Unreal Engine, like Unity, right now has to be used with Autodesk products like 3DS Max or Maya, or at least that is the case with the large majority of professionals. This is an impediment to using Unreal Engine and thus makes Epic Games sort of dependent on Autodesk.
There is probably no coincidence that the new general manager of Unreal Engine is an ex-Autodesk SVP, Marc Petit, who used to be in charge of Maya, 3DS Max, etc for a decade (https://www.awn.com/news/marc-petit-leaves-autodesk). He definitely sees the value of building up Blender to remove the dependence of Unreal Engine on expensive tools from other vendors.
3D modelling software has had a long history of expensive, esoteric, difficult-to-use applications. Epic, hoping to make money from large numbers of Unreal Engine licenses, would like to see that situation change. Ideally (for them), 3D modelling software would be free and easy to use for everyone, thereby lowering the barrier to entry for those who might make 3D games using UE.
By donating to Blender, Epic hopes to achieve two things:
1. Blender becomes better and easier to use, allowing more developers to enter the market and buy UE licenses.
2. Increased competition puts downward price pressure on commercial 3D packages such as Maya, thus making it more affordable for small-to-medium sized game studios, potentially freeing up budget for more UE licenses.
> Unreal Engine, like Unity, right now has to be used with Autodesk products like 3DS Max or Maya
I am asking why they think this is the case. Not the circular reasoning of "most professionals use it" but I am interested in what they feel the Blender is missing.
You have to experience trying to actually get something accomplished to truly understand how incredibly unintuitive everything from names to button placements to object movement to component selection etc truly is. I've used a lot of different 3D programs in a lot of different domains and nothing comes close to blender. It almost seems to go out of its way to make the most basic excersize a puzzle for the user.
The money for commercial programs is completely necessary in a professional context, since time is money and people have expectations.
But with the 2.8 RC news on HN last week I recently tried it again and I have to say that now it seems like the UI is pretty decent. Left click select, a move tool that works like every other move tool in paint, a clear hierarchy view with obvious nestable layers - it seems like many of the weird things I remember (like "drag the divider between the viewport and the menu bar down to see a secret option pane") are gone or redesigned now.
Check out https://www.youtube.com/watch?v=jBqYTgaFDxU for some example usage or https://www.youtube.com/watch?v=lPVpg4_POww for a preview of the UI/UX changes.
This is the circular reasoning I'm trying to avoid.
I'm not looking for any one thing either, I'm looking for any shortcomings OP can think of. I've used Blender+Unreal for a number of years and Blender itself for over fifteen years so I could rattle off issues I've dealt with in integrating Blender into the pipeline, but many have been dealt with so I'm curious about OP's particular pain points.
If you already know that `Tab` switches between object mode and edit mode, then it's quick and easy. If you don't, then you have to click the drop down mode selection menu, which is one of the dozens of widgets on screen with highly variable levels of usefulness (seriously, what the hell is 'Grease Pencil'?). And if you hover for a tooltip, it doesn't even tell you that tab is the shortcut!
Or when in the UV workspace, there's literally nothing in the interface window except the vertices/edges. How do I select all vertices? How do I select multiple via box? How do I do numeric input? None of this is obvious.
The defaults are also terrible - eg click to set 3D cursor, instead of select.
Overall, Blender seems to suffer from a classic case of open-source programmer driven development. If you already know how to use it, it's very powerful and customizable - the python scripting is great! But because no one really cares about making it into a product, the UI/UX suffers.
I haven't used Maya/3DSMax, but I have used Fusion 360 (another Autodesk 3D modelling package, but focused on CAD/CAM) extensively, and it is much, much more user friendly. There are easily navigable menus which surface key functionality and respond contextually to the workload - ie if you're modeling, creating components and performing extrusions/rotations is front and center. If you're sketching, applying constraints (and the various types available) is all right there.
Of course, none of this means you CAN'T use Blender with UE, you totally can. And if you're a Blender expert, then none of this probably bothers you, because you don't care about discoverability of functionality since you already know it all. But if you're fresh out of Digipen or whatever, and you have to learn a single software package, then the refined products from Autodesk are going to win.
edit: In fairness, I hear that 2.80 is going to fix a lot of this, and I'm looking forward to trying it.
Somewhat amusing comment seeing as though Blender was a product before it was open sourced, and from all I've heard the UI/UX has improved since it became open source.
However, shortcuts are well-configured and when using pie menus, you only need to know a handful of them to get started.
> If you don't, then you have to click the drop down mode selection menu, which is one of the dozens of widgets on screen
Tab is one of the first (and only) shortcuts you need to know to get started. The same options can still be accessed from the quick menu however. I wouldn't call this a discoverability issue.
