URL shorteners

[swansont]

We've been having some discussions and have concluded that URL shorteners aren't consistent with a good user experience; too often they mask a spammer's attempt to get you to visit a (possibly malicious) site that you wouldn't be inclined to visit if you could see the actual link.

So: don't use them here. We have no post-length limits that might justify them. We will assume that new people using them are spammers and react accordingly. If you have a track record, we'll delete the link and give you a chance to modify your habits.

[Peterkin]

What is a URL shortener?

[StringJunky]

@PeterkinIt shortens the url to only contain the basic address. As far as justifying them goes, they limit tracking information that ad servers can use when one is site hopping.

[Peterkin]

25 minutes ago, StringJunky said:

@PeterkinIt shortens the url to only contain the basic address. As far as justifying them goes, they limit tracking information that ad servers can use when one is site hopping.

Well, I kind of puzzled out what it does, but how do I know whether I'm using one? 

[Genady]

2 minutes ago, Peterkin said:

Well, I kind of puzzled out what it does, but how do I know whether I'm using one? 

If you copy and paste the URL address from your browser, then you are not.

[Peterkin]

Thanks! So, it's still okay to superimpose it ^^ per link function into the text of a post?

  Just don't want to fall afoul by accident.

[Genady]

1 minute ago, Peterkin said:

Thanks! So, it's still okay to superimpose it ^^ per link function into the text of a post?

  Just don't want to fall afoul by accident.

Yes, it is.

[TheVat]

You can shorten URLs, without removing the identifying stuff, if they are larded with junk after the "?" symbol.  I do that occasionally when there's a mass of tracking stuff and formatting specs that aren't needed.  I agree the tinyurl site, and others, should be avoided.  

 

[swansont]

33 minutes ago, Peterkin said:

Well, I kind of puzzled out what it does, but how do I know whether I'm using one? 

If you are getting the link from the site, and not using a URL shortener service (e.g. bitly or tinyurl, etc.) then you are fine.

IOW, if you can read the site's name in the URL and that's where it's taking you, there is no problem. 

[Peterkin]

Thank you all! I have been clueless all this time.

[swansont]

1 hour ago, StringJunky said:

@PeterkinIt shortens the url to only contain the basic address. As far as justifying them goes, they limit tracking information that ad servers can use when one is site hopping.

It's not this - it changes the address to where it doesn't include the original. 

 

e.g. tinyurl.com/ye29ahp7 will take you to this thread. But how would you know that without clicking on the link? (I've omitted the https:// so it's not a clickable link) There's no indication of the actual site's name. with www.scienceforums.net/topic/126922 you know where you are going, and can make an informed decision about whether to click.

Such a service has its utility, but it's inappropriate to use here.

[Sensei]

1 hour ago, StringJunky said:

As far as justifying them goes, they limit tracking information that ad servers can use when one is site hopping.

?? You're thinking about query-string arguments?

 

Domain A (e.g. the JS script hosted on it) creates cookie/supercookie if it is not present ("session id").

 

Website X uses the code from domain A (cookie/supercookie is set to initial value).

X has link to Y.

The user goes from X to Y (a reference URL is sent with it, so Y knows who referred it).

Website Y does not use the code from domain A.

The user goes from Y to Z (a reference URL is sent with it).

Website Z uses the code from domain A (cookie/supercookie is already sent).

 

"A" knows that you visited X, Z directly from analyze of the logs.. because the same cookie/supercookie is found..

 

[StringJunky]

42 minutes ago, swansont said:

It's not this - it changes the address to where it doesn't include the original. 

 

e.g. tinyurl.com/ye29ahp7 will take you to this thread. But how would you know that without clicking on the link? (I've omitted the https:// so it's not a clickable link) There's no indication of the actual site's name. with www.scienceforums.net/topic/126922 you know where you are going, and can make an informed decision about whether to click.

Such a service has its utility, but it's inappropriate to use here.

Gotcha, cheers. Yes, that is very dodgy for normal forum linking. Never used one like that. I  use a url shortener but it doesn't do that. It removes the extraneous additions to the site url that cause redirects through adservers so they can't track a user. Actually, it's called a 'url tracking stripper' that I use... it's effect is to shorten it but that's not the addon's actual aim. I used the wrong description, it turns out.

Edited March 28, 2022 by StringJunky

[Sensei]

1 hour ago, TheVat said:

You can shorten URLs, without removing the identifying stuff, if they are larded with junk after the "?" symbol.

Part after "?" is called query-string. https://en.wikipedia.org/wiki/Query_string

1 hour ago, TheVat said:

I do that occasionally when there's a mass of tracking stuff and formatting specs that aren't needed.  I agree the tinyurl site, and others, should be avoided.  

...if you received an "interesting link that you want to share with someone".... you should ALWAYS remove tracking data from the query-string, whether you send it to a family member by mail, text message, or provide it on a forum in raw form or shortened by URL shorteners.

[Peterkin]

2 hours ago, Sensei said:

you should ALWAYS remove tracking data from the query-string,

Do you leave the question mark?

[iNow]

No

[Sensei]

4 hours ago, Peterkin said:

Do you leave the question mark?

Depends.

 

If you have a link such as:

https://www.example.com/article?id=1234&uid=19182737774902&[.......]

The id is the article identifier, which is an essential argument to the script, so it cannot be removed.

So after deletion, the new link will look like:

https://www.example.com/article?id=1234

 

If you have a link such as:

https://www.example.com/article/1234?uid=19182737774902&[.......]

The id of the article is integrated with path part (which might be just an illusion given by rewrite-engine)

So after deletion, the new link will look like:

https://www.example.com/article/1234

Should be okay.

 

After clearing the URI/query-string portion before sharing, try the new URI string in a private browser session to see if the link still works.

 

Edited March 29, 2022 by Phi for All
requested correction

[Peterkin]

Thank you. This is all new to me; I've never done anything but copy and paste. 

[Sensei]

9 hours ago, Peterkin said:

Well, I kind of puzzled out what it does, but how do I know whether I'm using one? 

A list of URL shorteners can be found here (incomplete list):

https://www.google.com/search?q=URL+shorteners+list

Basically, the idea is to have as few characters as possible in the domain name (e.g., instead of google.com it's goo.gl) and as few characters as possible in the path (i.e., immediately following the hostname are letters and numbers that make no sense to humans).

 

Full URLs are detailed because of the way search engines work.

Therefor webmaster SEO (Search-Engine Optimizations) tend to put important search-engine keywords into URLs. Like the full title of the article with all the details of the thread being discussed.

e.g. website selling electronics, 3d printers with brand XYZ, with model ABC, want to generate URL with format like:

https://www.example.com/products/3dprinters/XYZ/ABC/

(which is usually handled by rewrite-engine https://en.wikipedia.org/wiki/Rewrite_engine at HTTP web server level, and converted to internal query-string)

 

Edited March 29, 2022 by Sensei

[Peterkin]

Thaqnks. I'll try to keep my inexpert fingers out of the machinery.