Jump to content

Paging Karl Berg...karl Berg, Please Pick Up The White Courtesy Phone...


1911 replies to this topic

#761 Cimarb

    Member

  • PipPipPipPipPipPipPipPipPip
  • Caladbolg
  • Caladbolg
  • 3,912 posts
  • Twitter: Link
  • Twitch: Link
  • LocationA hop, skip and jump from Terra

Posted 19 May 2014 - 05:49 PM

View PostWintersdark, on 19 May 2014 - 04:23 PM, said:

In live matches, I always fight.

I was just giving you a hard time - sorry if that didn't translate.

#762 Wintersdark

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • 13,375 posts
  • Google+: Link
  • Twitter: Link
  • LocationCalgary, AB

Posted 19 May 2014 - 05:57 PM

View PostCimarb, on 19 May 2014 - 05:49 PM, said:

I was just giving you a hard time - sorry if that didn't translate.

heh no worries, you know how things can be on forums. I always assume people do mean such things as a habit, because unfortunately they often do. Unfair in this case, as I know you're a fairly reasonable sort, but it's habit :(

#763 Cimarb

    Member

  • PipPipPipPipPipPipPipPipPip
  • Caladbolg
  • Caladbolg
  • 3,912 posts
  • Twitter: Link
  • Twitch: Link
  • LocationA hop, skip and jump from Terra

Posted 19 May 2014 - 07:59 PM

View PostWintersdark, on 19 May 2014 - 05:57 PM, said:

heh no worries, you know how things can be on forums. I always assume people do mean such things as a habit, because unfortunately they often do. Unfair in this case, as I know you're a fairly reasonable sort, but it's habit :(

I seem to be missing my mark lately, so totally my fault.

Karl, any luck on getting some of your compatriots to share the workload of keeping us happy? Not that we don't love you, but you have exhausted most of our questions already...

#764 Gizmoh

    Member

  • PipPipPip
  • Survivor
  • Survivor
  • 78 posts

Posted 19 May 2014 - 08:05 PM

You know what would be cool? Dennis de Koenig dropping by to tell us a story. :(

#765 Shlkt

    Member

  • PipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 319 posts

Posted 20 May 2014 - 04:38 AM

Re: heat retention
I've investigated a number of claims regarding undocumented heat issues or bugs, and in each case the heat system was working correctly. Say what you want about our heat system but it's at least predictable. The math lines up perfectly with in-game behavior.

#766 Cimarb

    Member

  • PipPipPipPipPipPipPipPipPip
  • Caladbolg
  • Caladbolg
  • 3,912 posts
  • Twitter: Link
  • Twitch: Link
  • LocationA hop, skip and jump from Terra

Posted 20 May 2014 - 06:28 AM

View PostShlkt, on 20 May 2014 - 04:38 AM, said:

Re: heat retention
I've investigated a number of claims regarding undocumented heat issues or bugs, and in each case the heat system was working correctly. Say what you want about our heat system but it's at least predictable. The math lines up perfectly with in-game behavior.

That depends on your definition of "correctly", lol.

I want a heat scale that works like TT. Where getting to 99% heat means you FEEL like you are about to shutdown, with actuators groaning under the stress, ammunition popping as it cooks off randomly, and your electronics crackling and smoking in agony.

Right now, shutting down in a mech feels like a car with a governor: sure, it stops you from going "too fast", but it also makes you oblivious to the consequences of going "too fast" (the scared feeling we all know from personal experience of going just a wee bit too fast that one time in high school...), so we just keep flooring the accelerator and driving as stupidly as always.

(Btw, I'm not arguing about the math - I'm glad it at least does what PGI designed it to - I'm just complaining about the poor choice in system to begin with)

Edited by Cimarb, 20 May 2014 - 06:29 AM.


#767 SgtKinCaiD

    Member

  • PipPipPipPipPipPipPipPip
  • Overlord
  • Overlord
  • 1,096 posts
  • LocationBordeaux

Posted 20 May 2014 - 06:36 AM

Even if the 3/3/3/3 is not coming shortly, can you consider turning the elo buckets back on ?

