|
|
Post by Chainsaw on Aug 15, 2023 6:34:27 GMT -5
LOL, by the way... I love these statements to Dave from HAL9000 Peer.  Shall I play Vectrex this evening, or watch this great movie again? |
|
|
|
Post by hcmffm on Aug 15, 2023 8:36:45 GMT -5
LOL, by the way... I love these statements to Dave from HAL9000 Peer. Shall I play Vectrex this evening, or watch this great movie again? HAL9000 or GCE3000 - that's a tough call.  |
|
|
|
Post by Peer on Aug 15, 2023 10:26:49 GMT -5
LOL, by the way... I love these statements to Dave from HAL9000 Peer. Shall I play Vectrex this evening, or watch this great movie again? HAL9000 or GCE3000 - that's a tough call. "Let me put it this way, Mr. Amor Müller. The 9000 series is the most reliable computer ever made. No 9000 computer has ever made a mistake or distorted information. We are all, by any practical definition of the words, foolproof and incapable of error." |
|
|
|
Post by Peer on Aug 19, 2023 0:59:39 GMT -5
Despite Happy Bird having some random hit detection glitch, … Just wanted to let you all know that I am in contact with the programmer who is already investigating and has some ideas what could be the cause here. Might take a one or two weeks before we can continue with this, as time is a bit tight right now. |
|
|
|
Post by playvectrex on Aug 20, 2023 12:41:32 GMT -5
Despite Happy Bird having some random hit detection glitch, … Just wanted to let you all know that I am in contact with the programmer who is already investigating and has some ideas what could be the cause here. Might take a one or two weeks before we can continue with this, as time is a bit tight right now. They also popped onto VFU, and I replied with some info  |
|
|
|
Post by hcmffm on Aug 20, 2023 12:53:56 GMT -5
^ Thank you for passing on the info about the bug to the Happy Bird programmer, Peer. Would be good if this random bug could be fixed. They also popped onto VFU, and I replied with some info You (Brett) were some minutes faster than me. Discussion in VFU can be found here. |
|
|
|
Post by Peer on Aug 20, 2023 13:31:04 GMT -5
… Would be good if this random bug could be fixed. I am confident that this will be fixed. As I wrote above, this might take some more days. I am currently traveling and do not have the means to test, or take a look at the code and assist the programmer, so I have to ask for your patience. I will investigate once I have returned. Also please note that I am not on Facebook and cannot follow any discussions there. |
|
|
|
Post by Peer on Sept 7, 2023 1:38:49 GMT -5
We haved fixed a bug related to the collision detection of project "Happy Bird". However, we are not 100% sure if this is indeed the very same bug which was bescribed here in this thread as "random behavior". Would be cool if those who have reported that issue could play and test again. The new bugfixed version is available on the "Happy Bird" project page.
Many Cheers, Peer
|
|
|
|
Post by hcmffm on Sept 8, 2023 1:06:22 GMT -5
Great that you've found and fixed a bug, Peer and Simon. I'll ping Brett ( playvectrex) via e-mail and am sure that he'll happiliy play-test Happy Bird. |
|
|
|
Post by Peer on Sept 8, 2023 1:51:39 GMT -5
Great that you've found and fixed a bug, Peer and Simon. I'll ping Brett ( playvectrex ) via e-mail and am sure that he'll happiliy play-test Happy Bird. Thanks a lot, Helmut! I have already pinged Brett and sent him the new binary some days ago, but have not yet gotten any feedback. |
|
|
|
Post by playvectrex on Sept 28, 2023 22:49:57 GMT -5
Hi! Sorry for the delay on testing. I just gave it some time tonight and I can still pretty easily get false hit detection somewhat far away from the pipes. Usually I can also go through the pipe quite a bit before it registers as a hit. Pretty much all you need to do to find the issue is try to flap so that you are always just touching the bottom pipe when you flap up. Within the first 10 pipes you'll see some kind of inconsistency very easily. I'll make a video, but it's going to take a moment to capture and post.
|
|
|
|
Post by playvectrex on Sept 28, 2023 23:22:58 GMT -5
Timestamps (in description as well) 0:07 1:02 (in review, this is the only one that you can actually see briefly touches... but during game play it doesn't feel like it should have counted.) 1:56 3:28 It almost feels like: - screen is drawn, position is updated, then hit detection is done, instead of - updating position, drawing the screen, then doing hit detection. Here's screenshots of the timestamped failures... 1:02 is probably ok, but most of the time going a bit through isn't an issue... so it's not very cleanly defined either way. In my opinion it shouldn't be so tight that you can't sail over the sharp edges of the pipe... otherwise the game is not going to be fun. But it also shouldn't be so sloppy you can go through the pipe a lot.     |
|
|
|
Post by Peer on Sept 29, 2023 2:00:58 GMT -5
Hi! Sorry for the delay on testing. I just gave it some time tonight and I can still pretty easily get false hit detection somewhat far away from the pipes. Usually I can also go through the pipe quite a bit before it registers as a hit. Pretty much all you need to do to find the issue is try to flap so that you are always just touching the bottom pipe when you flap up. Within the first 10 pipes you'll see some kind of inconsistency very easily. I'll make a video, but it's going to take a moment to capture and post. No problem at all. Thanks a lot for testing (again)!!! That was very helpful. I do not have the code in front of me, but I have a new guess of what could be happening here. It is not the move-draw-check order, I already verified that last time. My new guess is that this is related to using different scale factors for drawing the bird and the pillars. In theory, SF == 2 * (0.5 SF), and the scaling should behave in a linear fashion. But this is usually not the case on a real console. Worse, it is slightly different on each individual console (the infamous calibration issue). So the collision detection might behave accurately in an emulator (and on the console we used for testing in class), but differently on other different consoles  I will check and see if (hopefully) something can be done about this.
Cheers, Peer
|
|
|
|
Post by playvectrex on Sept 29, 2023 11:01:03 GMT -5
My new guess is that this is related to using different scale factors for drawing the bird and the pillars. In theory, SF == 2 * (0.5 SF), and the scaling should behave in a linear fashion. But this is usually not the case on a real console.
Cheers, Peer
Oh for sure that could be the issue! When my son and I were working on Track Attack for our Vectrex Homeschool project... we got to a point in the development where hit detection of the shot bouncing off of walls was super flakey. Sometimes it would detect a hit, sometimes not. I always suspected it was related to different drawing scales for walls and shots, but it effectively killed the momentum on game development. We still had a lot of fun though working on it together. |
|
|
|
Post by Peer on Oct 3, 2023 9:01:13 GMT -5
Just to keep you informed: I am in contact with the programmer of "Happy Bird", and we think that we found an issue (likely "the" issue), which is indeed related to scale factors. More on this soon.
|
|
I will check and see if (hopefully) something can be done about this.