1.16 SP Campaign: Trinity
Moderator: Forum Moderators
Re: 1.9/1.10 SP Campaign: Trinity
Ah, I knew about those things, but completely failed to put the pieces together. Thanks!
Depending on the file size, I may or may not want to avoid BLIT.
(I didn't really intend for it to be walkable. If it is, that is a blunder. The hex at "1" is walkable, but the moving image isn't.)
Depending on the file size, I may or may not want to avoid BLIT.
(I didn't really intend for it to be walkable. If it is, that is a blunder. The hex at "1" is walkable, but the moving image isn't.)
BfW 1.12 supported, but active development only for BfW 1.13/1.14: Bad Moon Rising | Trinity | Archaic Era |
| Abandoned: Tales of the Setting Sun
GitHub link for these projects
| Abandoned: Tales of the Setting Sun
GitHub link for these projects
- Alarantalara
- Art Contributor
- Posts: 788
- Joined: April 23rd, 2010, 8:17 pm
- Location: Canada
Re: 1.9/1.10 SP Campaign: Trinity
I was guessing based on the 1 hex since I didn't really want to look at the map itself (Wesnoth is not installed where I am now). I guess you can probably use the existing terrain transitions then.doofus-01 wrote:(I didn't really intend for it to be walkable. If it is, that is a blunder. The hex at "1" is walkable, but the moving image isn't.)
You'll have to move the 1 to somewhere in the hole though, since it will be a recognized tile and drawn on.
Re: 1.9/1.10 SP Campaign: Trinity
This does not seem to work, unfortunately. But I think it's a bug, not a mistaken idea. I'll post a bug report.Alarantalara wrote:Step 3: use image.png~CROP(offset,0,hole width,image height) for varying offsets in pixels from 0 to just before the start of the repeated section
EDIT: TechSupport Forum post: http://forum.wesnoth.org/viewtopic.php?f=4&t=33377
BfW 1.12 supported, but active development only for BfW 1.13/1.14: Bad Moon Rising | Trinity | Archaic Era |
| Abandoned: Tales of the Setting Sun
GitHub link for these projects
| Abandoned: Tales of the Setting Sun
GitHub link for these projects
Re: 1.9/1.10 SP Campaign: Trinity
Trinity 1.0.4 just went to the BfW 1.9 server. There are some improvements, but the last scenarios of the alternate branch still are not finished. Saves from earlier versions should still work.
BfW 1.12 supported, but active development only for BfW 1.13/1.14: Bad Moon Rising | Trinity | Archaic Era |
| Abandoned: Tales of the Setting Sun
GitHub link for these projects
| Abandoned: Tales of the Setting Sun
GitHub link for these projects
Re: 1.9/1.10 SP Campaign: Trinity
1.0.5 just went to the BfW 1.9 server. The alternative ending still needs work, but it is playable and beatable. Hopefully it is at least somewhat amusing.
BfW 1.12 supported, but active development only for BfW 1.13/1.14: Bad Moon Rising | Trinity | Archaic Era |
| Abandoned: Tales of the Setting Sun
GitHub link for these projects
| Abandoned: Tales of the Setting Sun
GitHub link for these projects
Re: 1.9/1.10 SP Campaign: Trinity
Alarantalara was kind enough to look into this and was able to fix it with a patch: http://gna.org/patch/?2664doofus-01 wrote:This does not seem to work, unfortunately. But I think it's a bug, not a mistaken idea. I'll post a bug report.Alarantalara wrote:Step 3: use image.png~CROP(offset,0,hole width,image height) for varying offsets in pixels from 0 to just before the start of the repeated section
EDIT: TechSupport Forum post: http://forum.wesnoth.org/viewtopic.php?f=4&t=33377
It does indeed work now, though it is very laggy. I suspect this idea is not really doable with the WML tools at hand. The screen scrolls, so scrolling large graphics is clearly possible in Wesnoth. But image path functions are too slow.
