DavidCrow wrote: Do you not think it pertinent to include the line number of the code this error refers to? We're not mind readers. Oh sorry, pointed lines are : real_T d; and ** real_T is the representation of double in this format, even if i write there 'double' appears again as error ** num.d = *uPtrs[0]; Thank you for your consideration! Cahit

ObjectCSyncSerialComm->ReadFile cje

In Windows NT and later, you cannot directly access serial ports. But check this article[^] See Part 5 for a link to the source to inpout32.dll, which can be used in NT + Alcohol. The cause of, and the solution to, all of life's problems - Homer Simpson

cahit23 wrote: union ifas { double d; char u8value[8]; //?????????????? }; Use unsigned char for u8value Alcohol. The cause of, and the solution to, all of life's problems - Homer Simpson

There really is no need for that. Unsigned char's will be placed on 8-bit boundaries on all common systems, and doubles will be placed on 64-bit boundaries. The union will be sized and oriented based on the largest size requirement (the double), so packing it serves no purpose. If you decide to become a software engineer, you are signing up to have a 1/2" piece of silicon tell you exactly how stupid you really are for 8 hours a day, 5 days a week Zac

oh sorry, i deliver my message in visual c++ forum then!