75 replies to this topic

[Kyle Polulak]

Tell us what you think of the November Technical Update from the Dev Team.

Spoiler

Matthew Craig, on 01 November 2013 - 01:01 PM, said:

Time for another Command Chair update to help shed some light on what’s currently taking up our time internally here at PGI, so without further ado what has engineering been up to?

DX11

As most of you are aware DX11 had some last minute critical bugs not entirely dissimilar to some of the black screen issues that have plagued production (it remains to be seen if the fixes for DX11 translate to DX9) we finally got to the bottom of these issues for DX11 this last week and it is moving back to QA to setup for a public test shortly.

Dynamic Pricing/Bundles

Every Friday and Monday we have to cycle the servers to accept new pricing information for weekend sales, we’ve desired for some time to not have to do this and this work is now completing to allow us to dynamically price items in the game without downtime. In addition there has been work done to allow us to sell bundles. These features allow for more dynamic and varied sales without any server downtime and will be released in conjunction with UI 2.0

UI 2.0

As many of you recently experienced UI 2.0 is moving along nicely there is still plenty to be done but Public Test has allowed many of you to experience it in its current state and provide feedback that will be incorporated into the final state.

Something we realized during the DX11 work is that the switch between windowed mode and full screen going into game opens up a lot of edge cases for the renderer e.g. handling you alt tabbing to a browser during the load into game.

We plan to remove this complexity in the upcoming UI 2.0 and keep the mode consistent front end to back, meaning that if you run windowed when you go into game the game will be windowed and similarly if you run full screen the Mechlab will be full screen and it will retain full screen as you transition into game.

Hit Detection

We are aware there are still some issues surrounding hit detection, particularly with SRMs and also certain Mechs hitboxes. As Paul recently posted hitboxes are being addressed and we are scheduling time for a closer look into why SRMs are suffering a loss in accuracy compared to other weapons.

Matchmaking

We know some high Elo players have been finding longer wait times to get into match, we’ve been monitoring the telemetry and there is no easy balance here if we loosen the values more to allow these players lower wait times they will start to match more readily with lower Elo players resuming the state we had before where new players would face more punishing competition than desired.

Ultimately we agree that the current status is not ideal and we are working towards addressing all these short comings for the new Pre Game screen that will be introduced shortly after UI 2.0 this will introduce tonnage limits and is being designed to handle faction play etc.

It will represent a significant shift in the way matches are made compared to the current system and as mentioned will incorporate all our learning’s so far, into what we hope will be a much more rewarding and balanced way to find a game.

Stability

We know there are still users experiencing freezing and black screen issues on production, we have been investigating but are still unable to reproduce this issue reliably or see any discernible pattern in the crash dumps we’ve seen.

If you can reliably get these issues and are willing to work with us to help resolve them as always please head to the Patch Feedback thread and post in the Crash and Stability Feedback thread we’ll do our best to work with you to get these issues resolved.

As discussed above a lot of stability work went into DX11 this work will be back ported to DX9 where applicable. You may also find that at least one render path or the other works better for your machine, we’d love to hear your feedback during the upcoming public test for DX11 so please come and join us when a date is fixed.

Disconnects

Since closed beta we’ve been tracking a mysterious slowdown that would occasionally get bad enough to backlog our users and cause users connections to drop it was finally resolved (hardware configuration issue in the data center) and since then we have not seen any recurrence of these issues.

If you continue to experience connectivity issues please post in the Patch Feedback forum in the Performance and Connectivity thread and again we will do our best to work with you to identify and resolve any issues that are game related.

Performance

We continue to monitor and work to improve performance, the next target for performance improvement is DX11. Our initial profiling has shown that out of the box the CryEngine DX11 render path is slower than the DX9 path so we have an uphill battle on this front. We’re going to be looking at what we can do to bring that performance above DX9 levels if possible (given the DX11 feature set we’d be surprised if this is not attainable).

If you have particular hardware or a setup you think is underperforming please notify us of it in the Patch Feedback forum and we’ll do our best to get it into rotation in our test lab and work to reproduce what you are seeing.


Hopefully this helps shed some light on what has been ongoing internally recently for us, we’re excited to get DX11 and UI 2.0 wrapped up and delivered to help lay the foundation for future work. These changes really do help to underpin a lot of future performance/stability/compatibility work not to mention improvements to the games visuals and new feature work.

