Destroyers Discussion what were you guys on?

[Der Teddy Bar]

I recently had a look at the end products for the thread "Destroyers Discussion (getting rid of pin point drops)" http://www.subsim.com/phpBB/viewtopic.php?t=45370

I am sadly amused to see the additions to the Sim.cfg files in the GW & CB versions. They show a total lack of proper testing and understanding of moding.

Now before anyone gets too bent out of shape about this lets just look at this sensibly.

A CFG file is a file that allows easy changing of variables. That’s it, end of story. It is not code, it is not 'run' by the program, it is simply read from and the relevant values taken and used by the program.

The use is very precise, that is, there is no wishy washy use it if it sounds like... The program says "if you find A use the value listed there, if you do not find A then use this", that’s it. There is no "oh OK, there’s no A, but I found a BA (or a AC etc) so I will use that one instead"

The idea that just because you take add additional 'data' under the group heading it will get used is just silly.

What I am referring too are these additions...
Grey Wolves
Already tracking modifier=20 ;[detection probability modifier]
Decay time=150 ;[>0] already tracking bonus decay, in seconds
Uses crew efficiency=true ;[true or false]

And in CB's one...
range factor=1 ;[>=0]
fog factor=0 ;[>=0]
light factor=0 ;[>=0]
enemy speed=0 ;[>=0]
aspect=0 ;[>=0]
sensor height factor=0 ;[>=0]
already tracking modifier=20 ;[detection probability modifier]
decay time=150 ;[>0] already tracking bonus decay, in seconds
uses crew efficiency=true ;[true or false]

Now in the situation of GW they have kept all the original Hydrophone entries with some small alterations, so in essence all they have done is add some useless values.

Now CB has really hacked it up. He has removed the original entires of Detection time, Sensitivity & Height factor. Then he has added a lot of useless values.

Unlike the GW version which keeps the formula intact, CB has removed 3 values Detection time, Sensitivity & Height factor from the formula.


I do not have the Escorts AI code, so I will use the SHIII formula Noise Factor (a U-boat sensor formulas) to show you what this means.

P noise = Noise Factor * (RPM current / RPM prag - 1)

If P noise <= 0 then object undetectable


Let’s use the value of 1 for the noise factor, RPM as 116 (or 0.40 of maximum of 290 revs) & of course RPM prag is 0.40.

P noise = 1 * (116/0.40 – 1) so in this case P noise=1.9


Yes CB did leave the ‘Noise Factor’ value in, but as this is an exercise to show the effects of his changes along with this is not the correct formula, I will substitute a 0 to represented the effects that his many omissions will have generated for the real unknown formula.

P noise = 0 * (116/0.40 – 1) so in this case P noise=0

So basically the formula has stopped working.


While this post is critical of the work and by default the authors, it is not meant to be anything more than looking at the facts in a hope to fix the escort detection for real this time and to highlight the issue of improper procedure & testing in the hopes that next time there will not again be thousands of hours wasted.


Cheers,

Teddy Bär

[Scorpius]

Got milk?

[Ducimus]

I thought my testing of wave/noise factors was fairly well done.

I cant help but feel someone just took a leak in my cereal bowl.

edit:
Ohh new forum avatar, must be neals way of saying i post too much. :roll:

[Cdre Gibs]

The question that needs to be asked is this - did you even try the files or did you just open your mouth before putting your brain into gear?

And an answer like - "I dont need to because i'm so good I can just tell and therefor totaly disregarding everybodys input on the subject" dont count.

I would suggest that before 1 makes an ass of 1's self that they at least TRY the supposed tweak, THEN give some postive feedback/fix's instead of being a jerk.

Ohh and before you say it, I have no vested interest in this DD fix 1 way or the other, I'm just appauled at your general attitude to several modding ppl/groups thats at least TRY and work as a community for the greater good.

For some1 who's suppose to know better, this has really shown your colours.

[Der Teddy Bar]

The question that needs to be asked is this - did you even try the files or did you just open your mouth before putting your brain into gear?

And an answer like - "I dont need to because i'm so good I can just tell and therefor totaly disregarding everybodys input on the subject" dont count.

I would suggest that before 1 makes an ass of 1's self that they at least TRY the supposed tweak, THEN give some postive feedback/fix's instead of being a jerk.

Ohh and before you say it, I have no vested interest in this DD fix 1 way or the other, I'm just appauled at your general attitude to several modding ppl/groups thats at least TRY and work as a community for the greater good.

For some1 who's suppose to know better, this has really shown your colours.
Gibs,
Side step tha actual facts and stay true to the cause :rotfl:

You are a classic :up:

Cheers.

[Der Teddy Bar]

I thought my testing of wave/noise factors was fairly well done.

I cant help but feel someone just took a leak in my cereal bowl.

edit:
Ohh new forum avatar, must be neals way of saying i post too much. :roll:
My post has no malice in it and is simply stating facts about the changes.

So I should have kept my mouth closed and let something that is incorrect go by because if I am not for you then I 'have' to be against you?

OK, the published solutions are great and , threee cheers, Hip Hip Hoorah, Hip Hip Hoorah, Hip Hip Hoorah!

Cheers

[Der Teddy Bar]

Got milk?
Denial is a beautiful thing...

[Ducimus]

It's not what you said, it was how you said it.

You could have taken a tone of,

"Guys, i think you got some stuff wrong, here's why"

Instead of a tone that sounds along the lines of:

"Are you guys on crack?! WTF are you thinking?! Why did i waste my time reading that thread! My way is the right way you idiots! You don't know your ass from a hole in the ground!"

While these arent your words, its how i personnally interpreted it. I intitally typed up a kneejerk response which i toned down because i didnt want to add to any friction that may generate in this thread.

[HEMISENT]

I cant help but feel someone just took a leak in my cereal bowl.


I thought my cornflakes were extra soggy too.

As Daniel Boone used to say "Some days you eat the bear and some days the bear eats you"

[Scorpius]

Got milk?
Denial is a beautiful thing...

I actually have no idea to whatever you refer to. I was just curious about this thread and that is the first thing that entered my mind. I have no idea why i posted that. I just did. :-j :rotfl:

[Reece]

I just gotta post! :D
I love your scarcasm Teddy Bar "OK, the published solutions are great and , threee cheers, Hip Hip Hoorah, Hip Hip Hoorah, Hip Hip Hoorah!" :rotfl:
Scorpius - "Got milk" what can I say ..... :rotfl: Lets hope that's why the cornflakes are soggy! :rotfl:

[Der Teddy Bar]

Got milk?
Denial is a beautiful thing...

