Why don't we use short urls anymore?

[HoangDinh]

Legacy: https://mangadex.org/title/24530

Now: https://mangadex.org/title/0cfadedb-99cf-4fe5-b117-dc624da32225

It's all hash man 😔

 

[kekesed97]

And, uhh, why don't we make it shorter then? Like, /t/24530?
What? You accidentally opened hentai from my code?

 

[HoangDinh]

???🤫

 

[HoangDinh]

The urls are unlike nhentai but more like e-hentai

 

[Bartolumiu]

Legacy: https://mangadex.org/title/24530

Now: https://mangadex.org/title/0cfadedb-99cf-4fe5-b117-dc624da32225

It's all hash man 😔

It has a lower computational cost on the longer term (I think) and also with hashes, the access speed is significantly faster than with a linear list...
Hash is O(1) usually in terms of speed, as it doesn't have to check every entry while linear lists are O( n ) as you don't know where the entry will be located

 

[RogueKitsune]

because the previous way allowed for iteration of resources. Which can be a problem when you don't want everyone and their dog scraping the entirety of your site wasting your network and compute resources. Also to a lesser extent there is a security aspect of being able to iterate over all the users of the site.

UUIDs mostly solve this, though as you can see they are not very "human friendly" but that was considered an acceptable trade off. Also UUIDs have an advantage of being a mostly built in supported data type not needing a bunch of code to make a custom unique id implementation.

 

[Manko-sensei]

It clearly works better in terms of perfomance, but for a person who bookmarks like myself the old way it meant changing a whole bunch manually so they didn't stop working. Some months ago it happened to me, and the only answer I got from the staff was that the title page was working properly but I was questioned for using legacy when they stopped using it. Basically, adapt, even I had to change from XP to Windows 7 and then to 10 so I could keep working.

 

[Teasday]

UUIDs are not hashes

 

[HoangDinh]

Which can be a problem when you don't want everyone and their dog scraping the entirety of your site wasting your network and compute resources

this