Note the poorly placed cast shadow placement under the roof poly on to the vertical poly below. Roof poly thickness = 4 units


Note the even worse shadow in this image. Roof poly thickness = 3 units


Note the same cast shadow in this image. (much better). Roof poly thickness = 6 units


As lee was so kind to inform me a sloped roof poly must be a minimum of 5 units in depth (thickness) to cast a more or less proper shadow.
The first image has roof poly thickness of 4 units the second has a roof poly thickness of 3 units the third has a roof poly thickness of 6 units.

Attachment 1 = 4 units. Attachment 2 = 6 units. Attachment 3 = 3 units

Good tip! Thanks!

Great post, thanks for that!

Well then the question is rather then a work around, will this be fixed for a future version, visually a too thick roof doesn't look great.Roofs aren't that thick .

@ wizard

I could not agree more. Being forced to use a rotated 6 sided box element with a height of 6 units to solve a real time shadow problem is not right.
Until there is a better solution it seem the only way to rid models of the appearance of a light leak between the roof and the structure.

My impression from lee's email to my question about the real time shadow problem was to light map everything and the problem will go away.
What I want my models to use is a simple single rotated surface plane for the roof and a fascia board to hid the surface edge.
Models using this method look very realistic until the terrible shadow happens.

The attachment shows the result when I do it the way I want too, trying to save on poly counts for larger house models.

Point is I am not going to make a corrugated sheet 6 units thick, to satisfy an obvious bug.I am sure it will get an overall in the near future, but definitely not the way to go about it. 6 unit thick corrugated sheet is not going to look great what so ever.

The only reasonable work around to the problem is to have the roof and walls made out of a single mesh, and just extend the bit at the end, quite a interesting oversight from the development team.

Agree with wizard. Great tip but alas not convenient for most cases. Needs addressing and I noticed a lot of visual comments coming up lately. But hey you all wanted to build easy so I can only say don't moan if building editor is on your vote list. If it's not I feel your pain

I never asked for build easy, what I was given was build easy. The post maybe should not have been "Hint" should maybe have been "Be aware of a problem". I see that Lee has seen the post and perhaps will have a solution (if possible) in the future.

But likely with all things visuals will likely just be thrown in a bottomless pit, for more pressing issues.

Sadly I will not be using the easy building editor, nor will I support it with models.....the end, we given feedback on what it should be, and they are doing exactly the opposite, so much for listen to users, requests for the users, but our ideas, isn't exactly a what users had in mind.

On topic with the building editor, this bug will likely creep over to the building editor, if and of course, it has better functions, instead of placing lego blocks, I am sure slanted roof will be available, 5 or 6 units thick.

I actually withdrew from the community, a few weeks ago, but I see I am back at venting anger, with that cheers, while I still do content, and prowl forums, I will not be posting on the forums here or steam, I am kinda fed up with seeing the same old problem not being address so meh....that it then

I disagree ... So far we have had as requested from the voting system plus extras thrown in on the way ...
Right now I don't think you can accuse them of not listening to the users / Community ..

Well, I agree that TGC is listening to the Community and the voting system is a democratic tool...

...which is why it is failing. As GameGuru has been marketed as "easy Game Maker" while its more the early access to a potentially powerful All-in-One engine, lot´s of customers and user have been attracted that lack the technical knowledge of 3D graphics, AI frameworks and behaviour and, most of all, tools creation.

Basically we have now a lot of mambojambo very high on the list while essentials that even low-cost, opensource or small team develoepd Engines have featured for ages are deep down or not listed at tall. And don´t forget that a huge backlog of essential fixes is still waiting to be done. An AI that deserves the name, having the nearest nearest light next to the player cast a shadow shouldn´t be a feature, it´s a must have, the lightmapper does only create great results when you set the ini settings so high you can´t bake an entire densely populated level anymore without having shadowmaps measured in GB instead of MB numbers and your Computer consume the monthly electricity demand of an Ethiopean village e.t.c.

There are some really useful things among the Top 10, Menu Editing, Bots navigation (which is still a fix, not a feature) - but asking for trading and crafting as well as a quest system is pretty much insane as long as the renderign basics aren´t completed and roofs have to have a certain thickness to cast proper shadows.

