Please refer to the Verification Guidelines before posting. Verifications are due by Feb. 17, 2014.
Please post your opinions about the run and be certain to conclude your post with a verdict (Accept/Reject). This is not a contest where the majority wins - I will judge each verification on its content. Please keep your verification brief unless you have a good reason otherwise.
After 2 weeks I will read all of the verifications and move this thread to the main verification board and post my verdict.
First off: All the files named "poop_HQ" - "poop_HQ 8" are the Atreides missions 1-8. "poop_HQ 9" is the second version of the Atreides mission 9, "poop_HQ 10" is the second version of Harkonnen mission 9, and "poop_HQ 11" is the first version of the Ordos mission 9.
Video: Looks great, except for Atreides & Harkonnen mission 9 map 2 and Ordos mission 9 map 1, probably encoded with the wrong aspect ratio, it's too stretched, and overall looks slightly fuzzier than the others, hopefully they can be fixed, they're still acceptable to me though if they can't be fixed. Audio: Sounds great Cheating: None noticed.
Important note: This would be more of an IL table as no saving etc is shown to take place. Nothing carries over between missions so an IL table should be very possible for this game IMO. But it raises the question if all map variants are completed...
Onwards to the run:
(First batch of files submitted)
Atreides: While there were a couple of mis-clicks and minor strategic mistakes, the missions (1-8, 9 map 1, 9 map 2) were solidly executed. Harkonnen: Mostly the same as Atreides, the 2nd map version of Mission 9 could have used a little better troop management during the annihilation of the Emperor. Ordos: In mission 9 2nd map there was an idle army for a while at the end that could have been put to work earlier, but it didn't waste much time.
While there is the issue with not using turrets and that turrets can save time during some missions, I find that these runs are of high enough quality that the fact of not using turrets can be overlooked.
ACCEPT.
(Second batch of files submitted)
--verification is not complete, awaiting the second batch of videos to complete the IL table--
This is currently missing some missions but the runner should already be working on those.
A/V
The videos called poop_HQ 9-11 all have a different aspect ratio to the other ones.
GENERAL NOTES:
Overall strategies feel smart. They're also really really varied, custom-tailored for each individual mission, so I can tell the player has played the game a lot and experimented with all sorts of permutations. Superb micro and macro.
One of the few things I can criticize the runner's playing for is he should stop macroing when it's time to fully focus on microing down the few last enemy buildings. It is not be the most important time saver but it's one that sticks out. In his defense it's not always clear-cut when you're past that moment.
We have to remember the enemy attacks are not strictly scripted, i.e. they'll initially happen with a similar timing at the beginning of each mission I think but it gets a bit random after that, so in each mission, especially the long and tough ones, the runner is forced to improvise a bit.
One last thing to note is that currently most of the missions forgo the use of turrets even though they could result in faster times in many cases. I am ready to overlook this because it's not an oversight by the runner, but rather the runs were not first made with SDA in mind. Such times would in any case be more than sufficiently fast for a massive IL table like this.
My personal request for the runner is not to send in improvements straight away if he feels he might still improve on the time later simply because there's a lot of runs passing verification most of the time.
Hi guys, thanks for taking time to verify this run.
Warepire: 1. About the file names - I have no idea how to change them, can someone advise? The original files are definitely not called poop 2. Run comments - I can do that, where is that typically submitted? In this thread or uploaded like the other files? 3. About the worms - there is also an interesting thing that sometimes the worms go different way and don't come to eat harvs, so sometimes it is a matter of getting lucky reset. I had a guy complaining in one of the later Ordos missions that this doesnt happen to him and his harvs get eaten up. That's because I did as many resets as was needed for everything to go decently smooth. And this can reduce the hard versions by as much as 10 mins. 4. Yes there are couple maps to choose from except the final ones, and the way I did it was I played one version of them on easy and the other one on hard, so if the same map loads for easy and hard, there is nothing to choose. I can try to make the missing ones if needed. What is an IL table? I'm not so familiar with all the stuff here. And if you want a link for the game, I can send it to you.
