Starisland is an open-sky Garry's Mod DarkRP map (note the "open-sky" - it's important), themed around Miami and GTA Vice City.
How to summarize approximately 2.5 years of obstacle-filled work... Let's start from the beginning. But first, I'd like to give you some context. :)
We're in 2016, more precisely in March (or April). This is the year I'm taking my middle school diploma, to give you an idea of how far back this goes. As I write this, I'm about to finish my master's degree, 5 years after high school. :')
After spending about 3 months trying to compile the Rockford V2(b) map, I finally managed to make my own modifications and share them with my friends. For me, this was the beginning of map making on Source Engine. Originally, I had absolutely no ambition. The idea was simply to modify Rockford to add ramps where my cousin and I could do car stunts. Yes, it's that silly. It could have been done via addons, I could have simply placed shapes and models provided by Garry's Mod without even needing to go through the Workshop, but I decided to modify the map. xD
In reality, I felt that greater potential would open up to me if I managed to compile the map. It took me ~3 months, it's the worst thing I'd recommend anyone do. However, I succeeded. :)
At this stage, the only manipulations I was able to do were fairly simple map modifications. Changing textures here, extending a wall there... However, as time passed, I learned to better understand the level editor, Hammer World Editor. I then started making increasingly advanced modifications. That's when I thought it would be a great idea to make a YouTube video explaining how to compile the Rockford map. 🤡
Yes, because at that time, I maintained a YouTube channel (which still exists) and it was one of my main hobbies until then. I then got many comments, many people wanted to contact me, which honestly was quite unprecedented for me. Today, I've deleted this video because honestly, it was a concentration of everything you should absolutely not do. I had even made a second video supposed to be a bit clearer for those who hadn't understood well. 😂
Anyway. Among these encounters was someone I got along well with and quickly became friends with (yeah I say friends), Loïc Dekeyser. :D Without going into details (because there are many 😂), he suggested creating a DarkRP server together where we could make our own map. And honestly, I was very excited about the idea. He provided the dedicated server and plugins, I provided the map. That's how my very first public map was born, Rockford Actu-Gamework (in reference to my website at the time). I won't present this map on this site because for me it's a career mistake, but it was a necessary step in my opinion to get where I am today. 🥲 I'll let you freely visit the link to my old site, with all its old 2018 salt.
However, this modified Rockford, mutant even since in reality it's a map reconstituted with several pieces of other maps, well it caused me more than one problem. Do you know the principle of intellectual property? Or perhaps the principle of respecting others' work? Because Sofiane from 2016 didn't know them too well. :)
The Workshop was for me a giant library where I could draw ideas, resources, everything that could be useful to improve the map. While it's true that this is indeed the very principle of the Workshop, at least for players, using it as such for content creation is dubious. The modders, who seemed to me to be inaccessible people often English-speaking, seemed much less distant when they came to complain about the map, after I decided to present it on the old MtxServ forum. 😂
The message was received, I had to start making my own map if I wanted to restore my honor. 🤣 (yes it's very chivalrous)
Well no, because here I'm talking about events that occurred in 2016. And if you paid attention to the map information, it was released in 2019. 🤔
Actually it's a bit complicated because, for several reasons, I had to stop creating this Rockford Actu-Gamework quite late. With Loïc creating the server, I couldn't leave him with an abandoned map on his hands. That's how I got to version v3d. You can find it on the internet quite easily, it had leaked.
All this to say that I really only started working on Starisland from the very end of 2016, during the Christmas holidays. The first builds I still have date from January 2017.
The original plan for the map was relatively simple and mainly relied on a map walled on all sides, a bit like Evocity v33x. Moreover, the first version was released as planned, in March 2017. But if you've been following, the map was released more than two years later. :)
What happened is that Starisland, in its original plan, didn't please me. I wasn't satisfied with the final work. Initially I was working jointly with Loïc on Starisland, which meant there were artistic disagreements (in addition to different ways of creating the map). So I quickly thought about the next version, Starisland v2, which I made alone. The first version was walled on all sides, the second would be more open. The first was rather small, the second opens up more space. And that's how two long years of frenzied creation began. I won't keep the suspense going, Starisland v2 never came out, nor v3, nor v4, nor v5... In reality, 6 versions were conceived. Moreover, I quickly stopped using the term "versions" because in reality they were much more prototypes.
