Forums Tyrant

SimTyrant - The simulator that runs from your browser! page 27

869 posts

Flag Post

Now the new error message in Firefox last stable version is:

SimTyrantJSc?maxworkers=4

assignment to undeclared variable field
on line 4530
for blob:3c2d97eb-f683-4480-a391-f9422a506f53


applying the ‘delete’ operator to an unqualified name is deprecated
on line 4529
for blob:2897b161-97ec-469a-b435-580963b47fe0

I used maxworkers=1 and gives me the same error.

 
Flag Post

I’m getting identical problems in the beta and non-beta versions of the simulator. Has the old version been changed too?

In Firefox: It does the first few simulations, then stops, and after a few seconds I get the following popup from the browser:

A script on this page may be busy, or it may have stopped responding. You can stop the script now, or you can continue to see if the script will complete.
Script: http://www.haileon.com/SimTyrantJS?deck1=QVDA+nHVIZgj&mission=267&sims=100000&autostart=1:4232

In Chrome: It does the first few simulations, then stops, and continues using the CPU without anything noticable happening. If I said I wanted to use more cores, it uses them (according to the CPU usage in the task manager).

 
Flag Post
Originally posted by djirtsdew:

I’m getting identical problems in the beta and non-beta versions of the simulator. Has the old version been changed too?

In Firefox: It does the first few simulations, then stops, and after a few seconds I get the following popup from the browser:

A script on this page may be busy, or it may have stopped responding. You can stop the script now, or you can continue to see if the script will complete.
Script: http://www.haileon.com/SimTyrantJS?deck1=QVDA+nHVIZgj&mission=267&sims=100000&autostart=1:4232

In Chrome: It does the first few simulations, then stops, and continues using the CPU without anything noticable happening. If I said I wanted to use more cores, it uses them (according to the CPU usage in the task manager).

Interesting… I can verify that the regular SimTyrant is not the multi-core version.

 
Flag Post
Originally posted by djirtsdew:

I’m getting identical problems in the beta and non-beta versions of the simulator. Has the old version been changed too?

In Firefox: It does the first few simulations, then stops, and after a few seconds I get the following popup from the browser:

A script on this page may be busy, or it may have stopped responding. You can stop the script now, or you can continue to see if the script will complete.
Script: http://www.haileon.com/SimTyrantJS?deck1=QVDA+nHVIZgj&mission=267&sims=100000&autostart=1:4232

In Chrome: It does the first few simulations, then stops, and continues using the CPU without anything noticable happening. If I said I wanted to use more cores, it uses them (according to the CPU usage in the task manager).

oh, i thought i was just fixing some debug stuff… ok, undid that thing with immobilize. should work again now.

 
Flag Post
Originally posted by Moraku:
Originally posted by djirtsdew:

I’m getting identical problems in the beta and non-beta versions of the simulator. Has the old version been changed too?

In Firefox: It does the first few simulations, then stops, and after a few seconds I get the following popup from the browser:

A script on this page may be busy, or it may have stopped responding. You can stop the script now, or you can continue to see if the script will complete.
Script: http://www.haileon.com/SimTyrantJS?deck1=QVDA+nHVIZgj&mission=267&sims=100000&autostart=1:4232

In Chrome: It does the first few simulations, then stops, and continues using the CPU without anything noticable happening. If I said I wanted to use more cores, it uses them (according to the CPU usage in the task manager).

oh, i thought i was just fixing some debug stuff… ok, undid that thing with immobilize. should work again now.

It’s working again now.
I did some quick tests on a quad-core laptop
(QVDA+nHVIZgj vs. Mission 183, as sss1 posted on the previous page, but with only 10k sims):

Firefox 19.0.2:
single-core (old version): 20.527 seconds (487.2 simulations per second)
dual-core: 15.192 seconds (658.2 simulations per second)
quad-core: 12.450 seconds (803.2 simulations per second)

Google Chrome 25.0.1364.172 m:
single-core (old version): 7.598 seconds (1316.1 simulations per second)
dual-core: 5.126 seconds (1950.8 simulations per second)
quad-core: 4.946 seconds (2021.8 simulations per second)

It seems like the first second nothing is happening in Google Chrome, after that it is ultra-fast compared to Firefox. Dual core it says around 2700-2800 simulations/second while it is running. I’m not sure what’s causing this initial delay, but it is a shame to only do 10k simulations if there is a full second offset in the times.
So 100k simulations:

single-core (old version): 68.681 seconds (1456.0 simulations per second)
dual-core: 36.309 seconds (2754.1 simulations per second)
quad-core: 30.020 seconds (3331.1 simulations per second)

 
Flag Post
Originally posted by djirtsdew:

It seems like the first second nothing is happening in Google Chrome, after that it is ultra-fast compared to Firefox. Dual core it says around 2700-2800 simulations/second while it is running. I’m not sure what’s causing this initial delay, but it is a shame to only do 10k simulations if there is a full second offset in the times.
So 100k simulations:

I think the initial second is the worker scripts loading up before the simulations start. You’ll probably only notice this on auto-start. If you stop simulations, and then press simulate again, does it have a delay there as well, or does it start right up?

Also, you’ll get more sims/sec if you do a higher number of sims. This is due to the fact that SimTyrant runs its simulations in batches, and then reports the results to the screen. It starts off with only one simulation in a batch, and then for every batch after that, it determines the batch size based on how many simulations per second the last run got. This is so that the user interface gets updates regularly, rather than having the progress be updated every few seconds.

Long-story-short, it doesn’t start off fast, but it ramps up speed quickly. The longer it runs for, the less impact that initial small batch has on the overall sims/sec calculation.

