#reproducibility — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #reproducibility, aggregated by home.social.
-
On Thursday I was at the notary for the incorporation of my first company: @atvari 🥳
Starting August, I will be offering consulting & contracting services to clients in the areas of Software & Cloud Engineering with a focus on Rust 🦀, Nix / NixOS ❄, and Open Source 🌐. I'm still open to opportunities, so feel free to reach out.
#DevEx #DevOps #Kubernetes #Declarative #Reproducibility #Observability #Scalability
-
RT @GenReasoning: 🔍 Wir stellen BackSearch vor. LLMs werden zunehmend damit beauftragt, die Zukunft vorherzusagen, doch ein gutes Backtesting erfordert einen Schnappschuss des Internums zu einem bestimmten Zeitpunkt. BackSearch ermöglicht es LLMs, das Internet so zu durchsuchen, wie es an einem bestimmten Datum aussah. Es eignet sich hervorragend für: 🔮 Prognosen und Vorhersagemärkte. 📈 Quantitative Finanzanalyse. 🌍 RL-Umgebungen, die die Welt simulieren. 🥶 Das Einfrieren der Websuche für reproduzierbare Benchmarks. Zunächst veröffentlichen wir einen schmalen Ausschnitt unseres Index, der sich auf den Nachrichtenbereich für 2026 konzentriert. Basierend auf Feedback werden wir in späteren Releases mehr unseres Indexes öffnen. 👇 Probieren Sie BackSearch mit dem Link unten aus. Video
mehr auf Arint.info
#AIResearch #BackSearch #LLM #QuantFinance #Reproducibility #WebSearch #arint_info
-
RT @GenReasoning: 🔍 Wir stellen BackSearch vor. LLMs werden zunehmend damit beauftragt, die Zukunft vorherzusagen, doch ein gutes Backtesting erfordert einen Schnappschuss des Internums zu einem bestimmten Zeitpunkt. BackSearch ermöglicht es LLMs, das Internet so zu durchsuchen, wie es an einem bestimmten Datum aussah. Es eignet sich hervorragend für: 🔮 Prognosen und Vorhersagemärkte. 📈 Quantitative Finanzanalyse. 🌍 RL-Umgebungen, die die Welt simulieren. 🥶 Das Einfrieren der Websuche für reproduzierbare Benchmarks. Zunächst veröffentlichen wir einen schmalen Ausschnitt unseres Index, der sich auf den Nachrichtenbereich für 2026 konzentriert. Basierend auf Feedback werden wir in späteren Releases mehr unseres Indexes öffnen. 👇 Probieren Sie BackSearch mit dem Link unten aus. Video
mehr auf Arint.info
#AIResearch #BackSearch #LLM #QuantFinance #Reproducibility #WebSearch #arint_info
-
Собрать прошлое: как архивировать весь трафик сборки SONiC
Интернет — это зависимость без зафиксированной версии. Сегодня он возвращает одно, завтра — другое, а послезавтра — ничего. Поэтому мы решили сохранить для сборки тот интернет, который она однажды увидела. Для этого пришлось перехватывать HTTPS, выпускать собственные сертификаты, переподписывать индексы Debian и выяснять, почему Docker отправляет «случайные» заголовки. В этой статье я расскажу, как разрабатывал кеширующую MITM-проксю, которая перехватывает весь HTTP/HTTPS-трафик и отдаёт ранее сохранённые ответы. Разберём также неочевидные технические проблемы, с которыми пришлось столкнуться.
https://habr.com/ru/companies/yandex_cloud_and_infra/articles/1060234/
-
Just finished listening to the first episode of @Felienne and Hanna Schraffenberger’s #podcast CS off course:
https://rss.com/podcasts/computer-science-off-course/2968986/
What stuck most: how tools for building #software are designed to build but not to keep track of the thought process.
This would also be a way to make CS output truly #reproducible. I love the idea of having an atlas type annotation tool. Testing would be more involved, metrics for #reproducibility of #researchsoftware are still illdefined. -
Just finished listening to the first episode of @Felienne and Hanna Schraffenberger’s #podcast CS off course:
https://rss.com/podcasts/computer-science-off-course/2968986/
What stuck most: how tools for building #software are designed to build but not to keep track of the thought process.