Each major change in the plan resulted in a new prototype. Thus, the first version of Starisland has almost nothing in common with the final version published on the Workshop... Except for a tree that I ironically named "Billy", long forgotten in a corner of the map and kept as an "easter egg" for those who followed the map's adventures. :) For those interested, it's the tree located in the inner courtyard of the fire station.
But here's the thing. If the first plan for Starisland was relatively simple and easy to achieve, the various improvements I made greatly complicated the map. If at first, optimization wasn't a problem, it quickly became one. The Source engine not being adapted to complex maps, and even less to open-sky maps, I quickly encountered serious problems continuing the creation of Starisland. To the point that I had for a time considered abandoning the project. Indeed, arriving at prototype #5, in its 7th version (yes I wasn't versioning on git at the time), I faced the impossibility of being able to add anything else to the map without exceeding the Source engine limitations. Usually I always managed to solve these optimization and technical limitation problems, but this was too complicated.
Fortunately, beautiful help from Facepunch Studio arrived shortly after noticing this problem. With the February 2019 update, Facepunch had doubled most of the technical limitations of the time. Some limitations like the total number of vertices hadn't moved much or at all, but at that time it wasn't a major problem. It was great. :D
This graciously offered update allowed me to have some respite on the map creation. While being aware that I was still very close to certain technical limitations that would cause me problems, I decided to finish the map plan by adding a small island, planned for a long time. However, the plans were somewhat modified. If initially it was agreed to create a second residential area, rather disadvantaged for the greater joy of gang rp, I had to cut back on the plans. For example, no habitable building as planned. The one that had been started was quickly made inaccessible. The power plant is also not accessible from the inside, only one store out of two will be open and the hotel will not have rooms upstairs, no service room and the second building will be condemned. A last normally accessible building will also be condemned.
It's a very sad fate that will be reserved for this small island that I named "Little Island", but which was necessary to allow the map to be released in a more or less final version. 🥲
And yes because we arrive in June 2019, and I'm seriously starting to have more than enough of Starisland. If I used to take every problem to heart, I must admit that this time, I sorted without trying to find good compromises. I went for the simplest. Moreover, you'll note the absence of a promotional video for the release of the final map. There was indeed a teasing dating from the beginning of the creation of the second island, but no announcement for the final map.
It's July 13, 2019, after a beta release in a small committee, I decide not without difficulty to release the map. Initially, we can say it works quite well. The illusion is perfect. 😁
The visuals, the general atmosphere, the plan, the fact that it's an open-sky map, all this makes the map work quite well with the public, without reaching the Workshop top (sniff).
But here's the thing, there's a problem that's omnipresent. There's a gigantic elephant in the room! Which one is it? But my god how awful the framerate is! And yes, it's terrible to say but the map will never be used by a single DarkRP server because of this problem. Some people will have tried, but they all failed.
The reasons for these poor performances are multiple. Already, as I said, the Source engine is not an engine designed for open world games. It's a corridor game engine based on the old BSP principle. A direct descendant of Doom! But what does that mean concretely? Well for optimization reasons, the game space is divided into rooms. Still for optimization reasons, the game loads all the content of a room. This is very good for small environments like in Valve games, but for maps this big with so much detail... With that in mind, you shouldn't expect much more. Especially since the map itself has more than questionable optimization, it's already very impressive to see how far the map creation could go.
Starisland remains for me a beautiful technical demonstration. It's a concentrate of everything you shouldn't do, nicely assembled and rather well documented. It's also a beautiful demonstration of the skill evolution of someone who started with meager mapping knowledge, and who finished with much more knowledge and experience. Clearly, this map is not a technical success. But for me, it's a beautiful accomplishment that I don't regret at all. :)
It was also an opportunity for me to make beautiful encounters, and to express my artistic vision through a beautiful work made in 3D. It's also ironically my only complete finished map, subsequent maps resulting in multiple abandonments. 😅