These releases should hopefully mark a transition to a higher level of fidelity for the game and set the framework we’ll be working within for the years to come.

[FuzzyLog1c]

If you want to improve wait time for high ELO players, try the following:

- Increase SHS and DHS dissipation and severely reduce the max heat cap.
- Remove the Gauss charge mechanic.
- Loosen up the hill speed mechanic.
- Remove "ghost" heat.
- Implement voice chat.

That's literally all that's standing between you and hundreds of thousands of dollars of immediate funding. By my estimation, over the past two months, forum traffic has dropped to somewhere between 1/6th and 1/8th what it was since you guys thumbed your nose at the people that made this game possible. If you stay the course, I predict that the average forum population will drop to below 1000 participants by the end of the year, and from there, all that's left is the inevitable quick slide and hard landing. You can already see the same old topics building steam with the second generation members... MWO won't survive the second communication apocalypse.

Also, if you're having problems with performance, start supporting SLI and Crossfire. Based on a survey of about 30 former MWO players, two-thirds of them have either SLI or Crossfire rigs. MWO works on both platforms using simple control panel overrides. The only downsides are minor graphical artifacts (occasional flickering, for instance). It shouldn't be that hard to finish the job.

Edited by FuzzyLog1c, 01 November 2013 - 02:38 PM.

[Jman5]

I would like to see more discussion on the upcoming weight limits. Specifically what your min/max numbers will be. In the ATD 45, Paul used this chart strictly for demonstrative purposes. Have you guys nailed down anything yet?

[Agent 0 Fortune]

FuzzyLog1c, on 01 November 2013 - 02:36 PM, said:

If you want to improve wait time for high ELO players, try the following:

- Increase SHS and DHS dissipation and severely reduce the max heat cap.
- Remove the Gauss charge mechanic.
- Loosen up the hill speed mechanic.
- Remove "ghost" heat.
- Implement voice chat.

That's literally all that's standing between you and hundreds of thousands of dollars of immediate funding. By my estimation

You almost forgot to mention that those changes also: cure cancer, burn fat, increase libido, build muscle, and improve memory. Oh wait, they wont do any of those things or any of the things you suggested either.

As for Matthew's comments: UI 2.0, Matchmaking, and Hit detection are all high on my priority list, I'm to hear you are working diligently on them. Although I am not sure what you can do about "high elo players not finding matches", if there isn't a large enough pool of high elo players, I don't know what you do but allow smaller team sizes or longer wait queues. Or as you mentioned, put them in general population (if they are solo), but if you do there really should be a hard limit on the lower end of the Elo they can be matched with/against.

[FuzzyLog1c]

Agent 0 Fortune, on 01 November 2013 - 03:51 PM, said:

You almost forgot to mention that those changes also: cure cancer, burn fat, increase libido, build muscle, and improve memory. Oh wait, they wont do any of those things or any of the things you suggested either.

I know. But someone occasionally needs to remind senior management (and I mean IGP, not PGI) that there's still a big, eager, wealthy fanbase out there. I haven't played MWO since the "this game isn't for you, Mechwarrior fans" post on principal alone. Judging by the lack of recognizable names on the Teamspeak servers (most of whom chipped in $200+), I'm one of many.

Like many others, I've been quietly observing the slow decay of the game, hoping that a change in direction will come. Or that the whole thing will fireball in an epic way, permitting a hungrier and sharper group of developers to take their shot at gaming glory.

Edited by FuzzyLog1c, 01 November 2013 - 05:39 PM.

[Deathlike]

That post was generally better than the overall quality of the "Creative Director" update and therefore I approve of it.

I'd rather see this more than the Creative Director update, because it shows a more clearer direction than the wait for the next hero mech and new map that we know we're going to get anyways.

[TexAce]

Deathlike, on 01 November 2013 - 05:38 PM, said:

That post was generally better than the overall quality of the "Creative Director" update and therefore I approve of it.

I'd rather see this more than the Creative Director update, because it shows a more clearer direction than the wait for the next hero mech and new map that we know we're going to get anyways.

If you ask me the 2nd staff (Thomas and Matthew for example) could take over all the communication from the 1st row staff (Paul, Bryan, Russ)...

They do a much better job at explaining problems and solutions to them.

Listening to Bryan, Russ or Paul is like listening to politicians.

Edited by TexAss, 01 November 2013 - 06:28 PM.

