Cirrus-ly Convenient Uploading

That was a cloud pun! Following our release two weeks ago, the Dash team is thrilled to present our newest functionality: you may now upload files directly from Box, Dropbox, and Google Drive!

Let’s get you publishing (and citing and getting credit for your data):

  • Using the “upload from server” option, you may enter up to 1000 URLs (and up to 100gb per submission) by pasting in the sharing link from Box, Dropbox, or Google Drive.

Screen Shot 2017-06-20 at 1.40.37 PM[2]

  •  Validate the files and your URLs will appear including the filename and size.

Screen Shot 2017-06-20 at 1.41.25 PM[2].png

  • Submit & download.
    • Box, Dropbox, and Google uploaded files will download the same as they were uploaded to the cloud
    • Google docs, sheets, or presentations will download as Microsoft Office word documents, excel spreadsheets, or powerpoint presentations.

We will be updating our help and FAQ pages this week to reflect our new features, but in the meantime please let us know if you have any questions or feedback. 

Manifesting Large and Bulk File Data Publications– Now A Reality!

The Dash team is excited to announce our June feature release: Large and Bulk File upload. Taking into consideration the need for large size and file numbers of datasets, as well as the practicality of server timeouts, we have developed a new feature that allows for up to 1,000 files or 100gb* of data to be published per DOI.

To accomplish this we are using a “manifest” workflow- which means that instead of uploading data directly from your computer, you may enter URLS for where your data are located (on a server or public site) for upload. Once uploaded, Dash will display the data in the same manner as direct upload. To reflect this new option for upload we have updated the Upload page to choose between uploading locally (from your computer) or via a server. Information about file size limits (2gb/file, 10gb total local or 1000 files any size up to 100gb*) are listed on this landing page.

Step 1: Enter URLs where data are located

Screen Shot 2017-06-07 at 1.01.59 PM

Step 2: Validated files will appear in Uploaded Files table with any other data files associated from current or former versions

Screen Shot 2017-06-07 at 1.02.19 PM

The benefit of using this workflow is that as a user you do not have to watch your screen for many hours as the data upload and instead your data will be uploaded in the back-end, without the involvement of your computer. This upload mechanism is also not limited to large file use- it can be an easy way to transfer your data directly from a server regardless of size.

A complication with this process is that you cannot upload local data and server-hosted data in the same version. Though this seems tricky- we would like to remind you that Dash supports versioning and after successful publication of the server uploaded data you could go back in and add local files (or vice versa).

While at the moment we do not allow for upload from Gdrive, Box, or Dropbox, we are investigating the sharing links necessary for integrating uploads from the cloud. If you have any feedback to make this feature, or any features more accessible or valuable for researchers please do get in touch. Happy Data Publishing!

Note: To utilize this feature and publish your datasets, your data will need to be hosted on a server. Many institutions, departments, and labs have servers used to host data and information (good examples across the UC campuses, MIT, University of Iowa, etc…). If you have any questions about servers on your campus or external resources, please utilize your campus librarians

*Size limits vary per institutional tenant- please check in with your UC Data Librarians if you have any questions

Make Data Count: Building a System to Support Recognition of Data as a First Class Research Output

The Alfred P. Sloan Foundation has made a 2-year, $747K award to the California Digital Library, DataCite and DataONE to support collection of usage and citation metrics for data objects. Building on pilot work, this award will result in the launch of a new service that will collate and expose data level metrics.

The impact of research has traditionally been measured by citations to journal publications: journal articles are the currency of scholarly research.  However, scholarly research is made up of a much larger and richer set of outputs beyond traditional publications, including research data. In order to track and report the reach of research data, methods for collecting metrics on complex research data are needed.  In this way, data can receive the same credit and recognition that is assigned to journal articles.

Recognition of data as valuable output from the research process is increasing and this project will greatly enhance awareness around the value of data and enable researchers to gain credit for the creation and publication of data” – Ed Pentz, Crossref.

This project will work with the community to create a clear set of guidelines on how to define data usage. In addition, the project will develop a central hub for the collection of data level metrics. These metrics will include data views, downloads, citations, saves, social media mentions, and will be exposed through customized user interfaces deployed at partner organizations. Working in an open source environment, and including extensive user experience testing and community engagement, the products of this project will be available to data repositories, libraries and other organizations to deploy within their own environment, serving their communities of data authors.

