Welcome!

By registering with us, you'll be able to discuss, share and private message with other members of our community.

SignUp Now!

Does TCC know I'm visiting the forums?

That's 2.1% for TCMD. That's reasonable if you're using it; unreasonable if you're not using it or if it was just started.

I'd expect TCMD to be the biggest CPU user.
As you can see it was running for 26 hours 37 minutes so I hadn't just started it and it was idle most of that time just running in the background.
 
Since I'm blind I'm probably missing something, but could TCMD only scan consoles while it was in the foreground, but not while it was in the background and also could it only scan the current active tab at least if it were maximized? Why keep the screen correct for a background window that you can't see?
 
Since I'm blind I'm probably missing something, but could TCMD only scan consoles while it was in the foreground, but not while it was in the background and also could it only scan the current active tab at least if it were maximized? Why keep the screen correct for a background window that you can't see?

I *could* do that, if you can guarantee that nobody will ever:

1) Resize the TCMD window
2) Create a splitter window
3) Create window groups
4) Autoattach consoles
5) Close a background tab window process
6) Show a modal dialog in any tab window
7) Change the default console screen buffer properties
8) Run anything in a non-current tab that changes the tab title (i.e., run anything at all)
9) Change the console font

(and probably a few more I can't remember off the top of my head).

Oh - and you can't run any command in a non-current TCC shell that uses IPC to communicate with TCMD.

If you can promise all that, I can probably shave 0.05% of the CPU usage for one core if you're running 10 tabs. For a couple of tabs, < 0.01%.
 
Well -- no, it can't.

But you can eliminate the console scans / updates if you minimize the TCMD window.
I minimized the TCMD window with no change in CPU usage so perhaps it's something else it is doing other than scanning consoles that is causing the usage I'm seeing. BTW is there a way to minimize from the keyboard or do I have to click the minimize icon? I couldn't find a way to do it from the keyboard.
 
Alt-Spacebar brings you a menu with options. Between brackets are the shortcut keys (will depend on your Windows language, though). On the screen these keys are underlined.

- Move (M)
- Size (S)
- Minimize (N)
- Maximize (X)
- Close (C)

So: Alt-Space + n will minimize your current foreground application


If you are on Windows 7 or Windows 10: Windows-key+arrow down will also minimize.

(BTW: I worked quite intensively with JAWS, a screenreader. I know how frustrating it is when the screenreader does not pick up on stuff (or scrambles the content).
 
Last edited:
Alt-Spacebar brings you a menu with options. Between brackets are the shortcut keys (will depend on your Windows language, though). On the screen these keys are underlined.

- Move (M)
- Size (S)
- Minimize (N)
- Maximize (X)
- Close (C)

So: Alt-Space + n will minimize your current foreground application


If you are on Windows 7 or Windows 10: Windows-key+arrow down will also minimize.

(BTW: I worked quite intensively with JAWS, a screenreader. I know how frustrating it is when the screenreader does not pick up on stuff (or scrambles the content).
I knew about alt-space, but forgot to use left alt rather than right alt and it didn't work because of that. What a silly mistake.
 
Now that I think of it (I know, I'm a little "slow" from time to time):

WINDOW /MIN in your TCC/TCMD program will also minimize.
Or straight from Process Explorer: Shift-F10 on the process, Window (W), Minimize (M)
Òr Windows' own Taskmanager: tab Application, Shift-F10 on the process, Minimiz (M)

In Process Explorer the current state will be grayed out, so you know the state the application is in. Taskmgr misses this functionality.
 
Have you got anything hooking the console session (or injecting code into it)?
Windows injects uiautomationcore.dll and oleacc.dll in to tcmd and conhost but I ran it with no screen-reader loaded to prevent this and had someone read the screen and it still happened. Does TCMD do any sleep for x MS like the thread that you changed in TCC? I assume a process dump that you could analyze wouldn't help you. Is there anything I can do to show you which thread it is which might help with debugging?
 
Windows injects uiautomationcore.dll and oleacc.dll in to tcmd and conhost but I ran it with no screen-reader loaded to prevent this and had someone read the screen and it still happened. Does TCMD do any sleep for x MS like the thread that you changed in TCC? I assume a process dump that you could analyze wouldn't help you. Is there anything I can do to show you which thread it is which might help with debugging?

I don't know that there's anything to debug -- the CPU usage you're describing is what I would anticipate TCMD & TCC requiring.

The change in TCC was specific to running a browser (i.e., Firefox) at the same time, where I added a kludge for a bug where the browser changed the timer resolution. That would not be relevant when running TCMD.
 
I don't know that there's anything to debug -- the CPU usage you're describing is what I would anticipate TCMD & TCC requiring.

The change in TCC was specific to running a browser (i.e., Firefox) at the same time, where I added a kludge for a bug where the browser changed the timer resolution. That would not be relevant when running TCMD.
Right and this kludge also helped my system. I run NTPD and it might change the timer resolution for the multi media timer. I was hoping that there was code in TCMD that would also benefit from this kludge, but you've already answered that there isn't.
 
Right and this kludge also helped my system. I run NTPD and it might change the timer resolution for the multi media timer. I was hoping that there was code in TCMD that would also benefit from this kludge, but you've already answered that there isn't.
Which NTPD and how well does it work?

As I understand it (Rex, please tell me if I'm wrong), at high time resolution, waits and timers will be more accurate, and a thread giving up its timeslice (for example with Sleep(0)) makes things worse because threads are visited more often. If there's more to it I'm all ears.
 
Which NTPD and how well does it work?

As I understand it (Rex, please tell me if I'm wrong), at high time resolution, waits and timers will be more accurate, and a thread giving up its timeslice (for example with Sleep(0)) makes things worse because threads are visited more often. If there's more to it I'm all ears.
[title]
 

Similar threads

Back
Top