I actually have no idea to whatever you refer to. I was just curious about this thread and that is the first thing that entered my mind. I have no idea why i posted that. I just did. :-j :rotfl:
No probs it just saw it in there will all the other posts which have nothing to do with the the issue at hand, that a lot of work on the AI_Sensor.dat is basically null & void.

NYGM can add sensors to the AI_Sensor.dat and had hoped to bring the community in on making a better escort sensor pack but everyone is more interested in 'having a go' so there goes that great idea....

[lurker_hlb3]

I recently had a look at the end products for the thread "Destroyers Discussion (getting rid of pin point drops)" http://www.subsim.com/phpBB/viewtopic.php?t=45370

I am sadly amused to see the additions to the Sim.cfg files in the GW & CB versions. They show a total lack of proper testing and understanding of moding.

Now before anyone gets too bent out of shape about this lets just look at this sensibly.

A CFG file is a file that allows easy changing of variables. That’s it, end of story. It is not code, it is not 'run' by the program, it is simply read from and the relevant values taken and used by the program.

The use is very precise, that is, there is no wishy washy use it if it sounds like... The program says "if you find A use the value listed there, if you do not find A then use this", that’s it. There is no "oh OK, there’s no A, but I found a BA (or a AC etc) so I will use that one instead"

The idea that just because you take add additional 'data' under the group heading it will get used is just silly.

What I am referring too are these additions...
Grey Wolves
Already tracking modifier=20 ;[detection probability modifier]
Decay time=150 ;[>0] already tracking bonus decay, in seconds
Uses crew efficiency=true ;[true or false]

And in CB's one...
range factor=1 ;[>=0]
fog factor=0 ;[>=0]
light factor=0 ;[>=0]
enemy speed=0 ;[>=0]
aspect=0 ;[>=0]
sensor height factor=0 ;[>=0]
already tracking modifier=20 ;[detection probability modifier]
decay time=150 ;[>0] already tracking bonus decay, in seconds
uses crew efficiency=true ;[true or false]

Now in the situation of GW they have kept all the original Hydrophone entries with some small alterations, so in essence all they have done is add some useless values.

Now CB has really hacked it up. He has removed the original entires of Detection time, Sensitivity & Height factor. Then he has added a lot of useless values.

Unlike the GW version which keeps the formula intact, CB has removed 3 values Detection time, Sensitivity & Height factor from the formula.


I do not have the Escorts AI code, so I will use the SHIII formula Noise Factor (a U-boat sensor formulas) to show you what this means.

P noise = Noise Factor * (RPM current / RPM prag - 1)

If P noise <= 0 then object undetectable


Let’s use the value of 1 for the noise factor, RPM as 116 (or 0.40 of maximum of 290 revs) & of course RPM prag is 0.40.

P noise = 1 * (116/0.40 – 1) so in this case P noise=1.9


Yes CB did leave the ‘Noise Factor’ value in, but as this is an exercise to show the effects of his changes along with this is not the correct formula, I will substitute a 0 to represented the effects that his many omissions will have generated for the real unknown formula.

P noise = 0 * (116/0.40 – 1) so in this case P noise=0

So basically the formula has stopped working.


While this post is critical of the work and by default the authors, it is not meant to be anything more than looking at the facts in a hope to fix the escort detection for real this time and to highlight the issue of improper procedure & testing in the hopes that next time there will not again be thousands of hours wasted.


Cheers,

Teddy Bär


My question to you is do you have a "better" solution ???

[Der Teddy Bar]

lurker_hlb3,
Yes and no.

We are slowly working out a close approximation of the formula. I have had no success in getting it off the Devs, different guys and some are not on SHIV so they are not there to get asked for it.

FYI the AI & the U-boat sensors were done by different guys which is in my opinion very obvious.

NYGM can add sensors to the AI_sensors.DAT and we can add NODES to the escort ships.


Cheers.

[Ducimus]

My question to you is do you have a "better" solution ???

Thats what i was wondering, i just didnt ask because i didnt want to be a smart ass.

In that thread i focused primarly on the AI's ablility to HEAR me. My thoughts have always been, the problem isnt pin point drops, the problem is getting the AI to not be so damn passive in behavior. Most of the AI is very passive, or in other words dumb. (exception being crew rating 4). I thought if i could increase their chance to be aware of my presence they would respond appropriately. And its a feasible thought, as the stock noise/wave factors - they are relatively deaf. Think about it, how often have you been detected by how much noise your making unless you deliberatly shove your boat into flank speed with an escort nearby?

To sum their behavior, i beleive is in need of work, but hardcoded. To use an old phrase, "you can lead a horse to water, but you can't make him drink". But then again. im on crack, and i don't have a clue. Must be all this peyote im smoking :rotfl:

[lurker_hlb3]

is this a valid formula

P noise = Noise Factor * (RPM current / RPM prag - 1)

if so where did you get it ?

[Ducimus]

All im going to say, is if i had a dime everytime i heard someone on a message board that says he knows the games developer, or is a game devloper, id be a rich man.

[Der Teddy Bar]

is this a valid formula

P noise = Noise Factor * (RPM current / RPM prag - 1)

if so where did you get it ?
http://www.mistari.com.au/SHIII/SHIII_Sensor_Formulas.htm. It was given to me by the Dev Team along with other fomulas that have been prviously posted for everyone to use

[lurker_hlb3]

is the 0.40 RPM prag value a constant or is define some where within the data files of the software ?

[lurker_hlb3]

is the 0.40 RPM prag value a constant or is define some where within the data files of the software ?

Disregard, I re-read your orginal post and see where the .40 comes from

[Observer]

My question to you is do you have a "better" solution ???

Thats what i was wondering, i just didnt ask because i didnt want to be a smart ass.

In that thread i focused primarly on the AI's ablility to HEAR me. My thoughts have always been, the problem isnt pin point drops, the problem is getting the AI to not be so damn passive in behavior. Most of the AI is very passive, or in other words dumb. (exception being crew rating 4). I thought if i could increase their chance to be aware of my presence they would respond appropriately. And its a feasible thought, as the stock noise/wave factors - they are relatively deaf. Think about it, how often have you been detected by how much noise your making unless you deliberatly shove your boat into flank speed with an escort nearby?

