[taylor]

also what would be nice are more people who add stuff to the wiki.

I was thinking the same thing.  There really just needs to be a person, or persons, who are part of SCP for the purpose of writing documentation.  Expecting a coder to write docs is not the sanest thing in the world.  Although we can do it, the docs may not be comprehensible, and it takes away from doing other things, like you say.  There are, or have been, SCP members who will test things and then those who code.  We just don't have anyone dedicated to the job of actually keeping up with all of the new features and document them.  I guess we need to try and create a "DocMaker" position for the project, and then trick someone into volunteering for the job.

[S-99]

The thing about making a new hud idea is very cool. But, as opposed to cosmetic changes, you can already flexibly customize the hud in the game. Color whatever you want, remove what you don't want on there, and i think even control transparency to preference(if i remember right). The real potential for changing the hud would be within making custom hud for different fighters. That would be sweet as hell. Something interesting for an interceptor hud would be a bomb tracking display that automatically locks onto the nearest bombs while having a standard display for locking onto everything else, even possibly another radar specifically for tracking bombs along with the standard radar(tracking everything on the single radar does get a little hectic with a big engagement). Another thing that would be cool for custom huds is something like secondary laser target reticles, for something that has offset gunpoints like the ursa per say.
But the great thing about the standard fs hud is that it's very flexible, and really comes with everything you absolutely need for everything you fly in fs. I don't see why the hud is outdated in any way, so far the only recommendations for the hud so far have been what seem to be minor cosmetic alterations. Another thing for customized huds is that yes it takes a lot of time that people don't have time for, and that what happens when someone makes a new fighter for a campaign? You'd have to make a new hud for pretty much every class of ship specifying where target reticles go and so forth, or what other capabilities you'd want. That'd be a hellstorm for a campaign that has tons of new ships like inferno. That's the nice thing you can forget about when making a campaign, you can forget about the hud layout, the aspect you need worry about is a good escort list. The fs hud is so far universal for fs combat, and flexible for different ships in the game.

[WMCoolmon]

There's so much possibility with what's already implemented, though. What's in there isn't much more advanced than what 90% of what the lab GUI uses. Or in other words, I believe that you could remake the GUI to a checkbox level, possibly even add in scrollbars and such. The current lack of keyboard input is a pain, restricting you to true/false values and spinners, and it's hardly the most optimized way to do things...but it's still very doable.

Most of the reason that these features don't get used isn't because they are poorly documented, it's because very few people are sitting down to do them. Things like "There's no documentation" or "There's no list of functions" are simply wrong in the case of scripting. There's now even the list of hooks that Wanderer put in, so you don't have to look that up on the forums. No, the one on the website is not necessarily up to date.

Short of copy-pasting in the 3-4 examples that I've posted, linking in the example HUD, and rewriting to the nth power, I'm not sure that there's much that could be done to make the documentation better, short of people actually reading it and trying to follow it and having problems with it, that are then addressed.


With hud_gauges.tbl, the most difficult thing to understand are the custom gauges, I think. You have to define those in the hud_gauges.tbl, then you can set preset values in the table, but will generally want to set the values using the SEXPs.

[DaBrain]

People wanted a way to customize the HUD.  WMCoolmon did a lot of work on hud_guages.tbl so that they could make changes to graphics, positions, etc.  Future work was planned to make it better as the community used it.  BUT NO ONE USED IT!

Really? You should know better.   

[taylor]

Really? You should know better.

Yeah, but we do know that now.  That's why we don't work more in many types of features.  I mean take the envmap code for OpenGL for instance.  At some point I may start working on it again, but even when it eventually gets into public hands there will be something found for everyone to complain about.  Either it's going to be too slow, or not look right, or everyone will get mad that I've never been able to figure out how to get -alpha_env to work properly, or it doesn't look exactly like the D3D version.  Someone will probably complain about documentation on that too.  It's just not worth the effort.

I'd rather just work on bug fixes instead.  Even though some do manage to complain about that for one random reason or another, it's a lot less stress on me. 
(no one caught that particular feature reference, right?  i'd put a :wink: face here or something, but it would be too easy to notice.)

[DaBrain]

Uhmmm I think 99% will love you for addin env mapping support again... the other people can just disable it.

Seriously. And I think it doesn't have to look exactly like the D3D env mappig. It just has to look good. 




About the scripting: Well, even if the mods don't go public with their ideas for a scripted HUD, the scripting is still getting used. If the first mods release something with scripted stuff and people get the chance to see it in action and get some (more) samples, it should get more and more popular. It's only a matter of time.

