Quote from andyb on May 26, 2012, 7:47 pm

I did a blind run of it.I did ok till near the end.I missed seeing a portal surface after getting distracted by something else and I spent 1/3 of the vid trying to get out of the last part.Not bad overall thou but sure could use some puzzles

http://www.youtube.com/watch?v=AEC-nKpS ... e=youtu.be

Quote from Groxkiller585 on May 27, 2012, 11:16 pm

You guys need to understand that students, who probably didn't have much, if any, Hammer experience, made these maps, so you can't expect expert quality. Also unlike most mods, the teacher probably had to include every student's puzzle, even repeats. I personally think that if he managed to organize it and it works at all, that is an accomplishment alone.

What might have prevented those bugs would be using instances, but then they wouldn't have learned, so that's why they might be different from one another: another person made it, so it works differently.

The only thing that's really killing it for me is texture misalignment - I am sure a day dedicated to that would have made a big difference, but your the teacher, not me. That's just a suggestion I'd make.

EDIT: Just finished. The visuals, as many people have said, are below par, but acceptable given they are students. (Whoever made the Mars map is cool ) The ending I liked, but it ends on a cliffhanger - what happens to our hero(ine)?

If you want I can help port this to a Sourcemod; From what I can tell it will work just fine there, you can launch it from Steam and no .bat needed!

Quote from josepezdj on May 28, 2012, 8:34 am

WOW Grox, that answer was ........

Quote from briandoyle81 on May 28, 2012, 11:00 pm

Thanks for the support, Grox!

You've pretty much summed it up. At the beginning of the project, I was the only one with any experience with Hammer. The students had to learn how to use Hammer and build the levels within a single semester. That's a pretty big task, and IMHO, the results are pretty great for a first time effort.

Another day to fix the texture alignment would be nice. As would another day to fix a bunch of things.

I'd be thrilled to have help getting it on sourcemods. I only learned that was an option from this thread and I haven't had a chance to research it yet.

Glad you liked the ending. If you'd prefer to not have a cliffhanger, try bringing the chair with you. It adds an interesting option

Quote from Groxkiller585 on May 29, 2012, 10:26 am

Ok, I have a sourcemod version working. Here is it and install instructions:

*Download Sourcemod Version Here* (this link isn't permanent so tell me you got it. Once you do, if you want to keep the sourcemod setup I reccomend you upload it to your site.)

* 1. Place the folder in *steam install filepath*/steamapps/sourcemods. (if this folder doesn't exist, create it)

* 2. Restart steam. (This is required so it knows about the new mod.)

* 3. Go to "Library" and look down the list for "Portal 2: Vitrified". Click it.

* 4. Now I'm not sure this will happen for you, (if I did it right it shouldn't) but the game might take a long time loading the first bootup. If this happens, let it go all the way until the indicator bar at the bottom right is yellow. After that you probably need to kill the program via Task Manager. This should only occur once.

* 5. Reload the game and voila!

A note: If you update the mod, you can do it as you've been doing for this mod, but do NOT delete cfg/autoexec.cfg as this is what loads the first map at start. Also if sounds are, for w/e reason, not playing (I think they should) add back in snd_rebuildaudiocache on a new line under the map line in cfg/autoexec.cfg. (this garentees a long load however. Might be better if you called that command from in the start map (point_clientcommand) shortly after it loads, it goes much faster.))

Quote from Nacimota on May 29, 2012, 1:33 pm

Brian,

Massive wall of text
[spoiler]UsCobra11 makes some solid points and after playing as much of this pack as I could, I'm inclined to agree with his general opinion. I can understand his trepidation at providing criticism for content that he believes would be beyond his own skill level; however I think that is irrelevant. One does not need to be a master chef, for example, in order to assess the quality of a meal.

I also follow Groxkiller585's remarks about this pack being the work of students and while I feel this explains a lot, it does not (in my opinion) necessarily justify some of the major design and technical issues present in the pack. It is not my intention to offend anyone on your team, simply to provide feedback which I can only assume is the reason for your post. Indeed, a great man once said "the cruellest thing you can do to an artist is tell him his work is flawless when it isn't".

Let's start with gameplay and level flow. The first thing I notice when I load the map is that the level branches out in several places in ways that strike me as unnecessary and almost confusing. In an above post, you said:

briandoyle81 wrote:

Most of the team thought that Portal 2 led you by the hand a bit to strongly, so we made a little more exploration necessary.

