![](\"https://images.unsplash.com/photo-1608408891486-f5cade977d19?crop=entropy&cs=tinysrgb&fit=max&fm=jpg&ixid=MnwxMTc3M3wwfDF8c2VhcmNofDR8fGRvZyUyMGZvb2R8ZW58MHx8fHwxNjgxNzQyOTYy&ixlib=rb-4.0.3&q=80&w=2000\")
I am a big fan of dog food, and I wrote about this topic already seven years ago:
Eating your own dog food is a slang term to describe that an organization should itself use the products and services it provides.
One of the major projects I am working on right now is the Rogue Scholar science blog archive that launched at the beginning of the month. As part of this work – but also because I am very interested in this – I read a lot of science blogs. And today I released an update of the Rogue Scholar that makes this easier.
People who know me know that I care about persistent identifiers for scholarly resources. I have worked for seven years for DataCite, a DOI registration to register datasets, software, and other non-textual resources. I was involved in the launch of ORCID (identifiers for researchers) in 2012 and ROR (identifiers for research organizations) in 2019. So it shouldn''t surprise anyone that I am officially announcing the Rogue Scholar identifier for science blogs today. Each blog that has registered with the Rogue Scholar is uniquely identified, e.g.
Persistent identifiers should not have any semantic meaning (e.g. the blog name) in them, as names can change over time. And they should not be linked to a domain name, (e.g. upstream.force11.org) as those might also change. The Rogue Scholar identifier uses a 7-digit random string generated by the base32 algorithm and a two-digit checksum (the Front Matter identifier for example was generated with the random number 16127113320). DataCite, ROR, and the repository Zenodo use similarly constructed unique identifiers. Their main advantage over UUIDs is that they are easier to handle because of their compact size – there are still more than three billion unique strings for the Rogue Scholar identifier. Finally, persistent identifiers should be actionable, which means expressed as URLs that a human or machine can follow.
Why did I not use International Standard Serial Numbers (ISSNs), well-established identifiers that also work for blogs (the Front Matter blog has ISSN 2749-9952)? Why ISSN registration can be easy and cheap, registration can become an issue, especially for new blogs that are just beginning to publish. And ISSNs have only the most basic metadata (e.g. title, country). And why not use digital object identifiers (DOIs)? They have traditionally been used for scholarly outputs such as journal articles, datasets, and blog posts. While you can register DOIs for serials such as journals, conference proceedings, or blogs, there is currently no standard practice to do so.
Persistent identifiers are not really useful without meaningful metadata. For science blogs, this means at least the following:
For the blogs participating in the Rogue Scholar, I am collecting this information and will make it available in the Rogue Scholar search. To not start from scratch, I am using the metadata available from most blogs via RSS or Atom feed. For some information, e.g. license or subject area, I need to ask additional questions to the blog editor.
RSS and Atom both use
XML, rather than JSON, which is much more pleasant to work with. Therefore
– after the initial conversion of RSS or Atom XML – I can use JSON
Feed to describe blog metadata, and the format can be extended to the
needs of the Rogue Scholar. To fetch the JSON Feed of a blog included in the
Rogue Scholar, use the identifier. Either by appending .json
to the identifier (e.g. https://rogue-scholar.org/h56tk29.json)
or by entering the identifier (https://rogue-scholar.org/h56tk29)
in your RSS reader. The reader will automatically find the JSON Feed via the
link tag in the page header:
The RSS Reader (assuming it supports JSON Feed, as most readers do) will subscribe you to the JSON Feed of the blog, simplifying the reading of science blogs. More work is needed to polish the RSS/Atom Feed conversion to JSON Feed done by the Rogue Scholar and streamline subscribing to multiple blogs at once, e.g. using OPML.
JSON Feed can also be used for the metadata and content of blog posts, so again I don''t need to use XML, e.g. Journal Article Tag Suite (JATS). For blog posts, I will continue to use DOIs, as they work well, and I am making progress with Rogue Scholar integration (see for example this blog using DOIs already: https://rogue-scholar.org/f4wdg32)
How does the above help with finding, reading, sharing, or otherwise reusing science blogs? The work released today should make it easier to find interesting science blogs via the Rogue Scholar and subscribe to them via your RSS reader of choice. Over time we will hopefully see evolving community standards regarding blog persistent identifiers and metadata, following the FAIR Principles, while at the same time pushing hard for Diamond Open Access, keeping the cost and technical complexity affordable.
","tags":["Feature"],"language":"en"},{"id":"https://doi.org/10.53731/z9v2s-bh329","short_id":"y2d1rjgr","url":"https://blog.front-matter.io/posts/rogue-scholar-open-for-business/","title":"The Rogue Scholar is now open for business","summary":"The Rogue Scholar science blog archive launched with limited functionality on April 3rd. Interested science blogs can go to the sign-up page, provide some basic information via the sign-up form, and then will...","date_published":"2023-04-04T08:43:36Z","date_modified":"2023-04-04T09:31:14Z","authors":[{"url":"https://orcid.org/0000-0003-1419-2405","name":"Martin Fenner"}],"image":"https://images.unsplash.com/photo-1575663620136-5ebbfcc2c597?crop=entropy&cs=tinysrgb&fit=max&fm=jpg&ixid=MnwxMTc3M3wwfDF8c2VhcmNofDR8fG9wZW4lMjBmb3IlMjBidXNpbmVzc3xlbnwwfHx8fDE2ODA1OTI3NTU&ixlib=rb-4.0.3&q=80&w=2000","content_html":"The Rogue Scholar science blog archive launched with limited functionality on April 3rd. Interested science blogs can go to the sign-up page, provide some basic information via the sign-up form, and then will be added to the Rogue Scholar archive within two business days.
To be included in the service, your blog needs to:
Blogs that have signed up for the service (more than twenty so far) are listed in the Rogue Scholar catalog of science blogs that launched last week. And since yesterday summaries of the latest fifteen blog posts of each blog are also available.
![](\"https://blog.front-matter.io/content/images/2023/04/Bildschirmfoto-2023-04-04-um-10.12.58.png\")
These summaries (precisely the information you get in the RSS feed) serve two purposes:
The screenshot highlights several considerations when using the RSS Feed to archive a science blog in the Rogue Scholar:
The Rogue Scholar is now open for business, and I hope the limited functionality (or minimum viable product) launched this week makes it an attractive service for blog readers and authors to try out. The next big milestone is the launch of the full-text index for searching and archiving, and that is planned to happen within the next three months. Followed by DOI registration for blog posts.
