• Zooville Critical Notice, For all Users!
  • User Permission and Privilege Changes (PMs and Forum Layout): Sept. 20th, 2025: Please read the following announcement: CLICK HERE!
  • Update, October 4th, 2025: The ability to create NEW groups in the social groups area of the forum has been disabled. Click the hyperlink in the notification above for more information.

1000 follower limit

And all Im saying is there are not a thousand people here with anything of value to say. You've been here a month. You just havent seen it yet.
 
OwnThatKinkGirl said:
It's a number they picked out of the air. I can say that as a fact. But I can also say that you're 100% right.
Click to expand...
It actually could have a reason for it though.
For example each ignored user adds a record into the database.
So each user of a large forum may have a 1000 database lines just for the ignore list and another 1000 just for the follow list in addition to anything else.

The forum has 261,851 members.
If all of them had 1000 ignored users, that is 261,851,000 database records.
 
pes said:
It actually could have a reason for it though.
For example each ignored user adds a record into the database.
So each user of a large forum may have a 1000 database lines just for the ignore list and another 1000 just for the follow list in addition to anything else.

The forum has 261,851 members.
If all of them had 1000 ignored users, that is 261,851,000 database records.
Click to expand...

First off, if I'm wrong, I want to know I'm wrong.

OK. You have a point, Generally, I'd expect a database table to have a unique identifier for each user, and the structure for ignored users could be as simple as a table of records of integers where each integer is of a known size, like an 8 bit int or a 16 bit int or a 32 bit int. Each one takes a known size to store and predictable amount of time to fetch in standard SQL. For example, if we used a 32 bit int, that would allow 4,294,967,296 unique values to be created so I assume we would not have that many users created in the lifespan of the site. Or we could just use a string of a known length as the type of the data for each record. We can use text encodings of UTF-8 (2 bits ber character) or UTF-16 (taking 3 bits per character to allow for things like Russian cyrilic text). Since I assumed we are just using numbers, we can just use UTF-8 encoding. If you're still with me, I'm just trying to show that I am aware of what we need to be cautious about. That said,1000 records is nothing.

I've done tests like what we are talking about 10, 20 years ago, but let's get current.

As I thought about this, I remembered that I was a lead on the database migration for a large healthcare company out of California for the entire state of Iowa. This database has many databases in it and a total of 2 million+ users. You saw above the detail, rigor and care I took when considering how to create the fastest most space efficient data structure for the table's records. You and I am speaking the same language.

That said, the table record values in this uber database of multi databases (9 I think), each containing many tables, simply used flat out plain old text records as their internal data type for almost everything. No attention at all to optimization for size and speed. None. Now, get this. We had a mirror of the production uber database that was good enough to the current one with all unique customer data replaced. A crapload of records. We could all run it locally on the computer we were working on, a consumer laptop. Our Mac or PC laptops with an SQL server app. Or on another server elsewhere. No special equipment needed. And we just pointed our app at it. And it was fine with regards to performance.

Back to your point, yes! And thank you for the detail. I needed to know that before making asumptions for now and for what the user count will be in the future. The next step is to know how many users are online at once and how many queries are there coming to the load balancer at once. My assumption here is the exclusion of values is being done as a filtering database operation at the time of the query, like so.

SELECT columns FROM table WHERE test_column [NOT] IN (value1, value2,...);

That is the rough SQL. This is a common operation in databases. I use it every day when my Netflix show list loads because of all the shows I have added to the parental control filter. But there are other ways to look at it. One of us could code a rough test up ourselves (naah) but ZooVille's website runs on Xenfora. Xenfora offers a developer install!! One that lets a customer install their own SQL server and web server. Either way, we could try out either option and auto generate test users and increase limits to see if we can make it fail. We could.

Or we could ask Google and Chat GPT.

what is a reasonable limit to the size of an SQL filter
or
creating large sql filters

The links returned give us insight into related issues.

If I remove one value from the IN clause the query succeeds after 24 seconds. For this incredibly simple query, 32767 is the maximum number of values that still allows the query to execute.

I see a hard limit on 32767 values but this suggests a way.

Every SQL batch has to fit in the Batch Size Limit: 65,536 * Network Packet Size.

You already found Erland's article on the topic of passing lists/arrays to SQL Server. With SQL 2008 you also have Table Valued Parameters which allow you to pass an entire DataTable as a single table type parameter and join on it.


Not assuming a join, we could have this.

SELECT columns FROM table_xxx WHERE desired_guids NOT IN (SELECT exclusion_ID FROM ignore_table);

And viola!

There are options out there and room for improvement. IMO, 1000 values seems small. I may need to try it ib SQL for real.

Fun. Let's keep talking. If I'm wrong, I want to know I'm wrong.

Cheers
 
saddlebum66 said:
And all Im saying is there are not a thousand people here with anything of value to say. You've been here a month. You just havent seen it yet.
Click to expand...
And I start off knowing that. I am the newb here, no matter what else I have done in life. I can only hope to become worth listening to over time.

For me, I'm more interested in filtering/ignoring what I don't want. Same type of list. But this is merely a mental exercise for me. Everything I've seen in software and recently with exclusion filters good and bad in Amazon Prime and Netflix say to me that 1000 should be no problem - depending on how many queries to the database are made at once.

Your words are noted. I hear ya.
 
I got curious and checked my profile. I'm not following anyone; I am ignoring a very small number of users (single digits). They were rude, obnoxious people who would not leave me alone and/or kept messaging me after I asked them not to. Plus another one who threatened me and has since been banned.

Beyond that, I simply don't waste time reading their posts.
 
What do I want with 1000 people following me? Are we on TikTok or Instagram here? I would rather vote that a follow request has to be confirmed before someone can follow me. There are users I don't want to be followed by...
 
This solves the problem:
0.jpg
Settings -> Privacy

They can follow you, but will not get any notifications or updates.
All following will do is add their icon to the list.
 
The first to follow 1001 users will be awarded a Virtual Trip to Bangladesh at their own expense.

And now a little more seriously.

1) If you follow 1000 people, then you care what they write. Which means you regularly read what they post. To follow the posts of 1000 people every day, you'd need a full-time job as, say, a "forum reader". And for 1001 or more, you'd have to regularly take overtime. That leaves open the question of who is going to give you a salary. Mare Lover 1975, pes, dogluver101, do you pay salaries? :unsure:

2) ... or are you just another virtual sucker who follows every sexy avatar pic?
 
UR20Z said:
Your own words and behavior.



I don't start shit. I have been known to finish it, however.


Huh?

I know it's difficult for you, but try to use your words. WTF is "ome.tv", and what does it have to do with "hashing out" anything?
Click to expand...
Well your wrong. How are you going to finish it? It's a website check it out. Convince other people to try stuff with a pet. I wanna talk to you on there.
 
You must log in or register to reply here.
Back
Top