> what the hell is 'Grease Pencil'?
Ask the documentation.
> when in the UV workspace, there's literally nothing in the interface window except the vertices/edges. How do I select all vertices? How do I select multiple via box? How do I do numeric input?
With the same exact gestures used to select vertices in edit mode. Blender's UI is highly consistent these days. Numeric input is as easy as calling an action and then typing in the number before confirming with Enter.
> The defaults are also terrible - eg click to set 3D cursor, instead of select.
I wouldn't call this terrible, just different-- and besides, Blender Foundation has just recently made the cursor defaults reflect the common defaults found in other 3D modeling packages.
> Overall, Blender seems to suffer from a classic case of open-source programmer driven development.
Have you used Blender in the last 7 or so years? These seem like antiquated criticisms. The Blender UX is vastly different and much more focused these days.
Every modeling suite has a learning curve. I'm interested in what Blender is sorely lacking. Blender's discoverability is actually very good and Blender itself might be my favorite example of how to design functional application UX. This wasn't true a decade ago but it's true now.
The grease pencil is Blender's 2D drawing system. Initially, it's purpose was to allow artists to sketch scenes out onscreen, in much the same way that CGI artists a few decades ago used literal grease pencils  to draw on their glass computer monitors.
Nowadays, the grease pencil is a full 2D art system, allowing for hand-drawn 2D art to coexist with 3D assets or tracked with camera footage. The Blender Foundation made a short film called "Hero"  to show it off.
> How do I select all (UV) verticies?
On the bottom edge of the UV editor window is a menu bar. From the left, the menus are editor type (shown as an icon), "View", "Select", "Image", and "UVs". Farther to the right in that bar is the widget that lets you set whether you are selecting verticies, edges, faces, or islands. In the "select" menu is "(De)select All", and a label indicating that the keyboard shortcut is "A". (This shortcut is universal to all of Blender's editors (except the text editor, where it is Ctrl+A).
If you have anything at all selected, then "(De)select All" will deselect everything. If you have nothing selected, it will select everything.
> How do I select multiple via box?
In Blender, this is called "border select". In the same "select" menu I described above, it's the bottom option (the one nearest your cursor when you open the menu). It also has a universal shortcut, indicated in the menu: "B".
> How do I do numeric input?
There are two ways I can interpret this. One is "given a number box, how to I change the value"? There are three ways that you can change the content of number boxes in Blender.
1. Click on the arrows on the left or right sides to decrement or increment the value.
2. Click in the middle, type a value or math expression, and then either click outside the box or type "enter".
3. Click and drag the box, from the middle, to the left or right.
The other way I can interpret the question is "how can I specify the coordinates of a UV vertex"?
Once you have one or more verticies selected in the UV editor, you should see some properties and editing widgets in a tall rectangular area on the right edge of the UV editor window. If you don't see it, you can reveal it by either clicking on the "+" button near the top of that edge, by going to the "view" menu and selecting "properties", or by hitting the "N" key while your cursor is over the UV editor.
In the top of the UV properties shelf I just described is a section called "UV Vertex". (It's called that even when you are in edge, face, or island selecting mode.) It has two numeric boxes in it labeled "X" and "Y". These have the coordinates of the most recently selected vertex. Changing the contents of those boxes moves the selected vertex to those coordinates, and moves all other selected verticies by the same amount from their previous locations.
An essential Blender video to watch is Captain Disillusion's presentation at Blender Conf: https://www.youtube.com/watch?v=1qSTcxt2t74
One view to look at it is that polygon based modeling, animation and material authoring can be done using a finite feature set, and once that feature set is complete, further enhancements are nice to have, but not critical.
This is definitely true, if the interface is refined. In this respect, blender doesn't just not hit the bullseye, it's as if someone held the arrow and shot the bow straight in the air.
But yeah - it doesn't seem like this should happen as often as it does.
It's the devil they know. It's not likely malicious or anything conniving, and just something they feel comfortable talking about when making the pitch for their integration.
Right now, this is what I usually do:
- Basic Modeling: Blender
- Sculpting + Details: ZBrush
- Painting: Substance Painter
- Animation: Maya
Last but not least, they should make more effort to improve their keybindings. They've been monkey patching it in 2.8 with the "industry-compatible" keybindings but when you use it a bunch of other things stop working.
I had pretty good success with it, and wrote a Python addon which builds on top to retarget mocap data from the Perception Neuron suit to the UE4 Skeleton.
