4 ms·

Interesting Twitter Hashbang Bug

24 points by edent 15y ago

statictype 15y ago
This doesn't really seem like a bug to me. It's like permalinks of the form /a/b/434-some-text-representation-of-object-434 (or like stack overflow does: /a/b/434/some-text ) It's purely cosmetic and meant for search engines (as noted in the article) and easy readability.
p4bl0 15y ago
This is not relevant, has nothing to do with the hashbang, and is not really a bug. It's an SEO feature. Most websites do this, including big names like Amazon for instance: http://www.amazon.com/The-Ultimate-Hitchhikers-Guide-Galaxy/dp/0345453743/ http://www.amazon.com/The-Ultimate-Hitchhikers-Guide-Galaxy/... is the same as http://www.amazon.com/dp/0345453743 http://www.amazon.com/dp/0345453743 (and also the same as http://www.amazon.com/The-Holy-Bible/dp/0345453743 http://www.amazon.com/The-Holy-Bible/dp/0345453743).
gibybo 15y ago
I suspect Twitter is fully aware of this and decided it didn't matter when designing it. It's a fairly common pattern to make URLs more readable. I do it intentionally on a couple of my sites as well.
rmc 15y ago
For the record, Twitter are getting rid of the hashbang
- p4bl0 15y ago
  The hashbang is an entirely separate issue.
hcedric 15y ago
Of course it is a bug. The tweet id doesn't belong to the username in the url. It should return a Page not found error.
- scotth 15y ago
  Why do two lookups when you can do one? Keep in mind this is an edge case.
  - hcedric 15y ago
    No lookup required, you load the tweet with its id then compare the usernames.
    - gokhan 15y ago
      You assume usernames are kept along with tweets.
      - p4bl0 15y ago
        The username is on the page so it is retrieved one way or an other.
aoprisan 15y ago
how is this possibly a bug? this is done for SEO reasons, many sites do this.. also, this has nothing to do with hashes, very misleading title/traffic trap.
raverbashing 15y ago
IMHO For SEO and UX the current scheme is good. So they should keep using it. But the problem is thinking that "we use REST so every darn resource on the site has to follow this pattern even though it makes no sense to use it" I don't think twitter is going to filter by user to see if it matches the url. The key for the tweet is the tweet_id and that's the key to their (hugely cached) DB
loceng 15y ago
I imagine it’s likely used for tracking what links / tweets you were the originator of sharing? Perhaps in the future you coud have fun with that by associating certain content with people’s Twitter account names. Not sure what affect it would have though. :)
tkellogg 15y ago
You can do this at github too. Really confused me the first time I saw a commit referenced for a user that didn't have the commit yet...

24 points by edent 15y ago

This doesn't really seem like a bug to me. It's like permalinks of the form /a/b/434-some-text-representation-of-object-434 (or like stack overflow does: /a/b/434/some-text ) It's purely cosmetic and meant for search engines (as noted in the article) and easy readability.

p4bl0 15y ago

This is not relevant, has nothing to do with the hashbang, and is not really a bug. It's an SEO feature. Most websites do this, including big names like Amazon for instance: http://www.amazon.com/The-Ultimate-Hitchhikers-Guide-Galaxy/dp/0345453743/ http://www.amazon.com/The-Ultimate-Hitchhikers-Guide-Galaxy/... is the same as http://www.amazon.com/dp/0345453743 http://www.amazon.com/dp/0345453743 (and also the same as http://www.amazon.com/The-Holy-Bible/dp/0345453743 http://www.amazon.com/The-Holy-Bible/dp/0345453743).

gibybo 15y ago

I suspect Twitter is fully aware of this and decided it didn't matter when designing it. It's a fairly common pattern to make URLs more readable. I do it intentionally on a couple of my sites as well.

rmc 15y ago

For the record, Twitter are getting rid of the hashbang

p4bl0 15y ago
The hashbang is an entirely separate issue.

hcedric 15y ago

Of course it is a bug. The tweet id doesn't belong to the username in the url. It should return a Page not found error.

scotth 15y ago
Why do two lookups when you can do one? Keep in mind this is an edge case.
- hcedric 15y ago
  No lookup required, you load the tweet with its id then compare the usernames.
  - gokhan 15y ago
    You assume usernames are kept along with tweets.
    - p4bl0 15y ago
      The username is on the page so it is retrieved one way or an other.

aoprisan 15y ago

how is this possibly a bug? this is done for SEO reasons, many sites do this.. also, this has nothing to do with hashes, very misleading title/traffic trap.

raverbashing 15y ago

IMHO For SEO and UX the current scheme is good. So they should keep using it. But the problem is thinking that "we use REST so every darn resource on the site has to follow this pattern even though it makes no sense to use it" I don't think twitter is going to filter by user to see if it matches the url. The key for the tweet is the tweet_id and that's the key to their (hugely cached) DB

loceng 15y ago

I imagine it’s likely used for tracking what links / tweets you were the originator of sharing? Perhaps in the future you coud have fun with that by associating certain content with people’s Twitter account names. Not sure what affect it would have though. :)

tkellogg 15y ago

You can do this at github too. Really confused me the first time I saw a commit referenced for a user that didn't have the commit yet...