LotBlind 1. For not finding the optimal routes - My first goal was to show gameplay and micro, then later I realized I might as well just keep the fastest attempts as that's the only measurable thing. Now like 2,5 years later I thought I could put it here and see what happens. I was prepared to redo things, but I read somewhere in the FAQ or so that the runs don't need to be perfect and that it is ok to see progress of the runs (history of doing it faster and faster). And besides it is 18 missions, it is quite time consuming to replay each one 50+ times to get it all perfect right away. 2. I would say the times should start once the in-game screen appears and end before the score screen. Either when it says mission accomplished or once the screen goes blank right before result screen. I'm not sure about the guidelines here and how this should be handled. I recorded rather more than less, because it can be always cut. 3. I think you are right, btw the missions #'s where you get to choose are same for all houses, not just atreides. 4. About Atr 6 - There is this little problem with playing Hard. And that is - you really want to try to get greedy at the beginning (in SC2 terminology), that is get your economy up ASAP. And this will pay in dividends couple minutes later on. But the thing is if you do this you will be dead 9 times out of 10 shortly after start as the AI can get incredibly aggressive and every mistake counts. And also I noticed sometimes when the AI feels you are taking heavy fire, they will send even more to finish you off. And then you will be happy for every attempt that goes through this stage. Yes the things you describe do matter, but that early advantage is far more important because that advantage snowballs. You can try it out on your own. But sure maybe after 20, maybe 50, maybe 100 attempts I will get a better time. In most games you won't make mistake somewhere, but will make it elsewhere. And maybe once in XXX attempts when all the constellations are right it will be all perfect (especially in the later missions). AI will attack whatever they see. And trying to set up too much into positions could make things slower. Sometimes it's just better to rush in and get the reinforcements in the mean time, because the units will inevitably die constantly, there is not much evasion to that.
Btw, Dune 2000 Atreides Mission 8 on Easy is screwed up, it can definitely be done faster, cuz I carelessly didn't click properly on one Harkonnen base silo that I had to return to after cleaning up the Ordos base. However, even after many attempts to redo it, I could not do it faster than with this mistake. Also I probably put the most effort into Hark and Atr mission 9 easy version 2 - the runs stalled around 20 min, but had to cut it down to 15 so I can put it on youtube (you get a 15 min limit as new user of youtube). And also Atreides mission 4 hard.
PS: now I realized I only upped the hard versions. Shall I upload the easy ones too?
PPS: I rewatched the mission 6, and ofc it can be done faster, but the problem was not retreating, but attacking with stronger army at the exact right time.
'IL' means 'Individual Level', it's only possible when performance on one level doesn't affect the next, so it's appropriate for many strategy games, but not say StarCraft II. We generally ask for every possible level in the game to be finished so we can have a full table of runs to put on the game page once published. If there's some missing then it would be cool if you could record those as well, I'm sure it's possible to add the files to the submission.
I've downloaded the current videos, I'll have a watch soon.
Run comments: usually submitted with the rest of the files. Send them to Uranium Anchor? Although at this point you could just dropbox them and give us the link. Personally I don't need super-thorough comments for this but I'm sure there's going to be a few things that I won't realize just by looking at the runs.
The easy missions: feel free not to submit them just yet because there's a lot of files to juggle already but yes it should be totally okay for this game to have both categories in my opinion.
I will probably write more "suggestions for improvement" in my verification but I can already tell there won't be anything you didn't already know. And yes it's definitely not ever a requirement for runs to be perfect, and like you say filling in an initial IL table is a lot of effort to begin with. You wouldn't even have had to do both hard and easy difficulties I'm fairly sure.
So you could just start recording the missing missions now and I'm really sure they will all pass with the same standards I've seen so far in the Atreides campaign.
EDIT: I also wanted to say I don't think I've ever verified an RTS yet but unlike in so many other types of games, you have to be "everywhere at once" instead of just doing A-B-C-D... linearly. In such games it's very easy to tell when the runner made a mistake but in a game like this it essentially comes down to the same elements as would win you matches in SCII for instance: micro and macro, generic decision-making skills and such. I realize that I could go on listing little things here and there in various levels but I only usually do so if I want to confirm with the runner that they've noticed it as well and also to keep track of how frequently they happen.
I have this question: Did you make a point about not using turrets or is this just a side effect of finding the fastest strategies?
LotBlind: Clicking and unclicking MCV's on starport empties silos, so it increases your maximum cash, which is good. Glitch on A9? You mean at 2:27? No glitch, just my infantry attacked that block, killing the concrete below it as impossible as it sounds. Ordos 5: I think the starport sells for like 100 or so, not sure. Wanted to keep it to empty the silos. Ordos 6: It's always better to eat 3 tanks than 3 harvs. Ordos 7: Nope, focus was to attack fast because had enough units to finish it and troopers at that point were 1) slow and 2) slows down controlling of units because u need to scroll the build menu.
The missing missions may be a bit problematic, because I'm experiencing slight flickering in the game since I upgraded my PC. So it might be worth to check them out as well for the quality reasons. I'm halfway through the comments and got some missing missions recorded.
I know about most of the things that could be done faster, but a lot of it is just winning the lottery - getting lucky and getting the sequence perfectly. And it is very time consuming, but I see it more like a work in progress, I would plan to regularly update missions when I feel like improving the time. It's something like in athletics where people improve the record over time stretching the limits.
