(Suggestion) Rating System

Idk if here is the correct section but:
Why don't ports have rating system so new users will know more popular ports and applications to use?
 
Assuming rating system as a difference in enhancement of design(mainly being better coded), it will always have a place to exist.
Since BSD is a license for freeware and shared sources, developers can work uniquely to improve the better one rather than starting a work that is already done in another software.
That is something already happening without us want, like let's say Blender is hands down best graphic designing software.
 
Take an email server for example. Which one is best, sendmail, exim, Postfix...? This depends entirely on how familiar you are with a product and on your use cases. All three are excellent products. Should you rate all three of them five stars out of five?

If you are new and need a mailserver, you should search the web, read articles on the topic, do some research, install all three of them on different test systems and decide yourself which you like the most.
 
Why don't <...> have rating system so new users will know more popular <...> to use?
Mainstreaming populism is the plague of the mostly uneducated masses.
Scoring and rating are the tools for keeping the mainstream mainstream.

Knowing what others prefer does not tell anything about on what state of mind a vote has been given and if the voter changed his mind meanwhile.

Popular things/votes have their very own risks and side effects and may be a good hint or even reason for avoiding/following just that.

Following on what others do may be convenient but it is a childish behavior that works out if a benevolent nanny is around.
 
I like it. Reminds me of google's play store.


It gives the developper a motivation to make a good product, like, get a high rating.
The number of ratings would show how popular an application is.

Maybe you should suggest this to freshports.org website.
 
The ratings are not actionable. If I go to for example the postfix developers and tell them that their rating is poor; they are going to tell me to fsck(8) off.

Again, the rating system for the port maintainers also seems unfair; they are already giving up enough of their free time to keep some software alive by a single thread.

Whilst I believe a staging area that is accessible for all users (such as Arch Linux's AUR or OpenBSD's jasperla/openbsd-wip) might have some merit (including feedback), I don't feel rating systems are appropriate for anything other than quick hype apps.
 
Looks like the TL;DR generation has made it in the FreeBSD forums like everywhere. They look on a statistics like on a MEME. They take it for given and might be the believers who will adapt to AI in lack of their own intelligence.

I admit that most people on earth won't even notice that statistics have biases and when told they even won't know what that means.

Reading yourself and thinking yourself protects from being deceived.

In history the elderly were attributed to have the most problems reading the small letters. Now the younger have that too, but it's not their eyes what causes problems, it's that what might be found behind.

http://bsdstats.org/ said:
This sample represents users of the given BSD operating systems that opted in to install a data collection program. It is not representative in any way of the total population of users/installations of the given BSD operating systems.

But those who made it to read this are required to start thinking.

Now ask yourself why OpenBSD is highly significant on the last rank?
Hint: look on the bold letters in the cite.
I won't give the answer here but encourage to start thinking yourself. Maybe you get healed from praising and demanding such metrics for unreflecting use.
 
Take an email server for example. Which one is best, sendmail, exim, Postfix...? This depends entirely on how familiar you are with a product and on your use cases. All three are excellent products. Should you rate all three of them five stars out of five?

If you are new and need a mailserver, you should search the web, read articles on the topic, do some research, install all three of them on different test systems and decide yourself which you like the most.

Well, yes, all of those to have 5 stars, and then decide on one of them for yourself, and for e.g. qmail you'd have 3 stars only, which would warn you that you shouldn't really try that option.

(To the qmail lovers: that's just an example ;) )
 
Why would I want to know what's popular? I don't run programs because of their popularity but rather because of their functionality.

OTOH, perhaps I should get a department store computer with MS-Windows pre-installed and sign up for Facebook. ;)
 
Why would I want to know what's popular? I don't run programs because of their popularity but rather because of their functionality.

OTOH, perhaps I should get a department store computer with MS-Windows pre-installed and sign up for Facebook. ;)

Woe to the fallen functionality of facebook: fictional account creation! Hackers fixed that bug quick! Also: dude, you're gettin' adele.
 
new users will know more popular ports and applications to use
We should not use a rating system for that but what about some kind of newsletter or a page with tips and tricks or something like that. In a newsletter you can learn about new and interesting projects (applications/ports) and on a tips and tricks page you could provide some good-to-know-about applications that deserve more attention.

My example of email servers would be a good reason against the rating system: you must give all of them five stars. But a tips & tricks page could let users know about (new) and obscure programs like nload(1). Then again, I learned about this tool (and other new and interesting projects) on the BSDnow podcast. In my opinion this is far superior to rating each and every project in the ports tree.
 
Why don't ports have rating system so new users will know more popular ports and applications to use?
I don't see any technical advantage to implement this into the ports system itself. It's already quite complex, no need to complicate things even further. Especially if there's no technical gain, what use is a rating system with regards to building and packaging ports?

You are, of course, free to set up your own website a la Freshports and include a rating system.
 
Think about the rating of Hotels and Restaurants in the Internet. While it was helpful in the very first beginning, it has become an ecosystem for deceiving the inexperienced new or staying dumb users with the benefit of being able to aggregate metadata and profiling and tracking users.
Also abusing such systems for revenge and making bad the local competitors is still common. As if that were not enough commercial influencers work hard to bias anything on earth.

Long story short: There are sheep and those who cut and slaughter.
 
Why don't ports have rating system so new users will know more popular ports and applications to use?

As discussed here, this is likely useless for users, but it may be useful for developers who decide to stop supporting legacy ports. The rating systems can be implemented as simple counter of installations w/o the need to manually rate ports.
 
It is the other way round, you design a rating system because you need information to help you make decisions for a specific purpose, so you also know exactly which criteria must be evaluated and how.

Having a rating "a la Amazon" is helpless: some people give 5 stars just because the product was delivered on time, others give only one star because they don't know how to use the product... Amazon tries and improve it with additional sub-ratings on specific aspects, not always relevant, meaningful or even applicable to the evaluated product.
 
Don't forget about the rampant abuse of pretty much all rating systems.
 
  • Thanks
Reactions: a6h
Back
Top