AFK_Weye Commentaries: History, Part 2
By Dwip April 2, 2019, 4:07 pm Comments (1) RSS Feed for this post

By the time AFK_Weye 1.2 came out in June 2009, I was pretty burnt out on AFK_Weye in specific and Oblivion in general. Furthermore, I had reached a point where I was pretty happy with what I had, and figured AFK_Weye 1.21 could probably stand as the definitive version of the mod. People seemed ok with it, and I was tired of dealing with it. So I went off and did other things. Played some Civilization 4 and some Fallout 3. Got addicted to Stargate SG-1. Had an actual life with a girlfriend who wondered what the hell I was doing this whole time. Ultimately, though, I eventually came back to Oblivion with a new character, and as part of that playthrough, went through AFK_Weye again for the first time in a long while.

Year 2 & Version 2

I wouldn’t go so far as to say I was appalled by what I saw, but certainly there was a lot of room for improvement. Part of it was that I had grown as a modder – By this point I had been involved in the community for over a year, and I knew better ways to do some of what I was doing, and how to fix some of the persisting bugs from the old days. Moreover, I knew how to actually improve on what I had previously built, tightening up some things, adding new ways to do some of the quests, etc. I had also spent enough time out of the Oblivion bubble to have been exposed (or re-exposed) to new ways of doing things – Fallout’s companion management and quest structure, quests with multiple endings, whole other ways of writing games. I knew I could make AFK_Weye more than it was, and with that idea planted, I couldn’t let go.

Although my initial work in October and November of 2009 was intended to be another bugfix version along the lines of the previous versions, I quickly started running up against scripting and quest stage problems that were holding me back from truly fixing what I wanted to fix. I made the decision to completely break compatibility with the 1.x versions (1.1, 1.2, and 1.21 had all been drop-in upgrades). That would require everyone using the mod to lose all of their progress, but gave me complete and utter freedom to do as I wished. That was a hard decision to make, as forcing everyone to destroy all of their progress is extremely annoying, but as I hope you will see, the decision I made was the correct one, and the tradeoff for complete freedom to expand and revise was worth it.

By December of 2009, I had reached the conclusion that I needed to make AFK_Weye 2.0 and not AFK_Weye 1.3. At this point I really started thinking on grander terms. “Can I fix this bug?” morphed into “If I added a new arc to this quest, or added a new ending, would that make it better?” “Would this benefit from a cutscene?” “How about a new dungeon?” “What can I do to make the house better?” My internal list grew at a very rapid pace, but I also decided early on that I should get some outside opinions. As a result, I wrote a couple of posts asking players what THEY would like to see in a new version of AFK_Weye. I had been getting requests all along, and some of these had made it into either earlier versions or my new 2.0 ideas list, but I wanted to make things explicit, and it worked quite well. I got a number of suggestions, and a player named Pvan went so far as to write several lengthy posts on possible improvements. Almost all of them made it into the new version in some fashion. Many of them were things I had never even considered, or made me think of features in new ways. Pvan’s help really improved the mod, and proved to me that it really pays to both listen to your players and also to get a fresh perspective from time to time.

Broadening Horizons

By early 2010, I was reaching my limits as a modder. I had figured out complex quests well enough, but some issues remained:

– I really wanted to get more life-like NPCs. That meant figuring out idle animations, which wasn’t particularly hard, and spawned my second mod, AFK_PrayerIdles, which I wrote as a test for making idles work and as a break from AFK_Weye.

– I also wanted my NPCs to talk to each other instead of just the player. That meant figuring out Oblivion’s conversation system, which is different than the dialogue system. Eventually, it was also going to involve voice acting.

– A lot of the models and textures I was using were heavily dissatisfying. To a certain degree, I could get around that by using freely available modders’ resources, and I used a great many (check the readme), but it was clear to me that I was either going to need to find somebody to do modeling and texturing for me, or I was going to need to learn how to do them myself. Since modelers and texturers were and are few and far between, I made the decision to learn both disciplines, and spent a great deal of the first half of 2010 doing so.

Of all of these, modeling and texturing was by far the hardest to work out. I had done some work in graphics programs in years past and could fake my own texturing with enough patience, but modeling meant learning Blender, and that was going to be a pretty daunting task to overcome. Assuming you could figure out how to get it installed (along with Python, NifSkope, NifScripts, and some other minor things), the interface alone was almost completely impenetrable and impossible to understand (Blender works like no other program in the world, and delights in this). Worse still, the documentation was either nonexistent or useless, and such tutorials as existed were usually either for earlier versions (which had a completely different interface), badly written, or worst of all, badly written for previous versions in a tone of smug condescension.

