To: Mike Winn who wrote (8124 ) 12/31/1997 2:41:00 PM From: Bonzo Read Replies (2) | Respond to of 31646
Mike thanks for taking the time to post some of your research on TPRO and Embedded Systems. My apologies if my initial response was less than cordial. There have been too many of late who have bashed TPRO without a single element of truth or fact. I was formerlly in Electronic Test for 25 years (having recently sold to my partner). Functional Test Programming (worked for GenRad years ago), application support, sales and marketing etc. My experience had been primarily focused at the Circuit Board (CCA) level and not at the system level. Having read your write-up, frankly, there isn't a great deal that I disagree with, except perhaps the perception of the overall magnitude of the problem. A couple of things you said I question: "Embedded systems don't depend on the clock" I assume you mean a Real Time Clock (RTC) and not the system (xtal osc.) clock. Generally speaking I would agree that this is true. However, we have tested MicroController boards (Embedded) with an RTC function incorporated in the design. As to whether the complete system actually used the RTC for time/date functions was less obvious in most cases. Depends on the application software written for the system and controlled and monitored by the Host. btw, we normally use CRC checks for Firmware verification (EEPROMs, internal ROM etc.), Normally a MicroController based board design would be emulated based on functional descriptions, theory of operation etc. Input stimulus applied expected output functions validated. You also said: "Everything revolves around the power up time and not the clock time" I don't fully understand or agree with that. The Power Up time is certainly a reference point for the initialization sequence to begin; Self-test diagnostics, reset functions etc. However, I wouldn't categorically state that every Controller function is referenced to that initial event. Subsequent events may initiate other tasks or routines to be performed - this has nothing to do with the power up time but is more a function of a sequence of timed events or completed tasks. You said: "TPRO does not have expertise with Assembly Language programming." Extremely unlikely. Most of thier Engineers are EE's, with expertise in Manufacturing, Process Control, Test. Assembly is a low level language but is important when working within these environments - it would make no sense for a company who prides themselves in providing "turnkey" Systems Integration support. Overall, I agree that the problems in the factory should not be as widespread as the IT side of the Y2K issue. However, they do exist. And regardless of how many years you have with Embedded Systems you cannot possibaly come accross every unique application that exists in Factory Automation and Process Control Systems - there are thousands. I agree there may be more problems with the application program controlled by the Host than the Embedded system itself. However, I believe TPRO has the expertise to remediate this side of the factory also - "turnkey" remember? If 95% of the factories and utilities are date compliant or date insensitive that does not mean that a Company would simply say: "I'm 95% sure we will not have a problem at rollover" and wait for the rollover with thier fingers crossed and attornies at thier doorstep. They MUST perform at least a complete Assessment in order to be 100% certain. TPRO will perform some of this and will license thier CD and toolset to Mfgs., and other SI houses who will perform the rest - WORLDWIDE. Remember the quick cash flow is in CD sales which will leverage thier core SI business. Look forward to your posts in the future. You are the first bonafide, at least imo, qualified dissenter that I have come accross on this board. Technically you are correct in what you have said but I don't think you have considered all the possibilities and ramifications of the Y2K Embedded Controller problem and how TPRO's business will benefit from this in several ways; The critical production processes that exist in many plants and the litigation potential, basically forces plant mgrs. to perform at least the Assessment phase because if they are one of the unlucky 5% they are in deep "Kimshee".