Microsoft has a lot to be proud of. They helped pioneer desktop computing. They have very successful platforms like SQL Server and Sharepoint. They are even working on a new mobile phone that doesn’t support copy and paste. But, I would argue their most popular product in use today is Excel.
Mayhaps ‘popular’ isn’t the best word here. Most reasonable people will probably agree that Excel is the predominant number crunching tool in the marketplace. When you last asked someone to send you some data, what format did you receive it in? Uh-huh.
When it comes to the database, people like to take data out of it. A lot. Every successful tool I have used for any database platform has
Knowing that people are doing this every day must seriously freak out anyone who is responsible for data integrity and security. Once data leaves the database, it’s almost impossible to protect. Let me make a very important distinction here – I am talking about production systems. If you are in R&D or Quality Control, you have little to worry about.
An emerging trend I have encountered while working with public sector accounts in the US is that users are being prohibited from saving data permanently outside the database. Writing to Excel is no longer a whiz-bang feature. Instead, it is something to cause one to worry about their job security.
I think it’s only a matter of time before this becomes the new de facto standard.
Thankfully there are many ways around this new paradigm. For one, you could easily crunch data in an instance of Excel, and not save the file to disk.
Do me a favor, go to your Documents folder and do a quick count of how many XLS/XLSX files you have.
Would you or your employer be nervous if your machine was lost or stolen? Many sites are going to the extreme of encrypting every laptop in an organization. I predict that will become the norm in the not too-distant future.
Why are people insisting on using Excel?
I think the most obvious reason is they are comfortable with pivot tables, excel macros, and the like than they are with SQL constructs like GROUP BY and relational algebra and group theory. These folks probably shouldn’t even be in the database to begin with. We wonder why their queries aren’t the most spectacular, well it’s no wonder. They have been placed in a position to grab data from a database, and no one has made sure they speak the language.
There’s a reason people write SELECT *, because they can. Steve Jones and I can talk till we’re blue in the face, but users will find the quickest way to their end goal, and damn the consequences. We can’t really blame them either. Business users and developers have tons of work to accomplish with inadequate resources, budget, and timeline.
What we can do is continue to educate the community and always be ready to give great advice when a user asks for help.
Speaking of those R&D folks that are using Excel to upload data to the database…
Save those files as a CSV and build a SQL*Loader control file to load your data to Oracle. You’ll pick up a new trick for your toolkit, and when you get to those 1,000,000 spreadsheets, the database will churn threw that data faster than you can imagine. And as @patch72 and @leight0nn reminded me yesterday, check out an External Table while you’re at it.