To say that things were hard is a gross understatement, and for a very long while, I was grasping at straws, seizing upon seemingly insignificant pieces of interface minutiae, trying to wring some aspect of comprehension of this Satanic beast of a program. It was a long struggle, and I often thought about just giving up and making do, but I KNEW I could do this, and I knew that if I figured this out, AFK_Weye would be leaps and bounds better than it had ever been. So I stuck it out, and bit by bit, piece by piece, I finally came to understand what I was doing.

It was a glorious moment, and from that moment all things were open to me. I don’t believe I’ve ever been so liberated as a creator. New stall for Thalonias and Elahai? No problem. New paths for the garden? A cinch. A massive new cavern to replace the laggy mess in Kerrach? I couldn’t even dream of doing anything like it a year ago, but now I pulled it off in a weekend. Things that had been completely impossible a few short months ago were now trivial obstacles, problems only so long as it took me to whip up a model to take care of it.

People tend to look on modelers as superhuman beings wielding impossible skills unattainable by mere mortals. People do it to me now, and I always try to kill that kind of thinking. 3D modeling isn’t magic, it’s a skill, and anyone can learn it if they just take the time and make an effort to try. I won’t pretend that it isn’t hard, and yes, the learning curve on Blender is excessively steep, but it can be done, and it’s worth the while.

And of course, that same thing goes for any skill I talk about. You can’t fake creativity, but if you have a story to tell and a world to build and just need the skills? So long as you’re willing to try new things and persevere, you too can learn those skills. It’s just a matter of time and effort.

That having been said, there are things that you cannot control. For me, voice acting was one of those things. I’ve mentioned a conversation system, and that was easy enough to figure out – it’s just like dialogue, really. Therein lies a problem, however. In order to get conversations working, you almost need to have voiced dialogue. To have voiced dialogue, you need people to voice it, and voice actors are even more rare than modelers within the community.

That said, I wasn’t against voice acting. Indeed, back in November WhoGuru did up a batch of lines for one of the cutscenes at Arthmoor’s behest, and I liked what I heard. What if I could get the whole mod voiced? Not very many mods had voice acting, and it would really be something if AFK_Weye were one of them. To be sure, the challenges were daunting – not only did I need voice actors, I had multiple thousands of lines of dialogue to deal with in a lot of different voices, and that was going to take some doing. Nevertheless, I went ahead with it.

The first group that expressed interested in doing some work bailed almost instantly for reasons only they know. Not worried, I turned to TES Alliance’s Voice Actor’s Project, where I was assured I would find plenty of people. They were there, but things quickly went south. I hate to be critical, but the whole experience was an unmitigated disaster – almost everyone bailed on the project because of real life obligations (understandable) or simply never got back to me (not so much), and only two or three people ever gave me anything at all. After months of wrangling, I finally gave up in disgust, with no other options open to me. I’m given to understand that many people have had much better experiences, but for me, the benefits weren’t worth the costs, and I abandoned all but a few lines of voice acting. If you turn on subtitles, most of the conversation system is still there, but I no longer have the heart to do anything with it.

This Is A Test. This Is Only A Test

Voice acting pain aside, by early April I felt that the mod was finally ready for alpha testing. The mod was basically feature complete, and I had done some testing work on my own, but much remained to be done. Thus I passed it off to Arthmoor and hoped for the best.

I said this before, but it really bears repeating – mods, never mind software in general, are hugely complex beasts, with tens of thousands of things that can possibly go wrong. In the case of AFK_Weye 2.0, it had perhaps double the content of AFK_Weye 1.21, but most of that new content was twice as complex. Since every interaction between those thousands of things contains the possibility for at least one bug, it’s utterly impossible to catch everything.

This is all to say that alpha testing lasted about two months, fixed a few hundred problems of various sorts, and involved more than one major rewrite of a quest or part of a quest.

At the beginning of June, I sent an email to Arthmoor, Hanaisse, and Pvan to begin the official beta testing period. In it, I announced that the greatest remaining problem was probably getting things ready for voice work.

It should hopefully be pretty obvious at this point that no such thing was actually the case. Beta testing actually lasted precisely a month, fixed all sorts of other bugs, and required a laborious re-modeling of the interior of Kerrach that I couldn’t even have done a year previously. As you’ll see, even at the end of beta testing there was more stuff that didn’t get caught.