To sum their behavior, i beleive is in need of work, but hardcoded. To use an old phrase, "you can lead a horse to water, but you can't make him drink". But then again. im on crack, and i don't have a clue. Must be all this peyote im smoking :rotfl:

Yes. This is the problem. It's also very difficult to sort out and get it right. To be blunt, this portion of the code is horrible. It makes it almost impossible for the escorts to ever be effective. Oh how I've tried...oh how I've tried.

[VonHelsching]

While this post is critical of the work and by default the authors, it is not meant to be anything more than looking at the facts in a hope to fix the escort detection for real this time and to highlight the issue of improper procedure & testing in the hopes that next time there will not again be thousands of hours wasted.
Cheers,

Teddy Bär

Your writing style does not help your stated purpose (ie to fix the escort detection). You are not writing this in order to achieve this purpose with the help of the community. I am sure that you have already in mind of a proposed solution for this problem, or at least you are very close in solving it. I think you just want to attract attention, using what I perceive as negative advertising.

caveat lector
.
.
caveat emptor

[Der Teddy Bar]

It is funny, as I read that as a straight forward statement of fact with no derogative undertones. You are also mistaken on the “just want to attract attention, using what I perceive as negative advertising”. It was a sincere statement regarding lost hours, as there is nothing worse when moding.

That said, I made my post regarding the previous ‘solutions’ as I wanted to get the ball rolling again regarding the AI sensor issue that for all intensive purposes has not resolved as I have demonstrated.

I re-opened this subject as I feel that I need the help of those people who put so much time and effort into the previous investigation.

The issue with the current Hydrophone configuration is that is does not scale well and often is an either or. The design is poorly thought out, in that the speed of the u-boat is not considered, only the ordered speed and the silent running works better than a clingon cloaking device.

On top of all that it does not scale well. So fix it for one situation and it is either uber or dumbed down in another.

NYGM have some ideas that we want to put forward as a possible solution and seek the advice and input of the people who had worked on this previously.

OK, onto the basic idea.

As noted, we cannot add or change the Sim.cfg file apart from the variables.

What we have been considering is to add additional hydrophone settings to the AI_Sensor.dat and then on top of that then add additional nodes to the escorts to place the additional hydrophones.

The additional hydrophones would then be specifically designed via the min/max range for the height, bearing, elevation & sensitivity of the sensor to meet a specific purpose.

For example, there would be several short range hydrophone sensors to cover several different situations. They would in some part overlap while also leaving gaps, again, this is a general approach that needs to be discussed and mapped out.

More specifically, to accommodate the Silent Running’s clingon cloaking ability the Hydrophone would be configured to detect a silent running u-boat at a range of 500 metres when the weather conditions are at a 10 metre per second wind. When the weather is glassy smooth then the maximum range would be limited by the new Hydrophones as described below.

We are looking at having a setup of Short Range Hydrophone and Long Range Hydrophone, or SRH & LRH.

The 1st SRH would have a configuration to hear at periscope depth to say 20 metres to a range of 1500 metres. The 2nd SRH would have a configuration as to be under the 1st SRH and would at the max range of 1500 metres leave a gap between the two, but in close, say to 500 metres they would overlap. The 3rd SRH would then essentially point very much in a downward direction and have a maximum range of 150 metres.

The issue is then the LRH will have the ability to hear a silent running u-boat at max range of 5000+ metres. So to combat this we would again add extra hydrophones & nodes to the escort and through shaping of the sensor limit its ability to hear. For example, put the LRH so that it hears only from 25 degree from dead ahead to say 35 degrees and to have a shallow arc so as to not detect the player in close.

Cheers.

[mikaelanderlund]

I have a dream, I have a dream GW+NYGM. We shall overcome :up: :sunny:

Mikael

[Dowly]

Funny to see, that after GW and NYGM decided to go 'independent', there´s been ALOT of crap flying everywhere.

After Kpt. Lehmann´s statement, there have started to be those "My mod is better than yours" type of threads.

I do think that Teddy should´ve though about the 'look' of his writings on this thread.

For example:

"They show a total lack of proper testing and understanding of moding."

Even a 2-year old can take that as an 'attack'.

[gouldjg]

My long speech :rotfl: :rotfl:

Whilst I think the slating of people’s time and efforts is a little out of order on your part, I do think that some points you observed hold valid and if you are going to produce better, I am happy to help with what I got from the DD thread. I am not going into the "I only wanted to provoke topic discussion" its lame and half of those to its directed are not here to defend anyway.

Ok people should never expect to see eye to eye when it comes to realism and gameplay but to me, publicly highlighting other modders flaws is what makes players stop playing the game period because they start to feel that the game is broke period and will never be fixed. As we know, many things are coming and are still to come, 12 months after release. Lets not start giving the perception that the game is broke and others are making it worse. Bad community spirit imho.



CB always stated he was not interested in releasing a claimed fix or claiming that he cracked the DD code. He just went with what was giving him more of a challenge compared to stock. People just asked him for his version and he therefore provided and provided ample warning about his views.

You will note when reading the full thread that I was not really into all the dramatic changes concerned with sim.cfg. After feeling like I was losing that argument and admittingly knowing I really had not enough knowledge of the historical aspects of DD sensors as well as the time to spend testing the campaign etc, I politely dropped out into the background but always keep my own little setup as to what I think adds to my personal gameplay a bit like what the other members did..

I cannot speak for GW team as I only did part time and left the modding scene in the middle of development, RL took over for a while. I do know that they followed the DD discussion closely and went with the best info to date. I think many players quite like it so they may not have been on anything except a view to pleasing the majority with what info was available.

You can very well say "what were they on?" but lets put credit where credit was due, things were discovered by every member and although it was no DEV team super piece of work, the whole thread was about discovery. At the time there was no alternative, no future patches, No Dev team contact. The shame is that it did split in the end and too many people were happy to accept the proposals because it did make their game a little more challenging for a while.

1. Vanilla game had Uber active sonar lock that in some cases was impossible to evade. Jungman

2. 1.4b patch bought the pin point DC, I think many players will notice that there is now enough time to evade DC as long as one was not lazy or unlucky when DD went abeam so steps were made good from the DD thread. Redwine, Myself

3. Hydrophone ranges were Just too toned down IMHO, (has always been my main nitpick) I just add 5000 meters to each Hydrophone and make all sensitivity = 0.03 which seemed to match what sim cfg prefered. That way I cannot really fly in from the side without running silent and in some cases having to manually adjust speed to 1 knot also. My view is I should always get head and be prepared or take the shot and then have to evade because that was the gamble then and it is now IMO. The others took the different road.