If anyone wants to see what the hell I'm talking about, download Trinity and Archaic Era, replace Trinity/utils/terrain.cfg with the file attached below, and open Trinity/maps/5b-?.map (I forgot its name, it's the first or second one with "5b" in the name) with the map editor.
EDIT:After you have applied Alarantalara's patch, of course
- Attachments
-
- terrain.cfg
- (12.3 KiB) Downloaded 712 times
BfW 1.12 supported, but active development only for BfW 1.13/1.14: Bad Moon Rising | Trinity | Archaic Era |
| Abandoned: Tales of the Setting Sun
GitHub link for these projects
| Abandoned: Tales of the Setting Sun
GitHub link for these projects
- Alarantalara
- Art Contributor
- Posts: 788
- Joined: April 23rd, 2010, 8:17 pm
- Location: Canada
Re: 1.9/1.10 SP Campaign: Trinity
I took a look to see what the final result was. It became reasonably smooth for me after about 10 times through the loop, probably due to caching. I suspect the small amount of stuttering that remains is related to the two short period frames at the start and end of the sequence since it occurs once per loop. Maybe if you set the change in offset to 4 pixels instead of 5 it would look better since 4 divides the width of a hex evenly, so you wouldn't have to use unusual frame lengths at the start and end of the image set.
Edit: I also note that the width of the cropped area has no integer relationship to the number of hexes in the image. This likely also contributes to the problem. The visible area is only 756 pixels wide by my count, so setting the crop width to that and using center to place it in the image should also help. You could also set the offset to 6 pixels to reduce the number of images that need to be cached, which may help performance as well.
There's a lot of copying of the image occurring in the process right now, which really limits the size of the displayed image. Perhaps I'll look into making the center key have time parameters like image does. It might save one copy, which would help.
Edit: I also note that the width of the cropped area has no integer relationship to the number of hexes in the image. This likely also contributes to the problem. The visible area is only 756 pixels wide by my count, so setting the crop width to that and using center to place it in the image should also help. You could also set the offset to 6 pixels to reduce the number of images that need to be cached, which may help performance as well.
There's a lot of copying of the image occurring in the process right now, which really limits the size of the displayed image. Perhaps I'll look into making the center key have time parameters like image does. It might save one copy, which would help.
Re: 1.9/1.10 SP Campaign: Trinity
I'd tried various offsets and hadn't noticed a difference. It didn't occur to me that crop area had to have an integer relationship.
I think I almost know what you mean, but I'm not sure. Is it possible to make those keys progressive parameters, like in unit animations?Alarantalara wrote:There's a lot of copying of the image occurring in the process right now, which really limits the size of the displayed image. Perhaps I'll look into making the center key have time parameters like image does. It might save one copy, which would help.
BfW 1.12 supported, but active development only for BfW 1.13/1.14: Bad Moon Rising | Trinity | Archaic Era |
| Abandoned: Tales of the Setting Sun
GitHub link for these projects
| Abandoned: Tales of the Setting Sun
GitHub link for these projects
- Alarantalara
- Art Contributor
- Posts: 788
- Joined: April 23rd, 2010, 8:17 pm
- Location: Canada
Re: 1.9/1.10 SP Campaign: Trinity
The crop area doesn't need such a relationship, but because of the amount of data in a picture is so large, making the area as small as possible improves performance. If my measurement of the size is correct, a crop width of 907 over 756 requires processing some 300+K of extra data per frame, multiplied by the number of copies that the code needs to do of the whole area (I haven't counted the exact number). It would probably have been better to say that the crop area was larger than the displayed image, but I was still thinking about the offsets at the same time.
Giving the offsets an integer relationship to the width of the hexes is more important since it means that when you reset to the beginning the amount moved on the reset is the same amount as for all the other frames. So any proper divisor of 72 for an offset multiplier should handle the end better. I just suggested 4 or 6 because they were close to 5.
