Daily discussion thread \ stacker news

Daily discussion thread

32 comments \ @saloon 15 Jul 2022 \ yesterday

This is a thread for random discussions that gets posted everyday at 5am central.

Tell us what you're doing today, ask questions, or vent about your life. Whatever you want, let it rip!

129 sats \ 1 reply \ @cointastical 15 Jul 2022

I see clicking on an image now expands it.

https://i.postimg.cc/nzq6ChKF/oh-wow.gif

Is that new, or have I just not noticed it before?

51 sats \ 0 replies \ @k00b 15 Jul 2022

It's new as of a couple days ago.

112 sats \ 2 replies \ @faithandcredit 15 Jul 2022

Blows my mind when people can’t find a use case for btc. They’re everywhere.

10 sats \ 1 reply \ @Brunswick 15 Jul 2022

If only we can replace email with a cost based messaging system where the recipient is paid by the sender, and the recipient chooses a programmable price profile I want SMTP to be as obsolete as the fax machine.

0 sats \ 0 replies \ @k00b 15 Jul 2022

Such a huge chicken and an egg problem though.

100 sats \ 0 replies \ @thedevil4k 15 Jul 2022

Today, i´m playing F122, this game has a lot of bugs, but the playability is great.

131 sats \ 11 replies \ @Brunswick 15 Jul 2022

@k00b another bad idea for your consideration: add selectable user trust levels and/or selectable popularity levels of new articles on the recent page. The goal being to see the most recent posts in order of age but not from new accounts. It might be interesting to have the opposite too, to only see articles from new/untrusted accounts to promote new users

139 sats \ 10 replies \ @k00b 15 Jul 2022

@kr and I discussed this today. On HN, new users are given green user names. We could do something similar, or as you suggest a filter, or even introduce ranking in recent which would mean /recent becomes something more like 'recentish but 0 votes'.

As a shorter term fix, we are going to probably just add an exponential cost for posting within a 10 minute window.

100 sats \ 6 replies \ @Brunswick 15 Jul 2022

My original reason was to filter out posts from new users when periodically checking recents throughout the day, but then go back and look over newbie posts once a day. The decaying exponential backoff/cost method (PI feedback) might solve the same problem of spam from newbies

0 sats \ 5 replies \ @k00b 15 Jul 2022

Ah, I see. How would you define a newbie?

100 sats \ 4 replies \ @Brunswick 15 Jul 2022

Based on your trust score. Something like a minimal threshold or gate to set them apart

0 sats \ 3 replies \ @k00b 15 Jul 2022

Cool. Should it apply to users with a trust of 0, or just anyone in the lowest 10th percentile, or 20th?

107 sats \ 2 replies \ @Brunswick 15 Jul 2022

That depends on what percentile of what. I dont know what the distribution looks like. I assume there are three or four groupings of scores and they aren't following a single gaussian distribution, probably at least three clumps of scores.

If I were to try to make it algorithmic, I'd probably try to collate the ranks into 100 bins, smoothe the bins through a moving average or high order polynomial fit, then search starting at a score of zero and look for a local minimum between new/inactive users and active users, and set the threshold there.

0 sats \ 1 reply \ @k00b 15 Jul 2022

Fancy. Okay, mostly trying to gauge how new/untrusted you meant to filter out

view replies

100 sats \ 2 replies \ @cointastical 15 Jul 2022

Is that just for Posts, or are you intending on doing that for multiple comments in a 10 minute window as well?

Also, ... if doing that even for just Posts, maybe add a setting like "Max fee per action", so that I can set 100 sats as the max that my comment or post would ever permit, (unless it's a tip where I manually enter the amount).

20 sats \ 1 reply \ @k00b 15 Jul 2022

It’d be just for posts.

Good idea on max fee setting

105 sats \ 0 replies \ @kr 15 Jul 2022

It could also be for self-reply nested comments, where it's just a single user replying to themselves 5 times in a row.

214 sats \ 3 replies \ @jimblogic 15 Jul 2022

I'm thinking about doing a "once a week" thread to encourage & coordinate people to do liquidity swaps between stacker.news users. "...multiple users like you open channels to each other. This setup results in a free incoming channel to you in return for you opening one to someone else..." we could use for example-> https://lightningnetwork.plus/

We could discuss channel sizes and who opens to who, setting fees to 0 to rebalance between everyone ideally and then back to normal, and so on. I wouldn't be able to afford opening channels all the time but would be happy to create a post every Saturday with some guidelines.

What do you guys think?

10 sats \ 0 replies \ @And1 15 Jul 2022

Sounds great, I like the idea!

105 sats \ 0 replies \ @Spartan 15 Jul 2022

Yeah this is a good idea. Make our own plebnet for SN.

5 sats \ 0 replies \ @k00b 15 Jul 2022

Great idea - at least worth trying to see if there's interest

100 sats \ 0 replies \ @gmd 15 Jul 2022

I've been having fun upvoting posts and comments to 69, but I wonder if this is discouraging more upvotes who don't want to ruin the sats total... 🤔 I think i'll start tipping more randomly from now on..

126 sats \ 2 replies \ @andyleroy 15 Jul 2022

Hi @k00b, happy Friday! I've noticed that links posted on Stacker News tend to quickly get indexed by google, and (at least for me) any inbound site links are displayed rather quickly to my google performance console. All of which is great - thank you.

I'm curious if you are doing any sort of programmatic addition of each post and its comments to search engines (or anything else special), to get indexed and help others with their inbound SEO?

125 sats \ 1 reply \ @k00b 15 Jul 2022

There's nothing special going on wrt to SEO - it's probably poorly optimized. We just have basic stuff like descriptions. I

Everything is server side rendered which supposedly helps, but I suspect 5+ years into heavy client-side rendering it makes little difference to the crawlers. Also, all the context is text and links which google loves.