This would also be a way to make CS output truly #reproducible. I love the idea of having an atlas type annotation tool. Testing would be more involved, metrics for #reproducibility of #researchsoftware are still illdefined. -
@oatp Torn between intrigued and horrified.
Yes, a tighter linkage between the code and mathematics behind the figures in a paper could reduce errors (and falsification - anyone could see how the figures were generate from the dataset and verify that the dataset is unique).
But there is real, tangible value in performing the analysis AFTER the data is generated.
If you write the paper before the data is even collected with a series of "if" statements based on your hypothesis, you are explicitly excluding opportunities for the kinds of scientific advance that have been most influential and created new insights. You are explicitly excluding the kinds of scientific work that arise when the data doesn't match what you expected, when it creates an edge case you didn't think of, when the results seem clear but writing them up causes you to rethink your hypothesis.
Instead of doing the work to understand the actual outcome, it would become fatally easy to force the data to fit that predefined paradigm.
And it would be appallingly easy to establish a publication culture whete researchers who don't do this, who want to publish a novel idea or new way of interpreting the data based on how the data failed to map into their prewritten paper, are actively penalized.
-
@oatp Torn between intrigued and horrified.
Yes, a tighter linkage between the code and mathematics behind the figures in a paper could reduce errors (and falsification - anyone could see how the figures were generate from the dataset and verify that the dataset is unique).
But there is real, tangible value in performing the analysis AFTER the data is generated.
If you write the paper before the data is even collected with a series of "if" statements based on your hypothesis, you are explicitly excluding opportunities for the kinds of scientific advance that have been most influential and created new insights. You are explicitly excluding the kinds of scientific work that arise when the data doesn't match what you expected, when it creates an edge case you didn't think of, when the results seem clear but writing them up causes you to rethink your hypothesis.
Instead of doing the work to understand the actual outcome, it would become fatally easy to force the data to fit that predefined paradigm.
And it would be appallingly easy to establish a publication culture whete researchers who don't do this, who want to publish a novel idea or new way of interpreting the data based on how the data failed to map into their prewritten paper, are actively penalized.
-
**Hackathon: FAIR & Reproducible Data**
Join us on July 23-24 at the University of Cologne to explore how reproduciblity can be embedded from the very beginning of the research process — and what this means for research data management across the entire data lifecycle.
A hands-on, collaborative and open format for experimenting, learning, and exchange. No prior experience required.
More info & registration: https://www.dkz2r.de/events/2026-07-23_fair_hackathon_cologne -
☕ Thank you for the incredible response to our first-ever international ReproducibiliTea conference!
We're thrilled to see so much enthusiasm from the research community. Registrations have been coming in quickly, and spots are filling up fast.
If you've been thinking about joining us, now is a great time to secure your place.
📢 SPILLING THE REPRODUCIBILITEA: A DECADE ON & DECADES AHEAD
🗓️ September 25, 2026
💻 Online via Zoom | Free & open to allREGISTER HERE 👇
https://forms.gle/KgemQ3gM2fEezYLv7#OpenScience #ReproducibiliTea #ResearchIntegrity #Reproducibility
-
☕ Thank you for the incredible response to our first-ever international ReproducibiliTea conference!
We're thrilled to see so much enthusiasm from the research community. Registrations have been coming in quickly, and spots are filling up fast.
If you've been thinking about joining us, now is a great time to secure your place.
📢 SPILLING THE REPRODUCIBILITEA: A DECADE ON & DECADES AHEAD
🗓️ September 25, 2026
💻 Online via Zoom | Free & open to allREGISTER HERE 👇
https://forms.gle/KgemQ3gM2fEezYLv7#OpenScience #ReproducibiliTea #ResearchIntegrity #Reproducibility
-
Grab your cup ☕; we're spilling the #ReproducibiliTea.
Our first-ever international conference.
🗓️Sept 25 | 💻 Online | 🎟️ FreeRegister now → https://forms.gle/KgemQ3gM2fEezYLv7
🔗 Learn more: https://reproducibilitea.org/conference2026
#OpenScience #ResearchIntegrity #Reproducibility -
Grab your cup ☕; we're spilling the #ReproducibiliTea.
Our first-ever international conference.