","tags":["News"],"language":null},{"id":"https://doi.org/10.53731/h4b6c-h1444","short_id":"j3ejvwep","url":"https://blog.front-matter.io/posts/feedback-for-blog-publishers/","title":"Feedback for science blog publishers","summary":"The Rogue Scholar science blog archive launched last week. Going forward the focus is on improving the service and adding more blogs. This includes giving blog authors feedback on how they can improve their...","date_published":"2023-04-11T12:31:40Z","date_modified":"2023-04-14T20:50:32Z","authors":[{"url":"https://orcid.org/0000-0003-1419-2405","name":"Martin Fenner"}],"image":"https://blog.front-matter.io/content/images/2023/04/Bildschirmfoto-2023-04-11-um-13.14.02.png","content_html":"![](\"https://blog.front-matter.io/content/images/2023/04/Bildschirmfoto-2023-04-11-um-13.14.02.png\")
The Rogue Scholar science blog archive launched last week. Going forward the focus is on improving the service and adding more blogs. This includes giving blog authors feedback on how they can improve their RSS/Atom feeds – used by the Rogue Scholar to collect and archive the blog content.
A good starting point is author information, which often can be improved. The first step is to support multiple authors and support their full (given and family) names instead of usernames. It is useful to include ORCID author identifiers, best done by using the author website field of the blogging platform. This information can then be included in the blog Atom feed, which works better for this than RSS feeds.
The blog (RSS or Atom) feed includes a link for each blog post but also an id (Atom) or guid (RSS). Ideally, this id/guid is globally unique, does not change over time, and can be used as a web link. DOIs are a perfect fit for this id/guid field, and several blogs included in the Rogue Scholar do this already (this blog but also Upstream). Many blogging platforms have a canonical_url field that can be used to store the DOI, separate from the URL.
Abstracts are useful for blog posts and widely supported. Unfortunately, there is no standard way to provide them in RSS or Atom feeds. A good practice is to use text and not HTML and to limit the total number of characters (the Rogue Scholar limits abstracts to 210 characters).
Feature images for blog posts are again widely used but there is no standard way to do this in RSS or Atom feeds. Examples of Rogue Scholar blogs using feature images are Chris Hartgerink, OA.Works and Syldavia Gazette.
This week I added basic statistics for the Rogue Scholar that give preliminary insights into the kind of science blogs covered by the Rogue Scholar. The category is the top-level classification of the OECD Fields of Science and Technology. Many blogs cover Natural Sciences, Engineering and Technology, Social Sciences – Health and Medical Sciences, Humanities, and Agricultural Sciences are covered less. Almost all currently included blogs are in the English language, please reach out if you manage a blog in another language. Knowing the blogging platform helps integrate the various RSS feeds into the Rogue Scholar, and the results are as expected. Wordpress is the most popular blogging platform, but science blogs also use a variety of other platforms, including Ghost, Medium, and Blogger. Another interesting key performance indicator (KPI) is the total number of blogs and blog posts included, but this needs more work as this information is not immediately available.
![](\"https://blog.front-matter.io/content/images/2023/04/Bildschirmfoto-2023-04-11-um-13.19.27.png\")
The Usage Stats for the Rogue Scholar are publicly available here. The numbers are still small and don''t cover individual posts, or usage numbers from the blog itself, both of which may come over time. The Rogue Scholar intentionally isn''t collecting any personal information or using any cookies, but the available public information can give important insights (e.g. the countries or referer pages where users come from).
![](\"https://blog.front-matter.io/content/images/2023/04/Bildschirmfoto-2023-04-11-um-13.18.09.png\")
![](\"https://blog.front-matter.io/content/images/2023/02/TalbotHound_Talbot_Shrewsbury_Book_1445.png\")
Talbot is a Python package I started working on at the end of 2022 and plan to release to the Python Package Index (PyPi) in March. Talbot converts scholarly metadata in various formats, including Crossref, DataCite, Schema.org, BibTeX, RIS, and formatted citations – the complete list of supported formats is here. Talbot is a Python version of the Bolognese Ruby gem that I worked on with my DataCite colleagues starting in 2018. After leaving DataCite in 2021 I wrote a fork called Briard that added important metadata conversions, namely writing Crossref XML for DOI registration and reading/writing Citation File Format (CFF) for software metadata.
Talbot, Bolognese, and Briard are all names for dog breeds, the naming convention I have used for most of the Open Source software I have written since releasing the Open Source software Lagotto for tracking article-level metrics in 2012.
My two main use cases for Talbot (and Bolognese) are DOI content negotiation, using DOI metadata to generate metadata in other formats such as BibTeX or as formatted citation in one of the thousands of available citation styles. The Python version will enhance the InvenioRDM Open Source repository platform, e.g. adding RIS and Schema.org JSON-LD to the supported export formats. The other main use case is supporting DOI registration via multiple input formats. Since 2021 the Briard gem for example allows me to register DOIs for this blog as well as the Force11 Upstream blog using metadata in Schema.org format. With Talbot I want to enable Crossref DOI registration in the InvenioRDM platform for use cases where this makes sense, e.g blog posts or preprints. Talbot will help register DOIs from RSS feeds as part of the Rogue Scholar blog archive I am launching in Q2 2023.
One lesson learned with Bolognese/Briard is that the platform/language matters. The InvenioRDM backend is written in Python (the Frontend is in Javascript/React). And while Bolognese/Briard can be used via the command line or in environments such as GitHub Actions that use Docker-based microservices where the language doesn''t really matter, having the scholarly metadata conversion available in a Python environment makes a huge difference. So I took the plunge of rewriting a fairly complex library in another language. I am fully aware that there are more languages used for writing scholarly infrastructure code, but for the next few years, Python addresses my needs and is hopefully useful to other infrastructure projects.
While the overall architecture for the evolving Talbot library looks rather similar to Briard, I am making some changes based on my experience over the last five years of working on generic scholarly metadata conversions:
![](\"https://blog.front-matter.io/content/images/2023/03/standards_2x.png\")
This week I launched Commonmeta, a new scholarly metadata standard described at https://commonmeta.org. Commonmeta is the result of working on conversion tools for scholarly metadata for many years. One conclusion early on was that these conversions are many-to-many, so it becomes much easier to have an internal format that is the intermediate step for these conversions.
Commonmeta is inspired by two initiatives: Codemeta and Commonmark. CodeMeta contributors are creating a minimal metadata schema for science software and code, in JSON and XML. The goal of CodeMeta is to create a concept vocabulary that can be used to standardize the exchange of software metadata across repositories and organizations. Commonmark is a strongly defined, highly compatible specification of Markdown, along with a suite of comprehensive tests to validate Markdown implementations against this specification.