Are you working in the data metrics space? Let’s collaborate.

Find out more and follow us at:, @makedatacount

About the Partners

California Digital Library was founded by the University of California in 1997 to take advantage of emerging technologies that were transforming the way digital information was being published and accessed. University of California Curation Center (UC3), one of four main programs within the CDL, helps researchers and the UC libraries manage, preserve, and provide access to their important digital assets as well as developing tools and services that serve the community throughout the research and data life cycles.

DataCite is a leading global non-profit organization that provides persistent identifiers (DOIs) for research data. Our goal is to help the research community locate, identify, and cite research data with confidence. Through collaboration, DataCite supports researchers by helping them to find, identify, and cite research data; data centres by providing persistent identifiers, workflows and standards; and journal publishers by enabling research articles to be linked to the underlying data/objects.

DataONE (Data Observation Network for Earth) is an NSF DataNet project which is developing a distributed framework and sustainable cyber infrastructure that meets the needs of science and society for open, persistent, robust, and secure access to well-described and easily discovered Earth observational data.

Announcing New Dash Features- April 2017

The Dash team is pleased to announce the release of our newest features. Taking in requests from users as well as standards in the field, we have now adapted the platform with the following releases: Private for Peer Review (Timed-Release of Data), ORCiD integration, email capture for corresponding authors, user friendly downloads, and a variety of search and view enhancements.

Private for Peer Review (Timed-Release of Data)

As mentioned in a previous post, this was formally referred to as embargoing data but we are releasing this feature in the context of keeping data private for the length of peer review. We have now implemented a feature to allow researchers to keep data private, for the purposes of peer review, for up to six months. If a researcher decides to use this option they will be given a private Reviewer URL that can be used by an external party to download the data.

This URL will redirect to the landing page with available data for download as soon as the data are public. If external parties have any questions or would like to request a download they will also now have the ability to reach the corresponding author.

Corresponding Author Email Capture & ORCiD Integration

Corresponding authors (and contributing authors) will now have the ability to enter their email address and ORCiD iD which will both appear on the landing page beneath author name. Just as article publications have, we believe Data Publications should have a corresponding author contact who can be reached with questions about the dataset.

User Friendly Downloads & Interface Improvements

What one uploads is what another may download. When choosing to download the data files, only the files uploaded by the corresponding author will be downloaded.

Some other fixes and features include:

  • the wording our our search filters and browse option
  • a checkbox at the file upload stage to ensure researchers are not uploading sensitive or identifying information 
  • explanatory information within the metadata submission for usage notes and related work
  • a preview of how large the dataset is on the download button

What’s up next?

  • Next Feature: large file upload and bulk file upload
  • Future Feature: a curation layer that will allow for administration capabilities

For more information or if you have any questions please check for updates on the @uc3cdl twitter feed, or get in touch at


Embargoing the Term “Embargoes” Indefinitely

I’m two months into a position that lends part of its time to overseeing Dash, a Data Publication platform for the University of California. On my first day I was told that a big priority for Dash was to build out an embargo feature. Coming to the California Digital Library (CDL) from PLOS, an OA publisher with an OA Data Policy, I couldn’t understand why I would be leading endeavors to embargo data and not open it up- so I met this embargo directive with apprehension.

I began to acquaint myself with the campuses and a couple of weeks ago while at UCSF I presented the prototype for what this “embargo” feature would look like and I questioned why researchers wanted to close data on an open data platform. This is where it gets fun.

“Our researchers really just want a feature to keep their data private while their associated paper is under peer review. We see this frequently when people submit to PLOS”.

Yes, I had contributed to my own conflict.

While I laughed about how I was previously the person at PLOS convincing UC researchers to make their data public- I recognized that this would be an easy issue to clarify. And here we are.

Embargoes imply a negative connotation in the open community and I ask that moving forward we do not use this phrase to talk about keeping data private until an associated manuscript has been accepted. Let us call this “Private for Peer Review” or “Timed Release”, with a “Peer Review URL” that is available for sharing data during the peer review process as Dryad does.

  • Embargoes imply that data are being held private for reasons other than the peer review process.
  • Embargoes are not appropriate if you have a funder, publisher, or other mandate to open up your data.
  • Embargoes are not appropriate for sensitive data, as these data should not be held in a public repository (embargoed) unless this were through a data access committee and the repository had proper security.
  • Embargoes are not appropriate for open Data Publications.

