Paul Rohr
@pevohr.bsky.social
Dad, startup guy. Ideas matter. Design matters. It's about we, not me.
hachyderm.io/@pevohr
hachyderm.io/@pevohr
Reposted by Paul Rohr
Recently we hired @jimray.bsky.team as a new devrel at Bluesky, and I wanted to give him an overview of the Atmosphere. We recorded the session!
I give you: Squeeky Board with Paul and Jim, a detailed overview of AT://
I give you: Squeeky Board with Paul and Jim, a detailed overview of AT://
Squeeky Board with Paul and Jim
YouTube video by Paul Frazee
www.youtube.com
November 11, 2025 at 11:33 PM
Recently we hired @jimray.bsky.team as a new devrel at Bluesky, and I wanted to give him an overview of the Atmosphere. We recorded the session!
I give you: Squeeky Board with Paul and Jim, a detailed overview of AT://
I give you: Squeeky Board with Paul and Jim, a detailed overview of AT://
"You can do anything with the PDS. It's just a NoSQL DB. Some may be recoiling reading this, but its true."
In this case, @wisp.place is a thin caching server to more efficiently host static sites on top of blobs in your PDS
In this case, @wisp.place is a thin caching server to more efficiently host static sites on top of blobs in your PDS
November 13, 2025 at 5:09 PM
"You can do anything with the PDS. It's just a NoSQL DB. Some may be recoiling reading this, but its true."
In this case, @wisp.place is a thin caching server to more efficiently host static sites on top of blobs in your PDS
In this case, @wisp.place is a thin caching server to more efficiently host static sites on top of blobs in your PDS
"Why not connect all my stuff together? AT Protocol provides everything I need—for free—and I retain full ownership and control of my stuff.
No vendor lock-in, no proprietary formats, no surprise pricing changes, just my stuff living in my data repo (hosted for now on Bluesky's PDS)."
No vendor lock-in, no proprietary formats, no surprise pricing changes, just my stuff living in my data repo (hosted for now on Bluesky's PDS)."
We can just do things. Thanks to the work of @atproto.com @atprotocol.dev @sri.xyz @juli.ee @leaflet.pub @awarm.space @schlage.town @cozylittle.house and @danabra.mov for the tools and inspiration to actually get this to work. I am merely a vessel for copypasta coding.
Using AT Protocol as a Headless CMS for your websites
wecanjustdothings.leaflet.pub
November 13, 2025 at 2:41 PM
"Why not connect all my stuff together? AT Protocol provides everything I need—for free—and I retain full ownership and control of my stuff.
No vendor lock-in, no proprietary formats, no surprise pricing changes, just my stuff living in my data repo (hosted for now on Bluesky's PDS)."
No vendor lock-in, no proprietary formats, no surprise pricing changes, just my stuff living in my data repo (hosted for now on Bluesky's PDS)."
Note to self
I'm only now realizing that most folks still look at #atproto architectures from the perspective of a single app. As though there's a reusable pile of orthogonal parts for efficiently building many such apps in parallel
All of which is true, of course, + eminently cool
I'm only now realizing that most folks still look at #atproto architectures from the perspective of a single app. As though there's a reusable pile of orthogonal parts for efficiently building many such apps in parallel
All of which is true, of course, + eminently cool
November 13, 2025 at 6:28 AM
Note to self
I'm only now realizing that most folks still look at #atproto architectures from the perspective of a single app. As though there's a reusable pile of orthogonal parts for efficiently building many such apps in parallel
All of which is true, of course, + eminently cool
I'm only now realizing that most folks still look at #atproto architectures from the perspective of a single app. As though there's a reusable pile of orthogonal parts for efficiently building many such apps in parallel
All of which is true, of course, + eminently cool
Reposted by Paul Rohr
A staggering statistic: "North American researchers were charged over US$2.27 billion by just two for-profit publishers. The Canadian research councils and the US National Science Foundation were allocated US$9.3 billion in that year." What are we doing?
We wrote the Strain on scientific publishing to highlight the problems of time & trust. With a fantastic group of co-authors, we present The Drain of Scientific Publishing:
a 🧵 1/n
Drain: arxiv.org/abs/2511.04820
Strain: direct.mit.edu/qss/article/...
