9/18/11

Gecko Backup

I am going to explain a few things about my Gecko Backup service I have started back in 2009. I just moved everything over to Gina Trapani's ThinkUp as I had no time to fix up my code and ThinkUp seems like it's getting better by the day. In the move, I tried my best to keep all data from my old system but because my old system didn't store everything there are some things that won't be carried over in the old posts, such as my bug with the posts being turned into scientific notation within PHP and re-tweet counts. I have done my best in the transition and everything appears to be fine.

Here are a few things to take note in my transition.
1. I will still allow people to join my crazy idea of backing up people's twitter accounts and providing access to their data until I decide to close and send out everyones data to them so they can move to another service or to their own setup for ThinkUp.
2. The system only crawls twitter every night at 12AM CST.
3. I will create a backup of everything once a week and store it in 2 places.
4. I will update ThinkUp once a week after backing up.
5. If I get tired from work and decide not to backup or update, it's up to me and not you.
6. If I forget, please feel free to tweet me about it and I'll try to do it when I can.
7. I promise not to use your data for anything except for maybe a view of stats on everyone at once and not specific data to you. Example, see how many tweets contain the word bacon.

mysql> SELECT COUNT(*) FROM `posts` WHERE `post_text` LIKE '%bacon%';
+----------+
| COUNT(*) |
+----------+
| 125      |
+----------+
1 rows in set (7.75 sec)

8. I promise not to sale your data.
9. I will be deleting follower data once a week and only for the previous week as there is a lot of it.

mysql> SELECT COUNT(*) FROM `follows`;
+----------+
| COUNT(*) |
+----------+
| 1226321  |
+----------+
1 rows in set (0.09 sec)

mysql> DELETE FROM `follows` WHERE `last_seen`<'2011-09-11';
Query OK, 130415 rows affected (72.86 sec)

mysql> SELECT COUNT(*) FROM `follows`;
+----------+
| COUNT(*) |
+----------+
| 1095906  |
+----------+
1 rows in set (0.01 sec)

10. I hope to work with you on any issue you may have.
11. My server is 733MHz so please be patient with it when loading the page.
12. If you want your account removed from my server, let me know.
13. I do support backing up FaceBook as well if you would rather do that than Twitter and will probably support backing up Google+ in the future as APIs become available.

If you want to use it, please visit http://u.webra.in/GeckoBackup. My server will default to my twitter account MrGeckosMedia and anyone else who makes their stats public to everyone if your not logged in. You can login and register at the top left of the screen and if you cannot register you can ask me for a code to register.

Hope this explains things for all of you and will allow you to enjoy keeping your tweets searchable and not lose them.