To embargo your data for longer than the peer review process (or for other reasons) is to shield your data from being used, built off of, or validated. This is contrary to “Open” as a strategy to further scientific findings and scholarly communications.

Dash is implementing features that will allow researchers to choose, in line with what we believe is reasonable for peer review and revisions, a publication date up to six months after submission. If researchers choose to use this feature, they will be given a Peer Review URL that can be shared to download the data until the data are public. It is important to note though that while the data may be private during this time, the DOI for the data and associated metadata will be public and should be used for citation. These features will be for the use of Peer Review; we do not believe that data should be held private for a period of time on an open data publication platform for other reasons.

Opening up data, publishing data, and giving credit to data are all important in emphasizing that data are a credible and necessary piece of scholarly work. Dash and other repositories will allow for data to be private through peer review (with the intent to have data be public and accessible in the close future). However, my hope is that as the data revolution evolves, incentives to open up data sooner will become apparent. The first step is to check our vocab and limit the use of the term “embargo” to cases where data are being held private without an open data intention.

Tagged , , ,

California Digital Library Supports the Initiative for Open Citations

California Digital Library (CDL) is proud to announce our formal endorsement for the Initiative for Open Citations (I4OC). CDL has long supported free and reusable scholarly work, as well as organizations and initiatives supporting citations in publication. With a growing database of literature and research data citations, there is a need for an open global network of citation data.

The Initiative for Open Citations will work with Crossref and their Cited-by service to open up all references indexed in Crossref. Many publishers and stakeholders have opted in to participate in opening up their citation data, and we hope that each year this list will grow to encompass all fields of publication. Furthermore, we are looking forward to seeing how research data citations will be a part of this discussion.

CDL is a firm believer in and advocate for data citations and persistent identifiers in scholarly work. However, if research publications are cited and those citations are not freely accessible and searchable- our goal is not accomplished. We are proud to support the Initiative for Open Citations and invite you to get in touch with any questions you may have about the need for open citations or ways to be an advocate for this necessary change.

Below are some Frequently Asked Questions about the need, ways to get involved, and misconceptions regarding citations. The answers are provided by the Board and founders of the I4OC Initiative:

I am a scholarly publisher not enrolled in the Cited-by service. How do I enable it?

If not already a participant in Cited-by, a Crossref member can register for this service free-of-charge. Having done so, there is nothing further the publisher needs to do to ‘open’ its reference data, other than to give its consent to Crossref, since participation in Cited-by alone does not automatically make these references available via Crossref’s standard APIs.

I am a scholarly publisher already depositing references to Crossref. How do I publicly release them?

We encourage all publishers to make their reference metadata publicly available. If you are already submitting article metadata to Crossref as a participant in their Cited-by service, opening them can be achieved in a matter of days. Publishers can easily and freely achieve this:

  • either by contacting Crossref support directly by e-mail, asking them to turn on reference distribution for all of the relevant DOI prefixes;
  • or by themselves setting the < reference_distribution_opt > metadata element to “ any ” for each DOI deposit for which they want to make references openly available.

How do I access open citation data?

Once made open, the references for individual scholarly publications may be accessed immediately through the Crossref REST API.

Open citations are also available from the OpenCitations Corpus , a database created to house scholarly citations, that is progressively and systematically harvested citation data from Crossref and other sources. An advantage of accessing citation data from the OpenCitations Corpus is that they are available in standards-compliant machine-readable RDF format , and include information about both incoming and outgoing citations of bibliographic resources (published articles and books).

Does this initiative cover future citations only or also historical data?

Both. All DOIs under a prefix set for open reference distribution will have open references through Crossref, for past, present, and future publications.

Past and present publications that lack DOIs are not dealt with by Crossref, and gaining access to their citation data will require separate initiatives by their publishers or others to extract and openly publish those references.

Under what licensing terms is citation data being made available?

Crossref exposes article and reference metadata without a license, since it regards these as raw facts that cannot be licensed.

The structured citation metadata within the OpenCitations Corpus are published under a Creative Commons CC0 public domain dedication, to make it explicitly clear that these data are open.

My journal is open access. Aren’t its articles’ citations automatically available?

