Comment on Suggestions for speeding up Alt tech adoption, the GETTR way
iamtanmay@wolfballs.com 2 years ago... you could save $$ if you only save wolfball user comments and votes on synced posts, but fetch the post itself from Reddit.com
Something like:
-
On website load, fetch posts from a default or user curated list.
-
Check if there are comments/votes from wolfballs users. Add them on top of the fetch
-
Present to user seamlessly, making it look like everything is hosted on wolfballs, but actually fucking Reddit.com over and saving nice $$
-
Profits???
masterofballs@wolfballs.com 2 years ago
Think I would just make a bot account, scrape reddit once a day for top post, post to wolfballs with bot account.
iamtanmay@wolfballs.com 2 years ago
Reasonable I think. Right now its still a small site.
That approach would start getting expensive if wolfballs became really big, and you'd have to pull and host TBs of data. But by that point, you'd probably also have a revenue plan to make the $$
masterofballs@wolfballs.com 2 years ago
It would be a small amount if data. 99% of reddit is garbage. Its only the top post on specific sub's that would matter.
iamtanmay@wolfballs.com 2 years ago
Aahhhh. Very clever my friend. Yeah, that would work for most subs like sports, movies etc
The subs I was mainly a member of where Coding subs. There, the reader would want access to all the posts not just the top ones, because almost all posts had some code snippets or techniques explained.
sj_zero@lotide.fbxl.net 2 years ago
While that would save you disk space, I think you'd end up more than making it up in bandwidth and CPU utilization. Every time someone visited the page the server would run to reddit and grab the posts you're looking at, then serve them to the user. Very quickly, a few terabytes of storage would become cheap compared to the bill you'd rack up doing all that extra processing to add some posts.
iamtanmay@wolfballs.com 2 years ago
I was thinking that the code of the wolfballs.com itself would instruct the user's browser to fetch the content from Reddit, i.e. the bandwidth and compute come from the end user's browser, not from a central server.
It would be a very cheap and efficient way to mirror huge amounts of content from all public websites, like Reddit, Wikipedia, Youtube, Twitter.
Only Facebook can't be mirrored like this, because they gatekeep their content from anonymous internet - you need a Fbook account to see what's there.
Using my suggested method, you would have a minimal server usage, only to get any comments/votes that wolfball users have posted on the synced content.