[Philadelphia Collins]

Fist of all thank you for this update it is appreciated. Would also like to say well done on the professional delivery seems like I enjoy reading every post Mr. Craig makes they are informative and straightforward.

I hate to say it but the overall the things discussed in the technical update seem like deja vu to me. First hints of UI2.0 was a long time ago and the wait is grating. It is what it is I guess.

Matchmaking has been pretty off since forever. Adding a cap to the number of preformed team members prevented more tactical and coordinated play but also reduced the number of outright 8-0 stomps (well sorta people just sync dropped and it happens all the time anyway). ELO kinda sorta matches people by skill except not really because as far as I can tell it is based on wins not individual performance. So if me and two other experienced players team up with and drop with a newer less experienced player and win a lot his ELO score gets jacked up and his solo pug play or any play with less experienced teamates becomes horrifying and depressing. Also all the skill matching seems to make it too difficult to keep tonnage matching or even weight class matching in good effect.

Hit detection I feel another round of host state rewind coming on. Guess we will see.

Stability for me is better than ever. Honestly in the last year this is the greatest improvement to the game. I rarely crash and neither do most of the people I play with.

[MischiefSC]

All good news.

SRMs... for what it's worth I would swear that when I run singles or fire them in singles I get better total hit percentage/more damage done. Are the missiles foiling each other? Trying to hit the same location at the same time and getting dropped as potential duplicates or errors? I just know that when I'm puking up a few dozen SRMs at someone they hit poorly. I chain them up and it seems like explosions drumroll.

Rest of it sounds great. Eager to see the hitbox results - there seem to be a lot of mechs who, due to these sorts of issues, lose their competitive edge. It'd be nice to have everyone on more even footing that way.

UI 2.0 - wants it. Got the Overlord package and just scrolling mechs is a hassle. A more flexible and nimble interface would be great.

Tonnage limits and any sort of control over matchmaking would be awesome. Don't want to feel like I'm playing with one hand behind my back when I'm pugging and not bringing JJs, PPCs and paired ballistics.

[DEN_Ninja]

FuzzyLog1c, on 01 November 2013 - 05:28 PM, said:

I know. But someone occasionally needs to remind senior management (and I mean IGP, not PGI) that there's still a big, eager, wealthy fanbase out there. I haven't played MWO since the "this game isn't for you, Mechwarrior fans" post on principal alone. Judging by the lack of recognizable names on the Teamspeak servers (most of whom chipped in $200+), I'm one of many.

Like many others, I've been quietly observing the slow decay of the game, hoping that a change in direction will come. Or that the whole thing will fireball in an epic way, permitting a hungrier and sharper group of developers to take their shot at gaming glory.

I will give you one god damned example of why you 'Let MWO die' folks need to [Redacted]. Command and Conquer just flopped. They straight up canceled the alpha despite months of testing and development because of negative feedback NOT UNLIKE THIS. Guess what, Victory games is now dissolved and those good people trying to make a C&C are jobless and either will filter into other EA studios or be jobless. They WANTED to pick up the C&C IP and make a C&C game but negative feedback from ardent fans that disliked the direction of the game killed it.

Who in the world would pick up and fund development for an IP that just flopped? You think that it is just so easy for the license to fall into the hands of passionate developers? Tell me one studio RIGHT NOW that would pick up Mechwarrior other than PGI.
Give me a name of a publisher that wants to fund or contract someone for a toxic fanbase like this?

C&C wasn't even a bad game. It had limited content because it was in damn alpha, It had a grind because the economy was not fleshed out. The gameplay was short and the same style as Generals which was more sleek and modern than RA or Tiberium. It was not a stale game and had much room for improvement but died earlier purely because the die hard fans disliked the newness of gameplay and that it did not go back to the original style of RA/Tiberium. The potential is lost.

You think Chris Roberts is making Wing Commander/Freelancer with just better graphics? His dream, is goal to make a better newer version of his games. He wants to improve upon and change the game because he did not have the technology back then to make it as he desired. It gives nod to his previous games but it increases the depth and content. He's got 12 or more months to add even more content.

Giving fans what they want is one thing. Give us too much power and you will never get the game you want. Every fan is different and has different visions of their dream game. When you push your malicious intent and it snowballs you get nothing.