These two specifications not only inspired the name but also the principles of how I want to see Commonmeta operate:
The website goes into a little bit more detail about why I didn''t pick any the existing standards but instead came up with a new metadata standard. This is a familiar pattern made famous by the XKCD comic shown above.
As I want this to be driven by real-world implementations and not committees, I also in the last few weeks launched commonmeta-py, a Python implementation of the standard available on PyPi. And in a few months, I hope to have tweaked the Ruby Gem that I originally wrote a few years ago to support Commonmeta as the internal format.
With testable specification, I mean both a JSON Schema to describe Commonmeta and many, many tests that validate the conversions with real-world data. The JSON Schema is available here, and will become stable once it reaches version 1.0. commonmeta-py comes with lots of tests, but I hope to further improve the test coverage.
Please reach out to me if you want to help with Commonmeta, in particular, work on implementations in other languages, such as Javascript, PHP, or Java.
","tags":["News"],"language":"en"},{"id":"https://doi.org/10.53731/eyf75cj-jsgv26c","short_id":"9memqjg2","url":"https://blog.front-matter.io/posts/building-blocks/","title":"Building Blocks for a Scholarly Blog Archive","summary":"Another follow-up post, extending three earlier posts (see references), on the Scholarly Blog Archive that Front Matter is building and that I plan to launch in the first half of 2023. I have been thinking...","date_published":"2022-12-21T14:23:47Z","date_modified":"2022-12-21T20:57:38Z","authors":[{"url":"https://orcid.org/0000-0003-1419-2405","name":"Martin Fenner"}],"image":"https://blog.front-matter.io/content/images/2022/12/James_Brown_-55208420--1.jpeg","content_html":"![](\"https://blog.front-matter.io/content/images/2022/12/James_Brown_-55208420--1.jpeg\")
Another follow-up post, extending three earlier posts (see references), on the Scholarly Blog Archive that Front Matter is building and that I plan to launch in the first half of 2023. I have been thinking about the building blocks that make this blog archive work:
Diamond open access (OA) is an open access business model in which no fees are charged to either authors or readers. German Research Foundation
Using this term sounds strange in the context of scholarly blog posts, but it means that scholarly blog infrastructure should be free to publish and free to read. One challenge with Open Access for publications, particularly in disciplines such as medicine and life sciences where there is a lot of money, is that there are no drivers for driving down cost, and subscription fees have often been converted to article processing charges (APC). And instead of technological advances making scholarly publishing cheaper over time, the costs for authors and readers (and their institutions and funders who ultimately pay for this) are only increasing.
There is of course already a lot of Diamond Open Access, and infrastructures for research data and research software also typically don''t charge authors or readers. This causes other problems in terms of sustainable scholarly infrastructure and innovation, but I think it is an essential building block for the science blog archive Front Matter is building. A lot of work is needed in 2023 to come up with a strategy for sustaining the Front Matter science blog archive in the long run, all I can say now is that it will not use advertising.
For content that is free to read we need a license that specifies that. The blog archive needs clear conditions for what it can do with the content, and the same is true for downstream users and services. History tells us that licenses should be clear and simple, so for scholarly blog posts I will aim to use the Creative Commons Attribution 4.0 License (CC-BY 4.0) for all content.
![](\"https://blog.front-matter.io/content/images/2022/12/cc-by.png\")
As I explained in a post last week, a central blog archive for blog content published in many different places makes the most sense for science blog posts – a model also used by PubMed Central for a free full-text archive of biomedical and life sciences journal articles. The InvenioRDM Open Source software is a good fit for this use case.
![](\"https://blog.front-matter.io/content/images/2022/12/Download--4--1.png\")
Starting a science blog is straightforward. There are plenty of cheap and free options available from Wordpress to GitHub Pages. You might run your blog as part of a larger platform, together with collaborators, or all for yourself.
DOIs are frequently used as persistent identifiers for scholarly content and are integrated into the InvenioRDM platform. The blog archive can either archive blog posts with DOIs, or it can issue DOIs for existing blogs not using DOIs. In the latter case it is important that the DOI resolves to the original content in the hosting blog platform, and redirects to the blog platform only when the original blog is no longer available.
DOIs (e.g. from DataCite or Crossref) have a required set of metadata that makes sense for scholarly blogs. Optional metadata that are desired for the blog archive are license (see above), abstract, subject area (using the 43 OECD Fields of Science and Technology), keywords, language, and persistent identifiers for the blog (ISSN), author (ORCID) and affiliated institution (ROR).
RSS is the standard protocol for distributing and consuming blog content. It is actually a group of protocols (Atom and multiple flavors of the RSS format), but they have been around for so long that the popular tools and services support the various protocols. RSS will be the standard way how content is ingested by the blog archive, and probably also how in turn content in the central blog archive is consumed, e.g. as an automated feed of all new science blog posts in a particular subject area and language.
![](\"https://blog.front-matter.io/content/images/2022/12/images--1-.png\")
Because RSS is so widely supported, other ways of registering content – e.g. via web form, API, or webhook – are less critical for the blog archive. Work is needed on the InvenioRDM software to add strong support for RSS feeds, but would allow the automation of a lot of the work needed to build and maintain the blog archive.
Markdown is a markup language popular with many blogging platforms. It is typically used for editing blog posts and other documents in online environments but is not really used for consuming blog content via RSS. Markdown has been extended to support features needed for scholarly documents, e.g. tables and references, but the uptake of this added functionality in science blogs has been slow.
PDF is commonly used for reading scholarly publications. The workflows for submitting manuscripts to journals and preprint archives in PDF format are broken because it is tricky to extract structured documents from PDFs. The blog archive will support PDF as an output format at some point but is not a high priority. Blog posts are typically consumed via blog reader or email (if the blog produces a newsletter) rather than as PDF printed out on paper. There is work needed on the InvenioRDM platform to display full-text content rendered as HTML.
Science blog posts typically see a lightweight review workflow before publication, and often receive feedback in the form of comments and/or social media mentions. For the Front Matter science blog archive, I want to keep that approach and not build any hurdles for inclusion. Some level of curation is needed, not only to check for quackery and hate speech but also to improve metadata that help with discovery, and to find blogs that should be included. Ideally we can build a community around the science blog archive, taking advantage of the communities (focussing on different languages and subject areas) feature recently added to the InvenioRDM software.
If reading this post feels like it is 2006 – the year James Brown (used for the feature image of this post) died – again with talk about blogs, RSS, Markdown, Creative Commons, and related technologies (I for example didn''t mention Zotero, XML, or Wordpress), you are right. This is intentional, these technologies are not as sexy as using artificial intelligence or cryptocurrencies to drive this, but I want the Science Blog archive to become a scholarly resource that is useful, open, and inclusive.
