Hard Light Productions Forums
Modding, Mission Design, and Coding => FS2 Open Coding - The Source Code Project (SCP) => Test Builds => Topic started by: chief1983 on April 23, 2008, 12:41:17 pm
-
Hasn't been a build posted for a while, so I thought I'd put a new one up now that I finally have a fully functioning compiling environment (that does FRED too!). Just checked it out from SVN, revision 4608. Since it's pure SVN, any bugs in it should likely be Mantised. Enjoy.
fso-20080423_CHIEF_4608.7z (http://swc.fs2downloads.com/files/fso-20080423_CHIEF_4608.7z)
MD5 (http://swc.fs2downloads.com/files/fso-20080423_CHIEF_4608.md5) (right-click, save)
Ok, I lied, it's not _pure_ svn, there was one tweak to stdafx.h and folderdlg.cpp to make FRED compile in VS2008 Pro.
Edit:
Slightly newer build, should probably use it since it has the targeting bracket fix. If no one makes a new windows build I'll try to get around to making another one this week, hopefully some more of the mantised bugs should be fixed. And there seems to be some changes to some multi code too.
April 26 Rev. 4609 (proper file name this time) (http://swc.fs2downloads.com/files/fso-20080426_CHIEF_4609.7z)
MD5 (http://swc.fs2downloads.com/files/fso-20080426_CHIEF_4609.md5)
-
yay! chief is a programmer!
-
We've had a bit of a run of people suddenly getting into coding for some reason. :)
-
We've had a bit of a run of people suddenly getting into coding for some reason. :)
Count me in, starting from next fall or so. I'm attempting to switch from process engineering to information engineering.
-
I didn't say I was getting into coding, although I am a Comp Sci major. But the project of this size scares me, I'd probably need someone to hold my hand and lead me into a small isolate chunk to get me started. I just figured since I could make builds I'd start filling the deadspace on this forum with trunk builds, since that's what it's for. I'd also like to suggest that a nightly build system be set up. It would encourage project files to be updated for one and make bug reporting go faster I would hope.
-
I'd probably need someone to hold my hand and lead me into a small isolate chunk to get me started.
This is exactly what I did a couple years ago, minus a lot of the hand holding ;) Actually it was more hand-slapping, "No! don't change that yet, it'll cause everyone's pilot file to explode!" :lol:
There is no way I understand most of this project... the graphics and AI sections do scare me... but thanks to some of Visual Studio's searching tools (Call Browser and Go To Definition are lifesavers) and patient help from the other members, I've learned enough to make my mark and change things in other areas that nobody would have time to bother with otherwise.
I encourage you to pick a topic or feature you're interested in, ask where to start looking, dig in, see what it does and then change little things here and there to see how it works. (I started with the gliding mode in the physics system, then spread out from there.) You'll probably end up starting over once or twice, especially if you plan to have your work end up being committed... but it's a great way to learn. Depending on what kind of computer field you're going into, this kind of learning to search existing code and change things is a valuable skill...
I'll be glad to answer questions and do the initial hand holding if you wish :)
-
From what i have learned during the short time i have been trying to do things... Stuff like creating doxygen (and keeping it updated) of the current SVN helps a lot... As well as things like WinGrep (even if it nags) for searching the whole database for something. But help from more experienced users has been the most important thing.
-
having played wit c++ for awhile i thought i might mess with c. possibly help fix bugs in scripting. but i dont think im skilled enough yet, and the freespace code still scares me.
-
You probably still know more about C than I did when I joined the project. Although I may understand Java well my C++ experience consisted of a "Teach yourself C++ in 24 hours" book.
The trick is simply to start small. There are dozens of tiny little things that need tweaking/fixing. Another trick is to start with an area of the engine you already know well as that means you already know what happens, you merely need to figure out how the code does it. Which is why I still consider SEXP.cpp the piece of code I most prefer working with.
I suspect you probably have enough of an understanding of the scripting system to do something similar with lua.cpp :) And I doubt WMC would mind helping you out at first as it means he wouldn't be responsible for every single scripting bug if he could leave some of the easier ones to you.
-
this is a regular trunck build right? I mean, it has fov fix and per-pixel shading?
-
Yes, straight out of trunk from SVN. So if those were in SVN at that point, then yes.
-
Interesting.
While taylor's Xt builds exhibit a "shrunken movie" bug, in which the rendered cutscenes are played at actual size instead of scaled to fullscreen, it appears that particular alteration has not yet hit SVN as this build does not exhibit those problems.
