Well...they didn't want to upgrade to CE5 because it was too much work and the engines were too different following their reworking...
But effectively did it anyway, which suggests their attempts to fix the NetCode in their CE3 engine were less than successful.
Other than that....I'd be interested to know what this mean for the game.
Is the NetCode now "fixed"...or are CIG limited in how the an modify it
Are CIG now stuck with Amazon?
What are the new instance limits? Will we get the 200 players per instance CIG were wanting or is that still some time away?
CIG had reportedly stopping using CE patches because the code base had diverged so much...how does that work for Lumberyard? When they say all the changes made have been moved to Lumberyard, does that mean it's been tested and intergrated smoothly as well or does it mean we are in for a year of "testing" instead of development and progress?
When they say they have moved to Lumberyard...could that mean they may simply have licensed the Netcode and effectively merged it into StarEngine rather than vice versa? That is, kept as much of their own engine intact and made as few other changes as possible in order to fix the netcode?
Does Lumberyard offer CIG any other advantages?
More importantly....does this mean work on the Engine has "finished"? Can we expect quicker progress? Can we expect SOME progress?
Lumberyard might be based on CryEngine, but it's Amazon's adjustments in it, not CIG's. Every custom modification - years of work - from both CIG and Crytek employees are now reset to zero.
The netcode was never "broken" per se - it's that the built-in CryEngine netcode could not support what CIG was trying to do - it's not an MMO engine. Lumberyard is also not an MMO engine. So the netcode work also starts over.
There's no "merging" either - the changes to create "StarEngine" were too deep. CIG's developers have gone on the record, in attempting to validate delays and lack of progress, saying that they've changed over 50% of the CryEngine code base. Now those same people are claiming that it's a simple process to merge their changes into Lumberyard because it's so "similar."
Yeah, similar except for the 50% they've been crowing about for years. They basically need to change 50% of the Lumberyard code now ("LumberStar?") and how long that will take is anyone's guess. Not to mention that the results of their work were embarrassingly bad, and there's no guarantee that Lumberyard will do any better. On top of that, there's no guarantee that Amazon will allow them to host a deeply modified version of Lumberyard on their servers without understanding the impact their changes will make to their infrastructure from a security and performance perspective.
If I was forced at gunpoint to contrive a positive spin on this, I would say that CIG's familiarity with their own changes to CryEngine over the years means that the migration to Lumberyard would take less time than it would to migrate to an entirely new engine, such as Unreal. That's what I'd say at gunpoint.
Not at gunpoint, I'd say LOL.
