> Thrift has been around since 2007, and is one of the first rpc protocols. Done reading there. Not sure if bad written or just not knowledgeable, please do your research.
Lessons from Building and Scaling Reddit's Ad Serving Platform with Go
121–130 of 132 posts
Re: Lessons from Building and Scaling Reddit's Ad Serving Platform with Go
#122Earlier quoted context omitted.
They were already using a third party service. It was slow and opaque, so they ditched it. At the scale of reddit, and because this is how they make their money , this is clearly something they want in house and something it would be a competitive advantage to control, if nothing else it gives them a much clearer conduit to their advertisers, and much greater control over things like ad selection.
I agree that it's important to keep the money-making component of a business free from external dependency to a large degree, but I think there's some liability that's being ignored here. A 3 person team built such an important piece of Reddit in a language they never used in production? I hope the team was at least experienced in ad-tech.
Re: Lessons from Building and Scaling Reddit's Ad Serving Platform with Go
#123Go-kit looks pretty nice, anyone else have experience with it?
It will not save you time typing: each route can have a transport request decoder and response encoder (encoder's are often shared), a request and response type, the actual business logic, a function that takes the request struct, applies it to the business logic function and returns a response object, a function that takes the previous function and connects it with the transport machinery (building the actual endpoint), and possibly middleware implementations.
However, by teasing all of these things apart from the very beginning it's very easy to come back later and add a circuit-breaker, or (as the article says) switch your transport from HTTP to grpc (or support both).
Re: Lessons from Building and Scaling Reddit's Ad Serving Platform with Go
#124Earlier quoted context omitted.
Where are people going to go this time? I feel like people usually copy their competitors which is why they consider redesigns worthwhile... but which Reddit competitor is Reddit copying here? I don't think there is one. Obviously Facebook/Instagram/Twitter/Snapchat kind of worry them, but then again, I feel like people are "done" with "social media" so I wouldn't be _really_ worried if I were Reddit. They have the a…
I mostly went to HN. My own migration was Slashdot → reddit → HN. I'm on lobste.rs too, which feels like it's in the growth stage. Fewer comments, but fewer bad comments still. HN is nowhere near reddit levels of decay, which is nice.
Re: Lessons from Building and Scaling Reddit's Ad Serving Platform with Go
#125Cool. Maybe these guys should get in touch with the redesign guys and get them fired. Also fire any executive that thought it was a good idea. Because all the hard work to serve ads more efficiently goes to waste when the site is a mess and unusable.
Some of the issues I have had with the new design is pretty amazing. For example, when you open a specific submission it's opened in a "lightbox" which basically contains all the comment. If you click outside of the lightbox it's automatically closed which I guess is OK. So some time ago they made a new release and broke it so that if you clicked on the scrollbar to scroll down in the lightbox to see what people had…
To put a cherry on top of it, the new layout now intersperses ads into the post lists and styles them to look like posts.
Additionally, ever since reddit decided to host it's own media, it no longer possible to directly link to a video or image. I don't want to send a link to my friend of comments about a gif. I want to just send him the damn gif. For this reason alone I'll continue to use imgur.
I'll be really bummed out the day I can't click "switch to the old layout", but I'm sure the time when they're removing that is coming.
It looks like we're finally seeing a series of decisions being made that caters stockholders instead of users of the site. They managed to hold out for this long, but I guess the day has come where it has finally happened. The big difference between reddit and all the other sites that have done this, is reddit used to be the hip place people went to get away from these kinds of user hostile moves.
The only question left now: is reddit so entrenched in it's position that it won't lose ground to a newcomer from the fallout of these decisions? Probably, but I know I'll definitely be keeping my eye out for competition that is gaining steam.
Re: Lessons from Building and Scaling Reddit's Ad Serving Platform with Go
#126Earlier quoted context omitted.
It seems biased. I had a similar situation with a client of mine. They went out looked at 3rd party companies, and everything looked decent on paper. Internally there were some napkin math and found out while the 3rd parties would save development time the returns of implementing features their customers were asking for now instead of waiting for 3rd party company paid for them to hire 3x the engineering staff the 3r…
> if they control everything and can drop a standard 90ms down to 10ms 10ms, after request processing, ad selection logic, and faster-than-light packet travel? I think that's far too optimistic of an improvement. As the post said, 30ms was the goal. My adserver delivers in 20-40ms. But since ads are routinely loaded asynchronously, is speed concern truly valid? The adserver can usually make a selection and deliver be…
I know it isn't the actual ad delivery itself but don't have public docs for that :(. Here is the ETL layer to back up claim: https://docs.google.com/presentation/d/13x_wdXA_C9ZymA2gcjsb...
Re: Lessons from Building and Scaling Reddit's Ad Serving Platform with Go
#127Earlier quoted context omitted.
> if they control everything and can drop a standard 90ms down to 10ms 10ms, after request processing, ad selection logic, and faster-than-light packet travel? I think that's far too optimistic of an improvement. As the post said, 30ms was the goal. My adserver delivers in 20-40ms. But since ads are routinely loaded asynchronously, is speed concern truly valid? The adserver can usually make a selection and deliver be…
Welp, guess I'm "Elon Musk" ;). I know it isn't the actual ad delivery itself but don't have public docs for that :(. Here is the ETL layer to back up claim: https://docs.google.com/presentation/d/13x_wdXA_C9ZymA2gcjsb...
Re: Lessons from Building and Scaling Reddit's Ad Serving Platform with Go
#128Earlier quoted context omitted.
The redesign is terrible for usability, but it does one thing very, very well: punish you for using an adblocker (sorry, "software blocker," lol). There are a few other dark patterns hidden in there too, all coincidentally pushing in a direction that's good for reddit. I suspect that's the real purpose of the redesign and that all complaints will fall on deaf ears until someone actually manages to disrupt reddit.
How does it punish you for blocking ads? Honest question.
Re: Lessons from Building and Scaling Reddit's Ad Serving Platform with Go
#129Earlier quoted context omitted.
It seems biased. I had a similar situation with a client of mine. They went out looked at 3rd party companies, and everything looked decent on paper. Internally there were some napkin math and found out while the 3rd parties would save development time the returns of implementing features their customers were asking for now instead of waiting for 3rd party company paid for them to hire 3x the engineering staff the 3r…
> if they control everything and can drop a standard 90ms down to 10ms 10ms, after request processing, ad selection logic, and faster-than-light packet travel? I think that's far too optimistic of an improvement. As the post said, 30ms was the goal. My adserver delivers in 20-40ms. But since ads are routinely loaded asynchronously, is speed concern truly valid? The adserver can usually make a selection and deliver be…
But yeah, it should be loaded async. Been a while since I worked with ad stuff, but I thought 10ms-20ms sounded more likely.
Re: Lessons from Building and Scaling Reddit's Ad Serving Platform with Go
#130Cool. Maybe these guys should get in touch with the redesign guys and get them fired. Also fire any executive that thought it was a good idea. Because all the hard work to serve ads more efficiently goes to waste when the site is a mess and unusable.
What are the actual complaints with the new design? I've been using it on and off and I can't say I've noticed any major issues. Though I'll admit most of my browsing is one on the mobile app these days.