#768 Iqfish

    Member

  • PipPipPipPipPipPipPipPipPip
  • 3,488 posts
  • Google+: Link
  • LocationGermany, CGN

Posted 20 May 2014 - 07:50 AM

View PostGizmoh, on 19 May 2014 - 08:05 PM, said:

You know what would be cool? Dennis de Koenig dropping by to tell us a story. :P


agreed, please do so!

#769 Bad Karma 308

    Member

  • PipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 411 posts

Posted 20 May 2014 - 10:10 AM

View PostKarl Berg, on 08 May 2014 - 09:14 PM, said:

..................
We're considering dropping the Cryscaleform wrapper completely, and just doing our own Scaleform 4 integration from scratch; but this is obviously a very large amount of work to commit to. So we'll have to see..


I've certainly seen that your optimization has improved over the last several patches. ASK THE DEVS - VLOG #3 also briefly mentions SLI/Crossfire implantation is being worked on. I'm also reading that this patch (20-May-14) (pre-released patch notes) will also include some performance enhancements which aren't specified.

So a quick question: Do you guys have a set design point when you hand off your work to either of the two GPU houses for driver optimization/integration? Or would the optimization fall under the parent CRY engine driver optimization?

Edited by Bad Karma 308, 20 May 2014 - 11:47 AM.


#770 Tekadept

    Member

  • PipPipPipPipPipPipPipPip
  • Knight Errant
  • 1,290 posts
  • LocationPerth, Australia

Posted 20 May 2014 - 09:55 PM

Posted Image
Keeping busy I hope... Any ideas on the point of failure for issues with the new matchmaker yet? is it the matchmaker algorithm, lack of people or "reasons"?

It really seems a bit like
Posted Image

#771 DeathlyEyes

    Member

  • PipPipPipPipPipPipPip
  • The Messenger
  • 940 posts
  • Google+: Link
  • Facebook: Link
  • Twitter: Link
  • Twitch: Link
  • LocationMetaphorical Island somewhere in the Pacific

Posted 21 May 2014 - 05:24 AM

How hard would it be to implement a dropship game mode? By a dropship gamemode I mean you drop in with up to 4 mechs with a set tonnage limit. You fight with the first mech until you are destroyed and then you are dropped into battle in a second mech of your choosing, so on until either you run out of mechs, win the match or lose the match.

#772 zudukai

    Member

  • PipPipPipPipPipPipPipPip
  • Trinary Star Captain
  • Trinary Star Captain
  • 1,707 posts

Posted 21 May 2014 - 11:14 AM

View PostSLDF DeathlyEyes, on 21 May 2014 - 05:24 AM, said:

How hard would it be to implement a dropship game mode? By a dropship gamemode I mean you drop in with up to 4 mechs with a set tonnage limit. You fight with the first mech until you are destroyed and then you are dropped into battle in a second mech of your choosing, so on until either you run out of mechs, win the match or lose the match.

what ever happened to that idea? you can not "ready" mechs anymore, nor is there a menu for selecting mechs in this manner.

#773 Mawai

    Member

  • PipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 3,495 posts

Posted 21 May 2014 - 11:16 AM

Hi Karl,

Thanks very much for your interactions in this forum ;)

I had two quick questions that you may or may not be able/allowed to answer ..

1) What happened to missiles in the last patch? Did someone put something into a base class that subsequently affected all missiles instead putting it into a clan missile derived class? (or something like that B) ).

2) Any update on the matchmaker? According to the announcements, there was a lot of good telemetry data obtained and I am just curious what sorts of issues are arising with the matchmaker? Is it solely a queue time issue or are there other problems? I'm mostly curious to be honest since the whole matchmaker is actually a pretty cool multi-variate optimization problem. If you do a good job you might even be able to patent the algorithm :lol:,

#774 Karl Berg

    Technical Director

  • 497 posts
  • LocationVancouver

Posted 21 May 2014 - 04:20 PM

View PostTichorius Davion, on 18 May 2014 - 02:42 PM, said:

Karl, more questions are surfacing with the HUD bug coming back.

The paper doll with the missile pods has been shown. Is there any functionality on this that you can tell us about?