You need to enable the addon as per usual along with UE4 Tools, and then once you have a model loaded and parented to the UE4 rig, and an Axis Neuron animation imported, it should be fairly easy to use. There is a "Bake Animation" control in there somewhere that should bake the animation out to an animation strip. I think you have to specify the name of the source rig (ie. the raw skeleton data imported from Axis Neuron), and there are buttons to enable/disable transferring finger/foot data as well.
I haven't used it in a couple of years however, so if it doesn't work feel free to hit me up with questions. It was just a quick hack, very rough and ready. No warranty :-)
Also any patches which improve the accuracy of the retargeting (hands can still be a problem in general) are very welcome!
This quick test clip I did the weekend I wrote it demonstrates the current quality of the retargeting, although it is rendered in Blender and not UE4:
I tried Blender a decade ago but since I had access to Maya & 3D Studio Max I didn't give it much attempt since the UI seemed more confusing & the material available to learn was lacking.
0) Blender is much more stable than Maya. Also it loads more quickly and is less than 10% the filesize. Although it is quite powerful, Maya is still very bloated with decades of technical debt and legacy baked in. Also blender is OSS which makes me prefer it on principle if I can use it.
1) Blender’s interface has a higher “skill ceiling” — it reminds me of Vim in that most common operations are a few single-key presses away. For example, “rotate the selected object 45 degrees around the x axis” can be executed via R->4->5->X->Enter. Many of these kinds of basic operations require the use of a mouse in Maya or a MEL expression. Of course, this also raises the “skill floor” of blender because you have to remember more key combos. However these days the UI is pretty discoverable (even before Blender 2.8).
2) For basic hard surface UV unwrapping, I like the workflow in blender much better. The interface is the same in 2D as in 3D, and the key combos all carry over, which I really like. Maya has more automatic UV unwrapping options, but I rarely reach for those anyway.
3) 100% of blender’s interface is scriptable. You can literally hover over any button and the corresponding blender python code to execute that button will show in a tooltip. This makes it easy to automate basically anything and has empowered a great community of plugin authors. Maya also has this with MEL, and more recently maya python, but those solutions feel tacked on whereas blender’s UX was built from the ground up to support python.
I should also mention that blender has already went through a major UI refresh, so if you haven’t tried in a while it would be worth trying it out again. I would say that the jump in accessibility from the previous UI refresh is the same jump that we see now. So I’d give blender a second chance if you haven’t already!
Ten years is a really long time though. Google's Go programming language didn't exist ten years ago.
That's what the little "F" button on datablocks is for. It adds a "fake user" to the datablock, which means that it won't be garbage-collected away if nothing is using it when you close Blender.
I’ve lost work this way as well; sure, it’s avoidable, but it’s most certainly a pain point.
Really? Did not know about it. Thanks! The number of users of a certain entity has always confused to me. I've used Maya a lot, and I can't but wonder why Blender just don't copy the got bits of other software when it comes to usability improvement.
This is taken directly from the EAC dll:
Linux is ridiculously powerful & versatile as I'm sure we all know...and having that much opportunity to mess with all the net/visual/input subsystems in a way the game can't even see (since it's trapped in WINE) is a major issue cheat wise.
Not ideal but I can totally see why a company might from a practical/commercial point of view just block it outright. Especially given player numbers of WINE players vs upsetting your entire user base due to cheaters.
If I wanted to aimbot in an FPS, I would just analyze the HDMI output from my computer (via an FPGA) and have it directly create the necessary USB transactions to move my mouse to the target's head. I would introduce randomness, "input lag", jitter, and even some misses, so that people looking at the kill cam wouldn't immediately think "aimbot" as they are apt to do whenever they die. (You also don't want to make your mouse movements statistically different from anyone else's, or it's the server that will pick up on your cheating... or at least a well-crafted MapReduce. Do game companies have MapReduce? I hope so.)
Executed well, the developer of the game could do nothing about it; maybe I'm really good, or maybe I'm aimbotting. My computer is completely normal, down to the device descriptors on my mouse and monitor. There's no wine, no virtualization, no additional software running.
In the end, trusting the client is crazy. If you want to decide win/loss based on things computers can do easily, your game is probably bad. And aiming is something computers can do in their sleep.
This is actually a major component of bot use for f2p games. A lot of companies try really hard to prevent anything that looks like playing in virtual envs. Having these bots fill up your games and play terribly is a bad look for any game that depends on multi-player interactions to keep people involved.
I believe you can get pretty affordable Windows VMs these days, however. I'm not sure that blocking wine really accomplishes a lot. (There are also companies that sell access to consumer Windows PCs on consumer ISPs these days. If you don't play Hearthstone, you probably don't mind $10/month in free money for letting someone bot Hearthstone on your computer. And it's not like Blizzard is going to say "look, enough Comcast, shut these people down or else we lock all Comcast customers out of battle.net". It's pretty tough to run a gaming client these days, I would imagine. So so much stuff is totally out of your control. I do not envy their engineers at all!)
