Tamriel Rebuilt has a fundamental choice to make in its design philosophy: do we want to design things to be better than vanilla, or do we want TR to feel like a seamless extension of vanilla design? Many players would likely prefer the former, should they be asked. Historically, TR has often gone with the former, too, but it has seldom turned out well. As a rule, Tamriel Rebuilt now opts to follow vanilla design sensibilities as much as reasonable.

The main issue is that what people consider "better than vanilla" is a rapidly moving target. If we develop the mod to adhere to the tastes of today, it may very well appear “moddy” and outdated in just a couple years, or it may bite us in the rear in terms of technical or organizational issues. Many of the recent trends in the Morrowind community make for cool mods that are certainly worth playing. But if such mods depart too much from vanilla sensibilities, they are unlikely to be something that most people would want to install for their every playthrough – at least not in ten years time. Whereas that is exactly what we strive to achieve with TR.

In contrast to these ever-changing modding trends, vanilla has stayed the same across the two decades of MW modding history, yet still attracts tons of new enthusiasts all the time. So, it's become clear to us that if we want to build something that lasts, we should strive to adhere closely to its design philosophy (of course, we don't take this to the absolute; vanilla has a lot of inconsistencies and laziness that we try to avoid).

To illustrate the point, some examples from each of our development departments follow.

Assets

Back in 2003 or so, the Morrowind modding community held nearly universally that the vanilla head and body models were awful (they do have a lot of technical issues) and that everyone should be using Better Heads and/or Better Bodies instead. So much so that the earliest alpha builds of TR that we have from that time literally bundle those mods – if you start the game in Seyda Neen with those versions loaded, you instantly get jump scared by Jiub with a Better Heads mesh. Likewise, older clothing assets made for Project Tamriel were made to fit with Better Bodies, instead of vanilla.

Of course, these days, most people vocal in the Morrowind modding community – and TR in particular – don’t prefer Better Heads/Bodies. Why would anyone replace the wonderfully expressive and iconic vanilla faces with something as transformative as Better Heads? Nor are the Better Bodies meshes considered much of an improvement over vanilla design anymore. So, obviously, we no longer bundle Better Heads and instead bash our heads against the wall trying to de-Better-Bodies-ify our old Imperial and Nord clothing meshes.

Exteriors

In the late 2000s and 2010s, making really great exteriors meant stacking a ton of rock meshes together in dense formations to create the kinds of elevation changes and visuals that vanilla didn't do. Likewise, cities were to be dense and vertical and had to get "creative" in how they use existing meshes to pull that off. A good example of this design trend is the mod Stonewood Pass, but this was also the governing design philosophy in TR for a while: places like the Mephalan Vales, Roth Roryn, and much of the TR_Preview Redoran lands that we've now stopped bundling with the mod. This approach can make quite impressive visuals, but is murder for performance (see Old Ebonheart and Roa Dyr), makes navigation difficult (especially if you have to escort NPCs) and instantly stands out as moddy to modern eyes – you can tell that the vanilla meshes were not meant to be used in the way they are in these exteriors.

Essentially all TR lands made in such fashion are slated to be heavily reworked in order to fix said issues. Instead, we now try to replicate vanilla design philosophy, using the terrain mesh to do most of the heavy lifting, and rely on static meshes much more sparingly (with some targeted exceptions, like the Shipal-Shin region that is built around bespoke cliff meshes). You can read more on our modern exterior design principles on the wiki.

Interiors

At roughly the same time, many interior designers opted to make what we call “all-rockers” – dungeons made not from the tiling caverns and passages that vanilla uses, but by piling together separate rock assets. This allows for much more liberty and variety in dungeon design and is used to great effect in dungeon mods like those by the great seelof. For TR, however, all-rockers come with several issues:

• Making dungeons this way is much more time-consuming. This is an issue in a mod where we have to fill out literally thousands of interiors, all hopefully within our lifetimes. We could reduce the number of interiors we include, but that's a disappointing compromise to have to make and would constitute a noticeable departure from vanilla
• There will appear an instantly recognizable style divide between the vanilla game and TR. Yes, we realize that this is the case in much of TR already, but it's something we want to move away from, not towards.
• All-rockers are impossible to do effective quality assurance on. It's a tall order to find all the holes and visible mesh back-faces among the hundreds of smashed-together rocks, especially since it tends to be difficult to view and edit a cave like this in the Construction Set after it's finished.
• It is tough to design an all-rocker to have good pathfinding for creatures and NPCs. All-rockers may look nice, but typically play poorly. Careful design can mitigate this in an all-rocker, but it's definitely much harder to get right.

Hence, we try to stick closer to vanilla dungeon design, although with quite a bit more consistency and quality assurance. Despite that overall direction, we do have quite a few huge and complex dungeons, and many more in the planning or being worked on. Kharom in the southern Velothi Mountains or the Necrom catacombs come to mind. The catacombs and sewers underneath Narsis are still upcoming, as of early 2024. So is the massive Hlaalu ancestral tomb just outside Narsis. Kemel-Ze will hopefully become that in the future. It's just that when we make these dungeons, we aim to make them look like they could have been made by Bethesda back in the early 2000s.

Quests

Starting in the mid-2010s, TR's questing practices developed in such a way that nearly every questline had to be a tightly woven 10+ quest long narrative with highly complex scripts. This was mostly driven by Rats' superlative work, such as the Old Ebonheart Thieves Guild, which every quest designer wanted to emulate and exceed.

This culminated in the initial planning stages for the Dominions of Dust expansion in early 2020, where the principle was applied to each of its dozen questlines. The result was a series of delays and is a large part of why it took a full three years for the expansion to be released. Making such questlines turned out to be a roughly two-year effort for most developers and more often than not resulted in them burning out for long stretches of time. It also dissuaded many new developers from even attempting to make them, as the idea of claiming a 12-quest outline was too daunting for most. Moreover, it becomes very difficult to do quality assurance – as complexity increases, the number of corner cases often becomes too great, meaning that these questlines still regularly produce bug reports years after their release.

Now, no one is saying that complex narrative questlines are fundamentally bad or that existing ones should be removed – just that it's not sustainable or desirable for every questline to be like them. Like vanilla, we now strive to tell our stories much more subtly. Vanilla faction questlines are much shorter and are typically presented as a series of separate jobs with very light instructions that allows for many different approaches to completing the task – perfect for roleplayers, who often complained about the railroading that TR's more complex questlines had. Only a subtle narrative through-line is woven into them, which comes to fruition in the later, high-level questlines. This change in design philosophy is possibly the main reason it only took a year to get the Andaram expansion out the gate.

A few other questing trends that we at one time adopted but have now abandoned are faction-less “settlement questlines" – like the one in Ranyon-ruhn, which is to become a proper Telvanni questline – or new factions that deviate a lot from vanilla tone or conflict with what’s implied in lore – such as the unfinished Molag Bal cult implied in the Ranyon-Ruhn Temple questline (although that one was never released).

—

As you can see, two decades of experience seems to offer a number of reasons to stick with vanilla design. Not the least of these is that Bethesda chose these practices so as to get things done – certainly a virtue in this two-decade old mod. Our shift back towards vanilla sensibilities seems to be working out for us so far. But it's possible that it’s just another trend in TR's ever-changing design philosophy – we’ll see.