105 replies to this topic

[toxie]

after quiet some searching in the mighty internet, i hopefully found the correct way to do it from within directx7 (which is what VP is still using), but unfortunately (from what i also read) the support from drivers for that specific directx7 feature seems to be 0. at least my nvidia board with newest drivers doesn't seem to do anything at all.

[BarryS]

[quote name='destruk' post='60671' date='Jun 7 2010, 09:45 PM']I've compiled VP with vsnet 2003, vsnet 2005, and vsnet 2008. Microsoft just released a free and FULL version of vsnet 2010 but I haven't tried it, so I don't know if it will work or not. It is a trial version, so it will work for 30 days - might as well look at it.

http://www.microsoft...;displaylang=en

nevermind... I think I got it...

Edited by BarryS, 11 June 2010 - 10:13 PM.

[toxie]

on a sidenote: i just tried visual c++ 2010 express, and this doesn't work due to the missing ATL support in all of the express editions.

[H4CK3R]

on a sidenote: i just tried visual c++ 2010 express, and this doesn't work due to the missing ATL support in all of the express editions.

ATI sucks, always has... stick to nvidia.

We should not be limiting what we can or can not do based on one video card company.

Reminds me of why EQ DIED ( Everquest ) They were running great for 10 years. But other games got better graphics and better drivers, making WoW.

Sony decided that they didn't want to upgrade the system specs, ( and force there customers to upgrade their CP's ) because they were afraid of loosing there user base. They lost more then that tring to stick with OLD tech. Everyone switched over to WOW. This is just an example.

Things evolve. So should we if it gives us better physics, graphics, and playability. Table designers are coming out with some great tables, but because of some people system specs, and getting VP on a single core, they turn out to be choppy, and unplayable.

If we are to keep things to a single thread, to please the 4/3 users, There should be a multi thread for FS table users and cab owners. Desruct, I hope maybe you could check out the HP forums, and maybe give some of us a clue, in what we can do to improve performance. Most of us are tring to get everything 100% perfect,

Evolve, or be stuck in the PAST!

Anyone still running there PC on win95? ME?

H4CK3R

[TheMcD]

Umm... wasn't that just about ATL support? Not ATI?

The McD

[H4CK3R]

Umm... wasn't that just about ATL support? Not ATI?

The McD

Ekk sorry, miss read that... NVM about the ATI thing, tho it still sucks =p

hehe

[destruk]

I told you 3 times already H4ck3r - Multithread or no multithread, it's not the cause of the slowdown. What you want is MultiCore, and I don't know how to change vp to make use of multicore.
VP currently works with ATI and NVIDIA, so you're a bit outdated there too - it doesn't work very well with onboard graphics or notepad chipsets from what I've seen. You're already using a program to run VP on one core, VPM on another, and HyperPin or UVP on a third - I'd like to be able to have VP run on all 8 cores I have in my machine, or any combination of processing configs to bring out the highest level of performance possible. Right now it sticks to the same core, while the other 7 are either idle or working on background processes.

Again, Multithreading was simply used for the keyboard and joystick routine, which DirectInput can handle just as well. You're not talking about the 'holy grail of game speedups' by changing it back. Toxie has done a bunch of optimizations to vp to make the current singlethreaded code base much faster than it was before and the compile option uses /MT - check that out.

[Shockman]

Please don't post or distribute any new exe's.

PM destruk and discuss it with him. We have a developers' forum all set up for sharing progress.

We just don't need a ton of different versions of VP going around until they've been tested and approved.

Thank you.

I don't understand this. VP is open now, and we do need improved builds. What does this mean anyway?, I just can't figure this out. No new builds? That ain't right. No new builds unless you approve of them? Why? I think it about time someone moves VP forward instead of just talking about it. We might not need a ton, but how many improved builds equal a ton? How many do we have? in number or in weight?
This just ain't right. Is this site going to black list improvements? Is the plan to have a team sponsored by the site and only their build will be available? Now that it is open source, this is the worse decision you could possibly make for this community.