I'd like to briefly note that while there's nothing wrong with explorative gameplay, there's nothing wrong with linear gameplay either; which one you go for depends on the game you're making. If you look back carefully at Valve's games, you'll notice that they are all extremely linear (not just the Portal series). A non-linear structure produces a large number of design issues, the most obvious being that players can become confused about where they are supposed to go and what they're supposed to do. Another problem is the state of mind it puts the player in.

For example, there's a red button in the first level that kills the player (with what I can only assume is neurotoxin gas). An instant kill mechanic (with no warning of its danger, very early in the game) is bad in and of itself, however it is made much worse by the fact that you specifically encourage players to indulge their curiosity through explorative level design and then inadvertently punish them for it with things like the aforementioned neurotoxin button. This is nothing short of obnoxious.

Additionally, most of the extra areas don't really achieve anything; they don't seem to provide further context to the story or alternative paths through the levels. They are, in essence, dead ends with absolutely no purpose whatsoever. If you want to encourage explorative gameplay, you should reward the player for it. Taking the original Portal as an example of this, there are several areas that players could access behind the panels of test chambers. These were also dead ends but aside from being obvious that they were off the beaten track, they provided things like notes and diagrams scribbled frantically on the walls which gave the player hints as to the overall story as a reward for their exploration. There were one or two areas in your map pack that I think did this fairly well, for example the hidden area behind some moving panels in the repulsion gel test that had a chair and an ashtray among other things. Overall, however, the exploration in your map pack seems pointless and I stopped bothering after the first few levels.

There's a lot of repetition of puzzle elements and nothing new or challenging is ever introduced; just mechanics from the previous Portal games. If you're assuming the player hasn't played Portal or Portal 2 then you need to introduce more player-training at the start of the game. If on the other hand, you're assuming the player has played the previous games, then you need to replace the puzzles with something different and more challenging. The first two tests are exactly the same puzzle (cube on button), as Logic notes above. Getting the cube also isn't necessary in the second test, not just because the door is gratuitously delayed as UsCobra11 said, but because you can shoot a portal on the other side of the door once it opens anyway.

I could go into several more examples but on the whole, the puzzles in the game all feel unimaginative and unrefined. There's no challenge or gratification to any of them (both key elements of a puzzle game) and they get pretty boring after some time. Several of the test chambers are unreasonably large given their contents.

Moving on to aesthetics, I think the pack is kind of hit-and-miss in terms of art direction itself. I get that you're going for an "old-Aperture blending into new-Aperture" look but I'm not convinced it works particularly well. It should make sense and feel like a natural evolution of Aperture but in places it seems very jarring and abnormal. It feels like you've just built the chambers using the 40s theme and just thrown some newer textures and props around the place haphazardly. There were a few areas where this was not the case though, for instance the big hall with the 70s Aperture sign (a convention hall, I think it was called...?) seemed to be heading in the right direction (artistically speaking).

Unlike UsCobra11, I actually didn't mind the voice actor who played Dr. Slater (actually, his voice reminds me somewhat of David Hyde Pierce). The problem is, of course, that his lines are occasionally unintelligible; sometimes it sounds almost as if he is mumbling and cannot be understood. I'm inclined to think this was more due to how his lines were implemented into the game than his performance itself.

In terms of technical execution of aesthetics, I think your team did very poorly. This is probably my harshest criticism of the pack. First, many of the chambers are too visually cluttered (this concerns not just prop placement, but textures and world geometry in addition to poor lighting). The aesthetic design of a level is not independent of its gameplay; it complements it. Poor visual design can inadvertently cause gameplay issues, as Valve's Nick Maggiore explains:

Nick Maggiore wrote:

Early versions of Portal featured more detailed, cluttered environments, much like Half Life 2. We quickly realized that unnecessary objects scattered all over the place distracted players to the point where it actually interfered with the portal training process. So we simplified the art style to favour clean, focused test chambers.

The lighting is generally quite horrible. Areas are either under-lit or over-lit but rarely well-lit. Most of the testing areas were far too dark and in fact there was one area that was deliberately pitch-black. I'm sure this sounded like a cool design idea to someone on the team but in practice it is hindrance to the game's flow. Last year when Valve had us run a map competition for them, there was an entry that had similar pitch-black lighting and it was unanimously panned by the judges for its poor lighting and said lighting's effect on gameplay.