4. Despite whatever the lost contact time was, the DD just fail to do any kind of Historical search patterns and just rush back to convoy. Even the hunter groups do this. Why the devs did not include this is why we all have different views and to some extent why people think the DD are dumb. I hope Sh4 make sure that DD do have a least a couple of search patterns and historical pinning tactics but will they really have players pinned for 11 hours etc. In the meantime, my best solution is to use the chaos file and randomise the above as it then adds a sense of human factors into what will become with any sensor model, a predictable game. I am holding onto the chaos file for a while because I need to determine sound base mods to which to place randomisation. Each maker mod need to produce what they believe suits the majority of their players without forcing them to use SH3 commander. I understand that, so I am trying to help GW get a Ship model up.




No what I gotta ask is this and please do not be offended by the questions,

What are you proposing (in English please)?

Are you not just suggesting yet another wild method that to some extent will only be similar to what CB discovered which is, the game is broke, we have to cheat the DD sensors to get variety and challenge?

What help do you need?

Is your proposal going to add variety, and actually produce a better DD?

And on a personal note, will I be able to randomise it without ruining a mathematical absolute formular that leaves little options for those of us whom isn’t so hardcore?

More importantly, Is this going to be a NYGM only product or will others be able to participate and benefit from it and have the right to attempt to tone it down on the realism aspect should it seem to detract from its equal share of gameplay.

I lay my cards on the table at this point,

I have a 100% interest ultimately with a future Chaos file used in conjunction with Sh3 commander.

I have diddled with bits of the game so much that I see the only way I will ever get a lasting buzz, is to have a slight "not knowing what to expect feeling but expect the worse" Chaos file will do that for me.

I need to be able to choose a sound base mod

I need to choose a well balanced ship damage model capable of being randomised up or down according to my realism taste.

I need to get a decent air mod, again capable of being randomised to make me think.

I need a sub mod that although should not leak as much as a fixed number solution provides (Hollywood), it should also not take the chance away for a exciting battle of repair and pannic. I.e. a randomised sub model.

I need a middle ground DD setting that again can be randomised to have lucky/unlucky feel and add a sense that a human may be in the DD.

Later as more doors open = more options i.e. a randomised secondary explosion mod.

All this is possible but we now have two teams whom do not see eye to eye = I have to wait longer :damn: :damn: :damn:

I really think a bit of humble pie on both sides is needed and then at last we can bash ideas of eachother to save hours and hours of testing time which is ultimately forcing players to wait more.

If thrown to the community and You and Kpt agree to disagree but decide to share knowledge and possibly modding ideas and inter mod options, I think we will get a really happy community. Is that too impossible to ask?

[HawkerT]

Herr Teddy Bar und Herr Observer!

I urge you deeply and sincerely to grab hold of your NYGM gem and follow the suggestions put forth above by Herr Gouldjg aka. Mr. Chaos. In my humble opinion Mr. Chaos is onto something really worth a second thought ... times 10 ... :D

The very few posts I have done on this forum have all been cheering for the following:

A) Please keep NYGM focussed (and I feel that you are) on improving the 'hard' parts of SHIII i.e. the core of the game engine. So engine modules like damage models, munitions effect, sensors etc. should fall into that category.

B) Now this is the subtle part ... please do A) so that it is prepared for randomization using the SH Cmdr concept created by JScones and for which Herr Hemisent aka. Mr. Thermal and Mr. Chaos have provided proff of concept.

A) and B) will in combination with sound and visuals from other mods really be something to move SHIII towards the skies.

You are doing a marvellous job, please give the thoughts from Mr. Chaos a second look!

All the best
Frans

[Observer]

Randomization is possible. It's entirely possible using whatever tool or method is available, in this case SH3 Commander serves the purpose extremely well. The DAT file will be the same; you will just have to define the appropriate hex offsets in order to adjust the values.

The goal of this project is to create a basic setup. Random modifications can then be applied (i.e. the chaos mods using SH3 Commander) as desired to the basic setup. In fact this works very well with the only drawback being the need to reload.

What we are looking for is community input on how best to build this thing. This isn't intended to be a flame fest, rather a recognition we need to go back to the drawing board and fundamentally rethink how we approach the escort sensor problem, and we are asking for help because this is such a big job. This is a task of innovation - blow away the old way of thinking about the problem (i.e. simple adjustments to the sim.cfg and the AI_Sensors.dat files) - turn it on its head and think about it from a fresh, new perspective.

Tonight I will post some conceptual images of what these hydrophone arcs and areas might look like. I think that will help clarify some of the confusion around this proposed solution. As with many other mods there will be some tradeoffs, but I think the benefits of this solution will outweigh the negative factors.

And just to be clear, let's put this topic back on track with constructive ideas rather than silly, non-constructive bickering. I for one refuse to participate in any form of flame fest. If this topic is not sufficient for the purpose of constructive ideas, then I will create a new one that is.

[gouldjg]

I am all ears and look forward to seeing your proposals Observer :up: .

1. Can you clone a number of complete new sensors cos if so, then yes, you will have options to making each ship class very much like its intended real life version which is a good thing in my book.

2. I think this was discussed on the DD thread or in a pm from someone.

3. Another idea discussed but not tested was that the cargo ships should also have a seperate hydrophones installed but set on very short ranges hopefully to ensure DD reactions once torped. One of the hardcode problems was the DD sometimes did and sometimes did not react to torps being fired. The Idea was to look for ways to ensure they became nosey after a ship was hit. Someone wanted to also place another visual sensor on cargo ships and have it set to -10 but again at very close range so ships could always see a torp in its last few hundred meteres. None of use were however capable of that sort of stuff then. Can you do that?

Food for thought, it certainly is,

[Ishmael]

I was on these 2:

USS Schofield DEG/FFG-3 Served aboard from 3/73-3/75

http://img101.imageshack.us/img101/6392/deg35ez.jpg (http://imageshack.us)

http://img101.imageshack.us/img101/6392/deg35ez.th.jpg (http://img101.imageshack.us/my.php?image=deg35ez.jpg)

USS Elmer Montgomery FF-1082 Served aboard from 11/75-9/78

http://img186.imageshack.us/img186/9669/ff10826wq.jpg (http://imageshack.us)

http://img186.imageshack.us/img186/9669/ff10826wq.th.jpg (http://img186.imageshack.us/my.php?image=ff10826wq.jpg)

[Observer]

I am all ears and look forward to seeing your proposals Observer :up: .