This is neither an accusation nor a rant, it´s simply the expected outcoume. If someone would add the Point "Modding Options like in Skyrim" to the list, it would rise through the ranks, as would "Vegetation and Riding Animals like in Farcry Primal". That´s to be expected.

Don´t get me wrong: I still stand behind the Engine, behind my Steam Review, behind my projects I have done and do with GG.

But yes, it is kinda frustrating when something that could be the #1 Prototyping tool for all mid-level game studios and a good entry for future Indies gets stalled in it´s completion by such an...

...interesting...

...set of priorities. Perhaps it would be better to have a high council of GameGurus that know stuff decide, some evil mighty shadow entities like Rolfy, Wolf, Wizard, M2, Science Boy e.t.c.
Just my thoughts.

Quote: "Well, I agree that TGC is listening to the Community and the voting system is a democratic tool...

...which is why it is failing. As GameGuru has been marketed as "easy Game Maker" while its more the early access to a potentially powerful All-in-One engine, lot´s of customers and user have been attracted that lack the technical knowledge of 3D graphics, AI frameworks and behaviour and, most of all, tools creation.

Basically we have now a lot of mambojambo very high on the list while essentials that even low-cost, opensource or small team develoepd Engines have featured for ages are deep down or not listed at tall. And don´t forget that a huge backlog of essential fixes is still waiting to be done. An AI that deserves the name, having the nearest nearest light next to the player cast a shadow shouldn´t be a feature, it´s a must have, the lightmapper does only create great results when you set the ini settings so high you can´t bake an entire densely populated level anymore without having shadowmaps measured in GB instead of MB numbers and your Computer consume the monthly electricity demand of an Ethiopean village e.t.c.

There are some really useful things among the Top 10, Menu Editing, Bots navigation (which is still a fix, not a feature) - but asking for trading and crafting as well as a quest system is pretty much insane as long as the renderign basics aren´t completed and roofs have to have a certain thickness to cast proper shadows.

This is neither an accusation nor a rant, it´s simply the expected outcoume. If someone would add the Point "Modding Options like in Skyrim" to the list, it would rise through the ranks, as would "Vegetation and Riding Animals like in Farcry Primal". That´s to be expected.

Don´t get me wrong: I still stand behind the Engine, behind my Steam Review, behind my projects I have done and do with GG.

But yes, it is kinda frustrating when something that could be the #1 Prototyping tool for all mid-level game studios and a good entry for future Indies gets stalled in it´s completion by such an...

...interesting...

...set of priorities. Perhaps it would be better to have a high council of GameGurus that know stuff decide, some evil mighty shadow entities like Rolfy, Wolf, Wizard, M2, Science Boy e.t.c.
Just my thoughts."

spot on ertlov i think it needs a council rather than the entire country.

sort of like politics, if we all in life voted and put in on every legislation policy etc, my word be a nightmare, need a trusted and viable and very sensible way of this to happen.

i think maybe they should have a 1st 2nd 3rd etc. priority list

as in what makes the engine and then bells and whistles and then extras

priority list 1 choose from that

ai lights efx speed up x 11 etc physics


priority 2

world extras better editing weather cycles lua commands

priority 3

dlc packs cars oc rift terrain textures building editor character editor weapon editor.

priority 8000000000000000

leaning left and right

( last 1 a quip )

makes better sense


i could make a very strong case to why voting aint working but i guess they started so now the ones who are happy with leaning going way up there would have a hissy fit so it is i guess too late to change although they changed with x11 for save and load


just read last bit ertlov. sadly i probably would not have the time to council but you are of fine calibre, so i pass the baton to you if ever.

M2 i was not having a go at anyone in particular, it was a general if they chose building editor before the lights etc or to be honest anything other than what is being done then they should not moan too much asin all honesty they chose what is to be up top, i dont think building editor is at all essential to the engine. its a bi product. but also in this case you never voted for it so i feel your frustrations.

2nd i do admire your optimism and work around though it is good to see you find alternative fixes and to share so i will retract any negative from your thread. except i do think the voting is not helping the way of gg. but maybe another thread for that so i do apologize for going off subject. and thank you for your findings

Been called many things but that one kinda tickles me. It was inevitable.....as stated if your going to let voting run it then your going to see a different set of priorities from the majority than those who want to go a little deeper into the rabbit hole. But your all right, we are starting to see more features being put in where some core needs are feeling unfinished.