Yes, my initial goal was to not use any turrets - to show micro and macro. I managed to do that except one of the Hark mission 9 Hard missions, which feels pretty difficult, that's also why it is the longest one - 45 min, you do one mistake and you get overwhelmed like nothing. I'm sure there definitely are ways to optimize the run with couple useful turrets. But this run is better than massing turrets and having no harvs. For example, I already managed to cut down the 46 min ordos mission by about 10 mins by using about 4 turrets and doing better unit exchanges in the middle.
Btw for the comments: I also don't remember all of the details I had to go through, but the memories would certainly come back once I'd start to all the ways that you can fail (going too greedy on eco, attacking too early or too late). I prefer if people just ask if there is something that doesn't make sense.
The comments about turrets are a bit problematic - we don't really care how you manage to reduce your time at SDA, just that you do. :p I realise you originally recorded these videos for a different purpose. If there is significant time to be saved by using turrets then those are the ILs that should get the most attention I think.
Watched Atriedes so far, seeing you accidentally cancel a building after finishing a few times was a bit painful. Other than that, I definitely agree with your overall strategy and you clearly have skill at the game.
Lag.Com: I think missions that can be done faster can be focused on later, can't they? You gotta start somewhere with the reference time. Besides I wouldn't mind if anyone else would try to get better times than me as well so I have some competition, and if it will feel impossible, maybe they won't be motivated. Nevertheless, the turrets are double-edged sword even in terms of speed running, they provide effective defense, but they don't protect any harvesters, and you do need good economy to attack, u can't move turrets, but you can move tanks, don't forget. When I hear others talk about how they did the last missions in couple hours or that they had to slow down the game speed, this would still be superior tactic. With all that said, my plan is once I start improving the individual times, I would try to include couple turrets for most efficiency.
About canceling buildings, there is a bug in the game that doesn't allow you to deploy a building and you gotta cancel it, it is much more noticeable in multiplayer if it lags.
Oh, yeah, I'm not just telling you to redo it, haha. I know turrets are very inflexible, I've played almost all of the C&C series, and the original Dune II as well. Getting good times in an RTS requires a lot of patience AND skill.
Glitch on A9 right: sorry I had the wrong time, it's a 5:40. You're trying to place a 3x3 slab and it won't let you so you end up rebuilding the slab. I see you've explained this in another comment.
Ordos 8: 23:10 - capturing a silo? Was the engineer a mis-click?
Did you know about the trick of sending over an engineer to capture some building, quickly placing down slabs around it and being then able to build anything else in that area. Maybe not useful in speed runs?
Ordos 7 13:35 (selling the barracks): it's just that you start rebuilding it after a moment later, around 14:00. Around a minute later you've already built like 4 more troopers...
I think for a massive IL table like this it's not really a problem even if we know there's many spots for improvement. Even in a 30-minute SS submission (just one of them) would there be some tolerance for mistakes. And when I asked the question about the turrets I was even ready to hear the answer that you'd simply not found them faster anywhere because like you said they're kind of passive. So I'm definitely not going to hold it against the runs. So instead of re-doing anything right now I think the only thing left for you to do is fill in the tables. We'll let you know if there's a mission that's for some reason a bit sloppier than the rest.
I'm doubtful about being able to find competition from amongst SDA-ers, though who knows, but have you posted in the Dune community forums? Because running this game fast is really just playing it well in an aggressive style, I'm thinking you'll have to place the challenge on another veteran player. Although is it really so different from other RTSes? Maybe if you took a look at which other RTSes have been run for SDA...
About graphical problems: yes your best bet is to post about it on the tech help forum section. Slight graphical glitching is not necessarily a problem. Especially if you're left with the feeling you can improve on the times later when the issue is fixed.
Ordos 8: I sent small group of units to finish off silos that die fast and there is no need for whole group to sit there when I can split them and gain less time played. About capturing, I know this very well, this is why engineers are forbidden in multiplayer, but as you said - useless in speedruns unless you are crazy korean that could multitask faster than me. But if you find one, bring him on, I would like a match with him And actually, even if you could multitask like that, it is beyond useless. What would you build there? Not turrets, not most buildings, possibly only another rax and more engies. It spends money you almost never have enough of in speedrun. If an engineer gets into enemy base, that mostly means that base was defeated already.
Ordos: Oh right, I see what you mean. This was done to unblock the path and to ensure units don't get jammed.
For the turrets, I actually believe they can be useful. If not for speed running then definitely for ease of play and avoiding resets. For the speedruns I believe they can be utilized as well. Hark 9 Hard is good example, where you get couple harvs, then static turret defense, save up for carryalls and then pump out as many units as possible along with more harvs. Also, if you get 1-3 missile turrets per entrance, backed up by siege, in case of attack you could just pull back harvs and defend with turrets / couple tanks. But it is quite difficult to pull off sometimes. Damage done to turrets can be repaired and they have more endurance than tanks.
