Ben
@bengamble7.bsky.social
Tech sommelier @ Ververica
Keeper of ducks, Builder of games
Mostly blind🦯 (its not a hiking stick),
He/Him,
All opinions my own
Keeper of ducks, Builder of games
Mostly blind🦯 (its not a hiking stick),
He/Him,
All opinions my own
Using Librdkafka in @ziglang has been completely doable, but it's causing me enough issues that I want a cleaner lib...
Time to build a zig native kafka sdk.
Will open source when I get the last few APIs implementation (queues)
Time to build a zig native kafka sdk.
Will open source when I get the last few APIs implementation (queues)
February 14, 2026 at 4:18 PM
Using Librdkafka in @ziglang has been completely doable, but it's causing me enough issues that I want a cleaner lib...
Time to build a zig native kafka sdk.
Will open source when I get the last few APIs implementation (queues)
Time to build a zig native kafka sdk.
Will open source when I get the last few APIs implementation (queues)
The fact that the new EU Digital Operational Resilience Act (#DORA)
Semantically overlaps with Devops Research and Assessment (#DORA) metrics is going to cause a lot of AI driven confusion.
The naming doesn't help😅
Semantically overlaps with Devops Research and Assessment (#DORA) metrics is going to cause a lot of AI driven confusion.
The naming doesn't help😅
February 13, 2026 at 10:02 AM
The fact that the new EU Digital Operational Resilience Act (#DORA)
Semantically overlaps with Devops Research and Assessment (#DORA) metrics is going to cause a lot of AI driven confusion.
The naming doesn't help😅
Semantically overlaps with Devops Research and Assessment (#DORA) metrics is going to cause a lot of AI driven confusion.
The naming doesn't help😅
I know databricks has revenue, but their current funding drive is kinda crazy. I wonder who they are buying up next....
February 10, 2026 at 10:53 AM
I know databricks has revenue, but their current funding drive is kinda crazy. I wonder who they are buying up next....
AI can do so many things, but it's ability to do ascii art on command line tools can not be under stated
February 6, 2026 at 2:21 PM
AI can do so many things, but it's ability to do ascii art on command line tools can not be under stated
I'm starting to use claude code differently as I run into the same issues again and again. The biggest one is file size bloat... It's totally fixable by hand but The AI can't deal with 6k lines of code...
I'm calling it a Smithers loop, as I'm giving it a different context each time we loop
I'm calling it a Smithers loop, as I'm giving it a different context each time we loop
February 5, 2026 at 9:36 PM
I'm starting to use claude code differently as I run into the same issues again and again. The biggest one is file size bloat... It's totally fixable by hand but The AI can't deal with 6k lines of code...
I'm calling it a Smithers loop, as I'm giving it a different context each time we loop
I'm calling it a Smithers loop, as I'm giving it a different context each time we loop
Writing a sql query engine is a compiler. Except the language is mostly pre-defined, the users are mostly other tools and everyone thinks it's a solved problem (use postgres)
February 5, 2026 at 9:18 AM
Writing a sql query engine is a compiler. Except the language is mostly pre-defined, the users are mostly other tools and everyone thinks it's a solved problem (use postgres)
Today's challenge :
Teach an AI how to use kcache grind.
So far my AI Dev tools can use dtrace, valgrind and perf. But we need to go deeper....
Performance is one of the few ways to reduce TCO cleanly so performance is everyone's problem.
Teach an AI how to use kcache grind.
So far my AI Dev tools can use dtrace, valgrind and perf. But we need to go deeper....
Performance is one of the few ways to reduce TCO cleanly so performance is everyone's problem.
February 4, 2026 at 9:52 AM
Today's challenge :
Teach an AI how to use kcache grind.
So far my AI Dev tools can use dtrace, valgrind and perf. But we need to go deeper....
Performance is one of the few ways to reduce TCO cleanly so performance is everyone's problem.
Teach an AI how to use kcache grind.
So far my AI Dev tools can use dtrace, valgrind and perf. But we need to go deeper....
Performance is one of the few ways to reduce TCO cleanly so performance is everyone's problem.
There's a new dresden files book.
It's one of those: "am I still reading it because it's good, or because I'm invested".
I enjoyed the earlier books at the time as a fun ride through a magical power fantasy, but on reflection I do wonder a bit.
Still bought it, but this is the last auto buy
It's one of those: "am I still reading it because it's good, or because I'm invested".
I enjoyed the earlier books at the time as a fun ride through a magical power fantasy, but on reflection I do wonder a bit.
Still bought it, but this is the last auto buy
January 30, 2026 at 1:54 PM
There's a new dresden files book.
