|Features||SEAD 1.5 (on the Medici Platform)||SEAD 2.0 Early Access (on the Clowder Platform)||SEAD 2.1 (on the Clowder Platform): We plan to migrate SEAD 1.5 users following this release.|
|Account Creation||Accounts need approval from SEAD staff. Users can login with Google or ORCID.||Account creation is not subject to SEAD approval. Users can login with Facebook, Twitter, and Google.|
|Project Space Creation||Teams need to make a request to SEAD to build each Project Space.||Users have the ability to create Trial Project Spaces on their own. Verified Project Spaces offer the highest level of functionality.|
|Project Space Customization||Users can add logos and banner images to customize their space. Admins have the ability to customize the set of metadata fields that appear for their space and to include definitions for each of the fields.||Users can add logos and banner images to customize their space. Admins can specify which metadata fields appear in their space and to include definitions for each of the fields.|
|User and Role Management||Project Space Admins can add users and assign roles. They can also create new roles with customized permissions for their space.||Simplified role selection and permissions. Admins can no longer create customized roles for their space.||More user-friendly process for adding existing SEAD users and inviting new. We will review (and possibly change) the default roles, contingent on user feedback.|
|Uploading Data||Files can only be uploaded as part of a project space and must be added to each project space separately. Users can upload a new version of a file that is linked to a previous one.||Individuals can upload files that only they see and can also add the files to as many Project Spaces as they have permissions to add data to. Datasets must be created first in order to upload files.||Improved bulk uploading capabilities.|
|Data Organization||Files can exist on their own or be part of collections.||Files must be organized into datasets and can be further organized with hierarchical folders. Users can create collections that link a variety of datasets together in multiple ways.||Users can bulk organize files. We will also create a diagram to help 1.5 users transition to SEAD 2.0.|
|Public Listing of Project Spaces and Data||All Project Space names are listed in the SEAD directory, and external users can request access from admins for a particular project space. Project teams can make data public if they choose to (outside of the publication process).||Verified Project Spaces will appear in the SEAD directory. Each verified space has a public landing page and teams can make datasets within their Project Space public (outside of the publication process).|
|Metadata Extraction and Editing||Metadata are automatically extracted from files and displayed in the interface. Users can apply custom metadata and geotag files individually or in bulk.||Metadata are extracted from files and displayed in the interface. Users can geotag files and datasets and can, in either the Early Access or 2.1 release, apply metadata to many files in bulk.||Improve how extracted metadata is displayed. Add ability for users to enter creator names in an order that will be preserved.|
|Deleting Data||Users can delete files and collections individually or in bulk.||Users can delete files, datasets, folders, collections, and Project Spaces. The bulk capability will be added in 2.0 Early Access or 2.1.|
|Social Tagging||Users can create and assign free form tags to files and collections individually or in bulk.||Users can create and assign free form tags to files and datasets individually. We will add ability to bulk apply tags in 2.0 Early Access or 2.1.||Improved display of tags.|
|Moving Data Around||Users can move files from one collection to another.||Within a dataset, users can move files from one folder to another. Users can also move files from one dataset to another.|
|Adding Relationships between Items||Users can specify relationships between files and collections (e.g. that one file is derived from another, one collection describes another, etc.)||Users can not specify relationships between items.||Users can specify relationships between items.|
|Sharing Data||Users can make subsets of their data public and can specify what kind of access these public users have.||Users can make subsets of their data public and can specify what kind of access these public users have. Additionally, datasets and collections can be copied to multiple individual Project Spaces.|
|Search and Browse||Users can search and browse within an individual project space. Files and collections can be sorted by name, upload date, or category. Browsing is possible via a hierarchical listing of files and folders, tags, a map overlay, and via a collection and file listing.||Users can search and browse all of the SEAD content accessible to them as well their own Project Spaces. Files, datasets, and collections can be sorted by date and title. Browsing is possible via datasets, collections, and tags listings.||Hierarchical display of datasets, files and folders within a project space.|
|Viewing Data||Users can view many common file types within the browser, without having to download the files. For image-based files, they can zoom and choose a full-screen view. For geospatial data, users can view the data overlayed on a map.||Users can view many common file types within the browser, without having to download the files. For image-based files, they can zoom and choose a full-screen view. For geospatial data, users can view the data overlayed on a map.|
|Commenting||Users can comment on files and collections and can use "@" to address particular people.||Users can comment on datasets and files.||We plan to implement the "@" ability to address particular people and create comment alerts.|
|Downloading Data||Users can download data as individual files or in bulk.||Users can download data as individual files or in bulk.|
|Publishing Data||Users can submit collections for publication to a SEAD partner repository from their Project Space via a submit button. A published data list appears publically within the Project Space.||Users can submit datasets for publication to a SEAD repository from their Project Space via the Staging Area within their Project Space. The Staging Area allows users to review their submission, make changes to the metadata, and get a matchmaker review that helps them pick the right repository and fulfill its requirements.||We will implement a public listing of all datasets published via SEAD.|
The code that does the extraction is a monolithic piece of code that is extremely fragile and hard to extend. Adding a new extractor requires a long time for a developer to get familiar with the platform before the extractor can be added.
Since there is exactly 1 extractor per medici instance the system can not scale and will not be able to handle many uploads quickly.
Each medici server has its own extractor running and can not share resources with other extractors, this requires each medici server to have resource set aside for the extractor that are not used when the extractor is idle.
Adding code to extract information from a new file format is trivial to do, a new application can be written to handle the extraction and be added to the bus.
The number of extractors will scale up and down depending on the number of files uploaded. This allows the system to handle many uploaded files quickly and show the extracted information
The extractors are shared with other instances requiring a smaller memory and cpu footprint for the clowder instance.