[Noah Fentz]

It makes perfect sense to not want dozens of builds floating around that we may or may not be able to provide support for, Phil.

Think about.

If you want to submit a build, contact destruk. We cannot provide support for any builds we are not familiar with. It's just not possible.

[PainkillerAlice]

I think what Noah is getting at is mostly to make sure that we are on the same page with development and as a whole improve VP for all. If so many people try to release VP with their own seperate improvements, how are the rest of us non-programmer pinball fanatics supposed to know which flavor of VP is the best? As far as "approval" goes, it is only just to keeps things organized, not to restrict anyone from wanting to help.
Also remember that many of the programmers are also making their own tables as well so their time is somewhat limited to devote to improving VP. I know there are improvements that can be made on VP, but I for one would rather the programmers take their time on this as to make sure it comes out right. Pinball is worth the wait!

Edited by Practicedummy, 14 June 2010 - 06:02 PM.

[H4CK3R]

Sorry Destruk, I got everything on my cab running 99.5 stutter free, by using the taskapp program. Even flipper lag is all but gone. We are just tring to figure out the best method of getting everything set up perfect, and seeing how you are one of the many talented people that help design this, What are your suggestion? What do you use, or don't use in your cab?

System specs, in FS, UVP, killed non necessary processes? ect ect. Just a base line system spec for a 2 monitor set up, a 3 monitor set up. Hints of what to do to improve system speed. Anything to help us get the most that we can out of our systems.

I Posted up for everyone to see what I am running, as to maybe help others get the best they can with there system. Trial and error. Would be nice to hear what an Expert has to say =)

Thanks again for all your hard work.

H4CK3R

[Shockman]

What I don't understand is why this site thinks it needs to support every build. VP is open source. That means anyone can make a build and support it themselves. If they can not offer it here, then here is not what I thought it was. If I made a build the LAST thing I would do is to submit it to someone else for approval, least of all, destruk.
Are you saying that this site is not going to allow better versions of VP to be offered unless you can assume total control over it? Contacting destruk reads that way. What would destruk have to do with a build he did not build?

How about a link at least to the new build?

[destruk]

H4CK3R - I'm not an expert. I use a custom coded frontend loader to switch between the Ultrapin standard frontend, the Ultrapin custom coded fronted, and HyperPin for Future Pinball. The frontend loader handles display profile switching, the Ultrapin frontends handle the tilt, gama correction, and game settings, and you know what HyperPin does already.
When HyperPin exits, it goes back to the main frontend loader so I can switch to another frontend without rebooting - this is done with the hyperpin settings file.
I use a 7600gs video card with xp embedded to get the most out of the system with the standard ushock board. I don't use FS tables, and I don't use UVP. My hardware in this cab is quite weak by today's standards (Pentium D 3Ghz with 2gb ram) so Future Pinball play is only acceptable to me with all settings on "Low".
For VP the only thing I disable is the ball roll decals - I never cared for them since they are a faked effect - even a ball that is sitting cradled in a flipper will still show the ball decal moving.
All my tables are spanned. I did make a couple changes to the sample rate for LOTR and the other stern whitestar games to get emulated sound working with minimal to no scratchiness.


You don't like me Phil, and that's ok. I'm basically only sticking around to make sure users have permissions to check in changes to the repository, and give simple feedback on builds to try to make sure nothing we currently have 'breaks'. I simply try a few of the changes that have been submitted for VP and if it breaks a bunch of tables I let the submitter know. If the tables still work, then the change goes through. I'm sure everyone else now helping with the core VP Development is doing the same thing that I'm doing. If you think I am excercising an iron clad gauntlet blocking what they want to do, you're mistaken - every change submitted so far has gone through, and any bugs that were introduced are being taken care of to provide a better program.