It's one of those: "am I still reading it because it's good, or because I'm invested".
I enjoyed the earlier books at the time as a fun ride through a magical power fantasy, but on reflection I do wonder a bit.
Still bought it, but this is the last auto buy
It's one of those: "am I still reading it because it's good, or because I'm invested".
I enjoyed the earlier books at the time as a fun ride through a magical power fantasy, but on reflection I do wonder a bit.
Still bought it, but this is the last auto buy
I've been writing a database from scratch over the last few months with claude code and it's definitely an experience.
The most telling thing is that it really doesn't know anything you don't tell it.
It's a good faster keyboard but it's even remotely good at stuff with lots of moving parts
The most telling thing is that it really doesn't know anything you don't tell it.
It's a good faster keyboard but it's even remotely good at stuff with lots of moving parts
January 30, 2026 at 9:18 AM
I've been writing a database from scratch over the last few months with claude code and it's definitely an experience.
The most telling thing is that it really doesn't know anything you don't tell it.
It's a good faster keyboard but it's even remotely good at stuff with lots of moving parts
The most telling thing is that it really doesn't know anything you don't tell it.
It's a good faster keyboard but it's even remotely good at stuff with lots of moving parts
An enduring CS memory: a uni computer lab, ~2007.
I was failing at printf and realising how weird it is.
Nearby, students a year ahead were learning Java. I overhear: “wait… that’s printf.”
Turns out one of them had implemented ALL of printf in Java—rather than just learning how println works...
I was failing at printf and realising how weird it is.
Nearby, students a year ahead were learning Java. I overhear: “wait… that’s printf.”
Turns out one of them had implemented ALL of printf in Java—rather than just learning how println works...
January 27, 2026 at 10:10 AM
An enduring CS memory: a uni computer lab, ~2007.
I was failing at printf and realising how weird it is.
Nearby, students a year ahead were learning Java. I overhear: “wait… that’s printf.”
Turns out one of them had implemented ALL of printf in Java—rather than just learning how println works...
I was failing at printf and realising how weird it is.
Nearby, students a year ahead were learning Java. I overhear: “wait… that’s printf.”
Turns out one of them had implemented ALL of printf in Java—rather than just learning how println works...
Claude code makes CMake only the worst thing in my day once a week rather than every day I work on a Cpp project.
January 25, 2026 at 11:28 PM
Claude code makes CMake only the worst thing in my day once a week rather than every day I work on a Cpp project.
It's weird what software tests can and can't catch. Especially with AI.
Unit coverage always feels like a vanity metric, as compilers and types should do most of that for you.
When it comes to feature level testing, you can end up catching all kinds of odd behaviour with randomised inputs
Unit coverage always feels like a vanity metric, as compilers and types should do most of that for you.
When it comes to feature level testing, you can end up catching all kinds of odd behaviour with randomised inputs
January 23, 2026 at 12:02 PM
It's weird what software tests can and can't catch. Especially with AI.
Unit coverage always feels like a vanity metric, as compilers and types should do most of that for you.
When it comes to feature level testing, you can end up catching all kinds of odd behaviour with randomised inputs
Unit coverage always feels like a vanity metric, as compilers and types should do most of that for you.
When it comes to feature level testing, you can end up catching all kinds of odd behaviour with randomised inputs
First trip of the new year! I'm in #NYC for the week, talking data, AI and all things distributed!
We're also hosting a meetup (link in next post) on Thursday so please come and say hello, or message for a coffee!
We're also hosting a meetup (link in next post) on Thursday so please come and say hello, or message for a coffee!
January 13, 2026 at 4:21 PM
First trip of the new year! I'm in #NYC for the week, talking data, AI and all things distributed!
We're also hosting a meetup (link in next post) on Thursday so please come and say hello, or message for a coffee!
We're also hosting a meetup (link in next post) on Thursday so please come and say hello, or message for a coffee!
The new year started out on a deep freeze and then getting sick.
It's been very un fun.
It's been very un fun.
January 7, 2026 at 12:14 PM
The new year started out on a deep freeze and then getting sick.
It's been very un fun.
It's been very un fun.
I'm in the wierd space of evaluating the costs of mutexs vs CAS.
Most of my time is spent removing mutexs, with CAS being the replacement. But you really shouldn't have too many spin locks...
Currently working back through my assumptions of what and where to be fast....
Most of my time is spent removing mutexs, with CAS being the replacement. But you really shouldn't have too many spin locks...
Currently working back through my assumptions of what and where to be fast....
December 21, 2025 at 11:49 PM
I'm in the wierd space of evaluating the costs of mutexs vs CAS.