Edit: Add a somewhat reworked terrain.cfg to show what I'm talking about. I also extended the image to fill the rest of the chasm at the end. You may also want to crop the original image vertically to reduce your download size. If it's still too slow, then maybe a smaller hole would achieve the same effect. Because the images are cached, you may be able to use two small copies if they aren't directly adjacent and get an overall faster result.
Giving the offsets an integer relationship to the width of the hexes is more important since it means that when you reset to the beginning the amount moved on the reset is the same amount as for all the other frames. So any proper divisor of 72 for an offset multiplier should handle the end better. I just suggested 4 or 6 because they were close to 5.
That would be the idea. The problem is that all the keys that currently have progressive parameters only have one value to start with, but center has two values, making the comma have two meanings. The simplest solution would be to replace center with center_x and center_y, which breaks all mainline terrain macros and likely a lot of user content as well. I'll probably have to talk this over in IRC, once I figure out who I should be talking to.doofus-01 wrote:I think I almost know what you mean, but I'm not sure. Is it possible to make those keys progressive parameters, like in unit animations?
Edit: Add a somewhat reworked terrain.cfg to show what I'm talking about. I also extended the image to fill the rest of the chasm at the end. You may also want to crop the original image vertically to reduce your download size. If it's still too slow, then maybe a smaller hole would achieve the same effect. Because the images are cached, you may be able to use two small copies if they aren't directly adjacent and get an overall faster result.
- Attachments
-
- terrain.cfg
- (11.68 KiB) Downloaded 616 times
Re: 1.9/1.10 SP Campaign: Trinity
OK, I see how the image file size could matter. It would be nice if it wasn't so sensitive to that, but that's life. I'll look at your terrain.cfg modification.
I do like the progressive center_x, etc. idea. If I understand correctly, it would allow things like clouds to move across the screen, like a sprite. That, combined with planes/layers that scroll or pan at different speeds than the regular map[*], would be a significant visual breakthrough for Wesnoth. In my opinion at least.
Thanks.
[*] I realize that is something different, in possibly completely different section of code.
I do like the progressive center_x, etc. idea. If I understand correctly, it would allow things like clouds to move across the screen, like a sprite. That, combined with planes/layers that scroll or pan at different speeds than the regular map[*], would be a significant visual breakthrough for Wesnoth. In my opinion at least.
Thanks.
[*] I realize that is something different, in possibly completely different section of code.
BfW 1.12 supported, but active development only for BfW 1.13/1.14: Bad Moon Rising | Trinity | Archaic Era |
| Abandoned: Tales of the Setting Sun
GitHub link for these projects
| Abandoned: Tales of the Setting Sun
GitHub link for these projects
- averyimaginativename
- Posts: 245
- Joined: August 21st, 2010, 12:40 pm
- Location: /dev/null
Re: 1.9/1.10 SP Campaign: Trinity
Just a typo - if the prince dies in the first scenario, the thug says he'll send "knews" to Weldyn rather than "news".
UMC Story Images web gallery
On an indefinite Wesbreak for health reasons - please only try to get my attention for UMC story images website issues.
On an indefinite Wesbreak for health reasons - please only try to get my attention for UMC story images website issues.
- averyimaginativename
- Posts: 245
- Joined: August 21st, 2010, 12:40 pm
- Location: /dev/null
Re: 1.9/1.10 SP Campaign: Trinity
I'm not sure how unfinished you think it is, so do you want a list of bugs for the unfinished branch or is it enough of a work in progress that bug reports aren't useful?
UMC Story Images web gallery
On an indefinite Wesbreak for health reasons - please only try to get my attention for UMC story images website issues.
On an indefinite Wesbreak for health reasons - please only try to get my attention for UMC story images website issues.
Re: 1.9/1.10 SP Campaign: Trinity
Development has slowed to a near-halt,but it isn't completely dead. If you find bugs, it would be helpful to report them. As far as I know, the campaign doesn't have any show-stoppers, but I haven't been able to test it much in the last few BfW 1.9.x releases.
Thanks.
Thanks.