1. Can you clone a number of complete new sensors cos if so, then yes, you will have options to making each ship class very much like its intended real life version which is a good thing in my book.

2. I think this was discussed on the DD thread or in a pm from someone.

3. Another idea discussed but not tested was that the cargo ships should also have a seperate hydrophones installed but set on very short ranges hopefully to ensure DD reactions once torped. One of the hardcode problems was the DD sometimes did and sometimes did not react to torps being fired. The Idea was to look for ways to ensure they became nosey after a ship was hit. Someone wanted to also place another visual sensor on cargo ships and have it set to -10 but again at very close range so ships could always see a torp in its last few hundred meteres. None of use were however capable of that sort of stuff then. Can you do that?

Food for thought, it certainly is,

Yes, this is possible.

[CB..]

oh dear...just get on with it..i m sure there's little benifit to be had in slagging me off!!!

[Der Teddy Bar]

3. Another idea discussed but not tested was that the cargo ships should also have a seperate hydrophones installed but set on very short ranges hopefully to ensure DD reactions once torped. One of the hardcode problems was the DD sometimes did and sometimes did not react to torps being fired. The Idea was to look for ways to ensure they became nosey after a ship was hit. Someone wanted to also place another visual sensor on cargo ships and have it set to -10 but again at very close range so ships could always see a torp in its last few hundred meteres. None of use were however capable of that sort of stuff then. Can you do that?

Food for thought, it certainly is,
The Devs said that the escort AI cannot hear a torpedo

We can add additional values i.e. sensors to the AI_Sensors.dat, we can add additional nodes to hang a sensor on to escorts i.e. the ShipName.dat

I would love nothing better than to have a more difficult to hit merchant ship, but I cannot see adding a Hydrophone to the merchants as achieivng anything positive and only offering a negative.

I just tested a underwater 'visual' sensor to an escort ship and it did not work. This to me confirms my opinion that the AI visual has a no see through/under water attribute.

[gouldjg]

OK so that does not work :-? thanks for confirming it. Well then, apart from actually making DD have more sensors, what exactly are you hoping to achieve.

A year by year DD sensor setting can already be done via sh3 commander. Thats why I spent the time to put all the info down and ready for when I start to do it or if someone already wanted to bash at it.

Now I am not going to block my ears to new ideas but they do have to offer something, I am waiting on Observers Idea/drawing but again the same questions remain:-

Do you have a theory or exclusive Dev team info that will resolve some of the hardcoded issues discovered those whom participated in the DD thread?

Can you get it so the sub is pinned underwater and DD always attack after every ship is hit?

If not, were back on page 1 of the DD thread IMO,

This is not a flame or anything but after spending hours and hours on the subject, it was obvious that the vast majority of problems seemed to be hardcoded and thus everyone made best of what info was available.

Unless some new behaviour is promised and it is not going to be one mod compatible only, I would much rather stick with randomising what we do know to get my variety through SH3 commander.

I await to see Observers idea

[HEMISENT]

oh dear...just get on with it..i m sure there's little benifit to be had in slagging me off!!!

Hey CB
Welcome back, just letting you know that after all the "wasted " time from the DD thread in question I need to credit you for one of your offhand comments leading to me putting together the thermal layer concept.

Thank You even if it is a bit late.

[CB..]

:D no worrys ..Cheers Hemisent!

remember here folks that there are lots of generic ships in the game which we kinda figured used the generic sensors listed in the dat...this coupled with named DD's using one asumes the named sensors allows for a little more variety..trouble here of course is just how madd you are on the old realism thing..trying to get decent AI behaviour whilst sticking to strictly realisitc sensor ranges senistivitys et al might be impossible...
testing is a pain the arse IMo because the game just doesn't play out the same in single missions as it does in a full campaign mission...plus the most relevant encounters with DD's are whilst you are approaching a convoy..and here the results are unpredictable because of the sheer number of ships involved...BUT that this is the best place to test the results simply because it was the closest you can get to genuinely reliable information...hugely problomatic of course because of the changing weather and so on and so forth but what you are after are changes which work as best as possible under ALL circumstances...so IMO testing in a campaign mission is the only way to go...which is a vison of hell..good hunting !! hands up those who wuld at this moment settle for a lowered graphical representation sme game poorer graphics BUT with uber realistic AI....we all know that once the novelty has worn of every game lives or dies by it's quality of AI...not by it's graphics..

[Der Teddy Bar]

trouble here of course is just how mad you are on the old realism thing..
As usual I am aiming for realism, but there is in this realism the fact that a U-boat was not invisible or the escort as all seeing, the situation as it is now.

My realism for the Hyd/Sonar sensors is one where you can play a cat and mouse game with the escort and with skill & some luck win.

trying to get decent AI behaviour whilst sticking to strictly realisitc sensor ranges senistivitys et al might be impossible...
Very true. That is whey we are looking outside the square as the saying goes.

testing is a pain the arse IMo because the game just doesn't play out the same in single missions as it does in a full campaign mission...plus the most relevant encounters with DD's are whilst you are approaching a convoy..and here the results are unpredictable because of the sheer number of ships involved...BUT that this is the best place to test the results simply because it was the closest you can get to genuinely reliable information...hugely problomatic of course because of the changing weather and so on and so forth but what you are after are changes which work as best as possible under ALL circumstances...so IMO testing in a campaign mission is the only way to go...
I am not sure if you were saying this or not, but testing in a single mission using the campaign allows for proper testing.

That is, you do not set up the convoy etc, you simply put your u-boat some where close to where a campaign generated convoy will pass by.

This gives a proper response where as I feel making a convoy in the single mission to attack is not going to give the best results.

which is a vison of hell..good hunting !! hands up those who wuld at this moment settle for a lowered graphical representation sme game poorer graphics BUT with uber realistic AI....we all know that once the novelty has worn of every game lives or dies by it's quality of AI...not by it's graphics..
Me...


Cheers.

[Observer]

Here is an image of the basic concept. Keep in mind it's still WIP.

http://i17.photobucket.com/albums/b57/Observer723/sensorsv4.jpg

[gouldjg]

:up: Thanks for that Observer

Looks like your determined to produce something in this area :hmm: .

A few questions spring to mind

1. How has it behaved in game upto now or have you not got that far yet.

2. Der Teddy earlier mentioned what I thought was the fact that you may be able to add other specific criteria on each node/sensor rather than them sharing the default sim.cfg settings. Does this mean individual detection times and possibly aspect of sub being taken into account.