Besides having to rely on real-time computer vision which is more expensive/resource consuming and harder, the results wouldn't be as good as with other techniques (although this might fit your argument of having some misses and jitter/lag/etc... to act as cover).
(I see two flagged posts asking the same question)
...pretty comfortable in saying that you're not up to speed on 2004 era hax.
It definitely it. Read up on anti cheating stuff...the windows stuff is down-right rootkit grade....to the point where there are privacy concerns from every the privacy tone-death crowd.
Very very different ballgame from "runs in a WINE container with no visibility beyond that".
It's an arms race, but whatever they can check, you can patch. Worst case, you can still virtualise everything.
There isn't a general way to prevent cheating on open platforms with this kind of game, so it ends up being a kind of cat and mouse game. Think of an aimbot (instantly moves your mouse cursor to an enemy as soon as the enemy is seen) or a wall hack that makes enemies visible behind obstacles. The only way to detect those is to check if your runtime environment is being modified by known methods. Opening a wine front (where you'd somehow have detect Linux processes/a modified wine or even a modified kernel etc.) is probably too costly compared to the number of non-cheating players on Linux.
The slippery slope of giving companies higher access is also seen in CSGO where (Faceit/ESEA, dont remember who exactly) put a bitcoin miner in their anti-cheat. Capcom put a rootkit into Street Fighter V to check for cheaters. Is that acceptable?
Incompetent, or rational decision in terms of bang-for-buck to reduce cheats?
Like another comment says, maybe it's easier to cheat using Wine, and thus banning the platform is faster/cheaper, even though they might end up catching a lot of legit users in this broad dragnet.
 e.g., https://en.m.wikipedia.org/wiki/NProtect_GameGuard
Of course, it's possible (if much harder) to go in and hack up the DLLs on Windows. Anti-cheat addresses that by whitelisting known-good MS hashes and/or requiring some type of signing. There's a lot more overhead involved in doing something like that for WINE, since every distro's build would be different, it's common for users to have their own builds with game-specific patches enabled, etc.
I've been out of the WINE scene for a while, but I'll say that I don't believe many WINE users are trying to exploit the platform to cheat. I think it's purely a cost-benefit tradeoff. The vendor doesn't feel supporting a legitimate WINE use case is worth the expense of figuring out a reasonable way to validate behavior, handle integration with distro builds, support debugging/custom patchsets, etc.
I would imagine Wine is in a similar predicament.
It seems far more likely to that most players are just using Wine for its intended purpose—to play Fortnite on Linux.
their method (blocking Wine) is questionable though. it reminds me of websites that block Central/Eastern European IP's "because of bots" (yeah, it happens). the goal is nice, but the method hurts many honest users.
Contradictory sentiments like these are what lead me to believe this grant is mostly for marketing purposes. (tin-foil hat mode)
However, I don't see the contradiction in Sweeney's statement. One can appreciate the advantages of open tools, libraries and platforms, and still want to publish a proprietary game on top of that technology.
It's kind of like an omnivore chef joining a forum full of vegans to work on improving seasonings for braised veggies. They might not be pleased when they find out that she wants to make a side dish to be served with steak; but that in itself doesn't mean her goal is exploitation or sabotage, rather than mutually beneficial cooperation.
..or Google or Apple. Or Nintendo. Or Sony.
"I don't give a rip whose money it is, mate" - Steve Irwin, when asked about accepting money from "unethical" groups and using it for conservation. https://youtu.be/-1gVkTFam1w?t=753
If it helps a good goal, what does it matter?
If it's done anonymously, then sure: you're quite right. But if it isn't then there is, in fact, a price. And it's not always a great deal.
But by accepting the donation you may be signaling that you're ok with how the donor made their money. By turning it down, you may also have a PR opportunity to spread a positive message.
I am merely pointing out that Epic's list of bad things outweighs their list of good things, regarding the open source community.
Assuming there are tax benefits for donating, it becomes a win-win.
I think he's operating with a different understanding of the word "open" than the rest of us.
Also, there's Wine and ReactOS, no? Aren't those two alternative (even if somewhat incomplete) implementations of Win32?
Using the word "standard", which he never does in that article, is quite misleading.
As it is, he argued that Win32 is more open than UWP or Android; regardless of whether you think the API is a standard or a platform.
Yes, it's not totally black and white. Android does kinda allow side-loading with shitty usability. MS did start its own store and lets not get into telemetry. Classical windows still gives the user a lot of control, though.