I also noticed a very heavy use of flickering/pulsating lights which really doesn't look as good as you might think (in addition to heavily inflating the size of the BSP) and is a pet hate of mine. It didn't even make that much sense in a number of places; at least twice I saw a series of fluorescent lamps slowly fading in and out as if they were malfunctioning or something. Fluorescent lamps don't work that way (unless you set them up with specialized ballasts or something to that effect) and even if they did, it's unlikely that all three lamps would malfunction in sync with one another. Besides not making sense, they're just generally annoying to look at. Flickering lights can easily be disorienting (sometimes even headache inducing) and should (in my opinion) be avoided. Even if you're trying to go for a chaotic and apprehensive atmosphere, you don't need flickering lights to achieve this. It didn't seem like that's why they were there though; I felt as if that they were tacked on for the sake of it.

Shadows were a big problem and often far softer than they should have been. This is due to a combination of the lightmap scale not being adjusted and VRAD running without -staticproppolys, among other things. It seems nobody on the team is familiar with the intricacies of Source's complex (and outdated) lighting system.

There were a lot of misaligned and poorly scaled textures which stuck out to me quite strongly. I also noticed a small area at the start of the game where floor tiles were being replaced and the wall was in the process of being painted over. I thought this actually looked pretty good but it was done with wilful disregard of the engine. The paint strokes were achieved using clipped brushes joined together, with the white paint on one side and the yellow paint on the other.

Not only would this have looked much better if a single face had been used with a custom texture, but it would have been much easier to put together and less expensive for the engine to render. Notice the large number of unnecessary vertices:

Since we're on the subject, most (if not all) of the levels in the pack suffer from a severe lack of optimization and on my machine the framerate dropped in several places where it shouldn't have. Visleaves in particular seemed almost completely without optimization which is probably the main reason for the performance issues. Visleaves were often needlessly cut up by world geometry that should have been either a static prop or a func_detail, as demonstrated by the floor tiles and carpet in the image below:

But you know, Brian, when everything is said and done I think the effort put in by your students is commendable and is something they should all be proud of regardless of my criticism (which, as we've established, is set at a very high standard of quality) and I hope nothing I've said has changed that. I think perhaps one of the points I am trying to illustrate here is that the Source engine is (in my opinion) not the most appropriate environment for an academic course. Source relies on some rather complex and outdated systems which are difficult to both teach and master. One of these is the BSP based level system itself; the other is the complex precalculated static lighting system. I feel like a lot of the lighting and optimization hassles your team has clearly gone through would not have been a problem in a more recent and approachable engine. I don't know if you are in charge of the curriculum but next time you might get better results with something other than Source.

Cheers.[/spoiler]

Quote from msleeper on May 29, 2012, 3:41 pm

:postoftheyear:

Quote from briandoyle81 on May 30, 2012, 12:41 am

Thanks Grox! I've got the file downloaded and I'll test it out first thing tomorrow.

Nacimota, thank you for the detailed criticism. One of the big problems in doing a short term project like this is that there simply isn't enough time to learn the tools, construct the world and then test it, refine it, optimize it, etc. This is the second time I've taught this class, and my biggest wish is always that I had another semester with the same group to tear it down and rebuild it from the ground up utilizing all the lessons learned and then optimize and refine it using more advanced parts of the tools, such as the lighting issues. It's definitely a difficult task to learn these tools, but all the pro tools are this complicated and they need the experience.

As for the dialog being difficult to hear, if anyone can point me in the right direction on how to properly implement it, I'd be eternally grateful. That took up a large part of my efforts and nothing I did seemed to help. Soundscripts seem to be ignored for custom sounds, regardless of if they're in the vpk or in the folder. I suspect that although it uses choreographed scenes, the audio isn't being played in the voice channel. Even the sound built into the doors, etc. seems to suppress it.

I think fixing this will help resolve the some of the stylistic issues you pointed out. The fact that it's often difficult to hear makes it hard to pick up on the fact that it all takes place in 71 while Aperture is undergoing a revival and some serious remodeling. In hindsight, I should have written that to be more explicit as well.

Quote from smash on May 30, 2012, 5:03 am

As a student and a team leader on this project, I would like to thank you folks here at ThinkingWithPortals.com not only for testing and reviewing our project, but also for your valuable help in creating this mod. Many of our questions were answered on this forum directly or through valuable tutorials and previous helpful threads, and it made our task much more obtainable. Without this forum, I'm not sure if our class could have created a mod of this magnitude and quality.

Another massive wall of text addressing the comments and talking about our project
[spoiler]Prior to this course, no one had any major experience with Hammer except our instructor, Professor Doyle. Many of us were skilled in other toolsets, but we had to learn everything from the ground up in 16 weeks. Also, students were not technically supposed to work more than 10 hours per week (the course allowance by credit load, set by the University). Some students did work much more than that, but it was an ambitious project and the students had varying degrees of talent, interest, experience and skill, so that was a huge factor.