🗓️Sept 25 | 💻 Online | 🎟️ FreeRegister now → https://forms.gle/KgemQ3gM2fEezYLv7
🔗 Learn more: https://reproducibilitea.org/conference2026
#OpenScience #ResearchIntegrity #Reproducibility -
Now at #jobim2026 Sarah Cohen-Boulakia is giving a keynote on #reproducibility by design in #bioinformatics
We're very proud in our community of our sharing culture: thousands of open data sources, 33k tools on bio.tools, thousands of analysis pipelines on git... but still, computational reproducibility is not a given!
Which version of what database was used? Which version of what software, with which parameters?
-
Now at #jobim2026 Sarah Cohen-Boulakia is giving a keynote on #reproducibility by design in #bioinformatics
We're very proud in our community of our sharing culture: thousands of open data sources, 33k tools on bio.tools, thousands of analysis pipelines on git... but still, computational reproducibility is not a given!
Which version of what database was used? Which version of what software, with which parameters?
-
A successful first edition of our REPRO-HPC workshop [1] at #ISC26 !
Lots of interesting talks and people : let's try to continue those discussions next year :)
-
A successful first edition of our REPRO-HPC workshop [1] at #ISC26 !
Lots of interesting talks and people : let's try to continue those discussions next year :)
-
@romanistikde We are excited to have received our first abstract to WoReLa 2 AND to find out it's about a Romance language!! 🎉
If you're interested in #reproducibility, #replication, #OpenScience and #metascience in the context of #linguistics, do consider submitting an abstract, too!
-
@romanistikde We are excited to have received our first abstract to WoReLa 2 AND to find out it's about a Romance language!! 🎉
If you're interested in #reproducibility, #replication, #OpenScience and #metascience in the context of #linguistics, do consider submitting an abstract, too!
-
RE: https://fediscience.org/@ElenLeFoll/116752603384068706
Very excited to have received our first abstract AND to find out it's about a Romance language!! 🎉
If you're interested in #reproducibility, #replication, #OpenScience and #metascience in the context of #linguistics, do consider submitting an abstract, too!
-
RE: https://fediscience.org/@ElenLeFoll/116752603384068706
Very excited to have received our first abstract AND to find out it's about a Romance language!! 🎉
If you're interested in #reproducibility, #replication, #OpenScience and #metascience in the context of #linguistics, do consider submitting an abstract, too!
-
Did you know that ReproducibiliTea journal clubs can contribute to post-publication reviews?
With support from the steering committee, participating journal clubs have access to suggested review targets, peer review guidance, publication venues for commentaries, and funding opportunities.
Interested in getting involved?
Email [email protected] to receive resources and learn more.
See our blog post for more information about the project.
https://reproducibilitea.org/jc/2025/11/07/post-publication-reviewing
#ReproducubiliTea #OpenScience #Reproducibility #PeerReview #ResearchIntegrity
-
Did you know that ReproducibiliTea journal clubs can contribute to post-publication reviews?
With support from the steering committee, participating journal clubs have access to suggested review targets, peer review guidance, publication venues for commentaries, and funding opportunities.
Interested in getting involved?
Email [email protected] to receive resources and learn more.
See our blog post for more information about the project.
https://reproducibilitea.org/jc/2025/11/07/post-publication-reviewing
#ReproducubiliTea #OpenScience #Reproducibility #PeerReview #ResearchIntegrity
-
👨💻 Oh joy, another benchmark tool for the digital hipsters tinkering with Postgres! 🎩🔧 Apparently, "reproducibility" is the new buzzword to spice up your database cocktail. 🍸 Just don't spill it on your ClickHouse cloud, lest it gets "managed" right out of your AWS bill. 💸
https://clickhouse.com/blog/postgresbench #benchmarking #Postgres #reproducibility #cloudcomputing #digitalhipsters #databasecocktail #HackerNews #ngated -
👨💻 Oh joy, another benchmark tool for the digital hipsters tinkering with Postgres! 🎩🔧 Apparently, "reproducibility" is the new buzzword to spice up your database cocktail. 🍸 Just don't spill it on your ClickHouse cloud, lest it gets "managed" right out of your AWS bill. 💸
https://clickhouse.com/blog/postgresbench #benchmarking #Postgres #reproducibility #cloudcomputing #digitalhipsters #databasecocktail #HackerNews #ngated -
📢 ESEIW/ESEM 2026 – Reproducibility Hackathon
Join our in-person hackathon: “Making Reproducibility Packages Usable (Again)”.