I'm easy to replace in that regard - you could download the source code and set up your own repository and instantly you'd have as much power as you think I do. You could even have it hosted on your own pc and thereby have full control over the actual SVN server, which is more power than I have.

If there are numerous private builds, then we should set up a support page with an email address/contact info for each person's specific build. I think supporting the released builds of VP8 and VP9 are already confusing to most of our user base here. I'm not against the idea of having hundreds of individual versions of visual pinball, but I don't see hundreds of people making their own personal builds either.
At this point, there is no stopping you from offering your own exe. As for private builds or test betas or whatever, the code is on sourceforge and it's current - download, compile, there you have a copy. When anything of true interest has been submitted, I think we can release a full fledged update - fixing one or two minor bugs and instantly releasing every week isn't very efficient IMO.

[Shockman]

Yeah, I don't like you. You can speak the truth after all.
I liked you before you started ragging on my contribution to VP though. So I liked you a great deal a few days ago. The downswing is not instant. It is not worse than a ball through the flipper, and should not be replaced with a strobe that says BUG!. And to say that it makes it nearly impossible to fix properly, without answering why, is the reason, not that you forgot who did it.

It however has nothing to do with wondering about the new builds.

[shagendo]

On topic :

I agree that there should be someone to overview and submit tested code to the repository.

If someone would manage to make an automated test that could be used to detect 'breaking' bugs in the program (which is quite unlikely), submission could be automated to a point. Untill that happens, we will need someone to manage the repository.

Unless someone else that is knowledable in the vpin field feels like taking up this task I think Destruk is the man for the job.
(I for one have some time to spend on programming VP but not enough to manage a repo)

And anyone is free to fork the code and take it anywhere they want to...
Open source rules!

Sander

[Shockman]

Oh, I agree as well. Improvement were announced and offered. I wondered what happened to them. As long as all improvements make it in, it's OK. I don't agree that I should not be able to offer up my own build without it going through someone else. And I can't think of an improvement in VP9 that did not break something else. The flippers don't have balls through the flippers, but they don't even show up if they have black rubber. The images could be higher res, but you can not trust the transparency anymore. The plunger is better, but nudging is broken. The ball is more a realistic size, but the stacking order is screwed up. I probably left out some improvements, but I did not list all the new problems either.

I would download a build that had any one of those things fixed. I would offer a build that had any one of those things fixed, if I had one.

[destruk]

For the Black Rubber issue - as well as any other pure 0/0/0 black issues - if you change it to 7/7/7 or 14/14/14 then it shows up fine, just like VP8. 0/0/0 is allowed to make the flipper disappear by design so you could use banana flippers, or invisible flippers in the table like was used on some Twilight Zone versions and The Who's Tommy. If you use earlier vp9 versions then they will automatically change 0/0/0 to 7/7/7 for you on table load, meaning you can't have a functional/collidable flipper that is invisible with those earlier builds.
Nobody is preventing you from offering up your own build on your own without it going through anyone but yourself.

btw Randy's VP8 supported images up to 4096x4096, so some things you mention there aren't vp9 specific 'improvements'.

Edited by destruk, 15 June 2010 - 12:54 AM.

[Shockman]

I know more about the VP editor than you seem to think I do.
I know all about 0,0,0 and the new 7,7,7 but I have to go higher on my card with 9.7. Actually till it looks grey. No biggie though, that' what I do. A weird thing about it. 14,14,14 works, and for each increment down, more of the tip is invisible, to 8,8,8 where it is just the round part of the tip. 7,7,7 and the rubber all around is invisible. So 13,13,13 leaves just the tip of the round part (one unit) invisible.
I know about 4096x4096, but until VP9 it was a striped blur on my card.

Unless I can decompile and recompile VP in Borland C++ for DOS, I doubt I could get anything done with it.

[destruk]

I'll ask Maddes to see if he can work on compiling with different programs like he did for PinMAME - he's really really good at that.

[Shockman]

I was just joking about using Borland C++ for DOS