3. Finally, general release or NYGM only? i.e. Do people have to have the negative boyant sub compatible version or are you releasing a DD code and providing the hex offsets for each sensor so others can then make it fit their preferences.

[CB..]

I am not sure if you were saying this or not, but testing in a single mission using the campaign allows for proper testing.

That is, you do not set up the convoy etc, you simply put your u-boat some where close to where a campaign generated convoy will pass by.

This gives a proper response where as I feel making a convoy in the single mission to attack is not going to give the best results.

which is a vison of hell..good hunting !! hands up those who wuld at this moment settle for a lowered graphical representation sme game poorer graphics BUT with uber realistic AI....we all know that once the novelty has worn of every game lives or dies by it's quality of AI...not by it's graphics..
Me...


Cheers.

plenty of common ground then...so much so one wonders why you had to start the thread in such in idiotic manner..digging up a VERY old thread purely for the simple reason of slagging of those members taking part in that thread mainly because you are now working on the same thing and of course your product will be vastly superior..

take a pill man...

seems i can't even wish you luck with-out having my thoughts analaysed point by point...

here you go bud

GOOD

LUCK!!!! :up: :up: :up: :up: :up: :up: :up: :up: :up:

[Der Teddy Bar]

CB...
Well I was actually hoping that you would come on board.

In hindsight I obviously said what I had to say in a more direct way than necessary............

Cheers,

Teddy

[Der Teddy Bar]

:up: Thanks for that Observer

Looks like your determined to produce something in this area :hmm: .

A few questions spring to mind

1. How has it behaved in game upto now or have you not got that far yet.
To date we have tested that the idea was possible. That is, we have added an additional Hydrophone to the AI_Sensor.dat, the Type150. We tested that the Type150 worked as advertised, which it did.

2. Der Teddy earlier mentioned what I thought was the fact that you may be able to add other specific criteria on each node/sensor rather than them sharing the default sim.cfg settings. Does this mean individual detection times and possibly aspect of sub being taken into account.
The tayloring of the Hydrophone etc would be done by shaping the the min/max range for the height, bearing, elevation & sensitivity of the sensor to meet a specific purpose.

The only value that we can specify in the DAT is the 'Sensitivity'.

3. Finally, general release or NYGM only? i.e. Do people have to have the negative boyant sub compatible version or are you releasing a DD code and providing the hex offsets for each sensor so others can then make it fit their preferences.
If we get community help then it will be a community project and product, for which we will also use for NYGM TW.

This is not a HEX edit, it is being done with the NYGM Mod Tool. The same one that allows NYGM to rezone ships etc. Basically we hang a new node onto the escort, add an entry into the ShipName.sns file for that node and then hang the newly created sensors onto that node.

This would be a stand alone mod that is not associated with the NYGM U-boat mod or require any other NYGM product. But would benifit from it for several reasons which I will say later as I have a time constraint.

Cheers,

Teddy

[CB..]

well i don't especailly mind that teddy rather you were honest as it were--
and i have to say i'm not sure how much i actually have to offer...as we were testing just about anything and everything as it came up -naturally the process was hugely chaotic...with so many dead ends it was a wonder folks carried on at all...
that much hasn't changed in spite of the new information...

it's the testing that remains the issue...it's still a endless and hugely tedious process with very uncertain results...made worse by folks seemingly having very different problems..one group were testing for the removal of pin point drops.. the other mainly had the opposite problem in that the AI for the most part completely ignored them...for which the idea of experiencing pin point drops would have been seen as a step in the right direction gameplay wise (albiet very temporarily!! no doubt about that)--have you managed to isolate the cause of the differing problems as this is a major issue?

as i was one of the ones that was finding that i had to allow the DD's to detect me rather than them them selves doing the job i needed a less than subtle soloution..

only after COL777's discovery that the removal of certain entrys from the sim.cfg resulted in the DD's hydrophones going uber..did i then have that un subtle answer..the problem there-in was how to calm them back down again to a playable level..this was sort of achieved by limiting their maximum range to very unreleastic levels (at least on paper- in game the percieved detection ranges were not that far out) and givng them vastly increased minimum ranges to give the player some sort of chance to escape once having been detected..there by creating a hole immediately around the DD into which the player could temporarily escape --a blind spot ---then by reducing the time before lost contact entry in the sim cfg to force the DD to break off the search and re-establish a new search if and when he detected you again- creating a give and take dynamic that gave some sort of positive game play experience..

and that was basically about it..
none of this was actually touching the AI at all as we didn't have access to that side of the game..it was just trying to force the AI into behaving in a more interesting fashion gameplay wise by hampering certain aspects of it's sensors..whilst ubering others..this was not terribly use-full for those experiencing uber detection rates and pin point drops-tho it did raise some interesting posibilitys..regarding the minimum ranges of the sensors and that blind spot at close range--

one intrigue-ing side effect of this was that in the game at night- i found that the DD's and or merchants would use star shells far more frequently- even whislt i was submerged...and had yet to launch any torpedoes..they seemed to be behaving in a "do you know where the damn sub is?" sort of fashion...tho i never did get much further..that in it self was a clue to some thing.and if nothing else a satisfyingly human sort of behaviour--tho as we know it's extremely easy to see one thing when the exact opposite is actually the case..

i ended up with a hugely unrealistic sytem (again on paper) where by within a certain distance of a convoy (that being purely a ball park realistic distance) those DD's that were furthest away actually had a better chance of detecting you than those who were closer (LOL)-- but it was those closest to you that did the attacking..--exactly how and why this was the case had to be some sort of function of the AI (rather than a purely sensor thing)--so a DD on the other side of the convoy would appear to be directing at least some of it's information to those closest to you..or at least saying "hey buddy wake up" to those closest to you..
so on the star shell thing a DD on the far side of the convoy might actually launch a star shell as you approached you would dive--and the DD closest to you would then lob a few star shells in your general direction...and begin to nose about in you general direction....the whole process seemed to build slowly up untill you found your self under allmost constant attack....the balance of this would create the gameplay--the thing was that tho you were ALLWAYS detected...but this didnt mean that you were allways ATTACKED straight away...so you could lose them and try again...this is speculation of course..as with out some sort of DEBUG system in game to monitor exactly what is happening -it is just what APPEARS to be happening ....and may be -just maybe thats what actually counts?

i changed this set slightly in the mod as it had one or two problems..but it's basically the same trick-- the ubered hydrophones giving those DD's furthest away a better detection rate than those closest...if this sounds like a maze of contradictions it's because it IS a maze of contradictions!!!

have you considered that in real life it might have actually felt pretty much like that?

