Twitter’s many problems often seem avoidable

Twitter regularly introduces new features—or feature updates—that are badly broken. Why is that? Why are they always acting first, thinking later?
UPDATE: Other people have noticed the same thing: “From top to bottom Twitter has made product mistake after product mistake, fundamental and obvious mistakes that have significantly confused and detracted from the simplicity of the service, for little or no gain.”

Minor Twitter update reverses a key security patch

Nearly a half-million Twitter user accounts were afflicted before Twitter “re-fixed” the security hole it had “unfixed” in the update. The official explanation. In other words, Twitter itself caused the security hole. This is the second time an exploit of this kind has been taken advantage of at Twitter. Then, two days later, a similar XSS attack hit Twitter.

Twitter changes usernames without consulting

An organization wanted to work with a Twitter user who had a name they had internationally trademarked. Their first step was to contact Twitter, who, without consulting either party, simply forced a name change. Neither the original user nor the organization was happy with Twitter’s heavy-handed approach.

Twitter’s OAuth implementation: a study in bad decisions?

Recently developer Ryan Paul detailed how Twitter “seriously botched its OAuth implementation and demonstrated, yet again, that it lacks the engineering competence that is needed to reliably operate its service.” Paul went on to say specifically that “Twitter should review the OAuth standard and take a close look at how Google and Facebook are using OAuth for guidance about the proper approach.”

Paul says Twitter’s approach is “a textbook example of how to do [OAuth] wrong.” He “received no response from Twitter after writing several posts outlining [his] concerns.” He points out that “The OAuth specification … says explicitly that implementors should not do what Twitter is trying to do.”

Moreover, Facebook and Google both do NOT take Twitter’s approach to the same situation. What’s an example of a kind of problem Twitter’s approach could cause? Hackers could put application makers (TweetDeck, HootSuite, et al.) “in a situation where their users are locked out for weeks when a key is compromised.” Paul demonstrates the problem by easily hacking a Twitter application. He also notes that Twitter’s approach to free and open source (FOSS) client software clients is even worse, a “really bad idea … because of Twitter’s misguided requirement.”

Beyond the problem with hacked applications, Paul points out there are a number of OTHER “bugs, defects, and inconsistencies that pose challenges for users and developers.” And this isn’t the first time Twitter has demonstrated problems in introducing new features:

Official tweet button slows sites…

Twitter first implemented its official buttons in such a way as to prevent or slowdown sites from loading if using the new Twitter button when Twitter is having problems. As reported by TheNextWeb, user @tschellenbach shows how you can modify Twitter’s code to prevent your site from being affected by Twitter problems. I have implemented @tschellenbach‘s change on this site.

…and crashes browsers:

More dramatically, within 24 hours of being released, Twitter’s new tweet button for websites was briefly crashing browsers like Firefox, as widely reported:

A brief problem is not much of  a problem, but who wants to have to keep checking to make sure the Twitter button isn’t crashing visitors’ browsers? Twitter acknowledged the issue in this tweet:

Twitter for iPhone crashes for a month

Most recently, Twitter for iPhone was updated…and version 3.0.2 promptly began crashing, failing to open, and having various other types of total failure. It took Twitter three weeks to provide a fix, which Apple took several days to approve. (The updated version is now in the app store here.)

Twitter for Android released with lots of bugs

Even as of early September these issues still were not resolved by Twitter:

1. Retweet works only from the home timeline and from nowhere else
2. The app force closes when scrolling the timeline
3. Cannot add @usernames in the Tweet. Says ”cannot get users names at this time, please try again latter” (later is misspelled as “latter”).
4. The “sync contacts” feature does not work.
5. Viewing private lists returns an error message.
6. Notifications don’t work and/or respect the notification volume.
7. Re-tweets from Private accounts can not be re-tweeted even if the original tweet is public.
8. Auto refresh does not work on Nexus ones and Droid with Froyo 2.1
9. Cannot delete search results on TwiAndroid

Facebook Blocks Twitter’s Way To Look Up Friends

This Facebook app made by Twitter was a big announcement, but as soon as they announced it…it stopped working. When the problem occurred, Twitter wasn’t in touch with Facebook for a couple of days. Eventually, Twitter posted

“Facebook has notified us that they have blocked the update to our application, and we are working on a resolution with them.” as reported on TechCrunch.

Why didn’t Twitter check with Facebook first, rather than later? The app turned out to violate a Facebook policy…that Twitter could have easily learned about beforehand.

Does Twitter have its act together?

This isn’t the first (or the second or the third) time they’ve announced a feature just before it stops working or has to be rolled back because it broke other things. Heck, even the second ad run by @EarlyBird had a typo in the discount code. Part of Twitter’s issue is admittedly that they are in the midst of an endless infrastructure change. The old joke about jumping off a cliff and having to build wings on the way down applies to Twitter pretty aptly. I agree Twitter can’t stop developing features that benefit its user base just because it has other teams working on infrastructure upgrades and problem resolution. But even though they have rolled a number of features out in limited release betas, they still end up too frequently having to go backwards due to issues found once they go into a full rollout. Also, as ReadWriteWeb points out, tongue-in-cheek: “Twitter has made numerous changes to fix its API. Those experiences have taught providers what mistakes not to make when launching a service.”

Is there a culture of carelessness at Twitter?

This is the company that got hacked because they let employees use passwords like “password” and “happiness” on important infrastructure—and the FTC went after them for being so easily hacked. Hopefully there is a key hire that has just not been made yet that will provide better oversight. Because to date, Twitter has a poor record. Since Twitter is already widely untrusted because the service is unreliable, you would think they wouldn’t introduce new situations that cause users to lose trust. But they at least appear to be avoiding problems like this one now: “Twitter timelines stopped updating hours ago. Why no word from the mothership?“