Further if the Oracle v Google suit does not get overturned MS could put an end to both Wine and ReactOS in a heartbeat
With UE4 they have created an engine that was extensible only via C++ and via their graphical scripting language. With the notorious instability and incompatibilities of ABIs of different C++ compilers, Epic didn't have a lot of options other than releasing the engine source so that their customers could recompile the whole engine with a single compiler version. And with Unity eating away at the indie market at that time, Epic needed a successor to the free, closed source UE3 UDK to keep a foot in that market. I don't think that Epic had a lot of other options at that point.
EGS doesn't have DRM, it is up to each game to implement their own if they want, but the store itself doesn't provide any.
Which is great since most games do not bother :-P (although i'd prefer it if they had a DRM-free stance like GOG, but Epic is too publisher-friendly to allow that).
As a slighty OT aside, I really hope blender's UI changes sloooow down. I've been using it for 16 years. 2.8 versus 2.7 is so different, It almost feels like they're making changes for change sake.
I've been trying to use their various 2.8 RC's so I don't have to say I don't know how to use blender anymore when it pops.
They made the following things toggleable at launch, but man these are some muscle memory breaking changes that are now the default.
- Changed selecting objects from right to left click
- Space bar no longer opens up the function search. The search is like 50% of my workflow, so this one hit me in the gut until I changed the setting.
Then there's some other things that feel renamed for no discernible reason.
- I can't search for 'remove doubles' anymore. There are certain geometries that I've come to build by snapping to axis, extruding, snapping to vertices, remove doubles. Now it's buried in a menu at mesh->clean up->'merge by distance'. It also makes a GUI element pop with the distance argument, and there's no obvious way to "apply". So weird.
- Ambient occlusion in the view appears to have been renamed 'cavity'.
- The tool panel lost its words, it's now just icons. Functionality has to be discovered by hovering. This is the worst UX habit from mobile, I wish it would stay out of my desktop tooling.
- The view/selection/snap/etc settings bar has moved from the bottom of the screen to the top. Why?
- Properties tabs moved to the side from the top. I can see why.
- I can't make objects unselectable but visible in the outliner. Why?!
- Layers are gone, they're now "collections" but they don't do the same thing at all!
This is the stuff I've hit in a about 2 hours of messing. I'm willing to acclimate to about 90% of these changes. So far it's just the layers / merge doubles thing that really sucks. Eevee is really pretty, so it's got that going for it, which is nice.
Tbh the UI being terrible was one of the main reasons I stayed away from Blender. The fact they changed it makes me have hope I can start using it seriously.
1) Mesh > Clean Up > Merge by Distance
2) F3 operator search. Start typing "mer..." and hit Enter.
3) Ctrl-V to show vertex menu, then Merge Vertices > By Distance
4) The quickest: Alt-M + B (show vertex merge menu, B to select "By Distance")
I scratched my head a bit first too, but frankly "merge vertices by distance" is far more descriptive than "remove doubles".
The UI that pops up is just the operator parameters that used to appear when you hit F6 (eg try adding a Circle, see the parameters that pop up) It's better this is visible to users, rather than hidden behind some secret hotkey. As with previous Blender versions, operators are "applied" by default - you can change the parameters via the UI until you like the result. Once you switch operator, move on, or change mode, the UI will disappear and the operation becomes undoable.
To make objects visible/selectable/renderable etc... open the filter menu on the Outliner (top right, looks like a funnel). This lets you add more toggles to the items.
To move the view/selection/snap settings to the bottom just right click, Header > Flip to Bottom/Top.
Collections replace layers and groups (eg you can instance a Collection), though I haven't completely got my head round them yet. But they look quite powerful.
One more thing on this - hit Shift-Space to show the available tool icons with labels as a context menu.
If you prefer this (I do), you can then hide the icons on the left via View > Toolbar (or equivalently, toggle via the T key).
Remove doubles can be done with Alt-M -> tap 'B'.
The 3D header (and any header in Blender) can be moved to the top or bottom: right-click the header, select "Flip to Top" or "Flip to Bottom".
In the Outliner, you can enable the "toggle selectable" button in the Filters popup. Click on the "Filter" icon, then click on the mouse cursor to enable the button. Now, you can click the cursor next to any object to make it unselectable, and it will remain visible in the outliner.
Collections can do everything Layers could do, and many more things besides.
Just engaging in the time honored tradition of hand wringing.
UE4's own tools (which are based around creating environments from existing models, placing actors, and realtime rendering) are WORLDS ahead of common Autodesk modelling tools eg 3DS Max and Maya (which Epic itself uses). UE4 is easy to pick up, but then you have to learn Maya and it becomes a grind.
