Cheating is not an issue at all. People can cheat better when there is no source code - by decompiling or so, they can still find exploits. However, server admins are powerless against that, as the game company no longer supports the game, so there are no further patches. With the source code released, on the other hand, the community itself can fix those issues. That said, game companies might be reluctant to release source code for this reason just because they're unaware that it's not an issue at all.
Removing references? If they had those in their code to begin with, then they were pretty poor coders... There is no reason to put that kind of information in your source code. It would be the equivalent of scribbling that information on a company's financial report - it might be used only internally in the company, but there is no reason for you to put that information there for others to read it to begin with.
Reused code and 3rd party IP are indeed the most important reasons not to release the code (Epic Games don't release their code for the former reason). That, and laziness.
You'd be amazed what corners are cut, and what things go against "proper" coding in the real world. There could be quite a bit in source code that shouldn't be exposed to the outside world, for one reason or another, and not due to poor coders or laziness. Real world trumps idealism here.
The end result is that even if the code contains no 3rd party ip, a review is required before publically publishing it.
I cant go into details but I have had a bit of experience with trying to get companies to release source code to old games. There are all sorts of reasons why they may not want to do it, including:
3rd party IP (sometimes they know what 3rd party code they are using, sometimes its a case where they dont know what code may be in there)
Code they are still using (sometimes they have code that they used in an old engine that is still in use in some form in a newer engine and releasing it may be an issue)
Cheating (even on old titles, if there is multiplayer, releasing the source may lead to cheaters being able to exploit the game in ways not previously possible such as forged network packets)
Copy protection (if the game links to APIs and libraries connected to copy protection, they may need to completly remove anything that depends on the copy protection so as not to reveal any of the secrets)
But in general the biggest issue is the amount of time it takes to examine the old code, make sure they have the legal right to release it (or remove the bits they cant legally release), make sure its got nothing in it that might be an issue (such as swear words in the comments, references to internal computer systems/email addresses/employee IDs/passwords/etc, details of unreleased games or ports etc) and make sure the source they release is usable (e.g. making sure they have all the bits they need)
While I agree with most of the points, don't you think investing some money into code review wouldn't pay in the end for EA?
I assume the positive feedback for open-sourcing some of the blockbuster mid- to end-90's game engines would certainly give
them an amazingly positive customer feedback and a massive press coverage. IMHO, EA is in dire need of positive feedback.
First off, I understand that those games are somewhat old, but still NICE TITLES.
ALSO, they are OPEN SOURCED. That's nice progress from EA. Quit crying!!!
Not as much progress as you'd think. SimCity was open sourced in *2008*, essentially as a donation for use on the OLPC (it is "SimCity" on the official version for OLPC, and "Micropolis" otherwise.) DUX software licensed the SimCity for Macintosh code from Maxis in the 1980s, and DUX then hired Don Hopkins to perform the UNIX port. He's kept it up-to-date ever since (at least through 2008.)
So, really, it appears to me that EA is really not doing anything new for open source, they are just highlighting what they HAVE done now that Valve is making big moves in that direction. All smoke and no fire.