Strange DNS behavior

I guess it will not be easy to reliably explain what I observed recently and what I am able to replicate whenever I want. Just sit down, and read:

Introduction:
Having two distinct networks driven by Mikrotik devices running ROS 6.18, 6.19 and 6.20 on different devices. Lets assume there is Site A and Site B. Both having one main router with internet access. Both sites have their own (different) internal IP ranges /24 and are interconnected by SSTP tunnel using third subnet. Routing between the sites is done manually by static routes thru this neutral subnet. No dynamic routing involved. Allowed devices are able to communicate freely from one site to another site.

Observation:
Having computer running Windows7 named “jarda-msi” connected inside Site A. Router of Site A - call it RouterA, has in DNS cache negative record with name “jarda-msi” and 0.0.0.0 data with unkwnown type. I do not know why it is there or how it got there, I did nothing for it, so it happened on its own. When I run “ping jarda-msi” command in terminal of Router A, I get error that the command was wrongly called:
[jarda@RouterA] > ping jarda-msi
invalid value for argument address:
invalid value of mac-address, mac address required
invalid value for argument ipv6-address
while resolving ip-address: name does not exist…no surprise, as expected.

Now I log in into RouterB of Site B and check the DNS cache - blank in terms of jarda-msi computer.

Open terminal and run ping to jarda-msi:
[jarda@RouterB] > ping jarda-msi
invalid value for argument address:
invalid value of mac-address, mac address required
invalid value for argument ipv6-address
while resolving ip-address: name does not exist…again, as expected.

Running resolve jarda-msi returns “dns name does not exist”, checking the neighbors, no info about the jarda-msi computer or its IP address.

So far nothing surprising or strange.

And now the miracle:
Open the tools-ping window and put the “jarda-msi” into the “Ping-to” field and hit Start button. What do you expect that happen? I expected that it should tell me something that dns resolution failed or some other similar error. But no. Not at all.

It sends the pings to correct address and gets the responses!

How is it possible??? Where it got the IP address of “jarda-msi” computer? Why it works in tools-ping window and not for common ping?

None is wondered by this?

Isn’t the dash the problem ?

I guess the winbox tools->ping part is able to resolve a name which is not in FQDN format, whereas ping in command line is not.

winbox uses your local (PC you are running on) DNS cache/whatever. As you have your PC configured with your host-name - of course it can resolve that name and ping it.

So does it mean that winbox does not only have control over the device, but also changes the user commands and injects its own modifications to the user commands?

I am not happy with that.

How then I can be sure that what I see is the result of processing in the device? How can I know what is really set and how it works if winbox modifies it?