I have some chips of this type that I have blowed someway the boot vector. The fact that Philips doesn't provide any information about parallel programming to the end user/developer is depressive.

If you install NoTouch losing the boot vector is not a problem so why worry about making a parallel programmer. The fact that so many know about NoTouch and do not install it and then, when it is too late, can not program their chips is depressing.

1) Other manufacturers (e.g. Atmel) have wisened up and dropped programming info from the data sheets for new chips. So no need to target just Philips. Someone, a while ago, posted a NoTouch for Atmel.
2) How on earth are you going to verify the endurance of programming realized by trial and error. The very reason for not including programming info is the endurance problems associated with kitchen table programmers..

Erik

NoTouch can be found at my user page.