Fenner, M. (2022, September 28). Starting Work on the Front Matter Archive. Front Matter. https://doi.org/10.53731/9z6rz5d-djbay0y
Fenner, M. (2022, December 12). Building an archive for scholarly blog posts. Front Matter. https://doi.org/10.53731/br9f5xa-a556w2t
Fenner, M. (2022, December 19). Launching the Front Matter Roadmap. Front Matter. https://doi.org/10.53731/cbdtfp1-1798beh
Fenner, M. (2010, October 6). Beyond the PDF – it is time for a workshop. Front Matter. https://doi.org/10.53731/r294649-6f79289-8cw7z
Fenner, M. (2013, June 19). Citations in Scholarly Markdown. Front Matter. https://doi.org/10.53731/r294649-6f79289-8cw1b
","tags":["Feature"],"language":"en"},{"id":"https://doi.org/10.53731/avg2ykg-gdxppcd","short_id":"j3ejvvep","url":"https://blog.front-matter.io/posts/need-to-fix-science-blogs/","title":"Do we need to fix science blogs?","summary":"Science blogs have been around for at least 20 years and have become an important part of science communication. So are there any fundamental issues that need fixing?Barriers to EntryBlogging platforms are...","date_published":"2023-01-25T15:14:17Z","date_modified":"2023-02-01T15:43:22Z","authors":[{"url":"https://orcid.org/0000-0003-1419-2405","name":"Martin Fenner"}],"image":"https://images.unsplash.com/photo-1585838017777-5003198884b5?crop=entropy&cs=tinysrgb&fit=max&fm=jpg&ixid=MnwxMTc3M3wwfDF8c2VhcmNofDMyfHxicm9rZW58ZW58MHx8fHwxNjc0NjUyMTEy&ixlib=rb-4.0.3&q=80&w=2000","content_html":"Science blogs have been around for at least 20 years and have become an important part of science communication. So are there any fundamental issues that need fixing?
Blogging platforms are mature at this point, and the technology is not imposing barriers to entry for most people. The user experience has greatly improved over the last few years and there are a number of affordable ways for hosting a blog that also work for science blogs, including free options such as GitHub Pages.
Science blogs have traditionally been free to read, but there is a general trend towards subscriptions for blogs (and related newsletters), as the advertising business model isn''t really working for niche content such as most science. How to sustain science blogging in the long run is an unresolved question, and charging authors (beyond a nominal hosting fee) doesn''t look like a path forward. Luckily the costs of publishing science blogs are very reasonable compared to journal publishing or hosting research data and code.
The basic functionality of formatted text with embedded figures and links is supported by many blogging platforms. The requirements of data-intensive science, e.g. interactive visualizations, can be a challenge, but that is also true for publishing journal articles. Interactive environments such as Jupyter Notebooks might be a better fit for these use cases.
Reference management is probably the biggest gap in science blogging, as handling more than a few references in standard ways is not easily done by hand.
Unfortunately a lot of the activities of scholars are driven by perceived Impact or Credit, and science blogs typically don''t score high in this regard – with the exception of some disciplines such as mathematics. There is probably no short-term solution, and I am not even sure it is a problem that needs fixing.
The long-term solution should focus on increasing the visibility and thus discoverability of science blogs to reach a larger audience. As I discussed in a previous post, my preferred approach is a central repository for science blog content originally published in many different locations (the PubMed/PubMed Central) model.
This leaves persistence as the other main problem with science blogs besides discoverability that needs fixing. Link rot (the resource identified by a URI vanishes from the web) and content drift (the resource identified by a URI changes over time) are well-known problems with digital content, from newspapers to scholarly content. There are mainly two approaches to address this problem:
Some science blog content is ephemeral and may not be worth archiving, but a lot of content is still worth reading years later (the first post of this blog is more than 15 years old), even if only to provide historical context.
In summary, we don''t need to fix everything with science blogs but rather focus on two aspects: discoverability and persistence. In doing that we also need to sort out better sustainability for science blogs, and as an added bonus improve their reference management.
Discoverability and persistence are an issue for all science blogs, and we are trying to fix them by launching the Rogue Scholar in the second quarter of 2023. If you are managing a science blog and care about discoverability and persistence, sign up for the Rogue Scholar waitlist. Particularly if your blog is no longer actively maintained, for example blogs hosted by grant-funded projects that have ended or are ending soon.
Today I launched the Rogue Scholar Documentation site, where I will document how to use the Rogue Scholar, e.g. what you can do to prepare your science blog for Rogue Scholar archiving. The site is written in markdown and hosted on GitHub, so feel free to ask questions or suggest additions via the links provided by the documentation site.
","tags":["Feature"],"language":"en"},{"id":"https://doi.org/10.53731/n7vvs-h6995","short_id":"zkevm5e3","url":"https://blog.front-matter.io/posts/rogue-scholar-releases-first-catalog/","title":"The Rogue Scholar releases its first catalog of science blogs","summary":"The Rogue Scholar blog archive today released its first catalog of science blogs, a total of nineteen science blogs that signed up for the Rogue Scholar via submission form and met the inclusion criteria: The...","date_published":"2023-03-29T20:46:54Z","date_modified":"2023-04-04T09:22:41Z","authors":[{"url":"https://orcid.org/0000-0003-1419-2405","name":"Martin Fenner"}],"image":"https://images.unsplash.com/photo-1662582632158-7f0f6e9a617b?crop=entropy&cs=tinysrgb&fit=max&fm=jpg&ixid=MnwxMTc3M3wwfDF8c2VhcmNofDMzfHxjYXRhbG9nfGVufDB8fHx8MTY4MDEyMTQ2MQ&ixlib=rb-4.0.3&q=80&w=2000","content_html":"The Rogue Scholar blog archive today released its first catalog of science blogs, a total of nineteen science blogs that signed up for the Rogue Scholar via submission form and met the inclusion criteria:
The Rogue Scholar will launch in the second quarter of this year, and this list of science blogs is an important step. The RSS feeds of the included blogs will be used to archive content and register DOIs, and they contain important information that I will include over time, including license, language, blog description, blog logo, contact person, and blogging platform.
![](\"https://blog.front-matter.io/content/images/2023/03/Bildschirmfoto-2023-03-29-um-22.38.08.png\")
The first Rogue Scholar catalog can be used as a starting point to find interesting science blogs, but more importantly, the catalog is available as an OPML file for download and can be imported (and modified) into any blog reader.
