An "OTE" is AB-speak for a standard coil, one that turns off when the logic driving it opens up (or goes to 0).
An "OTL" is what AB (Rockwell) calls a latch. It gets set when the logic driving it goes hi (or 1, or true. However you like to express it). It will remain on until a separate rung containing an "OTU" (what we would call an unlatch coil on different PLC platforms) gets energized.
What makes things confusing (and potentially dangerous) is that in certain PLC platforms, you can put your unlatch rung anywhere you want in the program, or leave it out entirely if that's what you want to do. Or have any combination of OTL/OTU rungs in your program.
Or, put the unlatch coil in a different subroutine. Or, if you're really nasty, leave it out, and use an indirect reference to reset the coil. The opportunities for abuse are endless.
But what latches are useful for, are for "remembering" events should you lose power to your PLC. However, if you lose power to your PLC, I would say that you had bigger problems than the improper use of latches.
Well, the way that AB (and the GE 90-30's for example) do it they can easily be abused causing the problems that the other guys here are warning about. So they do have a point.
And I think that the asynchronous aspect of the I/O on the early AB PLCs (meaning that inputs show up in your program whenever they darn well please) led AB to create their Latch/Unlatch (OTL/OTU) coils this way.
So, there is reason for concern, but the good carpenter doesn't blame the tools, right?