BfW 1.12 supported, but active development only for BfW 1.13/1.14: Bad Moon Rising | Trinity | Archaic Era |
| Abandoned: Tales of the Setting Sun
GitHub link for these projects
| Abandoned: Tales of the Setting Sun
GitHub link for these projects
- averyimaginativename
- Posts: 245
- Joined: August 21st, 2010, 12:40 pm
- Location: /dev/null
Re: 1.9/1.10 SP Campaign: Trinity
There's quite a few, but only three game breakers.
Illusions isn't beatable because of the event at line 1,040. There are no Trinity Whirl's on the map. Plenty of Trinity Flames, but no whirls. I used debug to beat it and that corrupted the replay, but I have a save at the beginning of the scenario if that's useful.
The Queen appears from nowhere in the epilogue and has code in Illusions. You've already marked it as "this doesn't fire, who knows why?" so I only mention it in case it's related to the above.
Another is in the finished path in Onslaught. If the Bomb units kill one of the heroes, you're not defeated, and you can carry on playing as normal, which obviously breaks the rest of the campaign.
Storming Weldyn - the bomb bug exists here too. If either of your leaders are in the blast radius, you're not defeated.
Minor bugs:
All the way through Khthon elves have a glitch when they're recruited. I'm not sure how to describe it...they sort of get lighter, pause for a bit, and then get darker. In 1.9 a few other units do something similar - drakes and bats in mainline - but with their animation pausing instead of colours changing in so possibly an engine issue.
Vector doesn't work in defence. It will damage, but not change the defender to a khthon. Khthon must be ridiculously hard to balance...is this by design?
3_H05_Encounter. Haldrad can recruit ghosts from now on. They're very, very useful, but they don't make sense. It appears to be a left over from a commented out section of with the removed/missing character Sadre.
3_K04_Storm. "Move Echidna past the southern region of the outer moat to get Ullien-Del to play his part." is ambiguous, and implies 13-31,35-40. "Near Daenyr's keep" would be clearer.
F1b_Wesmere - Objectives state a lose condition of "Death of Prisoner Wagon", however, you lose if they're attacked and lived, even if the attack comes from elves who don't have vector. Also, the AI is weird. It's easy to see what's going on if you look at it, but harder to put it into words - units flee as Nemesis approaches, but newly recruited units ignore this, advance on the first turn they can move, then flee the turn after.
F4a_Hera - not a bug, but it's ridiculously easy with nemesis + horseman spam.
F5b Illusions - there is a die event for the wagons, but the wagons are now in your recall list, and it's pointless to recall them. Their pictures also distort the recall list due to their size.
There is a depreciated tag warning in the Finale when Hera first speaks to the player, then leaves for the shroud to fight Nemesis. It's still working for now though.
Illusions isn't beatable because of the event at line 1,040. There are no Trinity Whirl's on the map. Plenty of Trinity Flames, but no whirls. I used debug to beat it and that corrupted the replay, but I have a save at the beginning of the scenario if that's useful.
The Queen appears from nowhere in the epilogue and has code in Illusions. You've already marked it as "this doesn't fire, who knows why?" so I only mention it in case it's related to the above.
Another is in the finished path in Onslaught. If the Bomb units kill one of the heroes, you're not defeated, and you can carry on playing as normal, which obviously breaks the rest of the campaign.
Storming Weldyn - the bomb bug exists here too. If either of your leaders are in the blast radius, you're not defeated.
Minor bugs:
All the way through Khthon elves have a glitch when they're recruited. I'm not sure how to describe it...they sort of get lighter, pause for a bit, and then get darker. In 1.9 a few other units do something similar - drakes and bats in mainline - but with their animation pausing instead of colours changing in so possibly an engine issue.
Vector doesn't work in defence. It will damage, but not change the defender to a khthon. Khthon must be ridiculously hard to balance...is this by design?
3_H05_Encounter. Haldrad can recruit ghosts from now on. They're very, very useful, but they don't make sense. It appears to be a left over from a commented out section of with the removed/missing character Sadre.
