Zugg Software :: View topic

Novice Joined: 09 Oct 2007 Posts: 36

typing "#cmd speedtest" spamms your screen and records how long it will take to process it, i think.
doing this with my current settup, (I'm sure its very laggy) i get 26.9
doing it with triggeres disabled i get 7.6.. HUGE difference..
anyways, I'm curious what everyone here runs about.. If you post both numbers I can get a better idea of how lagged my client is

And then can change things to be quicker! :P

Apprentice Joined: 13 Aug 2007 Posts: 166

With Triggers: 45,7
without: 2.7

Posted: Mon Mar 31, 2008 6:25 pm

Pressing Ctrl+Q will also do this.

Before you start comparing numbers, it's important to spam the command a few times to fill up the buffer. With an empty or partially empty buffer, memory has to be allocated to store the new lines, which will be slower than with a full buffer. In normal play this isn't significant because lines come slowly, but Ctrl+Q prints many lines at once and will cause your numbers to be different with empty and full buffers.

It's also important to remember that the speed of your computer and the way you've built your regexes matters immensely to the speed of the Ctrl+Q test, and that the way you've written your scripts (which can be vitally important in real-life situations) matters very little. It's not a perfect indicator.

Wizard Joined: 03 Mar 2001 Posts: 1127 Location: London

I don't play any muds atm, but on a clean session it's 1.7secs. If I disable triggers (even without having any created) then it drops to 0.9

SubAdmin Joined: 18 Nov 2001 Posts: 5182

Guinn, that is why I reccommend using the #IGNORE command within an OnLoad event for captutre windows that do not require triggers. Also the reported time is of the ratio 1=.1 second, so a reported number of 1.7 means .17 seconds.

The test is 1000 lines, so that is going through all the triggers 1000 times. The use of #IGNORE is checked before any triggers are checked for, which means that setting it properly cuts the work in half when there are no triggers in the session.

A small final note: you can also make another small speed improvement for windows that don't use a connection by properly setting them to no connection.

Posted: Mon Mar 31, 2008 10:07 pm

Maybe I'm missing something here, but where does it show an output of the speed? All I get is the a-z1-0 output...

Charneus

Enchanter Joined: 05 Mar 2003 Posts: 593 Location: Canada

The results are displayed on your status bar
If you already have something there, it might vanish before you have a chance to read it.

Posted: Mon Mar 31, 2008 10:15 pm

Yeah, I wouldn't worry too much about that kind of speed difference. Even at 26.9, that means it took 2.69 seconds to process 1000 lines of text. That is 2.69 milliseconds per line from the MUD. That is still very fast. Even if you received a screenful of text you are hardly going to notice this processing time.

So, while Ctrl-Q is a great way to tune your scripts, I wouldn't lose sleep over your values. If you are noticing large lag from your MUD, that is usually caused by packet loss on your network connection. Windows has a default retry timeout value of something like 5 seconds for a lost packet.

Oh, and not only does processing power matter, but just going from XP to Vista will probably double your speedtest values. That's just because of the slowdown in how Vista handles it's screen updates. To get rid of this slowdown you have to disable the Vista Desktop Manager, which also disables most of the "pretty" parts of Vista, such as the transparency.

Posted: Mon Mar 31, 2008 10:19 pm