Oligopoly: direct.mit.edu/qss/article/...
a 🧵 1/n
Drain: arxiv.org/abs/2511.04820
Strain: direct.mit.edu/qss/article/...
Oligopoly: direct.mit.edu/qss/article/...
November 12, 2025 at 1:58 PM
A staggering statistic: "North American researchers were charged over US$2.27 billion by just two for-profit publishers. The Canadian research councils and the US National Science Foundation were allocated US$9.3 billion in that year." What are we doing?
Come for the royalty rates ($611,911/40,000 = $15.30/copy), stay for the Mongols
Fascinating deep dive on what it takes to reach this level of success with a self-published tech book. Effectively a lean startup so carefully targeted that it's now all the rage in the Mongolian tech scene
Fascinating deep dive on what it takes to reach this level of success with a self-published tech book. Effectively a lean startup so carefully targeted that it's now all the rage in the Mongolian tech scene
Knowing such exceptions can exist might help someone decide that it’s worth spending all that time writing a standout book!
More:
newsletter.pragmaticengineer.com/p/the-softwa...
More:
newsletter.pragmaticengineer.com/p/the-softwa...
The Software Engineer’s Guidebook: a recap
Reflections on publishing The Software Engineer’s Guidebook two years ago, which has sold around 40,000 copies. Also: an unexpected trip to Mongolia to visit the startup which translated it
newsletter.pragmaticengineer.com
November 12, 2025 at 1:40 PM
Come for the royalty rates ($611,911/40,000 = $15.30/copy), stay for the Mongols
Fascinating deep dive on what it takes to reach this level of success with a self-published tech book. Effectively a lean startup so carefully targeted that it's now all the rage in the Mongolian tech scene
Fascinating deep dive on what it takes to reach this level of success with a self-published tech book. Effectively a lean startup so carefully targeted that it's now all the rage in the Mongolian tech scene
"My science is different." -- Ken Goodman
Yikes. This 2019 intellectual history of "three cueing" pedagogy shows how it disconnected from developmental psychology theories in the 1960s + missed out on cognitive science results from the 1970s on
Living in Thomas Kuhn's world is a trip
Yikes. This 2019 intellectual history of "three cueing" pedagogy shows how it disconnected from developmental psychology theories in the 1960s + missed out on cognitive science results from the 1970s on
Living in Thomas Kuhn's world is a trip
how did i go so long not knowing we were literally teaching people to read wrong. i knew something wasn't right but jesus christ www.apmreports.org/episode/2019...
How a flawed idea is teaching millions of kids to be poor readers
For decades, schools have taught children the strategies of struggling readers, using a theory about reading that cognitive scientists have repeatedly debunked. And many teachers and parents don't kno...
www.apmreports.org
November 11, 2025 at 10:06 PM
"My science is different." -- Ken Goodman
Yikes. This 2019 intellectual history of "three cueing" pedagogy shows how it disconnected from developmental psychology theories in the 1960s + missed out on cognitive science results from the 1970s on
Living in Thomas Kuhn's world is a trip
Yikes. This 2019 intellectual history of "three cueing" pedagogy shows how it disconnected from developmental psychology theories in the 1960s + missed out on cognitive science results from the 1970s on
Living in Thomas Kuhn's world is a trip
Search is down at the moment, but none of the current status.bsky.app endpoints -- mostly mushrooms, AFAICT -- seem to be noticing this
( devops for distributed systems is never fun )
( devops for distributed systems is never fun )
November 11, 2025 at 8:52 PM
Search is down at the moment, but none of the current status.bsky.app endpoints -- mostly mushrooms, AFAICT -- seem to be noticing this
( devops for distributed systems is never fun )
( devops for distributed systems is never fun )
Reposted by Paul Rohr
If you care about how Congress works, this newsletter is always a must-read. The analysis of how Pelosi's consolidation of authority in leadership both was the secret to her very real policy successes and laid the groundwork for our current troubles is smart. open.substack.com/pub/firstbra...