TGC of course are caught between a hard place and a rock when they need to be seen to satisfy the voting board, I got over ranting a long time ago when I realised they got a very full plate trying to satisfy everyone. Having said that if they don't finish up whats in there now they are going to see a lot of work down the line which they will need to back up to fix.

Maybe priorities seem skewed at the moment but that could be down to more folks now getting to a stage with their level builds where some issues have went unnoticed or unvoiced, the more people get to this stage the more voices will add to the clamour.

I think most of this could be down to how they test new builds of GG where most of it will be focused on whatever new is going in and whether it affected anything already in place. I have often seen Lee fail to do something in his Twitch broadcasts which he expected to work out of the box and he was surprised by it, he doesn't do much actual design with his product, too busy writing it. If made aware it can be fixed, but with a voting system at odds with the more serious user base whose demands are for an engine that works as it should and have a good idea of what to expect from the product, you are going to see a little progress here and there but nothing major for a while on the graphics.

What I don't get is why thickness is fixed in stone like this, a plane will cast shadow, heck even a plane with semi transparent animated scrolling texture will cast shadow so whats the problem with sloping roofs?

@science boy
Don't worry about offending me.. I can handle it, really. Just trying to keep the thread going.

@ rolfy


That is a very good observation which I believe to be an actual fact. When I have taken the time to give him a very detailed example post he tries his best to respond with a reasonable explanation why or why not.

I would like to suggest, at least, a temporary lifting of the voting process commitment and spend some direct time resolving the graphic issues once and for all.

@Ertlov
I am in total agreement with everything in your post. If there were ever to be such a thing as a High council, your name is on it. I'm not sure I would meet the criteria.

Yeah so I raised the point about the voting board several times and surprised its taken this long for others to get onboard.

As rolfy said, the most important aspect of a game engine is to have consistent results and this means fundamentals such as lighting need to work as you expect them to. This doesn't mean having the best AAA lighting on the planet but there definitely shouldn't be workarounds being used for models to cast and receive shadows properly.

I will say TGC have gone further with GG than ever before to try and keep the community aware of whats going on but there are some things which remain unclear like whether we need to wait until the DX11 upgrade before the lighting system will be looked at again. If so it could be a long time as this is a way down the list and Lee already confirmed the initial implementation will not focus on delivering any new features.

There is also the question of expectations. For experienced developers some of these inconsistencies would really put you off spending any significant amount of effort on a project but if the majority are happy to live with them and want an RPG quest system then your stuffed.

We will never get around the fact TGC have a small development team and have a relative large community to try and satisfy. GG would be much better served as an open source project in my opinion so Lee can maintain the official code base and others can work off that and make contributions.

I think a reply is in order.
What about how it's implemented, how often do they listen there ?

When I see threads like this, blood pressure rises, and some times loose it, honestly I am done with bickering and moaning, what I don't see doesn't bother the mind.Rather just sit on the sidelines, it's better for my health.

There is no way to really to disagree with this statement. What is the harm in TGC taking back just a little bit of the responsibility for what gets developed and when? Voting on what is wanted by the user is not a perfect workable concept. There are times when Lee should be able to just say "I think there are things that should be addressed now", no mater the highest vote item. It may be a small development team but I think lee is capable with out the pressure of trying to please everyone all of the time, including me.

Thumbs up to that m2design. Lee does indeed try to please everyone at once sometimes.

But he also adds little bonuses here and there and gets them started. ( FBX import ) Not anywhere close to the top of the voting board, but very important to a lot of people.

There is a lot of planning on when things should be added verses the voting board, verses bug fixes, verses core feature that should be there anyway.

This is actually what I am most interested in and most curious about. As I am right now dayjob-working as Project Manager / Producer on this small indie game here...



...which is due to its episodic nature actually quite comparable to the release strategy of GameGuru, the ongoing struggle of weighing priorities in Bugfixes, Polishing Tasks and Feature requests against each other is basically the largest part of the job. And no feature request and no polishing task could ever outweigh a needed fix for a P1 or P2 issue. (Not talking about blockers, these are taken care off by Lee and team carefully).
So basically, a battle plan (some kind of mid-term planning for let´s say, 3 months) would be needed that acknowledges that:

- all tasks are divided in Prio 1 to Prio 4 (P0 reserved for 100% reproducible show-stopping issues)
- every missing core feature is actually a P1
- every fix for a core feature that currently in no way produces results as expected is at least a P2, if not P1
- non-core feature requests can never exceed P3
- polishing tasks can never exceed P3

And then you break it down to weekly or biweekly sprints in order of priority; if same priority is popping up, add the layer severity. When a large bunch of P1, P2 and P3 issues could be tackled by a leap jump task over several sprints (this could be as example DX11 support, but not sure about that), you have to tackle that one first.

