Blizzard DotA - Page 7
| Forum Index > General Games |
|
Bosu
United States3247 Posts
| ||
|
Pewt
Canada201 Posts
On October 29 2010 23:52 Judicator wrote: While netcode could be an issue on bnet, what are you specifically referring to with "the shitty trigger limitations of the map editor (hello triggered heals)"? Obviously the editor has limitations, but in my experience the things that people claim as "limitations" are usually just those people not knowing how to do them.Nobody likes the War3 DotA technical limitations, if anything, that's the one thing holding back DotA. The shitty netcode of B.net, the shitty trigger limitations of the map editor (hello triggered heals) along with the overall pretty crummy interface. Everyone will jump from War3 DotA to Valve DotA the minute it gets released, except some of the Asians, they'll most likely look for ways to pirate the shit. (Some obvious ones, such as keyboard/mouse detection, aside) | ||
|
Judicator
United States7270 Posts
On November 09 2010 05:14 Pewt wrote: While netcode could be an issue on bnet, what are you specifically referring to with "the shitty trigger limitations of the map editor (hello triggered heals)"? Obviously the editor has limitations, but in my experience the things that people claim as "limitations" are usually just those people not knowing how to do them. (Some obvious ones, such as keyboard/mouse detection, aside) Let's use the example I gave, Triggered Heals. Instead of true invulnerability (which in itself creates issues), heroes have triggered heals to simulate invulnerability. Meaning instead of taking no damage from a X damaging event, you heal for X damage and then the X damage. Net result is 0 damage, so it's all good right? No. X can't be greater than your total HP and X can't put you over your max HP. That means you can't have pseudo-invulnerability at full HP, and you can't survive a skill that deals more damage than you have HP. Another side effect is that if X equals your current HP exactly, you will die and the triggered heals will not occur. Then there are fucked up interactions depending on the source of the X damage. The triggered heals bullshit mostly comes out of the hardcoded concepts of invulnerability. The limitations comes from the hardcoded concepts already in the War3 engine and having to code around those "rules". That's just triggered heals, there are MASSIVE issues with the RNG, specifically Blizzard's take on it. Hi CM's old ultimate, any bash-based or critical strike-based skill. It doesn't matter much for stuff like ORPGs and what not, but when it comes to games like DotA, you have to break open the book and figure out how abilities interact. A lot of top players still don't understand how skills interact. | ||
|
ZappaSC
Denmark215 Posts
On November 09 2010 06:56 Judicator wrote: Let's use the example I gave, Triggered Heals. Instead of true invulnerability (which in itself creates issues), heroes have triggered heals to simulate invulnerability. Meaning instead of taking no damage from a X damaging event, you heal for X damage and then the X damage. Net result is 0 damage, so it's all good right? No. X can't be greater than your total HP and X can't put you over your max HP. That means you can't have pseudo-invulnerability at full HP, and you can't survive a skill that deals more damage than you have HP. Another side effect is that if X equals your current HP exactly, you will die and the triggered heals will not occur. Then there are fucked up interactions depending on the source of the X damage. The triggered heals bullshit mostly comes out of the hardcoded concepts of invulnerability. The limitations comes from the hardcoded concepts already in the War3 engine and having to code around those "rules". That's just triggered heals, there are MASSIVE issues with the RNG, specifically Blizzard's take on it. Hi CM's old ultimate, any bash-based or critical strike-based skill. It doesn't matter much for stuff like ORPGs and what not, but when it comes to games like DotA, you have to break open the book and figure out how abilities interact. A lot of top players still don't understand how skills interact. How about infinite armor? | ||
|
Judicator
United States7270 Posts
That's hardly a problem, the only issue it creates is with Kunkka's Tidebringer and the only time that ever comes up is Guardian Angel. Cleave itself is hardcoded to ignore armor. | ||
|
Pewt
Canada201 Posts
Edit: And if I recall correctly the RNG is weighted, which is a good thing (to avoid extremely long chains of criticals etc). | ||
|
zbedlam
Australia549 Posts
On November 09 2010 04:58 Bosu wrote: I think these customs are more of a tech demo then anything. There are a lot of popular customs currently that are just not very deep or polished. These will show how amazing customs can be if people put in extra work. The sc2 map editor has a lot of potential, pity its severely hampered by sc2's custom game system. There is no way anyone could play a fps or fighting game or anything innovative like that with 200ms+. Most people wouldn't even want to play dota 2.0 with that kind of ping. If blizz replaced bnet 0.2's CG system with wc3's then I could see sc2 custom game scene going places other than nexus wars and tower defense. | ||
|
Pewt
Canada201 Posts
| ||
|
Judicator
United States7270 Posts
On November 09 2010 13:54 Pewt wrote: Easy to get around. Use a dummy ability based on the Amulet of Vitality (or whatever it's called) item's ability to temporarily increase the hero's life. Yes, it's annoying, but easy to do. This was solved years ago. (You can also just code most/all damage from scratch, which has become more popular in the last year or two. It's actually way less of a hassle than it sounds since good/interesting abilities tend to be triggered regardless) Edit: And if I recall correctly the RNG is weighted, which is a good thing (to avoid extremely long chains of criticals etc). You're kidding right? Don't comment if you don't grasp basic EHP concepts. If you increase the HP based on incoming damage, you open it up to shennigans with armor increasing and decreasing during the duration (which like it or not does happen). RNG weighted is good? So that's why they recoded Ogre Magi's ultimate...yeah. | ||
|
JoFu
United States25 Posts
| ||
|
Goshawk.