With that story in mind, a few observations about testing. I don’t figure these are particularly original, but:

– Two heads are better than one. More is better. Turns out that one of the problems of a small testing group is that over multiple playthroughs of a large mod, testers tend to become locked into certain ways of doing things. Sometimes this is for good reason (you need to test one particular way), but nevertheless it does happen. The more people you have, the less you run into the problem. Too, different people think and play differently, so it’s more likely you’ll get different feedback on different things from each person. Any way you look at it, more people will catch more stuff.

– There’s going to come a time when you’re going to run into something major. Maybe it’s a quest that really does need another method of completion, maybe a location isn’t working, but whatever it is is going to take some pain to fix. You’re going to want to fix it now, or you’re going to have to fix it later. You may as well do it now, and not bitch about it too much. This goes for the little stuff, too. Yeah, that AI bug is pretty minor, but it’s going to stop feeling minor after every single one of your players gripes about it.

– There’s also going to come a point, and remember that by July of 2010 I’ve been working pretty solidly on AFK_Weye for about 8 months straight, where you’re going to get sick and tired of looking at this thing, and it’s going to become a chore. You’re going to need to push through this, but it’s also worth realizing that it’s ok to stop and take a break for a little while. Not a long while, because then it becomes hard to get back to the project, but a little while.

– Bug tracking is really important. AFK_Weye’s testing had dozens of emails, IM conversations, etc, and there’s no way you can keep all that in your head. What I did was keep an MS OneNote file with a list of bugs and a checklist. As bugs came in, they got added to that list, and as bugs were fixed, they were marked off. I further grouped each set of fixed bugs, so that as sets of bugs were fixed I could tell the testers which ones to look at, and keep them separate from new bugs I needed to fix. If something turned out not to be fixed, it went right back into the queue.

A New Era

AFK_Weye 2.0 released to the public on July 2nd, 2010, one month to the day after the start of beta testing, almost three months after the start of alpha testing. Unlike when AFK_Weye 1.0 was released, by the time 2.0 hit I’d had released mods for over a year, and was reasonably active in the community. I’d spent a lot of the last year helping other modders, and I knew I was good, and I knew what I had with 2.0 was leaps and bounds ahead of where any of the 1.x versions could even dream of. I pretty much figured people were going to like it.

What I didn’t really expect was to be a very close runner-up for the file of the month vote at Nexus for July 2010, nor, having done that, did I expect to actually be one of the files of the month for August. By mid-September, AFK_Weye actually got reviewed in an article online. Download counts were soaring, and I was pretty happy.

That said, even after the heroic amounts of testing, there remained plenty of bugs in 2.0 that only became apparent after release to the public and thousands of people. Who knew that killing Maxentius while wearing the Cowl of Nocturnal was such a problem? Certainly not me. Thus, version 2.1 followed about a month after release, with 2.2 three months later.

I tried to take a slightly different approach with these patches than I had with the 1.x line. Where those patches concerned themselves solely with fixing bugs, 2.1 and 2.2 actually added a few things – a small new quest, some new meshes in 2.1, lots of small cosmetic changes in 2.2. This was mostly a case of my inner perfectionist running amok, but I was happy to do even little things to make my creation that much better. I tend to enjoy mods where the authors take the time to do this sort of thing, and figured I probably wasn’t alone.

After 2.2 released in November of 2010, I let things lie for another 8 months, and honestly thought 2.2 was probably going to be the final version of AFK_Weye. By mid-2011, however, it became obvious that there were a few little things that still remained to be done, and so 2.3, 2.31, and 2.32 were released in July, August, and November of 2011 to fix just a few more issues, and add a few more things. In particular, people had been asking about a stable area since the early days of the mod, and now in 2.3 I gave them one, and a small sidequest to boot.

And here we are. The bugs are fixed, Skyrim is here, and I’m satisfied that I’ve told the story with AFK_Weye I set out to tell, in a form that people can enjoy without trouble. It’s time to think about other mountains to conquer.

Hopefully this dissection of the lifecycle of a mod has been of help. For more nuts and bolts detail, future posts in this series will look at the individual quests of the mod in some detail, which will hopefully give some insight into how I went after designing and building them.


Computer Games - Elder Scrolls Series Comments (1) Trackback URL for this post RSS Feed for this post
Comments on AFK_Weye Commentaries: History, Part 2
avatar Pingback by The Ravings of Demented Rabbits » AFK_Weye Commentaries: Introduction #1
April 4, 2019 at 4:24 pm

[…] Part 1 History, Part 2 Main Quest, Part 1 Weye Manor Main Quest, Part 2 Main Quest, Part 3 Kerrach Side Quests […]

Leave a Comment