A problem with the desktop-oriented printer support in Unix and in Windows is that it has the underlying assumption that the person at the desktop should be in unfettered control of the ability to print (with lip service paid to an overall page quota). Where this breaks down for children is that . kids love tangible output, so unless someone does something about it, they will print everything. . in a school the printer may be somewhere else than in the classroom, and kids aren't allowed to roam the building picking up their output. So even more than printer users in an office environment, they have no clue as to why things don't show up as "printed", or why the print queue is long. (Could be a printer jam. Could be a graphics-intensive prior job, or their own job being graphics-intensive, could be an unwitting print of a 30 page job on a slow printer, etc., etc.) Just as in the "hit the launch button again" phenomenon when activity launching is slow on the XO, this leads to printing the same document again and again. I'd like to show you the scars from the arrows in my back after spending a day a week in elementary school classrooms over a period of several years, but it would scare the children. So here is a scenario that I think would be better, especially in an environment such as a third world school with few printers and scarcity of paper/toner/ink. Vamsi's default pdf printer is great...no consumables, no remote print server, nice way to preserve the visual effect of a piece of work for posterity. But for printing real output, I think a more draconian control system is needed. One possibility: submit pdf files to a Moodle-mediated "turn it in" queue. This ensures that an artifact has been produced that the child can review and is motivated to review before giving it to the teacher to decide if it warrants printing. The teacher can then browse the turned in work, printing those examples s/he thinks warrant the expense. Should it be possible to configure a normal CUPS printer queue where kids could "just print it"? Yes, there are probably circumstances where this is appropriate. But even where kids make the decision about what to print, I think having some kind of per-document filter, such as "has this document already been printed in the last n hours?", "is this document longer than x pages?", "has this child submitted more than y pages in the last time period" is crucial. And so it would be good to build that into the journal print interface. I don't think the ordinary unix print quota mechanism does this, but I admit I haven't looked at it in detail recently.