There needs to be something as powerful as Maya with a better UX to stop modelling being a blocker for new UE4 users. Being Free As in Beer and Free as in Freedom help too.
Sure there is.
- UE4 needs to move a viewport around 3D space, so does Maya. UE4 uses game controls. Maya you need to hold down something or tweak it to do that.
- UE4 involves creating and manipulating objects in space for BSP. Maya involves creating and manipulating objects in space for modelling.
Maya has a UI that's 20+ years old with minimal changes. UE4 does not.
> They are built to do very different things.
No, they're built to do different but related things. I'm not sure why you would think that modelling and environment creation are very different. For the purposes of gamedev or archvis, they're both tools to support making the environment. And one is a barely maintained cash cow.
Not knocking the donation per se but it's a bit easier to be a good guy when you're sitting on the world's biggest pile of kid's pocket money.
Several friends have stories along the lines of their children taking not spending any money on their summer holidays because they want to save it for Fortnite hats or whatever. It's carefully designed smack for kids and there has to be a certain amount of moral push-back on that.
You could easily replace "Fortnite hats" with "Pokemon cards", "Yugioh cards", "Gameboy games", "MMO items", "Starcraft expansions", etc.
The difference with Epic is that they've managed to take a much larger chunk of the market, which is why parents have started to notice it more.
I guess it's a tricky one as toys are not really much different despite having a physical manifestation. I would have happily spent every penny I could get my hands on on Star Wars figures when I was of that age. Is marketing a useless blob of plastic to kids any different to marketing a useless bunch of bytes?
The physical blob is, well, physical. It exists for longer, and 20 years later you may dig it out and relive some of childhood memories. Bytes in the cloud tend to have much shorter life, usually defined by how long the vendor bothers to keep the server applications up.
EDIT: Funny I managed to post an exactly opposite opinion to 'DonHopkins at the same time :). Let me just clarify: I prefer no marketing of useless anything, but for toys, I think physical beats digital because digital is ephemeral.
Especially when they use some of that money to support useful bunches of bytes like Blender!
(That said, I sure loved my Star Wars action figures!)
Epic isn't really doing anything predatory they don't use RNG or other such things.
On top of that Sweeney is also doing quite a lot personally both as a conservationist and philanthropist.
To the kids, the cost is worth it, bringing them happiness, social experiences, and a fun childhood to look back on. As long as it the amount doesn't get unreasonable relative to the means of their parents, it may be a good investment.
do you mean Steaming piles of cash? :)
It doesn't look like it has many (any?) strings. Aside from being split over three years in Blender's case.
It is worth considering that this is in Epic's best interests as Blender + Unreal Engine is a common platform for startups/indie devs.
Making Blender better may make the games that people develop using Unreal Engine better, which might mean higher license income for Epic. So it is a win for Epic and a win for Blender/the community.
End users get new features they may actually use, massive OSS gets funding. Win-win for all parties.
This is a 2D animation generated from a 3D environment. Mind blowing ->
My first experience with the Epic store was to get Satisfactory. And while it was annoying not being able to have it on steam, it really became a non-issue since everyone I play with uses Discord. Again, I am curious to hear other opinions on this because I really do not see the big deal in my experience.
That's the thing though, console exclusives at least saved developers the resources needed to port to another console. Especially in ye olden days, that was a non-trivial cost. The storefront exclusivity on PCs is completely artificial.
But also, yeah, my computer is full of enough bloated useless-ware. At least GoG and Itch don't make me use a client.
I was against Epic's aggressive moves, until I realized just how much cash they're injecting into developers. In the video game industry developers are the ones that bring the most value to customers.
Valve has been a dragon hoarding its gold. The services they provide are better than Epic, but they really don't reflect their near-monopoly position for over a decade. 30% out of the majority of PC game sales, and what do we have to show for it? Valve doesn't even seem willing to compete and lower its cut.
In the end going with Epic will guarantee the studio can make it to the next game. That's worth downloading another client. That and the free games.
Steam links, Steam controllers, Steam on Linux, Proton, and the OpenVR SDK to name a few. Sure several of those things were failures, but it isn't like they aren't trying.
> In the video game industry developers are the ones that bring the most value to customers.
That's some trickle-down theory right there. I'm not convinced that what's good for developers is necessarily good for gamers or the gaming market. Case in point: loot boxes.
Studio closures are common, game services shut down, mass layoffs, and project cancellations. An exclusivity deal is much better than a publisher going in and changing monetization or game design to fit their strategy.
Where do you get off telling me how I should feel? The way I feel is that exclusivity that has nothing to do with technical constraints and everything to do with corporate sumo wrestling is not and never will be good for gaming as a whole.