","tags":["News"],"language":"en"},{"id":"https://doi.org/10.53731/d6vdvbt-tffmezj","short_id":"5ldw65eo","url":"https://blog.front-matter.io/posts/rss-atom-jsonfeed/","title":"RSS, Atom, JSON Feed","summary":"As I discussed in a recent post, RSS is an essential building block for the upcoming Rogue Scholar Scholarly Blog Archive. RSS makes it easy to import blog posts (both metadata and content) automatically and...","date_published":"2023-01-16T16:57:54Z","date_modified":"2023-01-16T17:06:53Z","authors":[{"url":"https://orcid.org/0000-0003-1419-2405","name":"Martin Fenner"}],"image":"https://images.unsplash.com/photo-1597092451116-27787c07901d?crop=entropy&cs=tinysrgb&fit=max&fm=jpg&ixid=MnwxMTc3M3wwfDF8c2VhcmNofDJ8fGFyY2hpdmV8ZW58MHx8fHwxNjczODg2NDI2&ixlib=rb-4.0.3&q=80&w=2000","content_html":"As I discussed in a recent post, RSS is an essential building block for the upcoming Rogue Scholar Scholarly Blog Archive. RSS makes it easy to import blog posts (both metadata and content) automatically and is supported by all blogging platforms. This kind of automation is critical to keep the costs of running the Rogue Scholar low, allowing it to scale to cover a substantial number of science blog posts, and hopefully becoming an important Open Science resource.
But there are also challenges with using RSS:
Two weeks ago I opened up the waitlist for the Rogue Scholar, and I am happy with the feedback I have received so far: sixteen submissions and a number of encouraging discussions. Consider adding your science blog to the waitlist, or learn more at the Rogue Scholar website. If you have questions, post them in the comments or join the Discord channel (renamed from Front Matter to Rogue Scholar).
It has not escaped our notice that the specific use of RSS we have postulated immediately suggests a possible mechanism for the archiving and DOI registration of other scholarly content.
","tags":["Feature"],"language":"en"},{"id":"https://doi.org/10.53731/88drdpz-znvdjr9","short_id":"qlgxvqdm","url":"https://blog.front-matter.io/posts/launching-the-front-matter-gazette/","title":"Launching the Front Matter Gazette","summary":"On Wednesday this week I am launching the Front Matter Gazette, a weekly newsletter that highlights exciting science stories from around the web. The linked content highlighted in the newsletter is published...","date_published":"2023-01-30T12:48:26Z","date_modified":"2023-01-30T12:48:26Z","authors":[{"url":"https://orcid.org/0000-0003-1419-2405","name":"Martin Fenner"}],"image":"https://images.unsplash.com/photo-1521134976835-9963f2185519?crop=entropy&cs=tinysrgb&fit=max&fm=jpg&ixid=MnwxMTc3M3wwfDF8c2VhcmNofDE2fHxqb3VybmFsfGVufDB8fHx8MTY3NTAxMzMwNA&ixlib=rb-4.0.3&q=80&w=2000","content_html":"On Wednesday this week I am launching the Front Matter Gazette, a weekly newsletter that highlights exciting science stories from around the web. The linked content highlighted in the newsletter is published elsewhere and is free to read whenever possible. The newsletter requires a paid subscription (available here), 5 €/month or 50 €/year with a thirty-day free trial and free subscriptions on request. The subscription fees help pay for the curation effort – finding and summarizing the most exciting science stories.
With the Front Matter Gazette, I try a new approach to addressing an old problem: information overload.
The approach traditionally often used in science has been to use journals as a filter. There are many reasons why this approach has failed, described for example in this 2021 post on the ASAPbio blog by Christine Ferguson and me. Three important limitations are:
Two alternative approaches to journals as a filter are automation and curation. In the ASAPbio blog post mentioned earlier, Christine and I discussed an automation approach we tried out in 2021, filtering relevant biomedical preprints by the attention they received on Twitter immediately after publication. We have not continued this activity beyond early 2022 for two reasons: a) I spent the first five months of 2022 in the hospital, and b) in November 2022 I left Twitter and moved to Mastodon after the change in Twitter ownership.
There are many initiatives in this space that try to use computer algorithms to find the most relevant scholarly content, but Christine and I felt that this was only the first step and that curation was key to finding what is interesting and relevant. Curation is what journal editors have always done, and what is helped with peer review since it became increasingly required in the 1960s, but when curation is used to find what is interesting and relevant, and not what should be published, there is no longer a need to leave the curation exclusively up to journals.
An Open Science approach to curation has many elements, but a newsletter feels like a good fit. It is a low-tech approach that works even for the busiest scientists, and it can be combined with the automation approaches discussed earlier. And curated newsletters about Science and Scholarship work with preprints, research data, source code, and other forms of scholarship. A related activity, no longer so low-tech, is science podcasts, which arguably are currently more popular than science newsletters.
There are two elephants in the room for paying for this activity: advertising and grant funding. Advertising is not only a frustrating experience for readers and authors, but also doesn''t really work in a niche market such as science. The current issues at the German scienceblogs.de are only the latest example of the difficulties sustaining science blogging infrastructure.
Grant funding is a well-established strategy to pay for Open Science activities, but has two major limitations: a) it is not a good fit for the long tail of science (Front Matter for example is not (yet) a non-profit organization because the time and money required to start a non-profit in Germany are far from trivial), and b) grant funding likes to pay for innovation and research, getting funding for open scholarly infrastructure is much harder.
Of course Front Matter is open for startup funding for the Front Matter Gazette, but it should not be a requirement to get the Gazette started, and I can not promise any financial returns for an investment.
Paying even a small fee of 5 € per month for a useful Open Science resource can be a hurdle, as Impactstory can attest. That is why we offer a no-questions-asked fee waiver, and why we start the Gazette as an experiment where we don''t know the outcome yet.
Only time will tell whether the Gazette can attract enough readers to become a sustainable operation, and I will work on the Gazette until 2024 to make that call. The Ghost publishing platform powering this blog since 2021 is for people who believe in this vision (mostly in domains other than science):
Ghost is a powerful app for new-media creators to publish, share, and grow a business around their content. It comes with modern tools to build a website, publish content, send newsletters & offer paid subscriptions to members. – Ghost Homepage
Future plans for the Front Matter Gazette in case of a successful start focus on expanding the coverage – five stories a week is not even the tip of the iceberg of what''s happening every week in scholarship.
The Rogue Scholar is a science blog archive that I am working on and plan to launch in Q2 2023. Making sure that science blogs can be found over time with the help of full-text search, DOIs plus metadata, and long-term archiving is the first critical step. Using this open content in creative ways is the next step, and curation is one important aspect that I try to start addressing with the Front Matter Gazette. The Front Matter Gazette will highlight all kinds of scholarly content, not just blogs, and not only content archived in the Rogue Scholar, but there are of course synergies that I will try to explore.