OH for the day when we can actually go into a store and buy a 4 GB "AI Card" to compliment all these flashy graphics cards endlessly parading before our eyes!!!

then we can have AI that behaves in a genuinely human fashion..and that will the most unpredictable experience one could imagine!! he he!!

whats the story on the Devs imput??

[gouldjg]

It certainly sounds interesting and quite a big job. I am sure you and Observer have tried the alternatives so I look forward to seeing the differences.

Basically what I took from the DD thread was the fact that Hydrophones needed pumping up one way or the other and there seemed to be a number of ways to do this.

In my own circumstances, simple addition of 5000mtrs and set all sensitivities to 0.03. Yes I do get to go silent and cloak at times but when close, even silent needs to be notched down to 1 knot and usually actives had taken over by then anyway. I do have to say, I have never really tested a full campaign on this setting because I tend to always play with things before I get to the bigger subs. Feedback at the time by other members seemed to indicate that Adding 5000 mtrs approx was making them get detected on silent all the times which leads me to believe either a extreme rare case or sub class or different people just have different game outputs which seems hard to believe but it did cause a good few short fuses back then.

The rest was down to personal preference. Either way, it always got predictable after a while so I jumped on randomisation when it became possible. Even doing the min visual heights so sometimes spotted when periscope is just out of the water sometimes not.

Sonar arcs, elevations etc all slightly jumbled within a reasonable scales so I am in a position of only knowing the basic limitations of DD. Add on that the chance to use Hemisents thermal layers, I am now at the opinion that the active arcs can possible be made somewhere back to their slightly uber setting occasionally as determined by a chance factor and this can pick the game back up a notch or two.

To be honest, I am quite happy at that and it is only the hardcode issues that get on my nerves i.e. DD not being nosey enough on attacks. DD not doing any search patterns after lost contact. Sometimes only 1 and sometimes 3 DD attacking.

DC's were either spot on or way off so I now just juggle between the two so I can never really get complacent and yes, I do occasionally die despite my best efforts at escape.

They are the issues that bug me at the moment.

Now as the game stands at the moment unlike in real life is the fact we

1. Can outwait the games DD because they do not suffocate us under the water.

2. The repairs of the sub are just too efficient and nothing much forces the game players to have to surface and surrender. In real life things went wrong and sub had to surface and this was not always after 13 hours it sometimes happened within the first hour.

If you remember, I wanted to find a way to get rid of the repair station all together. I was looking at possible ways to get flooding that was unfixable under water and thus leaving a limited amount of time to stay submerged should one be hit by a close DC. I have not quite figured on the best way to do that but I know that it is a must have for me. I know the conning tower is a possibility if I make it also have a floatability factor on the sub but we get no visual representation of how much time left. Much better to use a crew rest quarters in a kinda clock is ticking way.

[TowerWatch]
Multiplier=1.000000
Flotability=100
HitPoints=330
Destructible=No
Armor Level=15
Critic Flotation=0.70
Critical=No
FloodingTime=2000
CargoType=None

In example above, should you be badly hit near conning tower, you will slowly lose depth and therefore require more engine power to stay boyant and eventually have to rise to repair or go to crush depth. It is my only solution to force surrender in the game. I would have much prefered it in the crew rest areas so I could see what time I had. Consider it a sort of outside the box view on unrepairable conditions that can occur.


Anyway getting back onto topic

I am interested if it adds gameplay and difficulty without going too far outside of the square. In other words, how far outside the realms of realism are you thinking of going (gouldjg shows his secret realism side here :cool: ).

[CB..]

one of the things that allways bugged me specifically about the AI behaviour was that on first getting the game installing it and playing the first few missions and starting a campaign--the AI was great..was one of the things i was enjoying the most!!! i even chatted about it with a neighbour i was so impressed!! (he is a PC gamer too)

DD's co-operating in a very satisfying manner..once two or more had reached the area one would creep along above and slightly to the side whilst another would hare round at top speed conducting the DC drops...looked and was very effective and interesting...and was unmistakable as co-operation....as the creeper would maintain it's position over head and slightly to one side and behind- matching both my course and speed exactly untill i had managed to lose it or it/they had killed me...the DD's generally had no problems detecting me at all and none of the issues that came up later were there at all...but after only a few hours of play this just quite simply stopped happening..i kept waiting for it to happen again..but it never did....(no mods at this point either)--all of sudden the DD's just went dumb...even a re-install didn't make any difference..even another re-install made no difference..summat somewhere regarding the AI behaviour just died.. i could never find any reason for this....i wonder if any one else experienced anything even remotely similar because that would be something relevant...in fact all the testing etc i did was really just me trying to get back some of that fun gameplay that had vanished after only a few hours of gameplay on originaly buying the game...it wasn't me learning how to evade etc...the DD's just stopped co-operating in the same effecient manner and began to behave more like a dumb version of SH2's DD's (attacking in pairs but with no real descernable or effective co-operation-that's when they detected me at all..i usually had to surface in order to wake them up- which was useless)

have to say i never had any real plan about this..just to get the game back!
any one else have any similar experiences of the AI abruptly going dumb?


on the additonal sensor thing (sound highly usefull!)
you could try giving the DD's two hydrophones--one which is slightly ubered but has a very limited arc- perhaps only 5 degress left/right of dead ahead and a much less sensitive one to cover the rest of the sensor arc...this should produce unpredictable yet relevant DD behaviuor..?

[Kpt. Lehmann]

Observer, please excuse me, but I wholeheartedly agree that a NEW DD agression thread should be opened. The GW team objects to the obvious slap in the face by Teddy Barr and allowing it to go un-answered is unacceptable.


Teddy Barr.

The GW team has had enough of your attitude problem and "holier than thou" view towards our efforts.

Here and now I am going to address your disdainfulness and clarify issues you have painted in an incorrect light for the community.

They show a total lack of proper
testing and understanding of moding.

The idea that just because you take add additional 'data'
under the group heading it will get used is just silly.

...all they have done is add some useless values.

Now CB has really hacked it up. He has removed the
original entires of Detection time, Sensitivity & Height
factor. Then he has added a lot of useless values.

...improper procedure & testing in the hopes that next time
there will not again be thousands of hours wasted.