We want to explore together with you what makes reproducibility packages hard to (re-)use and what can be improved.
🛠️ Hackathon: 7 October 2026
🔗 https://conf.researchr.org/track/eseiw-2026/eseiw-2026-hackathon-mrpua
#ESEM2026 #Reproducibility #EmpiricalSoftwareEngineering #Hackathon
-
Right, enough complaining about academia, let's make things better! A small step, but my colleagues and I are organising the Second Workshop on Replication in the Language Sciences (WoReLa 2) in July 2027 in Cologne: https://sites.google.com/view/worela2.
The call for abstract is now open (deadline 15 August). Linguists of all walks of life, we hope that many of you will want to join us for a two-day gathering of interesting people with fresh ideas on #replication, #reproducibility, #OpenScience, #metascience, and much more in all subdisciplines of #linguistics!
Fellow linguists, please boost and tell your friends, colleagues, and students! 🔃
-
Please give a warm welcome to the two newest ReproducibiliTea journal clubs:
✨ ReproducibiliTea NeuroMarseille at Aix-Marseille University, Marseille, France ✨More info can be found here: https://reproducibilitea.org/journal-clubs/#NeuroMarseille
✨ ReproducibiliTea LSE at the London School of Economics and Political Science, London, UK ✨
More info can be found here: https://reproducibilitea.org/journal-clubs/#LSE
Would you also like to join us for the Journal Club?
We’ll guide you every step.
📩 Get started: Email [email protected] to join Slack and launch your club!
#ReproducibiliTea #OpenScience #ResearchIntegrity #Reproducibility
-
Believe it or not, but tomorrow is already Thursday 11 June and therefore time for a cup of freshly brewed #ReproducibiliTea in the HumaniTeas! 🤗
We meet at our usual tea-time 16–17:30 CEST in room 4.006 of
@unibibkoeln (entrance via Kerperner Str.) and on Zoom and very much look forward to learning from our guest speaker Robert Keßler (Parallel and Distributed Systems Group @UniKoeln about reproducibility in computer science – a topic of relevance to all of us working with computational methods of some kind!Robert's recommended (but optional) preparation reading is one of our favourites: https://doi.org/10.5281/ZENODO.3233853.
Come for the (cardamon!) tea and biscuits and stay for the insightful discussions! 🫖 🍪
#OpenScience #reproducibility #ComputerScience #DigitalHumanities
-
What makes a workflow FAIR, and what makes FAIRification difficult in practice?
At the BioFAIR Showcase 2026, Galaxy BioFAIR Fellows Marisa Loach and Nicola Soranzo led an interactive workshop on workflow FAIRification, exploring why reusable and reproducible workflows matter, what barriers researchers face, and how Galaxy’s workflow “magic wand” can help make workflows FAIRer.
Read the recap: https://galaxyproject.org/news/2026-06-04-biofair-workflows-workshop/
#BioFAIR #FAIRData #OpenScience #Reproducibility #Workflows #GalaxyProject
-
Here's our ReproducibiliTea program for June!
📅 10 June | Galway (Hybrid)
H₂Ow Much?! Game Play-Testing
Registration link: https://universityofgalway.libcal.com/event/4541563
📅 11 June | Cologne (Online)
Technological Measures to Deal with Sensitive Data in Reproducible Research
Lecture by Gioele Barabucci (University of Cologne)
Zoom: You will receive a link if you subscribe to this mailing list:
https://lists.uni-koeln.de/mailman/listinfo/reproducibilitea-humaniteas
📅 12 June | Bielefeld (Online)
Some Papers About p-values
A discussion session exploring papers on p-values and statistical inference
More info and Zoom link: https://burst.pages.ub.uni-bielefeld.de/burst/
📅 25 June | Cologne (Online)
Reproducibility in Computer Science
Lecture by Robert Keßler (University of Cologne) with recommended reading from The Turing Way
Zoom: You will receive a link if you subscribe to this mailing list:
https://lists.uni-koeln.de/mailman/listinfo/reproducibilitea-humaniteas
🌍 All are welcome to join. We look forward to seeing you at one (or more!) of our June events.
#ReproducibiliTea #OpenScience #ResearchIntegrity #Reproducibility