 
Flag Post

v1.143
• Official multi-core support (you can select your desired number of cores on the top of the page)

 
Flag Post
Originally posted by Moraku:

v1.143
• Official multi-core support (you can select your desired number of cores on the top of the page)

Currently, disabling multi-core support. I seem to have run into problems with multiple versions of TheSench’s code and is causing conflicts with my own changes. I will be re-enabling multi-core support after figuring this all out.

 
Flag Post

Sorry to all of you Opera users, Opera doesn’t support Blob URLs (which is how the worker threads receive their JavaScript to run). Until they add support, you most likely won’t be able to use the multi-core version. To see if your browser supports the multi-core scripts, check these two pages:

http://caniuse.com/webworkers
http://caniuse.com/bloburls

 
Flag Post
Originally posted by Moraku:

you made two identical links to caniuse webworkers

Fixed. Thanks.

 
Flag Post

v1.144
• Multi-core support re-enabled. (Select your desired number of cores on the top of the page.)

 
Flag Post
Originally posted by Moraku:

v1.144
• Multi-core support re-enabled. (Select your desired number of cores on the top of the page.)

And… disabled again, due to issues with raids. Bear with us >.>

 
Flag Post

Beta-2 Multi-core support

I’m holding off from replacing the production version until I have this thoroughly tested this time around…

http://haileon.com/SimTyrantJSd

 
Flag Post
Originally posted by Moraku:

Beta-2 Multi-core support

I’m holding off from replacing the production version until I have this thoroughly tested this time around…

http://haileon.com/SimTyrantJSd

To anyone who uses a browser that does NOT support multi-core, please play around with the beta as well. There’s logic in there that is supposed to run the standard single-core version if your browser does not support it, but I’m not sure I caught all of the cases.

 
Flag Post

14.473 seconds (6909.4 simulations per second)
100,000 simulations. A marked improvement.

 
Flag Post

@Moraku Refresh skill used by commander does not count in achievement. Please fix it. Thank you.


4137 if (debug) echo += debug_name(current_commander) + ’ refreshes
‘;
+ // Skill achievement counting
+ if (current_unit[’p’] == ‘player’) achievement_skill.push(‘refresh’);
4138 }

Thanks Cid84 for reporting it.

Originally posted by Cid84:

There seems to be something wrong either in TO or SimTyrant with Gift of Giving 1:
RgOtJOF5-AgMeBoMBEE gives ~26% with TO (100k trials) and only ~8% (10k trials) with SimTyrant.
More specific values from SimTyrant:
Achieved 836 8.4%
= Winter Solstice 7 10000 100.0%
Used heal ≥ 8 4297 43.0%
Used refresh ≥ 8 1293 12.9%
Used supply ≥ 8 3846 38.5%

 
Flag Post

v1.145
• Commander Refresh now counts towards achievements

 
Flag Post
JavaScript error:
TypeError: current_unit is undefined
on line 4140
for http://haileon.com/SimTyrantJS
Browser CodeName: Mozilla
Browser Name: Netscape
Browser Version: 5.0 (Windows)
Cookies Enabled: true
Platform: Win32
User-agent header: Mozilla/5.0 (Windows NT 5.1; rv:19.0) Gecko/20100101 Firefox/19.0
SimTyrant version: v1.145
Deck hash: P8BoEDB6C6D2AoMCA/gJgi
Enemy deck hash: RgBoEDB6C6D2AoMCA/gJgi
Sims run so far: 28
 
Flag Post

v1.146
• Fixed Commander Refresh crash

 
Flag Post

BUG:

http://haileon.com/SimTyrantJS?deck1=P3AIAKAZAYALB2ffAhB3AF&ordered=1&mission=29&achievement=166&sims=10000&autostart=1

There is a bug with the achievement Common Theme (Defeat Mission 28 or higher after playing 10 Common cards), whatever cards that i use the achievement counter always says:
“Played common ≥ 10 | 0 | 0.0%”

Obviously the achievement winrate finishes in 0% when it is 39% (checked with Tyrant Optimize).

Deck in fansite says 0% too:
http://tyrant.40in.net/kg/ach.php?id=166#sdecks:did=14518

 
Flag Post

Some new code must have broken it. Old version works. Will look at this later.
http://haileon.com/SimTyrantJS_v1122?deck1=P3AIAKAZAYALB2ffAhB3AF&ordered=1&mission=29&achievement=78&sims=10000&autostart=1

 
Flag Post
Originally posted by DarkBlood1:

BUG:

http://haileon.com/SimTyrantJS?deck1=P3AIAKAZAYALB2ffAhB3AF&ordered=1&mission=29&achievement=166&sims=10000&autostart=1

There is a bug with the achievement Common Theme (Defeat Mission 28 or higher after playing 10 Common cards), whatever cards that i use the achievement counter always says:
“Played common ≥ 10 | 0 | 0.0%”

Obviously the achievement winrate finishes in 0% when it is 39% (checked with Tyrant Optimize).

Deck in fansite says 0% too:
http://tyrant.40in.net/kg/ach.php?id=166#sdecks:did=14518

Works now, but it’s only giving me about 37% achievement rate.

 
Flag Post

v1.147
• Fixed bug where Rarity-based achievements stopped working after v1.137

 
Flag Post

v1.148
• Chaosed Chaos/Jam/Weaken now properly targets Blitzing Assaults.
• Fixed glitch where Assaults that are Jammed after activating at least one activation skill (without suffixes) continued to activate the rest of its activation skills. (ie. via Paybacked Jam or Jam on Death)

 
Flag Post

Thank you for fixing Rarity-based achievements! The winrates are fine, similar to Tyrant Optimize, -2% or +2% is nothing.