But I guess it is too late, you let people follow that train. People posted numerous reviews on metacritic that were not even objective. You have some goonies filtering their news and reviews about MWO into their friends. Your friends take your opinions as truth and avoid the game. I can go on reddit and other forums to see hate being spewed against MWO.

if your wish comes to fruition and no one picks it up you can go play tabletop for the many coming years. I mean there is Mechwarrior Tactics but I don't get to pilot a mech in that.

Edit: I am jetlaged, tired, and utterly tired of what I consider Bull[Redacted].

Edited by Tichorius Davion, 02 November 2013 - 12:00 AM.

[Fuzzbox]

Thank you for the update, October was a bit quiet.

Sounds great that DX11 is progressing, and that you found whatever was blocking. I was about to ask as why there had been no mentions about it.

No mention on how the Community Warfare is progressing? This is the only thing I am awaiting, so any news here would be nice, even if there are delays.

[Kattspya]

This is how these are supposed to go. A matter of factly and most important of all a professional tone. GJ on that now hurry up with implementation.

[christophermx4]

Good stuff - I'm excited about the DX11 integration into the game.

[BigTaeng]

With regards to Windowed and Full Screen mode, will we be forced to run MechLab in Full Screen if that is how our game is setup?

A lot of us currently like the Windowed MechLab to make it easier to multi-task while we wait for our group to ready up. Alt-tabbing when in Full Screen mode usually takes longer, especially on lower end machines. We would like the option to be able to choose Windowed or Full Screen mode for MechLab separate from match launch.

[XphR]

I agree with BigTaeng,
It is very useful having a windowed application jump to full screen for the drop. It helps give access to VOIP applications, web pages and other tasks when someone in the lance has a quick errand or tweak to their Mech. Any way to remove the bug and not the feature?

Does this have anything to do with the mechlab running video cards to max heat if its left selected (I launch the game to cool my card back down..)?

[EoRaptor]

Thomas,

You mention unifying the the resolution and mode between the 'lobby/mechlab' and the in game experience. I ask, in the strongest possible terms, that you not do this until you can provide equivalent features in game that are now found in external tools, such as:

A global chat system with rooms, player hierarchy, and persistent messaging.
An in game voice chat system with equivalent functionality to mumble or teamspeak.
Mechlab upgrades that provide performance numbers for mechs 'as equipped' combined with ease of use improvements (see: smurfy, mechromancer).

These are the minimum features you need to support before you lock out easy access to them by jailing the mouse cursor or grabbing fullscreen. If you don't, people will continue to use third party tools and access them by alt-tabing or ctrl-entering the main game. This forced mode switching will have far worse problems than the 'scheduled' mode switching currently implemented, and you'll see more crashes and worse player experience.

I know you are trying to close off a path for bugs to be encountered by players, but this path needs to be there, and by erecting a gate, you are just encouraging people to jump it, not preventing the issues 'traveling' that path causes in the game engine. You need the carrot, not the stick.

Thanks.

P.S. To get 'Games for Windows' certified you need to support mode switching (alt-tab, ctrl-enter) anyway, so you might as well do the work and fix it now.

[Chopsaw]

I'm with FuzzyLog1c. I've been lurking for months waiting for the game to recover. I never downloaded the patch that included ghost heat and haven't played since that fateful Tuesday. I'm now going to give up lurking. Three other guys that I played with from my office have quit too.

I love how fixing a tool tip makes the patch feedback highlight list. Seriously? You need to dig that deep to make the list not look anemic? How about basic stuff like magic convergence and collisions? No mention of those being fixed in this technical update. [redacted]

Edited by miSs, 10 November 2013 - 11:07 AM.
nope.

[Troutmonkey]

Niko Snow, on 01 November 2013 - 02:00 PM, said:

We are aware there are still some issues surrounding hit detection, particularly with SRMs and also certain Mechs hitboxes.

*cough* Spider *cough*

Edited by Troutmonkey, 02 November 2013 - 10:55 PM.

[ShadowDarter]

Always great to hear how the Technical side of PGI are handling the multiple problems that a MMO faces month to month, please keep the player base informed.

[Appogee]

Good update, thanks.

Please prioritise MatchMaking. As it stands, I am sitting here browsing the forum instead of playing the game, because I am tired of participating in ROFLstomps.

PS: I don't mind waiting a long time for a good game. That's a LOT better than waiting a short time for a bad game. The months in which the MatchMaker expected me to be able to prop up a half dozen incompetent players were the worst of all.