Cocoon) for the user interface

- Database: by default, this is the open source database system Postgres, but some instances run on Oracle. We advise to run Postgres, but even some of our clients run Oracle.

++ Staff Knowledge Requirement

- Of course, it depends on which tasks you want the staff to perform. The DSpace web user interface is intuitive, in the sense that people experienced with web interfaces can easily get into the role of "repository administrator" (creating & administering collections, access rights, web-content in the system such as collection homepages, ...). Same goes for staff, responsible for submitting new content.

If you want to train internal IT staff to perform installation or customization of DSpace, the XML user interface has a steep learning curve for people who haven't developed in the Cocoon framework before. ( http://cocoon.apache.org/ )

+++ Adoption factors

- Comparing to other repository platforms, DSpace has by far the highest installed base and most lively community. The non profit DSpace foundation, who's being merged into the Duraspace organization on the 1st of July, organizes a global outreach committee and has a technical director who manages the roadmap.

- The number of support providers is larger than for any other open source repository platform and continues to grow: http://www.dspace.org/index.php?/Service-Providers/Service-Providers.html

++ Documentation

- There's a very extensive set of powerpoints called "the DSpace course": http://www.dspace.org/index.php?/New-User/New-User-Training.html

End User Interface

+++ Surrogate/Splash

- Browse features are very elaborated and configurable, as is the case with comprehensive metadata.

- The basic DSpace distribution contains a limited package for statistics, but @mire has a very comprehensive add-on module: http://atmire.com/USB/resources/reporting_suite.html . Furthermore, Statistics is one of the main themes that will be heaviliy improved, even before DSpace 2.

- Exporting metadata to different file formats from the web interface is not built in, but will be very related to export citation (mentioned below), and will require a limited amount of customization.

+ Interaction/ Web 2.0 Tools

- automatic emails to authors & RSS feeds are included

- tagging, comments, export citations, ... are not present in the basic DSpace, but are features that will require a limited amount of customization.

+++ Additional features

- File conversion is not built in in DSpace, but we have a module that does the trick: http://atmire.com/infocon.php

Requirements

Metadata export

- Exporting metadata to different file formats from the web interface is not built in, but will be very related to export citation (mentioned below), and will require a limited amount of customization.

+++ Search

- Full text search, Browse/clustered and indexed by search engines: all natively present and supported in DSpace

Version Control

- One of the more difficult customizations in the current versions of DSpace. This is typically a usecase//requirement that can be implemented much better in DSpace 2.

+ Complex Object Model

- Which use case is actually meant here ? Currently, DSpace has items, that can contain multiple bitstreams (attached files). Metadata per bitstream is limited, as the metadata is managed on the level of the item. But we have done pretty complex metadata structures, within these limitations

+++ Usability

- easy to deposit - including customizable workflow of who needs to review//edit

- Easy to retrieve/find by human users

- Easy to find/retrieve by search engines and harvesters: especially true because DSpace is OAI-PMH compliant out of the box, so registering with initiatives like OAIster is only a matter of configuration.

Taxonomy

- Depending on how you want to manage these controlled vocabularies, this might require some customization instead of configuration. The DSpace inputforms allow you to construct your own selection lists//vocabularies, and make them work as a list where users need to pick from. If you want to connect to a thesaurus or taxonomy, offered by a webservice (such as the FAO Agrovoc ( http://www.fao.org/aims/ag_webservices.jsp ), you will need some customization.

Standards

Bibliographic citation formats

- Not implemented in DSpace out of the box, but we have experience with all of these at the moment except for RefWorks

+++ Metadata

- DSpace is configured to support Dublin Core out of the box, but has a configurable interface to support other metadata schemes.

+++ Interoperatbility

- Z3.50 is not natively supported by DSpace, but we have experience in this type of customization

- ReDIF: open source scripts are available to make this work: http://ideas.repec.org/c/rpc/script/dspace2redif.html

- OAI-PMH - Out of the box supported

++ Syndication

- Atom is not (yet) out of the box supported

- RSS is present

++ Resource Identification and Localization

- DOI can be present as a metadata field

- integration with HDL supported out of the box

- I know we recently did an OpenURL customization, but I'm not aware of the details. Let me know if you want some info on that.

Adaptability

++ Document & Metadata Migration

- the file format registry allows administrators to stay in control which filetypes are being uploaded so these file types can be managed, and migrated if necessary. If you need further information on this, I should get more detail from our technical folks.

- our metadata quality module, allowing batch edit for metadata, is relevant here: http://atmire.com/quality.php

+ Platform Independence

- In theory, DSpace can run on any Java application server and Postgres database. However, I don't know many people who run it on for example Solaris or other platform.

Programming language independence

- DSpace is pretty much tied to Java.

++ Modular Architecture

- thanks to the Maven build system, add-on modules and isolating functionality has become a lot easier over the last versions and is still being improved further. (as you can imagine, because we sell add-on modules, we have a big stake in modularity as well)

+++ Metadata

- see earlier defined metadata advantages

System Management

+++ Security

- not sure which of these is currently out of the box supported, but I know we have experience with Active Directory//LDAP, Domain specific restrictions and Access limitations on file access.

++ Submission

- Approval process // workflow is elaborated in DSpace

- Once approved, there is no real infrastructure for versioning present, as indicated before, this is probably the most challenging customization you'll be looking at

++ Document Upload Page

- Dropdown metadata fields are possible

- the inputforms are elaborated and easily customized

- SWORD compliance is included in DSpace 1.5.2

++ System generated usage statistics and reports

- as mentioned earlier, not so elaborated in DSpace out of the box, but very elaborated in our statistics module.

with kindest regards,

Bram Luyten

@mire - http://www.atmire.com