Yay!! Got message from James Monroe (programmer of Strife)

Hey guys, here's what I had sent to him:

From: Bruno Vergílio [mailto:brunodeovergilio@gmail.com]
Sent: Tuesday, June 02, 2009 8:06 PM
To: GRP RAVENSOFT www
Subject: A letter for JAMES MONROE, from a big fan, I beg you for reading

Hi Guys. My name is Bruno. I'm from Brazil and I'm 21 years old, and I still have a big love for old games... one of them is strife... I know it's NOT YOUR RESPONSIBILITY, but, I know that james monroe was the programmer of strife, and when I discovered that he was still working with games, I told myself to take my chances and try to contact him... so James, this letter is for you:
I'd just like to have a little support about strife... I mean, we know that the source has been lost, but as a programmer, you never forget some lines of code hehe, so, I'd like to ask you to help the guys at doomworld forums (us) to decypher some lines of code and some functions, just in case you remember them... we have made such a big progress porting strife, so we would like to see it fully compatible with the old exe... I mean, demo recordings and all the other functions and codepointers... We would love to hear something from you, even if you answer saying that you can't help, but please, I beg you, talk to us... you love games as much as we do, otherwise you wouldn't be programming games for all this time, so please help us to keep alive what was a great game in the old days, help us port them to windows and other systems... I hope you read my message, and sorry for the bad english and for the missunderstandings... THANK YOU IN ADVANCE !!!!

BRUNO DÉO VERGÍLIO

Here's his answer:

fromMonroe, James <JMonroe@ravensoft.com>
tobrunodeovergilio@gmail.com

cc"Peters, Nathan" <NPeters@ravensoft.com>

dateThu, Jun 25, 2009 at 3:15 PM
subjectRE: A letter for JAMES MONROE, from a big fan, I beg you for reading

Hi. I'm glad you're still loving Strife. It was a long time ago, but I'd be happy to try to answer any questions you have.
It was based off the doom source code, so there is a lot of info out there about the code structure.

Now it would be nice if we could collect some questions from the guys who are messing with strife (Graf Zahl, Kaiser, Quasar etc), and send to him... he may knows some of them =D !!!

Neat.

Please note that I have already spoken with Mr. Monroe as of two days ago to clear up some issues relating to the message you sent him, which although I know you meant best, was misleading as to the nature of the problems we have with Strife.

Our reverse engineering of the source code is nearing completion. Asking him for help was never what we needed to do, as we've done very well at this ourselves. 99% of the functions have been identified, and we are even starting to crack the tough nuts like P_CrossSpecialLine, which is munged up by the compiler into a gigantic jump table which Hex-Rays can't figure out (Kaiser on the other hand has figured it out quite nicely).

In the reply to my message, James told me the following things:

A) At the time the Raven sources were released, he inquired to John Carmack about the possibility of releasing the Strife source code, and Carmack approved of the idea at the time, providing that the source could be found. However, Rogue's computer systems had already been disposed of and James did not have a personal copy, so he had to give up. That being said, however, he believes that rights in the code reverted to id when Velocity shut down, as per some preexisting agreement, and is currently looking for correspondence which points in that direction. I expect him to email me again if/when he finds it, but it appears already based on what he has told me that the way is 100% paved for a GPL publication of the reverse engineering results.

B) He thought ACC was already included with the Hexen source code, and the fact that it is still under a restrictive license is a mistake. He intends to rectify this when he gets time.

Please don't pester him about these two issues, as we're working together currently to have them resolved. I'll post more news when everything is final.

Interesting. Same guy who worked to make the Raven sources public, with the one who programmed Strife.

Ok Quasar, thank you very much for these explanations, I will just reply him saying thanks for the good will, and won't bother him with anymore details...

printz said:

Interesting. Same guy who worked to make the Raven sources public, with the one who programmed Strife.

Yeah, there seems to have been a lot of interbreeding between id-Raven-Rogue.

Anyway, this is really good news (ie. Quasar's comments). I agree with Quasar, though - please don't harass the guy.

I'm glad to hear that ACC will likely get a full source release. It will be interesting to see how it works and it will present the possibility of developing an accurate decompiler (sans whitespace and comments of course). Anyone else interested in a near-as-damn it ACS source for Hexen?

I think we already have the ACC source; it's just under a restrictive license like Hexen itself was until last year.

I didn't realize there had been a non-binary release of ACC (me Googles)...

EDIT: http://www.3ddownloads.com/showfile.php3?file_id=7902 EDIT2 Seems that link now points to something else entirely (might want to update the Source Ports page: http://www.doomworld.com/classicdoom/ports/index.php?platform=1)...

You can find it here.

There's also already a decompiler, De-ACC, by Luc Cluitman, which seems to work just fine. It's integrated into Wintex, and so clicking BEHAVIOR lumps there will display the scripts. A better one could probably be written, but you really wouldn't need the compiler source to create one. Just the interpreter and a spec for the bytecode format (which is pathetically simple).

My reasoning for the compiler source being useful in creating a decompiler for compiled ACS data is that it would allow for reversing any optimizations which the compiler might be applying (thus obtaining a more accurate script source). However, maybe I'm being a bit optimistic in thinking that ACC is doing anything complex at all optimization wise.

I don't think ACC has an optimizer at all :) That's one reason why decompiling it is so simple compared to most languages. Small for example could be disassembled, but decompiling it would be just as much work as it would be for binary C programs.

Decompiling the old format scripts was pretty easy. A number of tools did it pretty successfully. I used to use the tool that is part of DeePsea which gives a very good, compilable ACS file to work with (I decompiled TQ Trust with it and recompiled it again and everything worked). Unfortunately, I'm not aware of any tools that can do the same for the newer Zdoom compiled script lumps. :/

Enjay said:

Unfortunately, I'm not aware of any tools that can do the same for the newer Zdoom compiled script lumps. :/

Didn't Graf release a deacc for ZDoom scripts a while back? If I recall correctly he said the only problem was it couldn't handle hudmessages.

He did? I remember him mentioning that he had a tool that did some kind of disassembly with the new format. I didn't remember that it was a full decompiler or that he had actually released it.

If he did release it, knowing me, I probably already have it and forgot about it. Oops.

There is that.

And, sure enough, I posted in that thread. :)