Wednesday, September 10, 2008

Reinstate user account in Mac OS X Leopard

A note for me on how to reinstate a user account in Mac OS X Leopard that I deleted without creating an account dmg but just a saved user home folder.

If you delete the home folder then the chances of recovery is nil.

Once you delete the account with saved home folder, the home folder is saved in ~/Macintosh HD/Users/user (Deleted)

Now, 
sudo mv "user (Deleted)" user

and get back to the system preferences and create an account with the same user name. Leopard will inform you that a folder of that name exists and will offer you a chance to use that folder as home folder for the new user. Accept it. Voila you get back the old user.

Saturday, January 19, 2008

Google database

Ref: research@google
One of data management issues that plagues anyone new to creating, saving, remote accessing and analyzing data is the confusion between what is the difference between microsoft excel and SQL and related such softies. Often times they tend to think that excel is for beginners and SQL is just the same but high end highly programmable sort of spreadsheet`ing'. This is not absolutely untrue but only 1/100 of the truth. This confusion primarily causes issues for scientists who deal with huge datasets, they often just store them in excel spreadsheets. SQL both server and on machine versions (like SQLite) can store the same datasets that the scientists want stored. Sometimes they can even very efficiently serve it and with a little bit of programming can serve specific bits and pieces of particular characteristics (say, data on a certain day or time, data exceeding or below certain values). SQL based databasing can also automatically collect data from user input on web or automated inputs from instruments. Researchers also want the data analyzed numerically (statistically). And they tend to believe that since it cannot be done directly on SQL databases, that such databasing is useless (useful only for businesses) for them and they prematurely decide to adhere to excel. If one had put in some more effort, he or she might notice that there is something called odbc (usually an intimate component of the SQL programming - I call it SQL command line). Google it and you will know that using odbc one can move SQL datasets between batabases and statistical analysis softwares such as R, Stata or even excel.

Now I notice that google is working towards a scientific database management tactic. Read about it at research.google.com. I am am also so sure that they will let us use a SQL like language to manage the datasets so we can move it between their storage space and google spreadsheet or excel or any statistical software. This would be a tremendous boost to researchers around the world. I am excited because now I can place my raw datasets online and code it such that users around the world will see it the way I want it seen. Cool job Google.