H/PC Trying, but not connecting.

Been going a few rounds on this, but not sure where to go on it. I have my VM (Win95) going, I have a serial port going. I have H/PC 1.1 installed. When I go to the PC Link on the HP 320LX it looks like its starting on the PC, but then keeps saying its retrying on the HP. (The connection times out after 5 sec)

I have checked the port settings on the VM, I've tried tweaking different combos of flow control, nothing. I left it at 19200/8N1. Trying to figure out what if anything I am missing.

I tried a Palm III on the same VM for sanity. That seems to Hot Sync just fine, goes all the way though. The best I can guess is that A) The H/PC connection *needs* the Hardware flow control and B) my USB-Serial adapter is not cutting it there. I may try and switch that up.

Any ideas on why it would start but not go forward? Is flow control needed? Is there another way to talk to this thing for sanity? At this rate I may be playing with serial software for it sooner than later.

Pulled open Hyperterm on the PC and Terminal on the HP 320LX and got the handheld to push some ATZ and +++ commands trying to dial out, so it seems that the serial connection is doing something.

At this rate I may be porting Hot Sync to the thing.
¯\_(?)_/¯

https://www.hpcfactor.com/support/cesd/s/0010.asp

On a 320, avoid going over 56k on CE2.0, they often don't like the higher speeds, especially 115k. CE1.0 hated going over 19.2k though!

I was keeping everything at the default 19200 to start. I'm going to have to poke around and grok the spec some more. The port is working as I can use it for other things and I was able to pass data from the device itself. I suspect it's in either the hardware flow or some deeper misconfiguration of the Win95 box. I need to read up on what its actually trying to do there. Apparently dial up networking was a critical factor in the early days of ActiveSync so perhaps that is part of the puzzle. The device is likely trying to establish some form of PPP or other connection with the PC and then use the network as the transport layer... which is slick if it works, but top heavy in practice. The VM was not wanting to connect out with TCP either so if there is something in that stack not working, that could be it. Worst case I dust off a Toshiba Satellite from the pile in the garage and try real hardware. I just need a working connection somewhere to have something to sniff.

The good news is that I have the cradle and I confirmed that the power works and the serial cable is functional. So that's a step forward.

1) uninstall the serial ports in device manager and reboot to reset them. Under 9x, never meddle with the hardware settings when using ActiveSync, it always had a habit of breaking it even if you put them back
2) follow my 95 patches and updates guide, specifically make sure to update DUN - https://www.hpcfactor.com/support/patching/win95/

Awesome, Thanks for the pointers. I really just need to get *a* system up long enough to get some traffic with a few scenarios to test things. I suspect once I get the transport layer hashed out I can hum a few bars and have the dumps come directly from a newer OS once I get the lower level hammered out. The reality is that there is likely a lot of things that where 'baked in' to the Windows stack at this point that they where trying to reuse (which I likely have to then 'bake into' my stack, but fine).

I get some of the logic on the handheld side, space was a premium. On the Windows side proper though it seems like Microsoft is always chasing its tail with this. It's was a serial, and then a USB driven chunk of code. i.e. it could have been made 'ever green' in its design, but I guess that, financially at least, was never in their best interests. It's part of why Linux was able to zip ahead on the driver front so fast once it got its speed. You don't need 50 drivers for 50 'different' devices if they all use the same chipset, etc. I know they got a bit smarter about things (especially USB) later in life with device classes.

I think the frustration (or fun, depending on the day and the mood) it was never the Microsoft 'style' to just explain what the hell it is the thing is trying to do. They did the same thing with code. They would rather force you to find a 'wizard' to template in VS than just explain how msbuild or the like works enough to understand how to build a simple DLL. i.e. all the real complexity is shrouded in an extra layer of artificial complexity. Partly why I want to just chuck some of this noise. I mean if you happen to be a Win95 or NT 4.0 fan by all means, go knock yourself out. But if all this is a barrier to a new crowd of users and developers discovering and using these devices, that is what is going to kill these things more than anything else.