Master of the House: The Pelosi Paradox
How the Strongest Speaker Made Congress Weak
open.substack.com
November 10, 2025 at 1:55 PM
If you care about how Congress works, this newsletter is always a must-read. The analysis of how Pelosi's consolidation of authority in leadership both was the secret to her very real policy successes and laid the groundwork for our current troubles is smart. open.substack.com/pub/firstbra...
Maintainers who wanted purely vibe-coded "solutions" could've just DIY
But they already decided that doing the additional engineering work themselves to review, fix, +/or replace generated code wasn't worth it for that feature
Don't submit PRs which effectively force them to do that work for you
But they already decided that doing the additional engineering work themselves to review, fix, +/or replace generated code wasn't worth it for that feature
Don't submit PRs which effectively force them to do that work for you
My first advice to junior contributors is to STOP using vibe coding for PRs. OSS is always about people more than about code. We don't need more code generated by LLM, we need more people who care.
November 10, 2025 at 3:42 PM
Maintainers who wanted purely vibe-coded "solutions" could've just DIY
But they already decided that doing the additional engineering work themselves to review, fix, +/or replace generated code wasn't worth it for that feature
Don't submit PRs which effectively force them to do that work for you
But they already decided that doing the additional engineering work themselves to review, fix, +/or replace generated code wasn't worth it for that feature
Don't submit PRs which effectively force them to do that work for you
Reposted by Paul Rohr
Here's our interview with the creators of Save Our Signs, a project to archive signs and placards from national parks as the Trump administration attempts to rewrite history by having many removed.
YouTube here, subscribe for an interview like this every week: www.youtube.com/watch?v=xrCE...
YouTube here, subscribe for an interview like this every week: www.youtube.com/watch?v=xrCE...
Volunteers Archived 10,000 Signs from National Parks (with Jenny McBurney and Lynda Kellam)
YouTube video by 404 Media
www.youtube.com
November 10, 2025 at 2:09 PM
Here's our interview with the creators of Save Our Signs, a project to archive signs and placards from national parks as the Trump administration attempts to rewrite history by having many removed.
YouTube here, subscribe for an interview like this every week: www.youtube.com/watch?v=xrCE...
YouTube here, subscribe for an interview like this every week: www.youtube.com/watch?v=xrCE...
Most engineers don't design protocols like @bnewbold.net does, but we know + love this moment:
«But some point, even if you don’t have the final solution, you can see how you’re going to solve it. And when you see it, you’re always like, “This is going to work.” That it is elegant enough to work.»
«But some point, even if you don’t have the final solution, you can see how you’re going to solve it. And when you see it, you’re always like, “This is going to work.” That it is elegant enough to work.»
The latest interview up at the Protocol Oral History Project is a @bsky.app hometown hero: Protocol Engineer @bnewbold.net. I'm so grateful for his time and care in sharing his experience building the foundation for not only this app but many more: protocol.ecologies.info/interviews/n...
Bryan Newbold: Protocol engineering
A protocol engineer at the social-media platform Bluesky describes building an open protocol around a fast-growing social media platform.
protocol.ecologies.info
November 10, 2025 at 1:11 PM
Most engineers don't design protocols like @bnewbold.net does, but we know + love this moment:
«But some point, even if you don’t have the final solution, you can see how you’re going to solve it. And when you see it, you’re always like, “This is going to work.” That it is elegant enough to work.»
«But some point, even if you don’t have the final solution, you can see how you’re going to solve it. And when you see it, you’re always like, “This is going to work.” That it is elegant enough to work.»
What can a panel of 50 kids (age 3.5 - 4.25) tell you about risk reduction?
Whether safety caps on a medicine bottle (or other child-resistant packaging) actually work well enough to be legally used in the US
For details, see 16 CFR § 1700.20
Whether safety caps on a medicine bottle (or other child-resistant packaging) actually work well enough to be legally used in the US
For details, see 16 CFR § 1700.20
16 CFR § 1700.20 - Testing procedure for special packaging.
www.law.cornell.edu
November 9, 2025 at 9:30 PM
What can a panel of 50 kids (age 3.5 - 4.25) tell you about risk reduction?
