The future of the AGS engine: STALEMATE!

Started by monkey0506, Sat 23/07/2011 11:10:40

Previous topic - Next topic

Alan v.Drake


Snarky

Quote from: Ryan Timothy on Mon 20/02/2012 12:44:22
On a side note, AdventureGameStudio.net  is available. .com seems to be held hostage by a company trying to sell the domain - I think. I'd do a whois domain inquiry but I'm on my cell.

I went ahead and registered adventuregamestudio.org and adventuregamestudio.net. Once we've worked out the server situation, one way or another, we'll have those domains to use. (I think they're preferable to adventuregamestudio.co.uk or adventuregamestudio.co.uk, personally.)

Ryan Timothy B


m0ds

CJ replied saying he will come and sort it (the awards noms) soon.

Try not to think of his "soon" as my FoY "soon" ::)

edit: YAY! CJ  :o

Privateer Puddin'



Pumaman

Ok I think I've sorted the noms page, and sent it to bicilotti.

Sorry for my absence around here ... I've just been really busy lately.

The other thing I really want to do is finish off the new website ... it's so nearly there, 95% complete and just needs the Search page and Post Comment form to be done. I hope I'll have the time to do that "soon", but is there anything else people are needing from me?

Ponch

Quote from: Pumaman on Tue 28/02/2012 23:22:39
is there anything else people are needing from me?

Can I have a hug? Or possibly a pony?  :-*

Calin Leafshade

The source of AGS.Native (in any state, we'll fix it up) would be great. A lot of people are really anxious to get a look at the compiler.

timofonic

Quote from: Pumaman on Tue 28/02/2012 23:22:39
is there anything else people are needing from me?

Yes, older AGS sources (better if since 1.x days, even better if releasing the VSS repository). Pretty pleaseee...pleaseeee, with sugar on top :)

ScummVM Team want it for easier support of legacy AGS versions, there's a WIP engine from the "fuzzie" team member.

fuzzie

Quote from: timofonic on Sun 04/03/2012 19:23:52
Quote from: Pumaman on Tue 28/02/2012 23:22:39
is there anything else people are needing from me?
ScummVM Team want it for easier support of legacy AGS versions

I would definitely put AGS.Native at the top of the list, myself.

There's no huge hurry for the legacy code, I think, especially since JJS has been doing a really good job of puzzling out the behaviour of the various 2.x AGS versions.

tzachs

Quote from: Pumaman on Tue 28/02/2012 23:22:39
is there anything else people are needing from me?

Sub-forum(s) for editor and engine development.

monkey0506

It's nice to see that absolutely everything that we were, as a community, doing about this has dissolved completely into some sort of distant cultural memory.

Respect CJ, sure, but the point of him open-sourcing the software was so that we could maintain it.

The source of AGS.Native would be nice, but it's not even essential. If we're refactoring all the code anyway, we could just write our own compiler.

I actually pulled back on the throttle of some of what I was doing due to certain comments of a similar nature to this one (about nothing being done), and I was somewhat upset at those people who apparently never realized how actively I had been working on this.

In any case, I've (personally) decided that SDL seems to be more (pro)actively developed, if nothing else, evidenced by the fact that Allegro only so very recently got ported to Android. Based on that, I've started learning SDL myself. Working on getting it to build in Code::Blocks using the GCC 4.7 compiler so we can take advantage of C++11. It would be ignorant to hold ourselves back from the great functionality of C++11 unless there was a specific reason for doing so (which to date, no one has presented me with).

I'm not angry at anyone in particular, but it does seem a waste of everything CJ did for all the momentum we built up to just die off without anything to show for it. I for one, am going to stand behind what I said, and keep contributing...however I can. Whether my efforts end up totally for naught, whether I end up refactoring the entire code base, or whether I end up writing my own engine based off of AGS 3.2.1....I'm just going to keep trudging blindly forward in the hopes that someday someone will find some use of my work.


-monkey

fuzzie

Quote from: monkey_05_06 on Wed 14/03/2012 18:00:38
The source of AGS.Native would be nice, but it's not even essential. If we're refactoring all the code anyway, we could just write our own compiler.

To be clear, the script compiler in Common/csparser.cpp seems to be up-to-date. But I don't think anyone has been suggesting refactoring the editor side of things? So improving the editor iteratively based on AGS.Native source (if available) seems like a far more productive idea, even if it isn't essential.

Quote from: monkey_05_06 on Wed 14/03/2012 18:00:38
It would be ignorant to hold ourselves back from the great functionality of C++11 unless there was a specific reason for doing so (which to date, no one has presented me with).

Portability (to e.g. console SDKs) is the usual problem with using C++11.

It's unfortunate that my rewritten AGS code is stuck with an unpopular license due to being a ScummVM engine, but I've written an awful lot of code in 90+ commits (and there's a lot of uncommitted work, it's difficult to find the time), and obviously I'm hoping that when it's mostly done (a few months, probably), it will be at least a nice reference for anyone trying to do refactoring of the original code, even if the license makes it unacceptable as-is as an ongoing engine for the community.

In the meantime, all my work is public, and there's a lot of people who have been very positive about it, both from the AGS community and outsiders, as well as various people interested in contributing code. JJS's work on backwards compatibility as well as portability and improvements has had a very clearly positive response too - I don't think that work can be discarded as 'without anything to show for it'.

