Both Madelyn and Zhao are terrific programmers. In case you wonder, they are not random people put there for reasons, like for example since Madelyn is a woman so there was some pressure to add she. She deserves this 100% and is one of the top Redis contributors of the last times, with deep technical knowledge in basically every aspect of Redis. The same for Zhao, he is simply impressive, can track bugs about complex conditions like almost nobody can. I'm happy to see them aboard. Btw they were both effectively part of the informal "core team" that was forming spontaneously in the recent times.
> In case you wonder, they are not random people put there for reasons, like for example since Madelyn is a woman so there was some pressure to add she
but like you say, these aren't nobodies. However, I am curious to see how the pace will change (if at all), with you no longer doing the lion share of work.
Interesting move, putting “SLA-driven” infrastructure engineers (i.e. people working at public clouds to deliver Redis-as-a-Service solutions) in charge of the project. (Not just these two from AWS and Alibaba Cloud, but the rest being from Redis Labs.)
It seems like, for one reason or another (Redis Labs because they want to sell their “core plus proprietary modules” service; the clouds because they want to push you to use their other products for use-cases that fit them better), all the new leadership has reason to want to not add any more developer-facing features to Redis Core. I expect the Redis Core you see right now, is the same Redis Core we’ll have 10 years from now, client-API wise. Redis, as a USP, is “done.”
Unless, that is, some third-party comes in with their own polished feature PR, and pushes really hard for it. In other words, Redis is kind of moving to the Linux kernel model, where “new use-case out of nowhere” features come in mostly not from internal development, but rather from external contributors petitioning the core-maintainership priesthood with reasons their patch should be upstreamed.
Either way, there’ll certainly continue to be plenty of ops-staff-facing innovations, bug-fixes and polish. That’s what gets this new core team up in the morning. I’m sure people running Redis in production are happy about that.
> I’m sure people running Redis in production are happy about that.
I sympathize with the main points you were making, but surely ‘people running Redis in production’ are the core use case for Redis, so ultimately this move seems good for the project’s overall direction.
One of my favorite bits of trivia: What would Git be like if Linus stepped away? Oh yeah, he stepped away after a few months and Junio Hamano continued running it for the next 15 years[1] (!!).
As with Yossi and Oran vis-a-vis Redis labs, it is not clear whether Madelyn and Zhao are in the core team in their personal capacities or as representatives of Amazon and Alibaba respectively, we know that Redis labs reserves the right to name replacements for Yossi and Oran, is the same true for Madelyn and Zhao and their current employers ?
antirez|5 years ago
miked85|5 years ago
Kind of a strange thing to call out.
the-dude|5 years ago
https://www.merriam-webster.com/dictionary/terrific
sdesol|5 years ago
https://imgur.com/QUPER9q
https://imgur.com/RnlzECa
but like you say, these aren't nobodies. However, I am curious to see how the pace will change (if at all), with you no longer doing the lion share of work.
rurban|5 years ago
derefr|5 years ago
It seems like, for one reason or another (Redis Labs because they want to sell their “core plus proprietary modules” service; the clouds because they want to push you to use their other products for use-cases that fit them better), all the new leadership has reason to want to not add any more developer-facing features to Redis Core. I expect the Redis Core you see right now, is the same Redis Core we’ll have 10 years from now, client-API wise. Redis, as a USP, is “done.”
Unless, that is, some third-party comes in with their own polished feature PR, and pushes really hard for it. In other words, Redis is kind of moving to the Linux kernel model, where “new use-case out of nowhere” features come in mostly not from internal development, but rather from external contributors petitioning the core-maintainership priesthood with reasons their patch should be upstreamed.
Either way, there’ll certainly continue to be plenty of ops-staff-facing innovations, bug-fixes and polish. That’s what gets this new core team up in the morning. I’m sure people running Redis in production are happy about that.
imron|5 years ago
I sympathize with the main points you were making, but surely ‘people running Redis in production’ are the core use case for Redis, so ultimately this move seems good for the project’s overall direction.
asperous|5 years ago
1. Those orgs succeed when redis succeeds. So they want to see it used more which means more features.
2. Developers think for themselves and aren't always "double agents".
3. The redis community still exists and open source communities are often very vocal about directing products
swagonomixxx|5 years ago
He was the heart and soul of the project.
It also makes me think, what would Linux be when Linus decides to step away? I kind of don't want to think about it.
Thank you to Antirez for all your hard work on Redis.
johnfn|5 years ago
[1]: https://github.com/git/git/graphs/contributors
seemslegit|5 years ago
yjftsjthsd-h|5 years ago