An important aspect of making OpenOffice work correctly, is utilizing your computer resources wisely. When we talk about resources we're talking about computer memory(RAM) and CPU resources.
One of the indications that you need to manage your memory is when your system continues to get slower and slower using OpenOffice, maybe even locking up. Usually, your system memory, RAM, is what needs to be adjusted. To adjust your system memory go to Tools on the Menu Bar, Options and choose Memory under OpenOffice.org.
Each of these settings in the window has an impact on how OpenOffice works. The Undo feature allows you to simply choose undo if you change something in your document and then you decide that you made a mistake and you wnat to change it back to the former state. The number of undos is stored in memory so the fewer undos that you store the less memory you will need. The default is 20, but most people will only go back 3-5 changes so a setting of 10 is reasonable.
The Graphics cache is the next area that may need changes. The total cache, or available memory from RAM, for OpenOffice is 9 MB. This setting is really dependent upon how much RAM your system has. If you have a high powered machine and you have 750 MB of RAM and you are only running OpenOffice, then you can increase this amount with no problem. However, if you are running a system with 128 MB of RAM and you are running several programs at the same time you may even have to decrease this amount of memory. It is all dependent upon the machine and the number of programs that you have running. Basically, use the defaults unless either your system gives you all kinds of problems, then decrease system memory or if you have lots of memory to spare on your machine and you are using lots of graphics increase the numbers you see here.
The Memory per object setting allocates memory for each object. The default is 2.4 MB which may need to be increased if you are working with high quality images or decreased if you are not really using images and you need to save on memory resources.