Calin Leafshade

fuzzie is wise, you should heed her words.

To say that "nothing has been done" is insulting to those that have done something. Fuzzie's repos is very active and JJS's work is invaluable.

What would be more accurate is that *you* have done nothing.

Dave Gilbert

Quote from: Calin Leafshade on Wed 14/03/2012 19:31:25
What would be more accurate is that *you* have done nothing.

To be fair, there are so many threads about this issue scattered all over the forum that it is difficult to asses what actually IS being done. I wasn't aware that anybody had been working on this either. It wasn't that long ago that I was asked to be put in charge of managing the development of the engine (which I eventually accepted), and while I don't mind that things are going in another direction, nobody ever told me that this was happening. Or if they did, it was in one of the many forum threads about this topic. I had just assumed that progress and/or interest had completely stopped.

So for the sake of everyone else like me: what, exactly, is being done for the future development of AGS? Who is in charge of it (if anybody)? What are the short term plans? How long is a piece of string?

-Dave


Calin Leafshade

Well the code is available for all. In reality anyone can set up a branch of the engine and move forward with it. The only person who could sanction any kind of "official" AGS development would be CJ and he hasn't done so.

So if you (or indeed anyone) wants to advance the future of AGS they are welcome to do so. Set up a branch, recruit some coders and have at it.

Since the community (and remember this is only the big blue cup forum community, AGS as a whole is bigger than that) is finding it hard to do anything concrete due to drowning in bureaucracy this might be the best way forward. If everyone moves to do their own thing then eventually the weaker projects will die out and the strong ones will merge or compete until one is the clearly better product. Perhaps we should leave this to market forces and allow those who wish to take matters into their own hands to do so without fear of a backlash.

monkey0506

To be clear, my post was more directly about what Dave is also referencing.

fuzzie has created a fork that if for no other reason than the license alone is completely incompatible with the trunk and official development branch. I don't mean to discount the work that's being done there, but it isn't what I was talking about, as I will explain.

JJS was already developing those ports before any of the aforementioned "momentum" of "managing the development of the engine", etc. set forth. Yes, what he's doing is good and beneficial, but it again is unrelated.

Calin, to say that I have done nothing is simply inflammatory without cause.

I've already made it clear that I don't feel I have the same technical proficiency with C++ as those like JJS or fuzzie. I have made that very clear (from the very first post in this thread!!). So, I don't have as much to show for my efforts as they do (if indeed anything at all).

Still, I was referring to the sense of direction that we finally had built up, that someone was actually going to step in as a lead of the project, that someone was going to take charge of the technical aspect therein. Individuals creating incompatible branches and forks is not what I was referring to.

Aside from the likes of JJS and fuzzie who are working, much like myself, but with greater proficiency in C++, as individuals with their own thoughts, feelings, ideas, and goals in mind, there is no single person currently taking charge of AGS as a whole to move the project forward.

Calin you posted while I was typing, and this "fear of a backlash" is exactly what I was referring to. If nothing else, organizing a team from the forum community would be better than leaving it to a few sparse individuals to bear the weight on their own. As AGS is open source, there's nothing preventing anyone from trying to take it in a new direction, but the fact of the matter is that the move forward has been so slow as to negate the work that has and is being done.

Dave Gilbert

#198
Yes, I meant something "official." Having a bunch of different branches with their own different features and no way to merge them doesn't seem like the best way to go.  What I was hoping for (and others, I'm sure) is a return to the Old Days - one official version of the engine that gets updated on a regular basis.

Sure,  I could start using one of the other branches (or make one of my own), but if another branch develops cool features it will suck that I can't use them. Or if one branch becomes ported to iOS and mine doesn't. Or... well, there are too many variables.

Is there a current branch of the engine that is the most stable/further along? Can we just make that the official branch and be done with it?

-Dave

fuzzie

Quote from: monkey_05_06 on Wed 14/03/2012 20:23:03
fuzzie has created a fork that if for no other reason than the license alone is completely incompatible with the trunk and official development branch.

Yes, mostly I'm just mentioning it since it could perhaps be useful at some point (any relevant bits could be dual-licensed under the Artistic License 2.0 too), but if the community isn't happy with the GPL for the main development branch then it indeed isn't too relevant.

Quote from: monkey_05_06 on Wed 14/03/2012 20:23:03
Still, I was referring to the sense of direction that we finally had built up, that someone was actually going to step in as a lead of the project, that someone was going to take charge of the technical aspect therein. Individuals creating incompatible branches and forks is not what I was referring to.

Aside from the likes of JJS and fuzzie who are working, much like myself, but with greater proficiency in C++, as individuals with their own thoughts, feelings, ideas, and goals in mind, there is no single person currently taking charge of AGS as a whole to move the project forward.

Speaking as an outsider (obviously): it seems like setting some goals in terms of bugfixes and small features might be a good start, since no-one seems to have both the ability and time to start any serious refactoring work, which I think has usually been the first step suggested in most of the discussion so far.

Quote from: Dave Gilbert on Wed 14/03/2012 20:37:15
Is there a current branch of the engine that is the most stable/further along? Can we just make that the official branch and be done with it?

Well, presumably JJS's branch would be the one to pick, since it has improved portability and bugfixes, but then the priority there is more about backwards compatibility than any new development (inevitably).

SMF spam blocked by CleanTalk