Not sure about competition neither, there is quite some people who are on dune forums or play multiplayer, but with all respect they can't do this (possibly they can do the early missions, not likely the later ones though). Perhaps someone from C&C, but no one has come so far. Or maybe some pro from Starcraft 2, but that's just not happening
About the glitch, it's not just concrete, it's all buildings and since the company that distributed it and had copyright went bankrupt and lost the source code, there is probably no one that could fix that glitch.
Currently uploading all the missions into "Final encodes". Does anyone know if I can delete already uploaded files?
UA, can you answer Shao's question in his post above?
The comments are looking great. You can just use the same ones for the game page comments once this stuff has been published so it wasn't even just for the verification.
In my verification post I have some more questions that you didn't see yet and I'll keep deleting the ones that you've answered.
Ordos 9 (version 1) 12:00 - Possibly, I attacked, because I had the support of mercenaries and I didn't want their units to die for nothing. What I killed wasn't factory, but palace, it made the deathhand go away for some minutes I believe. Tried this mission multiple times, but this is the best time I got. If I wanted to further reduce the time, I would probably attack later, though.
28:55 - Very simple, once you get full production there is nothing to build. So you can just sit around or do something. The refinery as you said gives you the harvester and costs almost the same as harvester itself. And there is so much cash later on that you better build something or else the money is not utilized. Also if you keep fighting, your harvesters will start dying both in single or multiplayer. If you order them in Starport, well, you order 5 and then ur done for another 10 min as you need to wait for restocking. Or you can build them with refineries. Besides cash (some multi games got me even 200+k cash at the end), the harvesters can be used for defense / attack. If you send 30 harvesters between enemy sonics, it will make some huge armies die. But you can't send your last 30, that's the thing. You need constant production. Btw in rare cases the infantry can help too, but not so much, here at 27:30 I defended nearly whole attack with harvesters and refinery infantry: - Also notice that Ordos is the weakest house.
37:20 - You could say I didn't need production at this moment, but I would argue about the outpost. I guess you haven't played this game extensively, but what tends to happen is if you kill the last refinery with leftover harvesters, they will stop and sit in the spice they have been mining from. And this could leave you with 2 additional minutes searching for leftover units. Something I definitely wanted to avoid. I understand you don't see this problem in my videos, but that's because I trained myself to check for the refineries while killing the base and kill the refinery only at the point where nothing came from / in it for some time. That's why I was a bit hesitant towards the end, but it could have been executed quicker, that's for sure. Also, if you look carefully at the map, it helps to see if there are any leftover units. What could yield also additional seconds is that I just keep building even when I no longer need any more units and could micro the base killing faster, but it is difficult, because I'm used to it too much to let go off. Also sometimes you could stop building to realize you don't have enough units
If I need caryalls (running out of spice next to the base) or save up money on better deals for Starport orders (it's cheaper than production), I build Starport. If I need fast repair to gain map control or attack, I get repair pad and then send damaged units to repair if I manage it.
Ordos 9-2: No It wouldn't, it's not like the engineers got cloak or something. They either can't get in or they can, but then you are already wiping the base with units. Turrets are the most annoying to be honest and also the fact that Ordos tanks are literally made of paper. Sadly, deviators are even worse. I did use some engineers on Hark mission 7 hard, but this is very much based on luck to get little advantage. About the Starport, for one it would get usually melted fast and for two - the 5 units will not maky any difference will it?
45+ min: Same what I wrote above in this post. Btw, I already got this mission rerecorded in 31:57 total time by attacking with better timings, repairing like crazy and using couple turrets. I knew this mission had a huge reserves for improvement, and when I had enough time I did significantly cut the total time.
Next time clarify what you mean by left and right. And also, if it was refering to the map selection screen, then rather state which part of the map is the starting location so I don't need to load the game to check.
More answers: MCV's: Their purpose is threefold: 1. Expansion, 2. Safety, 3. Faster production; I haven't noticed ordering MCV in those Ordos missions, but I remember using them in Last Hark mission
Well, I uploaded the missing missions under the "final encodes" some time ago, but it doesn't feel like anything is happening. Verifiers can't finish verification if they can't see the files.
These files http://v.speeddemosarchive.com/dune2000/ (first post of this topic) were uploaded initially under "raw files" or something. Now I reuploaded everything including Easy difficulty under "final encodes". The list of files should correspond to this list: http://griffin65.webpark.cz/dune/list.txt Since everything is well arranged and there is no "poop" files (I hope) that I don't know where they came from, I would use all the files I reuploaded with the exception of files starting with (faster), those should replace the other ones in the list. I hope that makes sense, if not, ask me.
I see, so next time I do run I should prefer mp4 if possible? Is there a prefered resolution? Yes, version 2 is alternate, in the campaign you need to only do one of the versions to finish it.
New files still aren't accessible to verifiers... If they come up in a few days I'll be away for a week but will get on them after that so it'll take a bit longer.