FAQ for the 'Twitter Group'

[RShankar]

From Dr. Shankar: How does twitter decide that someone with a 'celebrity' hashtag is really that celebrity?

[maugus33]

Question from 'Paypal' group to 'Twitter' group: We noticed that you mentioned that the Streaming API falls back to the REST API if the Streaming API ever fails. We were wondering under what condition would the Streaming API ever fail? Maybe an example?

[deniseygrace]

Answer from the 'Twitter' group to the 'Paypal' group:

Since Streaming APIs are live updates of tweet data, a fallback to REST may be needed when one's WiFi connection is unstable. The example that Twitter provides is that the REST API fallback could be used when a user's connection changes to a mobile network connection. To summarize it all, REST API fallback is used when connection speeds are low or if there are problems connecting.

Sources: https://dev.twitter.com/streaming/userstreams https://twittercommunity.com/t/user-site-streams-vs-falling-back-to-rest-clarification/8308/2

[Jnesbeth]

From "Github" group to "Twitter" group: Since Twitter updated it's privacy policy and the person signing up for Twitter can opt out of being tracked. Is it still possible to still track and get information from the private user that opted out of being tracked?

[gbstoner1]

Answer to 'Dr Shankar' from 'Twitter' group. Twitters uses a few processes for verifying account identities. The proactive process is media companies reach out to Twitters media relations department and will literally provide spreadsheets of various Sports, or Celebrities that are clients. They will also use verification companies like gracenote.com. On the reactive side of things they will scan accounts with famous names and do a manual verification process which sometimes involves seeing if the account is linked to that famous person, group, company etc has tied twitter to their official website. http://www.gracenote.com/press/03/14/2012/ https://support.twitter.com/articles/119135-faqs-about-verified-accounts

[gbstoner1]

Answer to the 'GitHub' group from 'Twitter' Group. By default now "location services" in twitter is turned off. As a general rule running APIs on a users tweets you would not be able to track them but there are some exceptions. Poorly written or nefarius 3rd party apps can override your privacy settings in twitter when you post something using a 3rd party app/website. Also hashtags and some keywords in your tweets can give away your location without even having geotagging even turned on.

[aasthadave]

Question from 'Netflix' group for 'Twitter' group. Concerning privacy, how would you improve security leaks when other developers are using the Twitter APIs? Are there any additional security measures you would introduce to the platform?

[Magicperf]

Question from "Amazon Cloud" to "Twitter" group: If Twitter service went down as a whole, how much will it impact the companies that depend on its online functionality? Such as some companies you mentioned in your presentation today.

[adrianas92]

Question from 'NOAA' to 'Twitter' group: Hows does twitter store the data related to all the millions of tweets post it everyday? Do they have just one system or distinct systems to store the transition and history of tweets?

[deniseygrace]

Answer from the 'Twitter' group to the 'Amazon Cloud' group:

If Twitter service went down as a whole, services using Twitter's social networking services such as Twitter Kit, the REST APIs, and the Streaming API would obviously stop working. However, these are not the only services Twitter owns. As you can recall, Twitter owns Fabric and all its APIs. Here are some possible scenarios that could happen to each of Fabric's services if Twitter went down:

Crashlytics - No more data would be collected, but the apps using Crashlytics would remain intact. Companies who formerly relied on Crashlytics would probably have to move to another similar crash reporting system.

Digits - People would not be able to sign in to their accounts for applications using Digits. Companies who formerly relied on Digits would have to find another way to let their clients sign in to their accounts.

Mopub - Without the real-time bidding system of MoPub, companies would not be able to bid for ad space or sell ad space. Because of this, some ad revenue may be lost. This service is arguably the most valuable and would do the most damage if lost.

[gbstoner1]

Answer to "Netflix" from "Twitter" Group. OAuth 2.0 would be a good start.

1. SSL is required for all the communications required to generate the token.
2. Signatures are not required for the actual API calls once the token has been generated -- SSL is also strongly recommended here.
3. Once the token was generated, OAuth 1.0 required that the client send two security tokens on every API call, and use both to generate the signature. OAuth 2.0 has only one security token, and no signature is required.
4. It is clearly specified which parts of the protocol are implemented by the "resource owner," which is the actual server that implements the API, and which parts may be implemented by a separate "authorization server." That will make it easier for products like Apigee to offer OAuth 2.0 support to existing APIs.

[deniseygrace]

Answer from the 'Twitter' group to the 'NOAA' group:

Twitter has its own database architecture, Manhattan. The main goal for their database is "strong consistency service," which is a fancy way to say that the same operation from the same user should always have the same result. Also, they have optimized their database to compensate for large amounts of tweets/second (6,000+) with their timeseries counter service.

If you so desire, there are entire blogs dedicated to it and a quick search of "Twitter Manhattan" will give you all the details you would ever want to know about it.

Sources: https://gigaom.com/2014/05/12/3-lessons-in-database-design-from-the-team-behind-twitters-manhattan/

https://blog.twitter.com/2014/manhattan-our-real-time-multi-tenant-distributed-database-for-twitter-scale

[GiBiT]

Question from 'Facebook' Group 9 to the 'Twitter' group:

How does Twitter handle celebrity accounts? I know many celebrities have personal twitter accounts but are they assigned a normal twitter page or is it different (notifications, privacy, messages, etc...)? Is there a verification/validation process to prove it's the official account of that celebrity?

[Bainshinds]

Question from 'Amazon Retail' Group to 'Twitter' Group: What do you recommend that Twitter do for improvement.