[WMCoolmon]

Communication has to occur, though. If nobody talks about a feature, and I don't know of any mods/campaigns that use that feature, I'm going to assume that no one is using or particularly interested in that feature, and probably move on to something else. Or simply not code, if I don't want to do anything besides that feature. No reason to code something if no one is going to use it.

Of course, on the flip side, having people request features but then not use them is just as annoying, because then I feel somewhat deceived, and have also wasted my time.

Good communication FTW.

[CaptJosh]

Personally, I don't care what the D3D evironment mapping looks like anymore, or what the OGL version will look like in comparison because I no longer play FS2O in Winduhs anyway. And I think I can speak for my fellow Linux users, and probably the Mac users who play this game as well that we would just like the feature there and working already. And DaBrain is right. It doesn't have to match D3D. It just needs to look good.

[taylor]

Personally, I don't care what the D3D evironment mapping looks like anymore, or what the OGL version will look like in comparison because I no longer play FS2O in Winduhs anyway. And I think I can speak for my fellow Linux users, and probably the Mac users who play this game as well that we would just like the feature there and working already. And DaBrain is right. It doesn't have to match D3D. It just needs to look good.

I was just screwing with everyone.  Envmap support for OpenGL will hit CVS this weekend.
The code has been in final testing for a little over a week.  DaBrain knows all of this, he's just playing along.

This is the only announcement there will be of the code though, I'm going to sleep for a week after it's in CVS.  The reason I'm telling you now is that if HLP goes down for maintenance this weekend then you might not hear about it.  Also, I just finished the last missing part of the code (which I didn't think I would code in so fast) so the only thing to hold it up now is two more days of private testing.  And as you probably use CVS I figure you would like to know as soon as it hits.  Envmap for D3D will be disabled at the same time though, I don't care to update it to be in line with all of the new general envmap code upgrades.

[Backslash]

Of course, on the flip side, having people request features but then not use them is just as annoying, because then I feel somewhat deceived, and have also wasted my time.

Good communication FTW.

Agreed, and I hope we can do this better, have a fresh start now that there's the (knock on wood ) working domain, forums (with search!), and wiki.  There's been a lot of miscommunication, and quite a bit of forgetting or info getting lost or something.  Nobody's fault, it happens.  I wasn't there, but I'm guessing part of the problem involves some of the downtimes in the past, or something... for example, hud_gauges.tbl had lots of discussions in 2004 and then suddenly went silent...

One thing about this: it's not too late!  Whether or not the requested feature hasn't been used yet, it still can be!  For example, I'm pushing the hud_gauges.tbl to the WCSaga team.  (and considering taking up your Coding Opportunity  )
And all it really takes is one cool implementation, because then some other mod group will go "Whoa, how did they do that" and copy the example and then tweak it from there.

I was just screwing with everyone.  Envmap support for OpenGL will hit CVS this weekend.
The code has been in final testing for a little over a week.  DaBrain knows all of this, he's just playing along.

Very looking forward to trying it out.  Yikes you guys scared me there for a bit!  Man, taylor if you gave up just because of the stupid people, I would have to go and hunt down Every Single One.  And that would take a while.  Don't make me do that. 

[CaptJosh]

I hope you realize that if you did set about hunting down all the stupid people, you'd mostly depopulate the planet.

[IPAndrews]

Envmap support for OpenGL will hit CVS this weekend.

This is great news. Thanks Taylor.

[neoterran]

WHAT !!! OMG I'm going insane, thank you taylor. Taylor is my hero. 

I sure hope someone does a build....

So um, Taylor, can you tell us what it looks like ? Or better yet can you take a screenshot to tease us while we wait these interminable hours ?

[taylor]

So um, Taylor, can you tell us what it looks like ? Or better yet can you take a screenshot to tease us while we wait these interminable hours ?

Screenshots don't really do it justice unfortunately.  The effect is a bit more subtle and blended in with the model than the D3D was (I think, never really saw the D3D version in action).  It does look really good, occording to the testers (and myself ), but the look may change a little when it does actually hit CVS.  Right now I have cmdline options in the test builds so that testers can adjust the intensity of the effect, but the defaults seem to make everyone happy so far and those cmdline options won't go into CVS.  Still, the defaults might change between now and when it hits CVS.  But at this point the defaults look best on the widest variety of mods and models.