In the February 1st issue I will talk about Neanderthal families, ChatGPT in science publishing, the Tidyverse, eradicating an infectious disease, and medieval manuscripts.
","tags":["News"],"language":"en"},{"id":"https://doi.org/10.53731/wa7k5-v4t16","short_id":"wneyvxe4","url":"https://blog.front-matter.io/posts/starting-the-rogue-scholar-opml-feed/","title":"Starting the Rogue Scholar OPML Feed","summary":"While the launch of the Rogue Scholar blog archive is still a few months away (happening in the second quarter of this year), I want to give an update on the ongoing work.The Rogue Scholar blog archive will...","date_published":"2023-03-22T10:42:17Z","date_modified":"2023-03-22T10:42:17Z","authors":[{"url":"https://orcid.org/0000-0003-1419-2405","name":"Martin Fenner"}],"image":"https://images.unsplash.com/photo-1611864581049-aca018410b97?crop=entropy&cs=tinysrgb&fit=max&fm=jpg&ixid=MnwxMTc3M3wwfDF8c2VhcmNofDQzfHxmZWVkfGVufDB8fHx8MTY3OTQ3NDc2NQ&ixlib=rb-4.0.3&q=80&w=2000","content_html":"While the launch of the Rogue Scholar blog archive is still a few months away (happening in the second quarter of this year), I want to give an update on the ongoing work.
The Rogue
Scholar blog archive will improve science blogs in important ways,
including
full-text search, DOIs and metadata, and long-term archiving. The central
piece of the underlying infrastructure is the InvenioRDM
open source repository software. Front Matter is one of the organizations
helping with InvenioRDM development. For the Rogue Scholar, the specific
work needed includes the following:
All blogs provide RSS feeds, which will be central to automatically fetching metadata and content for the Rogue Scholar. RSS is not built into InvenioRDM and is not needed by most organizations planning to run InvenioRDM. I will therefore build a separate service for this functionality, integrating with InvenioRDM via its REST API. For a blog to be archived and indexed in the Rogue Scholar, users will use this RSS service, providing basic information such as RSS feed URL, language, license, and contact person – basically the information collected for the Rogue Scholar waitlist (feel free to sign up your blog if you haven''t already).
Next Tuesday I will publish an OPML (Outline Processor Markup Language) file with all blogs on the Rogue Scholar waitlist. OPML is the standard for importing and exporting lists of blogs, e.g. when switching from one RSS reader to another. It is a natural fit for managing blogs in Rogue Scholar, and hopefully helps people sign up for interesting science blogs they want to read. If you are on the Rogue Scholar waitlist, please make sure your RSS Feed URL and Home Page URL are correct, and – if you haven''t done so already – pick one (and only one) of the top-level categories from the OECD Fields of Science and Technology:
The OPML file (and your RSS reader if you import that file) will group science blogs into these categories. Many blogs fall into more than one category, but that isn''t supported by OPML.
There are several ways to run InvenioRDM repository software, obviously depending on the resources available at the hosting organization, and the size and complexity of the repository. A small data repository for a university department has different needs than Zenodo, one of the most popular generalist repositories with almost three million records. The Rogue Scholar sits in the middle, a small to medium-sized repository, anticipating 2,000 to 20,000 blog posts twelve months after launch. InvenioRDM relies on Docker and Kubernetes for running production services. This makes sense for large instances such as Zenodo but adds unnecessary complexity to smaller instances such as the Rogue Scholar.
After a substantial amount of deliberation and discussion, I decided to use a different approach for the Rogue Scholar, and this might potentially be of interest to other organizations planning to use InvenioRDM:
This will be the focus of my work in the next three months, and luckily I have learned a lot about infrastructure automation in my previous jobs at PLOS and DataCite.
By default, InvenioRDM uses DataCite DOIs, but Rogue Scholar will use Crossref DOIs for blogs that don''t already use DOIs. The Crossref pricing is much more favorable for startups such as Front Matter, and for annual DOI registration numbers that at least initially will be in the 100s or low 1000s. I spent a good part of January and February writing a Python scholarly metadata conversion library that I released two weeks ago (commonmeta-py). Among other things, commonmeta-py can read and write Crossref metadata and can enable Crossref DOI registrations in InvenioRDM – which is written in Python (and Javascript for the frontend).
As always, reach out to me with questions and comments.
","tags":[],"language":"en"},{"id":"https://doi.org/10.53731/cbvm43q-qdk3s1s","short_id":"nodz2pdp","url":"https://blog.front-matter.io/posts/science-blog-archive-waitlist/","title":"Sign up for the science blog archive waitlist","summary":"The science blog archive that I have started to work on (see previous posts) finally has a name: the Rogue Scholar. I picked this name because I liked the description in the Urban Dictionary.A person with...","date_published":"2023-01-02T11:31:52Z","date_modified":"2023-01-02T11:31:52Z","authors":[{"url":"https://orcid.org/0000-0003-1419-2405","name":"Martin Fenner"}],"image":"https://images.unsplash.com/photo-1577046823799-58b2d217d508?crop=entropy&cs=tinysrgb&fit=max&fm=jpg&ixid=MnwxMTc3M3wwfDF8c2VhcmNofDZ8fGhhcHB5JTIwbmV3JTIweWVhcnxlbnwwfHx8fDE2NzI2NTY4MzQ&ixlib=rb-4.0.3&q=80&w=2000","content_html":"The science blog archive that I have started to work on (see previous posts) finally has a name: the Rogue Scholar. I picked this name because I liked the description in the Urban Dictionary.
A person with extensive knowledge pertaining to various subject matters that extends beyond formal education. This person often gathers knowledge from various sources, such as media, friends, casual reading or the internet.
And I started a waitlist for people interested in having their science blog archived in the Rogue Scholar. There is still a lot of work to do, but I hope to launch the archive in the second quarter of 2023 with these core features:
The form to sign up for the waitlist is available here.
","tags":["News"],"language":"en"},{"id":"https://doi.org/10.53731/a0d9m3n-n7r8h0m","short_id":"3ng2zrg1","url":"https://blog.front-matter.io/posts/guidelines-for-scholarly-blogs/","title":"Guidelines for Scholarly Blogs","summary":"These guidelines are recommendations for authors of scholarly blogs to help with long-term archiving, discoverability, and citation of blog content.They are modeled after the publication A Data Citation...","date_published":"2023-02-06T11:52:24Z","date_modified":"2023-02-06T11:52:24Z","authors":[{"url":"https://orcid.org/0000-0003-1419-2405","name":"Martin Fenner"}],"image":"https://images.unsplash.com/photo-1584631277142-0ca0cfc76aec?crop=entropy&cs=tinysrgb&fit=max&fm=jpg&ixid=MnwxMTc3M3wwfDF8c2VhcmNofDZ8fGd1aWRlbGluZXxlbnwwfHx8fDE2NzU2ODM0NDc&ixlib=rb-4.0.3&q=80&w=2000","content_html":"These
guidelines are recommendations for authors of scholarly blogs to help with
long-term archiving, discoverability, and citation of blog content.