We will no longer tolerate your efforts to discredit the
GW mod on a case by case/file by file basis.


GW has stated that they will not use any NYGM mods as per
Kpt's post.
I believe that this stems from NYGM's request that GW
does not make changes to any of the NYGM work that GW
includes as it would then possibly lead to NYGM mods
being percieved incorrectly.

Not only have you mis-quoted me, but you have twisted
this statement to meet your needs.

The joint merger was cancelled/terminated by the GW team
because you denied us access to the NYGM/GW merger files
you posted on FileFront.com.

Furthermore,regarding your concern that the use of the
NYGM ship damage model as used and or tweaked by the GW
team potentially showing a bad light on the NYGM team...

The same can be said of your mod by the GW team using the NYGM ship DM's. If it didn't need fixing, we wouldn't have needed to
tweak it.


A case in point is the Tug Boat that GW supplied had been
moded by GW and as a result had no damage zones, at all
and is why it would not sink. NYGM discovered this for GW
and let them know.

Here! Eat this cookie while I play the violin for you.
It was a simple data entry error that even you could
make. Thank you for helping us fix it... Now come down from the mountain!



As the above example demostrates though, it was the NYGM
Ship Damage Mod that was made to look bad as well, its
the bit that makes the ships sink the way that they
do.

Again,regarding your concern that the use of the NYGM
ship damage model as used and or tweaked by the GW team
potentially showing a bad light on the NYGM team... The
same can be said of your mod by the GW team. If it didn't
need fixing, we wouldn't have needed to tweak it.

You also do not recognize that the success that GW has
enjoyed, has also brought POSITIVE recognition to the
NYGM ship damage models time and time again.


Kpt said that he will do what ever they pleases with
NYGM's Ship Damage Mod... I objected... and thus the NYGM
& GW merger is a thing of the past.

NEGATIVE! Again, We (The GW team) made the decision to
cancel the NYGM/GW merger as you refused us access to the
files you posted on FileFront.

Again you mis-quoted me. What I said was this:

"Please note, the GW team may alter any included file to
mesh as required by the needs of the GW Supermod for the
sake of (file) cohesion or to achieve a desired effect.
This is how discoveries and better mods are made."

The above is a necessary statement which allows us to
attempt on-the-spot fixes in ANY file where necessary.

You act as if this was tantamount to something
SEAWOLVES/X1 might do... which you have ACCUSED US OF
BEING SIMILAR TO IN PRIVATE SIMPLY BECAUSE WE DECIDED TO
USE YOUR SHIP DM!!! LOL, not only did we credit you as
the builder/designer, we paid you high praise.

You forget yourself. Just as you do, we spend our free
time modding SH3 attempting to improve the game for
players that wish to use mods in their games. Where GW
is concerned, NO OTHER mod team has had the courage to
build and release a package that addresses so many
bug-fixes, realism issues, sound and graphics
improvements... in the same package.

Where CB is concerned... he had the courage to address
the DD behavior issue and we followed that thread. We
included the best information available to us at the time
of release. I could go on and on here. Not only are
you insulting CB... but everyone who contributed to that
thread.

About ten days before the release of GW 1.0 DTB, you
urged me to combine our efforts with yours in order to "not
factionalize the community..."

DTB... your thoughtlessness and total lack of respect for
others feelings and efforts shines through with peerless
clarity. If anyone has "factionalized" the community, it
has been you... as demonstrated by your total disregard for the efforts of others.

Excercising a little tact will not only serve you well in this
forum... but in real life as well.

I suggest that you do whatever you must to learn to
co-exist. The GW mod team isn't going anywhere and GW
WILL REMAIN (more certainly now than ever) a STAND-ALONE
mod offered to the community as an installable option.

Having freedom of CHOICE enriches the community... It does not fracture it!

Not only do you owe the GW team an unconditional
apology... but you owe one to CB for trashing his
efforts, and to your own team for embarrasing them.

Ende.

[caspofungin]

there's plenty of data in the intial dd thread about sensor capabilities for those willing to trawl thru it all. i personally have my own set up that i'm satisfied with, but for me the bigger issue is the scripted ai.

what i mean is, eventually we can get (or have already gotten) the sensors to work as you'd expect them to -- decent ranges for sonar, s certain noise level for hydrophones, beam geometry, etc. all that can be tweaked,

what can't be fixed is what the ai does once you're actually detected and you're position is pinpointed. it's always the same -- the escort come towards you, starts weaving, drops a few charges, circles and tries again. if there's more than 1 escort, they both circle -- sometimes that gives the impression of the famous "creeping" attack. other than that, ther's not a lot else. no patrol lines for when they lose contact, no firing of multiple starshells all over the convoy whena ship's been hit (sometimes the dd's will fire a starshell over your expected position even if you're submerged, but hardly the barrage of light that's meant to catch a sub on the surface).

the bbc has a history website with a portion devoted to the battle of the atlantic -- it's got a few examples of the allied tactics -- plaster, raspberry, etc -- standard tactics for standard situations. you should check it out.

you'd think that wouldn't have been too hard to code in -- but nevertheless, it's going to be impossible for us to simulate without dev input.

so, fixing the sensors may be a great idea, but people shouldn't assume that it's going to be some panacea that will revolutionize the game. sorry to be so down, but that's how i see it.

i've plenty of historical data on allied sonars -- if anyone wants it, just ask.

[Observer]

There's nothing more constructive to be gained here.

Mods: please lock this thread.

[Kpt. Lehmann]

There's nothing more constructive to be gained here.

Mods: please lock this thread.

I and the GW team agree and concur as long as Teddy Barr has viewed/is aware of the details of our response.

His behavior is unacceptable and will not be ignored.

[The Noob]

any one else have any similar experiences of the AI abruptly going dumb?


I had a Few Months ago... :damn:

[CWorth]

I think this thread needs some proper theme music.. :roll:

Click Here (http://www.tv-timewarp.co.uk/midi_files/ExternalLinkFilter/DaysOfOurLives.mid)

[JScones]

I think this thread needs some proper theme music.. :roll:

Click Here (http://www.tv-timewarp.co.uk/midi_files/ExternalLinkFilter/DaysOfOurLives.mid)
<spits out coke all over keyboard> :rotfl: :rotfl:

[The Noob]

<spits out coke all over keyboard> :rotfl: :rotfl:

Same happened to me. Seriously!

[Onkel Neal]

Please, let's keep the tone friendly.