Certainly. Our systems are build with support for 12 individual components, since this was the maximum number we were told we would ever have to support within the battle tech universe. We track health, armour, and equipment for all 12 of those components. UI did the same when they developed the paper doll display for the HUD. Whether or not design decides to utilize those 11th and 12th components at some point is a completely separate question though. As far as I am aware, there are no current plans to activate those two remaining components for Clans.

View PostFeatherwood, on 18 May 2014 - 02:16 PM, said:

Thank you, sir!


Sorry sir, I haven't had a chance to get answers for you yet with all the patch madness taking place. I'm actually out of office right now, but I will try my best to have answers for you by tomorrow.

#775 FactorlanP

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,576 posts

Posted 21 May 2014 - 04:34 PM

View PostKarl Berg, on 21 May 2014 - 04:20 PM, said:


Certainly. Our systems are build with support for 12 individual components, since this was the maximum number we were told we would ever have to support within the battle tech universe. We track health, armour, and equipment for all 12 of those components. UI did the same when they developed the paper doll display for the HUD. Whether or not design decides to utilize those 11th and 12th components at some point is a completely separate question though. As far as I am aware, there are no current plans to activate those two remaining components for Clans.



Sorry sir, I haven't had a chance to get answers for you yet with all the patch madness taking place. I'm actually out of office right now, but I will try my best to have answers for you by tomorrow.



Those two remaining components should be used for two more legs, if they ever introduce the quad mechs.

#776 Karl Berg

    Technical Director

  • 497 posts
  • LocationVancouver

Posted 21 May 2014 - 06:07 PM

View PostHammerhai, on 19 May 2014 - 12:37 PM, said:

Karl, Please look at the following post:
http://mwomercs.com/...7#entry3351077.

And please excuse the emoting there. I was annoyed. I do think anybody would be if his laser alpha potentially has ghost heat and 20 actual damage over time, not 36 as advertised

The questions arising from this:
Is there a laser hitreg problem as well? Disclosure: I believe this to have been and still be the case since closed beta.
My point: At 100% hits the damage per shot for large lasers is half of what it should be. Namely 5.

In the alternative: Does getting one tick per shot on target count as a "hit". The fact that LL's register 5 tells me that a mean is being calculated here, not a damage per tick sum. In any case then it should be closer to 5.5 plus secondary damage.

So I would like to know: Exactly how is that number being calculated?

The way I came across this was with Sjur War Eagles Improved statistics program, which showed the number of hit per weapon pretty much as the client reports it to the stats page, but you get to see the individual game, not the aggregate. Sadly, the tool does not work anymore, and development has ceased.

If I am in error, I think you are my last resort at finding out from somebody in the know. Putting my threads in the morgue for a potential current issue will not make me go away, moderators. (***Looks in the direction of offending moderators in general.***) As always, if you cannot answer, thank you for at least reading about it.

And I would like to remind us all to be gentle on Karl, he does have a private and business life to run as well


Hi! Lasers and other trace-fires have the simplest hitreg of any weapons, short of server side homing weapons such as LRM's and streaks. Lasers are done as a set of traces, where each trace is an instantaneous check done against fully rewound positions on the server to determine whether or not a hit has occurred. If a trace hits, it deals that amount of damage integrated with respect to the amount of time that particular trace represents compared to the whole fire event. This is very classic latency compensation found in most other game engines.

Underlying this is the fact that laser damage is a quantized process, based on the number of ticks the fire event performs. Unfortunately I don't know how you performed your damage output tests; so I can't tell you why you're seeing differences between expected and actual output. Hopefully the description above gives you some insight.

#777 Karl Berg

    Technical Director

  • 497 posts
  • LocationVancouver

Posted 21 May 2014 - 06:13 PM

@Tekadept

Indeed; a fairly brutal day. Things are still fairly hectic, but I wouldn't be surprised if a writeup gets released on our plans for the matchmaker.