Most of my time is spent removing mutexs, with CAS being the replacement. But you really shouldn't have too many spin locks...
Currently working back through my assumptions of what and where to be fast....
Most of my time is spent removing mutexs, with CAS being the replacement. But you really shouldn't have too many spin locks...
Currently working back through my assumptions of what and where to be fast....
The main reason I actively hate c++ is dispite all the issues with it, there's no limit to perf you can get out of it.
Even with the horrible ergonomics, the foot guns, the bonkers approach of the committee and the bazantye over engineering everyone seems to commit to, you can go so fast.
Even with the horrible ergonomics, the foot guns, the bonkers approach of the committee and the bazantye over engineering everyone seems to commit to, you can go so fast.
December 14, 2025 at 11:36 PM
The main reason I actively hate c++ is dispite all the issues with it, there's no limit to perf you can get out of it.
Even with the horrible ergonomics, the foot guns, the bonkers approach of the committee and the bazantye over engineering everyone seems to commit to, you can go so fast.
Even with the horrible ergonomics, the foot guns, the bonkers approach of the committee and the bazantye over engineering everyone seems to commit to, you can go so fast.
With uber publishing cool ideas around moving to flink and IBM playing catchup by scooping up confluent, here's my take : www.linkedin.com/pulse/being-...
LinkedIn Pulse
www.linkedin.com
December 14, 2025 at 10:56 PM
With uber publishing cool ideas around moving to flink and IBM playing catchup by scooping up confluent, here's my take : www.linkedin.com/pulse/being-...
With the CFPs for current london and bengaluru running, wondering what to submit:
Would ppl prefer to hear about AI & memory, consensus and replication, performance and compute, apache Fluss and data lakes, a how to on common usecases, or something entirely different?
Would ppl prefer to hear about AI & memory, consensus and replication, performance and compute, apache Fluss and data lakes, a how to on common usecases, or something entirely different?
December 8, 2025 at 1:16 AM
With the CFPs for current london and bengaluru running, wondering what to submit:
Would ppl prefer to hear about AI & memory, consensus and replication, performance and compute, apache Fluss and data lakes, a how to on common usecases, or something entirely different?
Would ppl prefer to hear about AI & memory, consensus and replication, performance and compute, apache Fluss and data lakes, a how to on common usecases, or something entirely different?
Reposted by Ben
Yeay... oomed my macbook again...
Seriously need to upgrade at some point 😅
Seriously need to upgrade at some point 😅
December 1, 2025 at 4:43 AM
Yeay... oomed my macbook again...
Seriously need to upgrade at some point 😅
Seriously need to upgrade at some point 😅
Some plane rides later, I'm working with our partners @factorhouse.io in #melbourne!
Time to talk about Streaming data and drink good coffee!
Time to talk about Streaming data and drink good coffee!
November 27, 2025 at 4:31 PM
Some plane rides later, I'm working with our partners @factorhouse.io in #melbourne!
Time to talk about Streaming data and drink good coffee!
Time to talk about Streaming data and drink good coffee!
The killer phrase in dangerous data systems :
Business critical spreadsheets 😅
Business critical spreadsheets 😅
November 26, 2025 at 7:51 AM
The killer phrase in dangerous data systems :
Business critical spreadsheets 😅
Business critical spreadsheets 😅
After another dig through an #agentic framework without a memory manager, here's my blog explaining why you need one, with links to an open source one I wrote :
www.linkedin.com/pulse/memory...
www.linkedin.com/pulse/memory...
LinkedIn Pulse
www.linkedin.com
November 26, 2025 at 1:27 AM
After another dig through an #agentic framework without a memory manager, here's my blog explaining why you need one, with links to an open source one I wrote :
www.linkedin.com/pulse/memory...
www.linkedin.com/pulse/memory...
Claude opus 4.5 is out? Not with a bang but with a whimper
I think we're deep into diminishing returns of brute frontier model performance at the moment, but we'll see...
The gemini release was wildly underwhelming so I've got minimum expectations
I think we're deep into diminishing returns of brute frontier model performance at the moment, but we'll see...
The gemini release was wildly underwhelming so I've got minimum expectations
November 24, 2025 at 9:05 PM
Claude opus 4.5 is out? Not with a bang but with a whimper
I think we're deep into diminishing returns of brute frontier model performance at the moment, but we'll see...
The gemini release was wildly underwhelming so I've got minimum expectations
I think we're deep into diminishing returns of brute frontier model performance at the moment, but we'll see...
The gemini release was wildly underwhelming so I've got minimum expectations