From using a vpn, I find that youtube adverts are different between countries.
I find the adverts aimed at Panama to be least displeasing, they get adverts for South American soft drinks that I'v not heard of before and real estate agents that talk rapidly in Spanish.
Personally, I find adverts in a language I don't understand at all to be more tolerable.
I figure that I'm being nice by not using an advert blocker.
You have asked for minimum development work so ...
"Write your email address legibly on a piece of paper and mail it with a US $1 bill to $ADDRESS."
Obviously 1) it will cost people 68¢ for a stamp, 2) won't really work for people outside the USA and 3) if you only sell fifty copies it's not a wildly silly suggestion.
My view on web forum spam from my experience a few years ago, so possibly slightly out of date:
Once a web forum has existed for six months the flood of spam will be greater than a single person can manually moderate without giving up or going crazy. You can't check your forum every eight hours every day for the rest of your life.
The stopforumspam.com list of spamming ip addresses is useful to reducing abuse to a manageable level. You need it fully automated. You want to use the api, downloading a copy of the blocklist every 24hours would be somewhat less effective as new ip's start spamming constantly. Their list of email addresses that are being used to to to create accounts on every web forum in existence can be useful.
There are boxes at large VPS providers that have been constantly attempting to post automated spam on every forum and wiki that they can find for years. Sending complaints to the hosting companies does nothing.
There will be occasional collateral damage from CGNATs getting blocked.
Every commercial VPN server and tor exit will be blocked and occasionally someone will whinge about that.
Restricting new accounts to only being able to post in one area of a forum and telling people they have to write a message introducing themselves and manually approving people seems to work reasonably well. If you are one person trying to run a web forum on your own then you will still find that you don't want to check it every day for the rest of your life.
Captchas can help but you need ip blocking as well. You often see the load of the signup page and the submission of the form from ip's in different countries, presumably captcha solving services.
My opinion is that for a web forum to exist in a useful state for a decade the minimum requirement is to have half a dozen volunteer moderators who visit regularly. (or it's a company forum where part of someones working day is checking the forum.)
The metal cables that held up the tower have been laying on the ground for long enough that plants have grown over them then died in the winter.
Three foot tall plants have grown in front of the door of the equipment cabin that would have been trampled down if anyone removed the radio equipment recently. The cabin looks like it has been derelict for years.
The electricity meter is missing and the place where it should been looks like it has years of corrosion.
It is a muddy field with no signs of recent footprints or gouges that would indicate the tower was removed recently.
Anyone looking at the state of the site would think that the medium-wave AM station has been off the air for years.
The radio station had an FCC license that required they operate an AM transmitter. Now that the FCC has found out that they are not broadcasting on AM , the station is not allowed to continue on FM.
I used to read a forum about Visual Basic programming.
A recurring theme was "I put an X byte message into a TCP connection and I have written the software at the other end with the assumption that all X bytes will arrive at the same time every time ..."
If the message is under 1500 bytes it usually works until it doesn't.
If ten thousand people have a copy of the source code for bash 5.1.4 then do you count that once or ten thousand times?
If you only count each version once then do you only count official releases or do you increase the count by 500MB when someone corrects a one character typo in their copy of the linux kernel source on their laptop hard drive?
How do you count the size of the git repositories that can check out a million different versions but only store the differences?
I think I have some very old CD-ROMs of compressed debian source code, does the compressed or uncompressed size count?
As far as I can tell from googling for a few minutes, in this context "RCU" is a linux kernel thing for arbitrating different threads accessing the same memory using Read-Copy-Update method.