United Kingdom5338 Posts
| ||
|
Pewt
Canada201 Posts
On November 10 2010 03:53 Judicator wrote: Not an issue in practice, largely because unless the armour modification is tied to the damage itself it won't occur while your code is executing (that is, between the pre-damage trigger and the post-damage timer). Have you actually implemented these systems or just heard about them?You're kidding right? Don't comment if you don't grasp basic EHP concepts. If you increase the HP based on incoming damage, you open it up to shennigans with armor increasing and decreasing during the duration (which like it or not does happen). Also, as I mentioned, you can code the map's items and spells with these corner cases in mind from the very start and thus have damage systems not need to reduce damage so much as only deal it after factoring in reductions. -- About the RNG, what was the issue with the Ogre Magi's ultimate (I take it it was that multicast passive? Haven't played DotA in ages)? | ||
|
Judicator
United States7270 Posts
On November 10 2010 04:42 Pewt wrote: Not an issue in practice, largely because unless the armour modification is tied to the damage itself it won't occur while your code is executing (that is, between the pre-damage trigger and the post-damage timer). Have you actually implemented these systems or just heard about them? Also, as I mentioned, you can code the map's items and spells with these corner cases in mind from the very start and thus have damage systems not need to reduce damage so much as only deal it after factoring in reductions. -- About the RNG, what was the issue with the Ogre Magi's ultimate (I take it it was that multicast passive? Haven't played DotA in ages)? No it doesn't, the mechanics sections at PlayDota.com have looked at this before. In practice it won't work because basically weird shit was happening in specific incidences namely with armor values. This was brought up before and the general consensus was no thanks. I will say that I am deferring to those with more knowledge of this than I do, but I trust them on it. It was the multicast passive, it consistently got extremely streaky, same with CM's old ultimate. There were models on both before dota-allstars.com went down, but the RNG sucked for that reason. | ||
|
Pewt
Canada201 Posts
As for the damage discussion, I don't trust the DotA people since I've seen them say stuff that I know is false (such as not being able to have orb effects stack thanks to engine limitations; sure, you have to code them manually, but that's a valid way of circumventing the problem). More importantly, I've implemented the overall damage handling systems myself and I could make and give you a demo map if you are interested which addresses the specific issue we are discussing and accounts for unit attacks, items, and spells. | ||
|
Judicator
United States7270 Posts
On November 10 2010 05:08 Pewt wrote: Strange that the RNG got streaky--if it is properly weighted it should specifically prevent streaking (hence the point of weighting it). Did it actually get streaky or did people only notice the streaks (and thus remove the possibility of causing them at all)? After all, true randomness is often streaky. As for the damage discussion, I don't trust the DotA people since I've seen them say stuff that I know is false (such as not being able to have orb effects stack thanks to engine limitations; sure, you have to code them manually, but that's a valid way of circumventing the problem). More importantly, I've implemented the overall damage handling systems myself and I could make and give you a demo map if you are interested which addresses the specific issue we are discussing and accounts for unit attacks, items, and spells. You could code them to circumvent but there's a very good reason why we don't. I know the work around that you are talking about with vitality, but again, in specific cases with some hero's skills it created issues in rare (but enough) occasions. Then again, this was during the pre-patch versions so my info could be outdated. As for Magi's ultimate, it doesn't matter if true randomness or the fake RNG was streaky, it got to the point where playing Magi was a complete crapshoot high % of Multicast or low % of Multicast, I don't remember the exact numbers but it was basically varied between the two extreme categories and very rarely in the middle. It got enough complaints where Icefrog did go ahead and recode the RNG in a very roundabout way. Also, I am not saying Icefrog is a good coder, there's been a few times where his coding got him in trouble, most recently the infinite loop between Centaur's Double Edge and an active Blade Mail. | ||
|
Pewt
Canada201 Posts
But yeah, you can definitely make triggered damage prevention (and anything else in DotA from the last time I saw it, and probably still) work perfectly fine with no problematic edge cases. You need to be fairly experienced with (v)Jass and the editor in general but it's certainly doable. I think many people would be surprised how much you can achieve with the editor when you're sufficiently bored (like myself and many other experienced (v)Jassers were/are). | ||
|
zzaaxxsscd
United States626 Posts
+ Show Spoiler + ![]() source: http://www.playxp.com/sc2/news/view.php?article_id=2353657 | ||
|
undyinglight
United States611 Posts
| ||
|
lepape
Canada557 Posts
Well that spoils the expansion campaings. | ||
|
GrazerRinge
999 Posts
| ||
| ||
![[image loading]](http://i52.tinypic.com/51ynvm.jpg)