As far as many of the problem, I very much agree; we needed more time to optimize and test. We didn't learn about func_detail, autosaves or soundscapes until quite far into the project because our scripters were spread very thin. Those features and optimizations are definitely something I'd take into account from the beginning if we could redo the project. I would very much like to add more autosaves and clean up some of the misdirection and confusion. The neurotoxin was part of the story element, though if we had more time, I would have liked to add some autosaves before the player encounters them.

As for repeating or simplistic puzzles, or borrowing ideas from Portal 2, that is the nature of a mod produced during a single semester. All the ideas had to be designed within two weeks in order to allow the level artists and scripters to begin, and that was, admittedly, a difficult method to employ in order to produce a linear and challenging map. However, I think our strength in this mod is that we attempted a story that is within the Portal universe.

Groxkiller - we planned to add an ending for the Hobo-Hero, but it had to be trimmed down due to time, sadly. Our character model artist (MJ Parker, aka Taralqua) also worked quite hard to give this mod its own unique character model, but in the end we just did not know how to implement the model despite seeking advice. I am not sure if that can even be done, though I think it would have helped "sell" our mod a bit more.

Nacimota, I agree with you about the virtues of linear play in a game like Portal; we very much tried to emulate the original game in many respects, though we could have refined this to be much more elegant. I think if we had another shot at it, we would design the levels in a more holistic way in order to make the game flow a central concern. The method of piecemeal ownership of level design was not conducive to a linear outcome. While that is not necessarily a bad design method, its product really depends on the designers, the resources and the type of flow targetted. Originally we did include many exploration easter eggs, but due to the concern that it would confuse the players, we removed them. This created a vacuum in some areas, but the player could still sense that something was supposed to be there (as noticed in andyb's blind playthrough of one of the levels - I left a comment on youtube explaining those issues). Additionally, some of the levels had novice designers with almost no testing, and that is just the nature of a class with no pre-requisites. We were able to not only include all of the levels that were designed, but also incorporate them into a coherent story and fully script and populate them visually - not a short order for our small class.

The one critique I will argue with is the voice acting. For the voice actor, we did realize that there was a major problem with the volume... after the class was over. We did not have much of a chance to correct this in time. UsCobra, to say that he is horrible is rather unfair. I have not seen many other Portal mods with any voice acting at all. We were lucky to have someone capable of performing the lines with such gusto, and if the sound volume was stabilized, I think it would be easier to appreciate. Nacimoto, I never noticed but I think you're right about the voice actor sounding a bit like Pierce. That makes it quite amusing now...

Back to major issues - lighting was a concern from the beginning. We wanted the levels to look like they were being used without Aperture's knowledge, and we wanted the rooms to seem dark because Slater didn't want them to notice a power drain on the system (which is probably the source of the lighting fades that Nacimota mentioned). I admit, this did not translate well. Nacimoto, you were quite right; no one on the team was familiar with Source's lighting system prior to the class, and by the time we learned enough about it, it was too late and we had too many other critical issues to deal with. The lighting was definitely a huge issue and I sincerely appreciate your thoughts and explanations about it - I wish we had more time to clean it up.[/spoiler]

Overall, I think our class performed quite well considering the obstacles with learning Hammer and creating a mod of this scale in such a short period of time. We learned more in this course than in any other design course in the program thus far. Most of us probably put more work into this class than any other course in the department as well. There were students with extremely variable skillsets, and like Groxkiller said, we needed to accommodate everyone. While Hammer might not have been the easiest and most straightforward choice for a class, it was definitely challenging and rewarding. I have not seen many mods of this scope produced by other teams, but I feel that our mod is an impressive product with much potential. Of course, I am biased.

Again, I truly thank you all for the comments. It's given us a lot to think about as a team. Many of us are taking a similar course with the same professor this fall because we enjoyed working on this mod so much. While we will use another toolset for another game, many of your suggestions apply to game design in general and I think they are very insightful. We will absolutely go over your comments in our initial meetings to help us look out for issues ahead of time. While Hammer is a powerful dev kit, like Nicomoto said, it's probably too complex/outdated to create such a massive mod in only one semester. However, I know that the vast amount of my classmates truly enjoyed creating this mod, and we value our experience as a development team.

Thank you all again for the comments, and thank you for the sourcemod version, Groxkiller. I will definitely check that out!

-Mandi, Vitrified scripting lead & team leader