Bob <[EMAIL PROTECTED]> sez: > > >I wonder if the problem was with the cameras or the CF cards since the >same CF card could be used in a Canon, a Nikon, a Cisco router, >etc..... Does the camera, or device that uses a CF card, just present >the file to the CF card? Each might have it's own way of formatting >the data so the logic in the CF card would have the job of knowing where >it is and where to put the next file..... I have never tried pulling a >CF card out of a Nikon and putting it in my 20D.
There's firmware in both. The CF or SD card has to look like an IDE drive, so that's its firmware. This means that it has to respond to block read/write/erase functions as defined in the IDE (now ATA) spec. The logic in the CF card can be either a general-purpose MCU with the necessary firmware, or an ASIC or gate array, with the 'firmware' in the synthesis rules. The firmware in the camera has to build a file system abstraction on top of that, just like any (other <];^) computer. The designer chooses the file system abstraction that is compatible with a favored type of host computer; namely, FAT32. So a substantial chunk of the camera's firmware is dedicated to that. Typically, an embedded firmware designer works on products in industry "A" with detailed knowledge in that domain. His company wants to invest in the 'family jewels' -- distinctive functionality in domain A. However, more and more, a product in any domain A needs functionality in domains B, C, etc; and many other quite different product domains do too. Examples of these supporting domains are Ethernet, WiFi, USB, Firewire, file systems, and operating systems. So the designer goes to the marketplace and aquires packages that give the functionality of domans B, C, etc. that his product needs. The packages have been written by (supposed) experts in each domain. They come in the form of libraries of software functions that his main code can invoke, once he figures out the (usually arcane) ways of doing so. The designer then tears out (what remains of his) hair getting them to work. He is constantly oppressed by the knowledge of Sodnagel's Second Law: "Given little enough time to do it, nothing is not impossible." [[ I'm in the embedded micro business (I hold a patent in techniques for making EEPROM behave like RAM). ]] * **** ******* *********************************************************** * For list instructions, including unsubscribe, see: * http://www.a1.nl/phomepag/markerink/eos_list.htm ***********************************************************