No. Although Open Access articles may be open and freely available to read on the publisher’s website, their references are not separate, and are not necessarily structured or accessible programmatically. Additionally, although their reference metadata may be submitted to Crossref, Crossref historically set the default for references to “closed,” with a manual opt-in being required for public references. Many publisher members have not been aware that they could simply instruct Crossref to make references open, and, as a neutral party, Crossref has not promoted the public reference option. All publishers therefore have to opt in to open distribution of references via Crossref.

Is there a programmatic way to check whether a publisher’s or journal’s citation data is free to reuse?

For Crossref metadata , their REST API reveals how many and which publishers have opened references. Any system or tool (or a JSON viewer) can be pointed to this query: to show the count and the list of publishers with public-references “: true .

To query a specific publisher’s status, use, for example: ery=springer then find the tag for public-references. In some cases it will be set to false.


You can contact the founding group by e-mail at: .

Describing the Research Process

We at UC3 are constantly developing new tools and resources to help researchers manage their data. However, while working on projects like our RDM guide for researchers, we’ve noticed that researchers, librarians, and people working in the broader digital curation space often talk about the research process in very different ways.

To help bridge this gap, we are conducting an informal survey to understand the terms researchers use when talking about the various stages of a research project.

If you are a researcher and can spare about 5 minutes, we would greatly appreciate it if you would click the link below to participate in our survey.

Thank you.

Data Publication: Sharing, Crediting, and Re-Using Research Data

In the most basic terms- Data Publishing is the process of making research data publicly available for re-use. But even in this simple statement there are many misconceptions about what Data Publications are and why they are necessary for the future of scholarly communications.

Let’s break down a commonly accepted definition of “research data publishing”. A Data Publication has three core features: 1 – data that are publicly accessible and are preserved for an indefinite amount of time, 2 – descriptive information about the data (metadata), and 3 –  a citation for the data (giving credit to the data). Why are these elements essential? These three features make research data reusable and reproducible- the goal of a Data Publication.

Data are publicly accessible and preserved indefinitely

There are many ways for researchers to make their data publicly available, be it within Supporting Information files of a journal article or within an institutional, field specific, or general repository. For a true Data Publication, data should be submitted to a stable repository that can ensure data will be available and stored for an indefinite amount of time. There are over a thousand repositories registered with re3data and many publishers have repository guides to help with field specific guidance. When data are not suitable for public deposition, i.e. when data contain sensitive information, data should still be stored in a preserved and compliant space. While this restriction is a more difficult hurdle to jump over in advocating for data publishing and data preservation, it is important to ensure these data are not violating ethical requirements,  nor are they locked up in a filing cabinet and eventually thrown out. Preservation of data is a necessity for the future.

Data are described (data have metadata)

Data without proper documentation or descriptive metadata are about as useful as research without data. If a Data Publication is a citable piece of scholarly work, it should contain information that it allow it to be a useful and valued piece of scholarly work. Documentation and metadata range from information regarding software used for analysis to who funded the work. While these examples serve separate purposes (one for re-use and the other for credit), it is important that all information about the creation of the dataset (who, where, how, related publications) are available.

Data are citable and credible

We’ve established that datasets are essential to research output and are an important piece of scholarly work- and they should receive the same benefits. Data need to have a persistent identifier (a stable link) that can be referenced. While many repositories use a DataCite DOI to fulfill this, some field-specific repositories use accession numbers (i.e. NCBI repositories) that can be referenced within a URL. This is one of the reasons data need to be available in a stable repository. It’s a bit difficult to reference and credit data that are on your hard drive!

If it’s so clear- why are there barriers?

Data publishing has become more widely accepted in the last ten years, with new standards from funders and publishers and a growth in stable repositories. However, there’s still work to be done and more questions to be answered before we reach mass adoption. Let’s start that conversation (you can be the questioner and I’ll be the advocate):

Organizing and submitting data are time intensive and in turn, costly

Trying to replicate a data set from scratch takes much more time (and money) than publishing your data (see robotics example here). Taking the time to search your old computer files or get in touch with your last institution to get your data is more complicated than publishing your data. Having your paper retracted because your data are called into question and you can’t share your data or don’t have it would take more time, money, and hit to your reputation than proactively publishing your datasets.