#778 Karl Berg

    Technical Director

  • 497 posts
  • LocationVancouver

Posted 21 May 2014 - 06:20 PM

View PostBad Karma 308, on 20 May 2014 - 10:10 AM, said:

I've certainly seen that your optimization has improved over the last several patches. ASK THE DEVS - VLOG #3 also briefly mentions SLI/Crossfire implantation is being worked on. I'm also reading that this patch (20-May-14) (pre-released patch notes) will also include some performance enhancements which aren't specified.

So a quick question: Do you guys have a set design point when you hand off your work to either of the two GPU houses for driver optimization/integration? Or would the optimization fall under the parent CRY engine driver optimization?


AMD, Intel, and nvidia all have representatives that consult on performance issues. I'm almost certain that these same companies send representatives to work with CryTek directly as well. If they do perform driver-side optimizations, they wouldn't really need to consult with us; as they could easily instrument D3D to tell them exactly which calls we make, when during the frame, and with what data.

View PostCimarb, on 19 May 2014 - 07:59 PM, said:

I seem to be missing my mark lately, so totally my fault.

Karl, any luck on getting some of your compatriots to share the workload of keeping us happy? Not that we don't love you, but you have exhausted most of our questions already...


Believe me, I've tried.. :ph34r: Dennis, Brian Buckton, and Fox have both popped in at some point to clarify specific points. Maybe if another thread were to be opened, paging some of those other dev's?

#779 Karl Berg

    Technical Director

  • 497 posts
  • LocationVancouver

Posted 21 May 2014 - 06:41 PM

View PostMawai, on 21 May 2014 - 11:16 AM, said:

Hi Karl,

Thanks very much for your interactions in this forum :ph34r:

I had two quick questions that you may or may not be able/allowed to answer ..

1) What happened to missiles in the last patch? Did someone put something into a base class that subsequently affected all missiles instead putting it into a clan missile derived class? (or something like that :unsure: ).

2) Any update on the matchmaker? According to the announcements, there was a lot of good telemetry data obtained and I am just curious what sorts of issues are arising with the matchmaker? Is it solely a queue time issue or are there other problems? I'm mostly curious to be honest since the whole matchmaker is actually a pretty cool multi-variate optimization problem. If you do a good job you might even be able to patent the algorithm :wacko:,


1) Indeed, some clan related refactors did make it in for this patch. The actual culprit turned out to be a knock on from server side explosions however. Stuff that's been in branch test for quite some time, but wasn't caught by QA before it went to production unfortunately.

2) The player base actually did an admirable job of weight queue balancing! Far better than I expected to be honest. Even with 3's disabled right now, the presence of that queue display is still strongly affecting player behaviour. The reasons for having to regress to the old MM came down to specific implementation issues, as opposed to design viability. Hopefully we will be able to go into more detail with a specific post on this topic to clear up exactly what our future plans are.

#780 Kageru Ikazuchi

    Member

  • PipPipPipPipPipPipPipPip
  • The Determined
  • The Determined
  • 1,190 posts

Posted 21 May 2014 - 06:53 PM

View PostHammerhai, on 19 May 2014 - 12:37 PM, said:

- snipped because it's quoted above -

My observation of laser statistics ... if you swipe your laser ever so briefly over an enemy 'mech, it counts as a "hit" for weapon statistics purposes (which is why lasers has by far the highest "accuracy" (hits / shots fired) of any weapon in my arsenal).

However, that brief swipe only does a small fraction of the potential damage ... unless you're on target for the entire beam duration, you don't do maximum damage (which is why "effectiveness" (damage / (max damage * shots fired))) would be a much better metric.

(Edit: I've seen very, very few players who can actually maintain the beam on target for anything close to max duration, most of the time. Your final point in your original, original post is correct ... this is why most of the top players strongly prefer ACs and PPCs over lasers ... in each hit, all of the rated damage is guaranteed to be dealt to one location.)

All that said, from reading both of your posts, I can't tell how you're doing your math (and if you're mad at the right things). Also, with your 300-ish ping, the potential exists for HSR to be a very small fraction of a second off, but in most cases, that shouldn't affect your effectiveness that much.

Edited by Kageru Ikazuchi, 21 May 2014 - 07:03 PM.






6 user(s) are reading this topic

0 members, 6 guests, 0 anonymous users