Why do mainframe greybeards refer to DB2/zOS as "he"?

If you ask a DB2/zOS engine DBA a question about DB2's behavior, the DBA will refer to the DB2 engine as "he" much the way a sailor uses "she" to refer to his ship.

For example: "On开发者_如何学Cce you fill the freespace, DB2 still wants to keep those rows in cluster order in the tablespace. That's why he'll split that page in half, and you end up with lots of half-empty pages. That is, unless the cluster key of the row you've just inserted is the highest in the table, in which case he makes a new empty page, and he puts just your new row into that page. So I wouldn't have to do this REORG if you would just sort your input like I suggested."

Does anyone know where this tradition comes from?

---

This is certainly not a settled topic. There was a recent discussion on the DB2-L mailing list on this very topic, and Phil Grainger tallied the over 100 responses like this:

• 29% said "DB2 is an it"
• 25% said "DB2 is just DB2"
• 21% said "Definitely a he"
• 20% said "DB2 is a she"
• Other 5% claimed that the question had never occurred to them

---

I'm one of those grey beards. It's been a while since I worked on mainframes, but I recall this tendency to 'personalize' the programs.

It is in part an acknowledgement of the person who wrote the program. But primarily, it is just a tool of speech, serving as a reminder that the program is designed to work in a given way, for specific reasons.

And it is generally easier than describing the program as a lifeless autonomatic machine, in pure functional terms ( when this happens, it responds with that, etc ).