Here is some (very) basic info on the changes though:
  - Both -env and -alpha_env work as advertised.
  - Testers report little to no speed difference between using -env and not
  - Works equally well in static and dynamic env modes
  - DDS loader and OpenGL texture code updated to properly handle cubemaps
  - Added mission option for modders to include pre-made, mission specific, envmaps
  - Added cmdline option to save static render targets to file after they are rendered, so that you can distribute them

Not everyone will be able to actually make the env maps themselves since it requires the GL_EXT_framebuffer_object extension, which not everyone will have support for.  Most everyone will be able to actually use envmaps though, which is why I added to ability to use them from a file.  This way mod makers can just include copies of the static envmaps for each mission if they wish.  There is also a default envmap, "cubemap.dds", which will be loaded (if it exists) and used as the default envmap in the case that the user's video drivers don't support creating envmaps and the mission doesn't specify one.  And a good reason for using files is that the render-target versions can't be compressed in-game.  This means that the two default envmaps, when they have mipmaps, will use over 6meg of video ram each.  But, since you can save those to file, a mod maker can compress them into DXT1 making the files need only about 1meg of memory instead and distribute those files with his/her mod.


Of course, what you should really be asking is not for screenshots, but why I let the secret out.  I can keep a secret till death, even keep it so well you wouldn't even know there was a secret, if I chose to do so.  For me to let you know all of this before the code actually hit CVS can mean only one thing...

I've got a better secret. 

[neoterran]

The effect is a bit more subtle and blended in with the model than the D3D was 

That's really good to hear, because it looked utterly ridiculous in D3D (but still cool, and i'd still turn it on) so it's really good to hear it's a bit toned down because that's exactly what was needed.

I've got a better secret. 

The new launcher ?!

This is going to be the BEST WEEKEND EVAR !!!! TAYLOR FTW !!!

[taylor]

I've got a better secret. 

The new launcher ?!

This is going to be the BEST WEEKEND EVAR !!!! TAYLOR FTW !!!

Nope, not the launcher. Everyone already knows that I'm working on the new launcher, so what kind of secret would that be?  (the launcher is a major upgrade to game code so it's not close yet)

The real secret thing, possibly a feature of some sort, will probably go unnoticed by just about everyone.  It's just that it's such a minor thing, something I whipped up one afternoon.  It happens to be one of those things that I like, but no one has ever complained about that thing or had issue with it.  I'm really only doing it for the one or two people who might actually enjoy it.  Since I like it it's something that I can get excited about but whether anyone else likes it doesn't really matter.  I'd compare it to a computer upgrade, it would be awesome to me (for a little while at least) but nobody else would care that I got an extra gig of RAM or something like that.  This secret is that type of thing.

[Mars]

Completly redone engine?
Redone AI?
Better handling of RAM?

[neoterran]

What ! What could it be ?  Damn, I can't wait for the weekend to get here. You've already improved my freespace experience a thousand fold, looks like you'll be doing it again...

*edit*
So uh, since the forum will be down this weekend for upgrading or whatever, what day can we realistically expect a new build if everything goes well with the testing ? Monday ?

[taylor]

Well there will always be speed improvements and better memory usage.  Lots of little things have been fixed in the course of this.  For instance, I found one bug where compressed DDS images loaded from disk/VP, when they have mipmaps, will use 30-35% more memory than they actually needed.  I cleaned up a few redundant math and state changes per frame.  The envmap squishing bug where the generated envmaps look freaky should be fixed (hopefully, as reported in Mantis with the D3D code).  Just various little things like that.

But, like I already said, the new feature (or whatever it actually could be) is rather minor.  The few people who notice it will enjoy the novelty, for about 5 seconds.  Then it will fade away and everyone can go back to criticising all the various little issues that don't really matter.  There are really only two things to know:  1) DaBrain may need medical assistance, chronic laughter and rolling around on the floor like that can't be good for you, and 2) I'm an evil, evil bastard.

EDIT:

So uh, since the forum will be down this weekend for upgrading or whatever, what day can we realistically expect a new build if everything goes well with the testing ? Monday ?

I was thinking about that too.  There will be one more private test build.  If it goes well then I'll just move it to a public location (such as http://icculus.org/~taylor/fso/nightly) over the weekend.  Whether I remember to actually do that is another question, but it may not hurt to look there Saturday afternoon or so.

[neoterran]

Okay, that would really be neat because the weekend is about the only time i get to sit down at my home theatre set up and play.

Let's hope all goes well, I'm interested to see how env / env alpha works on OpenGL and if I can figure out what the "secret" addition is.