Exclusivity can be good if it results in more games being developed, at a better quality, cheaper. Do you disagree with that?
Gamedev isn't cheap, especially for indies and smaller studios.
I'm happy about their revenue share and was excited when they got started - until it became obvious that their tactics didn't align with their PR.
Epic's play is to try and see if being a better looking deal for game devs will translate to being better for gamers or other groups (like linux users, open source advocates) too. We'll see. Valve has already demonstrated their goodwill towards all the groups multiple times, even if it's motivated by keeping their platform dominance.
Yeah, it became the industry standard because it rakes in mind-blowingly huge revenue for the app store owner, e.g. https://www.cultofmac.com/601492/app-store-google-play-reven.... From the article: "It’s no wonder Services — which includes App Store revenue — has become an increasingly important business for Apple as hardware sales have slowed.".
While you're correct that it costs money to run Steam, it's not _that_ much money.
https://youtube.com/watch?v=stxVBJem3Rs Here's a talk from a very old indie company, there's one section talking about how they existed pre-Steam (and other stores they use since Steam's not exclusive) and appreciate what it's enabled.
Steam isn't so much better than the competition that no one can compete with them--they just got into the market at the right time and network effects have taken over.
I mentioned probable economies of scale in a cousin comment, but I don't think they offer all that much. Do you have a % in mind that you intuitively think could be knocked off from the stores of Valve, GOG, Google, Apple, Microsoft, Salesforce App Exchange, and Amazon Appstore, all taking around 30% before you get into the fine print?
Assume a $50 game. 100,000 sales. 5 gigs.
Storage and bandwidth = $28,000 (per Amazon's calculator)
Cost of doing business in N countries = use paypal = 2.9%+0.30/sale = approximately $1.75/sale = $175,000
Costs = $203,000
Revenue = $5,000,000
Ratio = 4.1%
So yeah, Valve is not the shining knight here and that 30% fee is unjustifiable. Even Epic's 12% share is still very much profitable for Epic.
The 30% fee is also nowhere comparable to wholesale pricing schema of the retail model (effectively a 40-60% fee), since that involved the sale of physical copies of goods, and the retailers had to deal with issues like shrinkage and the risk of unsold inventory.
This example is also using s3 retail pricing and paypal retail pricing, which already factors in profits for Amazon and Paypal.
None of the big marketplace's costs are anywhere near 30%.
I hate Epic for their Exclusivity, but lets not pretend that 30% Fee is nothing more than a money grab by Apple, Google, Valve and the rest of these platforms
The 30% figure is a bit deceiving. If you sell a lot on Steam, you can get a better cut from Valve, but you don't even need to sell a single copy to get a 100% cut - Valve charges nothing for Steam key generation. You could sell keys on your own website and only pay payment processing fees, or get a decent cut from Humble, GMG, etc and not have to manage a storefront. One-third of all Steam game registrations are key activations , so this isn't some insignificant revenue stream.
What HN gets annoyed with is exclusivity and lies. Valve never required exclusivity and did not try to become a monopoly (you could sell on GOG, Origin, and any other storefront alongside Steam if you wanted), they just sucked less than the competition (GFWL et al) and developed consumer loyalty through their sales.
Valve is far from perfect, but they're decent and aren't anti-consumer, so when Tim Sweeney starts handing out $2 million for exclusives  to undermine everyone else and lies through his teeth on Twitter, people get mad. Epic doesn't want fair competition or they'd focus on making a much better launcher and sell their platform on its merits rather than its exclusives. Tim doesn't care about introducing competition, because it already exists (GOG, Origin, UPlay, etc), and people are still happy to stick with Valve.
>That 30% is coming straight out of gamers' pockets
This implies gamers care about the cut like it's a tax, and that if Valve lowered their cut, games on Steam would be cheaper. I did a quick check and saw the new Watch Dogs: Legion can be preordered on EGS for $59.99. Logically speaking, if the 12% cut is priced at $60, buying from Ubisoft directly (where there's a 0% cut to EGS) should be even cheaper, but nope, it's the same price. The lower cut translates to more profits, and executives won't lower prices just because they're getting more money. The reason for this is that video games are intellectual property, not commodity goods where people can directly compare prices for equivalent products (that is, while I can pick and choose a generic paper towel brand over Bounty and get similar utility for less money, choosing Need for Speed over Forza will result in a very different experience, such that you can't really compare games on price).