As an important side note: Data Publications do not need to be linked to a journal publication. While it may take extra time to submit a Data Publication in proper form, if used as an intermediate step in the research process you can reduce time later, get credit, and benefit the research community in the meantime.

What’s the incentive?

Credit. Next question?

But beyond credit for a citable piece of work, publishing data as a common practice will shift focus from publications being an end point in the research cycle to a starting point and this shift is crucial for transparency and reproducibility in published works. Incentives will become clear once Data Citations become common practice within the publisher and research community, and resources are available for researchers to know how (and have the time/funds) to submit Data Publications.

Too few resources for understanding Data Publishing

Many great papers have been posted and published in the last ten years about what a Data Publication is; however, less resources have been made available to the research community on how to integrate Data Publishing into the research life cycle and how to organize data to even be suitable for a Data Publication. Data Management Plans, courses on research data management, and pressure from various funder and publisher policies will help, but there’s a serious need for education on data planning/organization (including metadata and format requirements) as well as awareness of data publishing platforms and their benefits. This is a call to the community to release these materials and engage in the Research Data Management (RDM) community to get as many of these conversations going. The more resources, answers, and guidance that institutions can provide to researchers, the less the “it takes too much time and money” argument will arise, the easier it will be to achieve the incentive, and the further we will push the boundaries of transparency in scholarly communications.

There’s no better time than now to re-evaluate what resources are available for research output. If we strive for re-use and reproducibility of research data within the community, then now is the time to increase awareness and adoption of Data Publication.

For more information about research data organizations, machine actionable Data Management Plans, or Data Publication platforms, please utilize UC3 resources or get in touch at

Ensuring access to critical research data

For the last two months, UC3 have been working with the teams at, Data Refuge, Internet Archive, and Code For Science (creators of the Dat Project) to aggregate the government data.

Data that spans the globe

There are currently volunteers across the country working to discover and preserve publicly funded research, especially climate data, from being deleted or lost from the public record. The largest initiative is called Data Refuge and is led by librarians and scientists. They are holding events across the UC campuses and the US that you should attend and help out in person, and are organizing the library community to band together to curate the data and ensure it’s preserved and accessible.

Our initiative builds on this and is looking to build a corpus of government data and corresponding metadata.  We are focusing on public research data, especially those at risk of disappearing. The initiative was nicknamed “Svalbard” by Max Ogden of the Dat project, after the Svalbard Global Seed Vault in the Arctic.  As of today, our friends at Code for Science have released 38GB of metadata, over 30 million hashes and URLs of research data files.

The Svalbard Global Seed Vault in the Arctic

To aid in this effort

We have assembled the following metadata as part of the Code for Science’s Svalbard v1:

  • 2.7 million SHA-256 hashes for all downloadable resources linked from, representing around 40TB of data
  • 29 million SHA-1 hashes of files archived by the Internet Archive and the Archive Team from federal websites and FTP servers, representing over 120TB of data
  • All metadata from, about 2.1 million datasets
  • A list of ~750 .gov and .mil FTP servers

There are additional sources such as Archivers.Space, EDGI, Climate Mirror, Azimuth Data Backup that we are working adding metadata for in future releases.

Following the principles set forth by the librarians behind Data Refuge, we believe it’s important to establish a clear and trustworthy chain of custody for research datasets so that mirror copies can be trusted. With this project, we are working to curate metadata that includes strong cryptographic hashes of data files in addition to metadata that can be used to reproduce a download procedure from the originating host.

We are hoping the community can use this data in the following ways:

  • To independently verify that the mirroring processes that produced these hashes can be reproduced
  • To aid in developing new forms of redundant dataset distribution (such as peer to peer networks)
  • To seed additional web crawls or scraping efforts with additional dataset source URLs
  • To encourage other archiving efforts to publish their metadata in an easily accessible format
  • To cross reference data across archives, for deduplication or verification purposes

What about the data?

The metadata is great, but the initial release of 30 million hashes and urls is just part of our project. The actual content (how the hashes were derived) have also been downloaded.  They are stored at either the Internet Archive or on our California Digital Library servers.

The Dat Project carried out a HTTP mirror (~40TB) and uploaded it to our servers at California Digital Library. We are working with them to access ~160TB of data in the future and have partnered with UC Riverside to offer longer term storage .


