Paul H

Yes, TTimeSpan on its own seems to do exactly what I need (and a lot more). Thanks.

I did say originally I wanted a counter with H-M-S support, which (to me at least) implies decimal input/output. Sorry if that wasn't clear to you.

Thanks David - I was hoping to find something with H-M-S support including setting / getting in decimals the way you can with TTime. It is probably no more than a couple of hours work to do something basic with a record helper, I just didn't want to duplicate something that already exists.

Thanks for both comments, but I don't need a timer, so TStopwatch or the system timer are't suitable either. I am building a time based numerical integrator, so real time is not important, I will be setting the integrator "time" manualy as needed. The Time Counter merely keeps track of where we are in the integration and provides the integration time step.

I need a millisecond counter (NOT a timer) that can operate on Hours/Minutes/Seconds with an upper limit of > 1000 hr. Delphi's TTime and TTimeStamp aren't what I need - does anyone know of anyhing suitable? Without too much effort I could probably put something together using a record helper with a simple Cardinal type, but if something already exists I don't want to reinvent the wheel.

I am totally tearing my hair out over this one. I have a largish VCL project (100+ forms and frames) and recently I have started to reegularly "lose" ALL the <FormType>dfm</FormType> tags in the drpoj file, so when I try to add a frame to a form all I get is an empty frame list. Re-adding all the FormType tags to the dproj file every week or so is a pain. Before I give up and rebuild the project from scratch, has anyone any idea what is happening with the IDE? I am using Delphi 12.2

OK - thanks.

I am getting some unexpected behaviour when declaring inline floating point variables using implicit typing. Not sure if it is a compiler bug or expected behaviour. See this (from Delphi 12.2): (if I insert as code the system thinks I am sending spam!!! Does anyone else get this? Anyway, hope it is clear) I would have expected the implicit type rules of inline variables to create doubles where the assigned type is a double, but this isn't happening.