Hi Jos,

Just one question, Is the problem occuring on clicking to open up the menu, or on clicking on the actual menuitem ?

We do have a tool, logmeinrescue to get at the client desktop and try to debug issue. If you go to www.logmeinrescue.com you can have a free trail (a month I believe). It also is allowing to transfer files from your end to theirs. What I've done in the past is to quickly make an executable drop it on the client machine and test it. It might help you to track the problem.


Walter,


>Hi Thomas and Walter
>
>There are practical issues involved that will not allow me to do these suggestions (a) the client is in Cape Town, 1400 km away :), (b) they are running real time trading systems on production work stations. I really can't be messing with them during the day even if I was down there.
>
>The claim from their side is that Excel, Word, and other applications work ok. Tough discussion, hence my desire to truly eliminate hyper threading which was just me grasping an an obvious commonality between the 5 machines.
>
>But it would appear to me that if there are no known VFP9 issues with hyper threading that there must be something else going on with those machines. BIOS upgrade is a possibility and that suggestion has been given to their IT guy. But it is also possible that there is something else running which is conflicting with the app. Who knows, an AV, some video driver, something... Only a process of trial and error elimination will find it and I will have to do that next time I'm down there.
>
>The weird thing - from a toolbar button I have in the click event "DO XYZ_PRG". From the menu option it says "DO XYZ_PRG". And even from an ONKEY routine I have "DO XYS_PRG". All three DO the same thing. But from the menu it flips out the cpu and the app goes non-responding.
>
>
>
>>Jos,
>>>Is it the consensus of UT members that VFP9 definitely has no issues with hyperthreading cpu machines? I am just trying to get a definitive answer because I have read about other applications having problems with hyperthreading cpu machines. Just looking to be sure before giving the client a categorical answer :)
>>
>>I would phrase the answer differently<g>. There were issues with double core machines known, most of them in the area of power saving states different in each core. Some of them persist today. Also the high perf counters are often not synched, so anything working of them is untrustworthy (I was bitten there, restraining to use 1 core helped <g>). HT was definitely slower in some SQL/PostgreSQL server machines on multi cpu machines and sometimes single cpu machines.
>>
>>HT cpu's also were at the time of introduction running very hot. I was bitten by probably temperature induced failures on an old 90nm 3.2 GHz HT cpu - on this machine I got errors not seent on Athlon's and a 2.4 PIV. The system was overheating during 5H of full cpu usage- running in winter with an open case I could finish without error. So ***some*** HT cpu's may be more prone to temperature errors.
>>
>>All in all I think Walter's suggestion of running under coverage is a very solid next step - perhaps with opened computer case and room window. To make sure it is not a pathing issue coupled with menu I'ld also add filemon into the tracking part<bg>.
>>
>>regards
>>
>>thomas