Whether safety caps on a medicine bottle (or other child-resistant packaging) actually work well enough to be legally used in the US
For details, see 16 CFR § 1700.20
Whether safety caps on a medicine bottle (or other child-resistant packaging) actually work well enough to be legally used in the US
For details, see 16 CFR § 1700.20
Since I didn't make it up to Montreal earlier this week, dedicating the rest of my sunny Saturday afternoon to catching up:
1. hit play on 2 hour video
2. skim mailing list, chat transcript, + other artifacts
3. ...
4. enlightenment!
1. hit play on 2 hour video
2. skim mailing list, chat transcript, + other artifacts
3. ...
4. enlightenment!
IETF 124: Authenticated Transfer (ATP) BoF Session
www.youtube.com/watch?v=x9tT...
From 2025-11-05 14:30 (Montreal)
www.youtube.com/watch?v=x9tT...
From 2025-11-05 14:30 (Montreal)
IETF 124: Authenticated Transfer (ATP) 2025-11-05 14:30
YouTube video by IETF - Internet Engineering Task Force
www.youtube.com
November 8, 2025 at 7:12 PM
Since I didn't make it up to Montreal earlier this week, dedicating the rest of my sunny Saturday afternoon to catching up:
1. hit play on 2 hour video
2. skim mailing list, chat transcript, + other artifacts
3. ...
4. enlightenment!
1. hit play on 2 hour video
2. skim mailing list, chat transcript, + other artifacts
3. ...
4. enlightenment!
Thinking more about how revolutionary the #atproto architecture is ...
What could you do if every record in your "locked open" structured datastore had a resolvable AT url? Because it already does
What could you do if every record in your "locked open" structured datastore had a resolvable AT url? Because it already does
By choosing to build on atproto, innovators have the advantage of starting with all the following pre-built:
- structured datastore
- distribution + moderation infra
- large existing user base
- trivial social integration
... allowing immediate focus on what's unique to their app or community
- structured datastore
- distribution + moderation infra
- large existing user base
- trivial social integration
... allowing immediate focus on what's unique to their app or community
Innovation here is exploding because of how bad other spaces have become: locking down reach with algos that don’t like links, or gating reach through platform payments.
On ATProto you can start with a custom data Lexicon and front end app… and that’s it!
Logins & user owned data included.
On ATProto you can start with a custom data Lexicon and front end app… and that’s it!
Logins & user owned data included.
November 7, 2025 at 7:11 PM
Thinking more about how revolutionary the #atproto architecture is ...
What could you do if every record in your "locked open" structured datastore had a resolvable AT url? Because it already does
What could you do if every record in your "locked open" structured datastore had a resolvable AT url? Because it already does
Reposted by Paul Rohr
I want my Machines to have more Loving Grace and do less All Watched Over
November 7, 2025 at 4:34 PM
I want my Machines to have more Loving Grace and do less All Watched Over
Is this a "just me" behavior, or are others doing this?
- turned off reposts to lower velocity of my following feed
- followed a high-quality reposter (who rarely posts)
Why? Hoping to increase the diversity of feeds like:
Quiet Posters
Mutuals
Popular With Friends
For You (the good one)
- turned off reposts to lower velocity of my following feed
- followed a high-quality reposter (who rarely posts)
Why? Hoping to increase the diversity of feeds like:
Quiet Posters
Mutuals
Popular With Friends
For You (the good one)
November 7, 2025 at 4:46 PM
Is this a "just me" behavior, or are others doing this?
- turned off reposts to lower velocity of my following feed
- followed a high-quality reposter (who rarely posts)
Why? Hoping to increase the diversity of feeds like:
Quiet Posters
Mutuals
Popular With Friends
For You (the good one)
- turned off reposts to lower velocity of my following feed
- followed a high-quality reposter (who rarely posts)
Why? Hoping to increase the diversity of feeds like:
Quiet Posters
Mutuals
Popular With Friends
For You (the good one)
Reposted by Paul Rohr
Wild that a Fed judge has to remind federal law enforcement that obeying the Constitution isn’t something they can call a “harm.”
Ellis: "Requiring the government to comply with its obligations under the Constitution, in particular the Fourth Amendment, is simply not a harm."
November 6, 2025 at 5:24 PM
Wild that a Fed judge has to remind federal law enforcement that obeying the Constitution isn’t something they can call a “harm.”