No bugs yet, I'm happy to say =)
-
I am getting black boxes around particles.
-
He knows about that bug, he had been working on that code I believe. Although I thought he said it should have been fixed before his 3/14 build and yet I still noticed it then.
-
That could also be a problem with the particle art itself. Unnecessary white alpha channel. It's not a bug, it's a feature -- now we can actually use alpha channels :D Give me the particle and I can fix it for you. Or if it's an old mediavp set, get the 710 alpha fixes.
Fov fix is definitely in SVN... per pixel shading sounds more like something in an Xt build.
-
nevermind, the images are bad, that's all.
-
I must have missed a post, earlier I was referring to Taylor knowing about the movie bug. Sorry for any confusion.
-
Found a bug.
http://www.hard-light.net/forums/index.php/topic,53511.0.html
http://scp.indiegames.us/mantis/view.php?id=1671
-
just a quick question , this is marked a 3.6.9 build , it has no normal maps and no jpa /tga textures but logs into 3.6.10 network in multiplayer can you please clarify what this build is as we are bug testing 3.6.10 multiplayer........
-
Another bug:
http://scp.indiegames.us/mantis/view.php?id=1672
-
Yet another bug. This one is multiplayer only.
http://scp.indiegames.us/mantis/view.php?id=1673
-
just a quick question , this is marked a 3.6.9 build , it has no normal maps and no jpa /tga textures but logs into 3.6.10 network in multiplayer can you please clarify what this build is as we are bug testing 3.6.10 multiplayer........
Sorry, guess I forgot to change the file name. The default in the build process is apparently to name them 3.6.9 still, and I didn't even notice it. It's just a checkout from SVN, stable branch, and then built as-is. It's only 'marked' a 3.6.9 build by the filename, in game it's still 3.6.10 correct?
-
just a quick question , this is marked a 3.6.9 build , it has no normal maps and no jpa /tga textures but logs into 3.6.10 network in multiplayer can you please clarify what this build is as we are bug testing 3.6.10 multiplayer........
Sorry, guess I forgot to change the file name. The default in the build process is apparently to name them 3.6.9 still, and I didn't even notice it. It's just a checkout from SVN, stable branch, and then built as-is. It's only 'marked' a 3.6.9 build by the filename, in game it's still 3.6.10 correct?
Correct. It shows as a 3.6.10 build.
-
It also connnects fine to the 3.6.10 server in MP, and plays nice enough when paired with Xt builds. We were able to test both of those extensively tonight.
-
Ryan, I uploaded another build that had a change done to hudbrackets.cpp, and might have been a bugfix just after I made the last build. Give it a shot if you want. Hasn't been any real activity on SVN since the 23rd though.
Build (http://swc.fs2downloads.com/files/fso-20080426_CHIEF_4609.7z)
MD5Sum (http://swc.fs2downloads.com/files/fso-20080426_CHIEF_4609.md5)
-
Yeah that fixes the targetinfo bug. I did that. My bad. :nervous:
-
No problem, I had just timed my first build poorly I guess, you squeezed that fix in just a few hours after I got it posted.
-
That fixed the targetinfo bug, but not the multiplayer bug (http://scp.indiegames.us/mantis/view.php?id=1673) or the big mysterious one (http://scp.indiegames.us/mantis/view.php?id=1671). Note that I've added a few more test scenarios and information over the original bug report in the case of the mysterious one.
-
what's the big mysterious one? mantis isn't working right now.
-
what's the big mysterious one? mantis isn't working right now.
If you have Derelict SCP, try playing the mission "The Sting" in this build. Whatever is causing the mess that results has been introduced between Taylor's 0314 XT build and now.
Essentially:
-A Sobek can drive right into/through an Orion model (which the FSU says is a code problem).
-The Sobek totally ignores its waypoints, which work fine in every other build.
-Fighters are not properly accounted for, so even when all fighters are killed you're left with a positive count of 1 that screws up the rest of the mission.
I have a theory on the fighter problem: I think that the game is not counting kills of a remaining fighter from the first wave of the wing once the second has arrived. Thus, even when both waves have been killed you're left with the number remaining from the first wing once the second has shown up.
I'm sure these are larger bugs that will appear in multiple places given the things they affect, this just happens to be the mission I discovered them in and which they are glaringly obvious.
-
Hey, thank you! :)