3_K04_Storm. "Move Echidna past the southern region of the outer moat to get Ullien-Del to play his part." is ambiguous, and implies 13-31,35-40. "Near Daenyr's keep" would be clearer.
F1b_Wesmere - Objectives state a lose condition of "Death of Prisoner Wagon", however, you lose if they're attacked and lived, even if the attack comes from elves who don't have vector. Also, the AI is weird. It's easy to see what's going on if you look at it, but harder to put it into words - units flee as Nemesis approaches, but newly recruited units ignore this, advance on the first turn they can move, then flee the turn after.
F4a_Hera - not a bug, but it's ridiculously easy with nemesis + horseman spam.
F5b Illusions - there is a die event for the wagons, but the wagons are now in your recall list, and it's pointless to recall them. Their pictures also distort the recall list due to their size.
There is a depreciated tag warning in the Finale when Hera first speaks to the player, then leaves for the shroud to fight Nemesis. It's still working for now though.
UMC Story Images web gallery
On an indefinite Wesbreak for health reasons - please only try to get my attention for UMC story images website issues.
On an indefinite Wesbreak for health reasons - please only try to get my attention for UMC story images website issues.
Re: 1.9/1.10 SP Campaign: Trinity
Well, that's embarrassing. I'll fix that this weekend. If you have the start-of-scenario save, that would be a big help. Thanks.averyimaginativename wrote:Illusions isn't beatable because of the event at line 1,040. There are no Trinity Whirl's on the map. Plenty of Trinity Flames, but no whirls. I used debug to beat it and that corrupted the replay, but I have a save at the beginning of the scenario if that's useful.
I almost remember that, it might have something to do with debug. Not sure.averyimaginativename wrote:The Queen appears from nowhere in the epilogue and has code in Illusions. You've already marked it as "this doesn't fire, who knows why?" so I only mention it in case it's related to the above.
Easy enough to fix. Thanks.averyimaginativename wrote:Another is in the finished path in Onslaught. If the Bomb units kill one of the heroes, you're not defeated, and you can carry on playing as normal, which obviously breaks the rest of the campaign.
Storming Weldyn - the bomb bug exists here too. If either of your leaders are in the blast radius, you're not defeated.
I think it has something to do with recruiting animations, or else it is a bug. I never quite had the energy to tackle it, but it may be fixable.averyimaginativename wrote:All the way through Khthon elves have a glitch when they're recruited. I'm not sure how to describe it...they sort of get lighter, pause for a bit, and then get darker. In 1.9 a few other units do something similar - drakes and bats in mainline - but with their animation pausing instead of colours changing in so possibly an engine issue.
Yeah, that's by design, clunky as it is. As you said, they are very difficult to balance.averyimaginativename wrote:Vector doesn't work in defence. It will damage, but not change the defender to a khthon. Khthon must be ridiculously hard to balance...is this by design?
The default behaviour of the AI was silly in that scenario, so I tried to make it flee from Nemesis, yet go after everyone else. It's not perfect, and I think something may be one step out of sync, but that is what is happening.averyimaginativename wrote:F1b_Wesmere - Objectives state a lose condition of "Death of Prisoner Wagon", however, you lose if they're attacked and lived, even if the attack comes from elves who don't have vector. Also, the AI is weird. It's easy to see what's going on if you look at it, but harder to put it into words - units flee as Nemesis approaches, but newly recruited units ignore this, advance on the first turn they can move, then flee the turn after.
Yeah, I think that's what happened.averyimaginativename wrote:3_H05_Encounter. Haldrad can recruit ghosts from now on. They're very, very useful, but they don't make sense. It appears to be a left over from a commented out section of with the removed/missing character Sadre.
Anyhow, thanks for the list, I'll try to get these fixed soon.
BfW 1.12 supported, but active development only for BfW 1.13/1.14: Bad Moon Rising | Trinity | Archaic Era |
| Abandoned: Tales of the Setting Sun
GitHub link for these projects
| Abandoned: Tales of the Setting Sun
GitHub link for these projects