If you break down the general "don't" into the reasons, you might get a better view of the topic.
The "no kitchenware" applies mainly on EPROM-based devices, where the programming voltages were applied directly to the EPROM cells and accuracy of this voltage and the exact timing had huge impact on future reliability of the programmed device. This has been changed in the FLASH-based devices, where programming voltage and timing is generated internally, with little left to spoil.

Partially, the bad reputation of "kitchenware" programmers is also due to the SPI-like ISP devices (such as the AT89S-series) programmed using the "rusty nail" approach. Ignoring peculiarities of clocked signals over longer (read possibly bouncing and crosstalking) cables, the badly defined PC-LPT port, problems with mixed voltages, inadequate power supply (e.g. power derived from COM port), problems with timing under Windows etc., plus the fact that some manufacturers simply cannot get the internal reset working properly (and the FLASH gets corrupted at powerdown just after the programming has finished), lead to the bad reputation of kitchenware, too.

The third "no"-reason for kitchenware is the lack of support... :-)

Nevertheless, if one is aware of all these problems, and has at least some experience with modern electronics, and is willing to invest slightly more time and money into it (not just to get the "rusty nail"), kitchenware can be get right and might be a valuable tool for hobby-grade work. I have not heard negative reports on the controller-based kitchenware programmers here nor elsewhere, except for the lack of support.

Some links here: http://www.8052.com/forum/read.phtml?id=124534
Peter Dannegger's programmer: http://home.tiscali.de/peterd/to.../index.htm


Well, that's a quite prohibitive reason... :-)


P89LPC92x by NXP (if you can live with 2 extra pins... :-))
The gotcha is, they are 3V only. Plus they might not be available in the shop over the corner...

Jan Waclawek