Given that Tim wants to force everyone to use his crappy launcher with no intention of Linux support, I can see why HN would rally around Valve. Middlemen are necessary for distribution, server hosting, and dealing with scammers trying to chargeback thousands of transactions, so we might as well have a company that isn't actively hostile to consumers doing it. Valve could lower their cut, but it would do nothing unless they stooped to Tim's level and started paying for exclusives, and Gabe has made it clear he wants PC gaming to remain an open platform, making paid exclusives unlikely. Doing so would result in a race to the bottom, and Tim knows that although Epic has tons of VC money to fund such a race with, Valve's refusal to take VC funds over the years would result in them losing.
One can also look at the possibilities by modeling more than just the $60 game in isolation. Include the cost, it took $X to create, and each $60 sale chips away at $X by some $Y so that after enough sales over some time window breakeven is reached (and profit after that). A new tax means that $Y becomes smaller, so you need more sales than before to reach breakeven within the same time. Except you won't get more sales than before, because the price is the same. You have to eat the tax. You could try raising the price to say $70 or whatever is needed so that $Y remains the same, and now you only need the same number of sales as before, let the customers eat it. Except you won't get the same number of sales as before, because the increase in the product's price will result in a decrease in customer demand, you'll have fewer sales than before within the same time frame. Compared to the difference with $X, it's still the company who has eaten the tax and has to deal with its consequences on their continued existence.
This is a pretty wild accusation to make with no lead-up or follow-through.
iTunes, Apple App Store, Google Play Store, even Amazon Kindle book store, all are missing carts. Because a one-click buy is way easier to navigate than adding to cart, going to cart, and then finally checking out.
They are generous with giving away steam keys, I agree with that. However the alternative for them is people selling exe on their websites instead of the steam key and losing potential customers over it. Also giving away free keys like that allows them to force devs keep the price same as steam prices
It's not a good approach to competition though, and only hurts users instead of motivating innovation. I agree that Steam could use some competition, but Epic isn't actually competing with them, they're just forcibly buying their way into the market.
People also ignore that Steam built itself up on exclusives too. Heck, Valve bought Turtle Rock Software a few months before Left 4 Dead's release to get an exclusive lock on the Left 4 Dead franchise in perpetuity, then froze the original developers out. But somehow gamers are more scandalized by Epic paying developers to release on Epic's platform a few months before Steam, despite that being less predatory in every possible way.
I think some things they do well are (which are admittedly important things):
* fast downloads
* consistent achievements between games
But I think they're lagging others (mainly discord) in chat and social. Their voice quality is comparatively horrible. They don't really have a good space for a group to have a server like discord does, which really helps with cohesion. So while I definitely don't think epic is killing it in any of these areas, I don't think steam has that much going for it outside their core competency.
Yes. Yes it is. Steam built the online games distribution market from basically nothing to where it is today. They helped usher in the indie game explosion, and gave players unprecedented tools for finding and assessing games. Not to mention the normalization of high-discount sales.
Oh yeah, and also Valve went ahead and made Linux gaming a more realistic proposition than it has ever been in history.
But that doesn't mean that steam today is the best experience for gaming, and I would probably go so far as to argue that their hold on gaming is harmful to the industry, the largest issue imo being their large take of sales.
Steam is where it is today because it first earned, and then kept, gamer trust. All the anti-Valve propaganda out there the past few years doesn't change that.
If you're buying the game to play with someone, you buy it on whatever platform the other person has it on.
Otherwise, you look at whatever platform you spend the most time on. If it's available there and the price is "reasonable" for you, then you buy it there. End of story. I don't believe that most people actively search out a new platform.
Otherwise, if it's not available or you think you can get it for lower cost elsewhere, then you go to the next store you think is likely to have it or have it cheaper.
Given that framework, it doesn't really matter how good a platform is for a consumer outside of being able to buy a certain game. The main factors for that platform are:
1. How many people are on it.
2. How wide the selection on the platform is.
So I would argue that not only does the platform compete for users on 1 (which I've argued is mostly based on 2 and pricing), but also for developers on 2. I think steam's behavior with their features is mostly oriented towards making the platform attractive to developers, and Epic is offering exclusives and $$$ to draw developers in.
As to whether Epic's exclusives are good for that gaming community, you can argue that they will allow for an ecosystem of epic subsidized developers, and lead to more competition from steam for holding developers.
Also, for a consumer there's a big convenience advantage to having a single store - from UX perspective (but probably not from the wider market leverage perspective, tragedy of commons yadda yadda) I'd very, very much prefer using a single service that's good enough to having to use three or more separate places that each individually are somehow better but each doesn't carry all the content that I want. The same applies to most other media - movies/shows, music, books.
IMHO the only way to get competitiveness would be to ensure that exclusivity deals are impossible/illegal, so that you might have multiple services each providing all content but competing on other qualities.