They
are modeled after the publication A
Data Citation Roadmap for Scholarly Data Repositories, where many of the
same guidelines apply, and where I was the first author and co-chair
of the corresponding Force11 working group.
These guidelines focus on the required or recommended work for scholarly blog authors. For scholarly blog archives such as the Rogue Scholar, additional guidelines are in development.
| Level | \n# | \nGuideline | \n
|---|---|---|
| Required | \n1 | \nThe full-text content must be made available via public RSS feed (in RSS, Atom or JSON Feed format). | \n
| Required | \n2 | \nEach blog post in the RSS feed must have a title, author(s), and publication date. | \n
| Required | \n3 | \nEach blog post must have a URL that resolves to a public landing page specific for that blog post. | \n
| Required | \n4 | \nThe full-text content must be made available via a Creative Commons Attribution (CC-BY) license. | \n
| Required | \n5 | \nThe blog must provide documentation about long-term archiving, discoverability, and citation. | \n
| Recommended | \n6 | \nEach blog post in the RSS feed should have a persistent identifier, description, language, and last updated date. | \n
| Recommended | \n7 | \nThe landing page should include metadata required for citation, and ideally also metadata facilitating discovery, in human-readable and machine-readable format. | \n
| Recommended | \n8 | \nThe machine-readable metadata should use schema.org markup in JSON-LD format. | \n
| Recommended | \n9 | \nMetadata should be made available via HTML meta tags to facilitate use by reference managers. | \n
| Recommended | \n10 | \nMetadata should be made available for download in BibTeX and/or another standard bibliographic format. | \n
The requirement for full-text content via RSS feed and with a CC-BY license comes from the need to make archiving and indexing as simple (and cheap) as possible. Dealing with multiple licenses, private feeds, and private content adds an extra level of complexity and is not supportive of Open Science.
Metadata via HTML meta tags and JSON-LD (using schema.org markup) are two main strategies to embed metadata in web pages, to support reference managers but also indexers. Schema.org is simpler to work with, e.g. for more complex author information such as separate given and family names, author identifiers such as ORCID, and affiliation information. On the other hand, reference managers and Google Scholar currently use HTML meta tags, and it is sometimes easier to add this information to a blog.
Registration of DOIs as other persistent identifiers for blog posts is something that I want to provide via the Rogue Scholar archive, as the effort required is not trivial. The information required (mainly title, author(s), publication date, and URL) is readily available via the RSS feed. Of course, displaying these DOIs on the blog is recommended, and for the DOIs to resolve to the blog itself rather than the blog archive at the Rogue Scholar or elsewhere.
The recommended or optional metadata for science blog posts is of course a big topic that needs more discussion. Description, language, and last updated date seem desired and readily available. References used in blog posts would be fantastic to be included in the metadata, but there is currently no easy and standard way of doing this. For better discoverability, it would make sense to provide geo coordinates and/or temporal information, and all blogs would benefit from using subject classification such as the OECD Fields of Science and Technology, but all this would require significantly more effort.
These guidelines are a work in progress and are made available as part of the Rogue Scholar Documentation. Feedback is greatly appreciated.
","tags":["Feature"],"language":"en"},{"id":"https://doi.org/10.53731/4nwxn-frt36","short_id":"1jgo8yel","url":"https://blog.front-matter.io/posts/does-it-compose/","title":"Does it compose?","summary":"One question I have increasingly asked myself in the past few years. Meaning Can I run this open source software using Docker containers and a Docker Compose file?As the Docker project turned ten this...","date_published":"2023-05-16T11:36:56Z","date_modified":"2023-05-16T11:36:56Z","authors":[{"url":"https://orcid.org/0000-0003-1419-2405","name":"Martin Fenner"}],"image":"https://images.unsplash.com/photo-1523351964962-1ee5847816c3?crop=entropy&cs=tinysrgb&fit=max&fm=jpg&ixid=M3wxMTc3M3wwfDF8c2VhcmNofDUzfHxjb250YWluZXJ8ZW58MHx8fHwxNjg0MjMyMTQ0fDA&ixlib=rb-4.0.3&q=80&w=2000","content_html":"One question I have increasingly asked myself in the past few years. Meaning
Can I run this open source software using Docker containers and a Docker Compose file?
As the Docker project turned
ten this spring, it has become standard practice to distribute open source
software via Docker images and to provide a Docker
Compose file to run the software together with other dependencies. The
Awesome Compose
project has collected many examples, and all you need is a docker-compose.ymlfile
and a recent installation of Docker, e.g. Docker
Desktop. Be aware that Docker Compose has evolved over the years. It started
out as a dedicated Python application but was later integrated into the Docker
application (written in Go) as Compose V2.
Docker and Docker Compose allow you to run pretty complex applications without first addressing a long list of requirements (which might conflict with other software you have installed), or needing a long and complex build step where many things can go wrong. For example a self-hosted instance of Supabase (a hosted Postgres database with additional features) that I installed last week following these instructions.
An important open source project that I am involved in is InvenioRDM, the turn-key research data management repository. InvenioRDM started in 2019, with a first production-suitable version in August 2021, and the next major goal is to have the large and popular Zenodo repository running on top of InvenioRDM. Zenodo turned ten last week, a few weeks after Docker. Interestingly, my personal tenth anniversary was last year in May as I became a full-time software developer and left academic medicine as a medical doctor treating cancer patients in May 2012.
Unfortunately,
InvenioRDM \"doesn''t compose\" yet. It is very close, but there are no ready-made
Docker images to download, and the installation
instructions start with installing a Python command-line tool (invenio-cli).
So if you have 1-2 hours to play with InvenioRDM and get a first impression,
there is no official solution from the InvenioRDM project yet. For this reason,
I started the docker-invenio-rdm
repository on Github. It contains a Docker Compose file that uses pre-built
Docker images, and using that file with a docker compose upcommand
on your local computer should give you a running InvenioRDM within 15 minutes:
![](\"https://blog.front-matter.io/content/images/2023/05/Bildschirmfoto-2023-05-11-um-10.37.55.png\")
I started this recently and obviously want to move forward in two directions:
But of course, Docker Compose is not the answer to all questions regarding running Docker-based infrastructure. For production environments, most people shy away from using Docker Compose. The reasons for that and the alternatives will be the topic of a future blog post (spoiler: there is exciting news).