Retain the vetting results for fundable proposals, too:
"Meanwhile, funders must repeat the review process already performed elsewhere, assembling new expert panels and recruiting external peer reviewers to judge the technical soundness of the research plans."
"Meanwhile, funders must repeat the review process already performed elsewhere, assembling new expert panels and recruiting external peer reviewers to judge the technical soundness of the research plans."
“Financial restrictions oblige funders to reject vast numbers of sound proposals. Why not retain them for other funders to consider?”
www.timeshighereducation.com/opinion/comm...
www.timeshighereducation.com/opinion/comm...
A common grant application database would cut the waste of resubmission
Financial restrictions oblige funders to reject vast numbers of sound proposals. Why not retain them for other funders to consider, asks Mikhail Spivakov
www.timeshighereducation.com
November 6, 2025 at 8:28 PM
Retain the vetting results for fundable proposals, too:
"Meanwhile, funders must repeat the review process already performed elsewhere, assembling new expert panels and recruiting external peer reviewers to judge the technical soundness of the research plans."
"Meanwhile, funders must repeat the review process already performed elsewhere, assembling new expert panels and recruiting external peer reviewers to judge the technical soundness of the research plans."
Thinking like a startup. Still in awe of what these metrics imply about the effectiveness of replacing inefficient legacy approaches like burning $$ on mailers or attack ads
Even if you only use the 93% of votes counted so far (1,036,051), the ratios don't change by much
Even if you only use the 93% of votes counted so far (1,036,051), the ratios don't change by much
Electoral thresholds surpassed:
104,000 volunteers
3,000,000 doors knocked
4,400,000 calls
1,100,000 votes (est.)
Ratios:
votes / volunteer ~= 10.6
(knocks + calls ) / vote ~= 6.7
2/5
104,000 volunteers
3,000,000 doors knocked
4,400,000 calls
1,100,000 votes (est.)
Ratios:
votes / volunteer ~= 10.6
(knocks + calls ) / vote ~= 6.7
2/5
November 6, 2025 at 8:02 PM
Thinking like a startup. Still in awe of what these metrics imply about the effectiveness of replacing inefficient legacy approaches like burning $$ on mailers or attack ads
Even if you only use the 93% of votes counted so far (1,036,051), the ratios don't change by much
Even if you only use the 93% of votes counted so far (1,036,051), the ratios don't change by much
Reposted by Paul Rohr
Advances in renewables and grid-level storage have just about made the issue of energy scarcity a solved problem. Now the real issue is energy equity and access. The problems now are politicial, not technological. Too many oligarchs are heavily invested in dusty relics of the past.
November 6, 2025 at 5:55 PM
Advances in renewables and grid-level storage have just about made the issue of energy scarcity a solved problem. Now the real issue is energy equity and access. The problems now are politicial, not technological. Too many oligarchs are heavily invested in dusty relics of the past.
"These are unprecedented numbers. [...] What all 104,000 of us have accomplished has rewritten the possibilities of mass democratic action."
-- field director Tascha Van Auken
1/5
-- field director Tascha Van Auken
1/5
November 5, 2025 at 7:08 PM
"These are unprecedented numbers. [...] What all 104,000 of us have accomplished has rewritten the possibilities of mass democratic action."
-- field director Tascha Van Auken
1/5
-- field director Tascha Van Auken
1/5
Reposted by Paul Rohr
Today, lots of pundits will write that last night confirms their priors. Ignore them all. The truth is that Americans are fundamentally good and when they vote, good candidates, who represent their values win. That's the basic bet of democracy and last night proved it
November 5, 2025 at 12:59 PM
Today, lots of pundits will write that last night confirms their priors. Ignore them all. The truth is that Americans are fundamentally good and when they vote, good candidates, who represent their values win. That's the basic bet of democracy and last night proved it
Reposted by Paul Rohr
A reminder that @chadkoh.com did a step-by-step write up of how you can join the ATProto #IETF124 BoF session at 9:30am EST this morning remotely. No cost registration.
November 5, 2025 at 12:19 PM
A reminder that @chadkoh.com did a step-by-step write up of how you can join the ATProto #IETF124 BoF session at 9:30am EST this morning remotely. No cost registration.