And now I put my Mipumi / Hitman project management hat aside and the Homegrown Games / CEO hat on:

Of course I understand that GG has to earn ongoing revenues on Steam, and I know that an engine is sold by:

- A promise
- The actual tools and accessibility
- The power of it´s showcases

The promise is right now "Easy Game Maker"
- which can only be taken serious if you put it like

THE EASY GAME MAKER *)
*) but expect horrible results unless you are already an experienced game creator or willing to study the work of others for months

The actual tools and accessibility:
All rants set aside, this is actually where GG partially shines already and could smash all competition. Easy prototyping at it´s best, the power and control over details has yet to come.

The power of showcases:

I know how many requests regarding FPSC I got after releasing Anderson and Into the Dark, I have educated guesses about how many people went to steam buying GG after inquiries on the Antinomy prototype and demos. And I also know many people who picked up FPSC back then after playing one of Wolfs creations on IndieDB. (and I guess they went into deep depressions afterwards when they realized you don´t need the engine alone, you also need a Wolf to get something like that).

Bottom line is: There are some really great showcases here in the board (yes, ScienceBoy, looking in your direction), but a full-fletched game out there that stuns every player or viewer of the games steam store page or IndieDB database is needed soon. Something where everyone says "Oh my god, I never thought this is possible with GG!"
And for that we need some visual functionality that almost every tiny engine already has.
Not easy at all

The shadow issue is caused by the resolution of the depth map that captures the rendering from the perspective of the light towards the direction of the object. The light source has to be high enough to shadow everything (the sun), but the light ray long enough to reach everything. With this, you have a lot of range and that puts a strain on the resolution of the depth map (32 bit). I also use cascade shadow maps to break this limit down into four phases, so objects closer to the player use a relatively higher density of resolution to get more detail in, but even then you can see the limits if you look really close to a shadow cast on a character at high shader settings. Happy to receive any tips on how this depth map limit can be overcome without resorting to something like 8 cascades or a 64-bit depth map, both of which extoll an extra performance hit on the graphics pipeline. The current requirement of a roof thickness of 6 units is to overcome the issue of a shadow being seen on the 'other side' of the polygon that casts the shadow (due to above depth resolution). Ideas to solve this without hitting performance most welcome.

The question is how much extra ? We already 3 or 4 generations GPU's later since game guru started.Is it not reasonable to have this on the highest settings, and medium and low reverting back to less accurate. I am sure it is not impossible to have two pipeline lines. How about adding a extreme settings, like almost all engines.

GTA V had a nice system of letting you know exactly how much video ram would be used with specific settings, nearly all of the gave fair warning that using. Graphic fidelity simply isn't there, throw caution to the wind might not be a bad idea.

Per-pixel lighting
http://www.flipcode.com/archives/Light_Mapping_Theory_and_Implementation.shtml

"Bleeding due to bi-linear filtering being turned on." I have seen this happen in a few instances

http://www2.imm.dtu.dk/pubdb/views/edoc_download.php/4115/pdf/imm4115.pdf



This some thing that needs fixing I can't have a 6 unit thick roof, it is not visually correct, users would have a field day.

I've been saying what Ertlov just typed numerous times now and I could not agree more (Also big thanks for mentioning my work with such high regard!).



Yes! A bit too little of the community work in progress is presented. There is some pretty creative stuff out there that screams at potential customers that they can get a version of their idea up and running in game guru...like what cybernesence is doing for example.



thanks! But to be fair, in its prime, there where a whole bunch of people with similar results. (TheK, Jonez, Bugsy, Doomster, Getsfile etc.) I just happened to have reached a slightly wider audience and released more or any games at the time.

Yes: Priorities! Most of us don't want anything groundbreaking in the visuals department but we want a stable and durable engine in the core department. We know that this is a huge project and quite a lot of toil but we all have some of our spare time invested in it and want to see it thrive.



-Wolf