Docker Compose also needs more work to be set up correctly for development environments. It is a common practice and a workflow I used while working at DataCite (where we launched Docker-based infrastructure in 2016), but for now, the easiest way to set up InvenioRDM development environments is using the invenio-cli tool with a local development environment.
Please reach out to me with feedback on running Docker Compose for InvenioRDM (use the discussions feature in the GitHub repo), or if you have questions about running InvenioRDM in production.
","tags":["News"],"language":"en"},{"id":"https://doi.org/10.53731/fawv321-14359c4","short_id":"56gl1qd9","url":"https://blog.front-matter.io/posts/announcing-commonmeta-ruby/","title":"Announcing commonmeta-ruby","summary":"Following recent announcements of the commonmeta standard for scholarly metadata and a Python package that converts several metadata formats (commonmeta-py), today I am happy to announce commonmeta-ruby, a...","date_published":"2023-03-20T14:54:00Z","date_modified":"2023-03-22T12:32:52Z","authors":[{"url":"https://orcid.org/0000-0003-1419-2405","name":"Martin Fenner"}],"image":"https://images.unsplash.com/photo-1676284572206-2501ff5c6956?crop=entropy&cs=tinysrgb&fit=max&fm=jpg&ixid=MnwxMTc3M3wwfDF8c2VhcmNofDUwfHxiaWtlJTIwbSVDMyVCQ25zdGVyfGVufDB8fHx8MTY3OTMyMTU4MA&ixlib=rb-4.0.3&q=80&w=2000","content_html":"Following recent announcements of the commonmeta standard for scholarly metadata and a Python package that converts several metadata formats (commonmeta-py), today I am happy to announce commonmeta-ruby, a Ruby gem and command-line tool to convert scholarly metadata using commonmeta as the internal format. commonmeta-ruby is based on the bolognese Ruby library that I started a few ago while working at DataCite, but is a major rewrite that uses commonmeta as its intermediary conversion format.
Originally planned for later this year, I decided to speed up the release as Ruby version 2.x (currently 2.7.7) reaches its end of life this month, and briard (the fork I wrote to support additional metadata conversions such as Citation File Format and Crossref DOI registrations) didn''t fully work with Ruby 3.x. In addition to supporting Ruby 3.x and validating with the commonmeta JSON Schema, commonmeta-ruby dropped support for DataCite XML. The DataCite REST API has always been a JSON API, and DOI registration using DataCite XML for many years has used JSON under the hood. Metadata conversion using XML is painful, and focussing on JSON metadata simplifies further development.
The next steps for commonmeta are:
![](\"https://images.unsplash.com/photo-1575663620136-5ebbfcc2c597?crop=entropy&cs=tinysrgb&fit=max&fm=jpg&ixid=MnwxMTc3M3wwfDF8c2VhcmNofDR8fG9wZW4lMjBmb3IlMjBidXNpbmVzc3xlbnwwfHx8fDE2ODA1OTI3NTU&ixlib=rb-4.0.3&q=80&w=2000\")
![](\"https://images.unsplash.com/photo-1585838017777-5003198884b5?crop=entropy&cs=tinysrgb&fit=max&fm=jpg&ixid=MnwxMTc3M3wwfDF8c2VhcmNofDMyfHxicm9rZW58ZW58MHx8fHwxNjc0NjUyMTEy&ixlib=rb-4.0.3&q=80&w=2000\")
![](\"https://images.unsplash.com/photo-1662582632158-7f0f6e9a617b?crop=entropy&cs=tinysrgb&fit=max&fm=jpg&ixid=MnwxMTc3M3wwfDF8c2VhcmNofDMzfHxjYXRhbG9nfGVufDB8fHx8MTY4MDEyMTQ2MQ&ixlib=rb-4.0.3&q=80&w=2000\")
![](\"https://images.unsplash.com/photo-1597092451116-27787c07901d?crop=entropy&cs=tinysrgb&fit=max&fm=jpg&ixid=MnwxMTc3M3wwfDF8c2VhcmNofDJ8fGFyY2hpdmV8ZW58MHx8fHwxNjczODg2NDI2&ixlib=rb-4.0.3&q=80&w=2000\")
![](\"https://images.unsplash.com/photo-1521134976835-9963f2185519?crop=entropy&cs=tinysrgb&fit=max&fm=jpg&ixid=MnwxMTc3M3wwfDF8c2VhcmNofDE2fHxqb3VybmFsfGVufDB8fHx8MTY3NTAxMzMwNA&ixlib=rb-4.0.3&q=80&w=2000\")
![](\"https://images.unsplash.com/photo-1611864581049-aca018410b97?crop=entropy&cs=tinysrgb&fit=max&fm=jpg&ixid=MnwxMTc3M3wwfDF8c2VhcmNofDQzfHxmZWVkfGVufDB8fHx8MTY3OTQ3NDc2NQ&ixlib=rb-4.0.3&q=80&w=2000\")
![](\"https://images.unsplash.com/photo-1577046823799-58b2d217d508?crop=entropy&cs=tinysrgb&fit=max&fm=jpg&ixid=MnwxMTc3M3wwfDF8c2VhcmNofDZ8fGhhcHB5JTIwbmV3JTIweWVhcnxlbnwwfHx8fDE2NzI2NTY4MzQ&ixlib=rb-4.0.3&q=80&w=2000\")
![](\"https://images.unsplash.com/photo-1584631277142-0ca0cfc76aec?crop=entropy&cs=tinysrgb&fit=max&fm=jpg&ixid=MnwxMTc3M3wwfDF8c2VhcmNofDZ8fGd1aWRlbGluZXxlbnwwfHx8fDE2NzU2ODM0NDc&ixlib=rb-4.0.3&q=80&w=2000\")
![](\"https://images.unsplash.com/photo-1523351964962-1ee5847816c3?crop=entropy&cs=tinysrgb&fit=max&fm=jpg&ixid=M3wxMTc3M3wwfDF8c2VhcmNofDUzfHxjb250YWluZXJ8ZW58MHx8fHwxNjg0MjMyMTQ0fDA&ixlib=rb-4.0.3&q=80&w=2000\")
![](\"https://images.unsplash.com/photo-1676284572206-2501ff5c6956?crop=entropy&cs=tinysrgb&fit=max&fm=jpg&ixid=MnwxMTc3M3wwfDF8c2VhcmNofDUwfHxiaWtlJTIwbSVDMyVCQ25zdGVyfGVufDB8fHx8MTY3OTMyMTU4MA&ixlib=rb-4.0.3&q=80&w=2000\")