You can download the metadata here using Dat Desktop or Dat CLI tool.  We are using the Dat Protocol for distribution so that we can publish new metadata releases efficiently while still keeping the old versions around. Dat provides a secure cryptographic ledger, similar in concept to a blockchain, that can verify integrity of updates.


If you want to learn more about how CDL and the UC3 team is involved, contact us at or @UC3CDL. If you have suggestions or questions, you can join the Code for Science Community Chat.  And, if you are a technical user you can report issues or get involved at the Svalbard GitHub.

This is crossposted here:

Government Data At Risk

Government data is at risk, but that is nothing new.  

The existence of, the Federal Open Data Policy, and open government data belies the fact that, historically, a vast amount of government data and digital information is at risk of disappearing in the transition between presidential administrations. For example, between 2008 and 2012, over 80 percent of the PDFs hosted on .gov domains disappeared. To track these and other changes, California Digital Library (CDL) joined with the University of North Texas, The Library of Congress, the Internet Archive, and the U.S. Government Publishing office to create the End of Term (EOT) Archive. After archiving the web presence of federal agencies in 2008 and 2012, the team initiated a new crawl in September of 2016.

In light of recent events, tools and infrastructure initially developed for EOT and other projects have been taken up by efforts to backup “at risk” datasets, including those related to the environment, climate change, and social justice. Data Refuge, coordinated by the Penn Program of Environmental Humanities (PPEH), has organized a series of “Data Rescue” events across the country where volunteers nominate webpages for submission to the End of Term Archive and harvest “uncrawlable” data to be bagged and submitted to an open data archive. Efforts such as the Azimuth Climate Data Backup Project and Climate Mirror do not involve submitting data or information directly to the End of Term Archive, but have similar aims and workflows.

These efforts are great for raising awareness and building back-ups of key collections. In the background, CDL and the team behind the Dat Project have worked to backup, itself. The goal is not only to preserve the datasets catalogued by but also the associated metadata and organization that makes it such a useful location for finding and using government data. As a result of this partnership, for the first time ever, the entire metadata catalog of over 2 million datasets will soon be available for bulk download. This will allow the various backup efforts to coordinate and cross reference their data sets with those on To allow for further coordination and cross referencing, the Dat team has also begun acquiring the metadata for all the files acquired by Data Refuge, the Azimuth Climate Data Project, and Climate Mirror.

In an effort to keep track of all these efforts to preserve government data and information, we’re maintaining the following annotated list. As new efforts emerge or existing efforts broaden or change their focus, we’ll make sure the list is updated. Feel free to send additional info on government data projects to:

Get involved: Ongoing Efforts to Preserve Scientific Data or Support Science – The home of the U.S. Government’s open data, much of which is non-biological and non-environmental. has a lightweight system for reporting and tracking datasets that aren’t represented and functions as a single point of discovery for federal data. Newly archived data can and should be reported there. CDL and the Dat team are currently working to backup the data catalogued on and also the associated metadata.

End of Term – A collaborative project to capture and save U.S. Government websites at the end of presidential administrations. The initial partners in EOT included CDL, the Internet Archive, the Library of Congress, the University of North Texas, and the U.S. Government Publishing Office. Volunteers at many Data Rescue events use the URL nomination and BagIt/Bagger tools developed as part of the EOT project.

Data Refuge – A collaborative effort that aims to backup research-quality copies of federal climate and environmental data, advocate for environmental literacy, and build a consortium of research libraries to scale their tools and practices to make copies of other kinds of federal data. Find a Data Rescue event near you.

Azimuth Climate Data Backup Project – An urgent project to back up US government climate databases. Initially started by statistician Jan Galkowski and John Baez, a mathematician and science blogger at UC Riverside.

Climate Mirror – A distributed volunteer effort to mirror and back up U.S. Federal Climate Data. This project is currently being lead by Data Refuge.

The Environmental Data and Governance Initiative – An international network of academics and non-profits that addresses potential threats to federal environmental and energy policy, and to the scientific research infrastructure built to investigate, inform, and enforce. EDGI has built many of the tools used at Data Rescue events.

March for Science – A celebration of science and a call to support and safeguard the scientific community. The main march in Washington DC and satellite marches around the world are scheduled for April 22nd (Earth Day).

314 Action – A nonprofit that intends to leverage the goals and values of the greater science, technology, engineering, and mathematics community to aggressively advocate for science.

Tagged , , , , , , ,