# Rockstar Got 13 Years for GTA 6. Your Team Got 18 Months to Replace 30 Years of Code.

> Thirteen years after GTA V and three release dates later, GTA 6 is out. What Rockstar did that big IT projects are famous for not doing, and how a rewrite dies.

- Author: Salim Ben Hassine (https://salimbenhassine.com/about)
- Published: 2026-11-19
- Topics: Legacy Modernisation, Engineering, Leadership
- URL: https://salimbenhassine.com/posts/gta-6-took-13-years

GTA 6 is out today. Thirteen years after GTA V, three release dates later, and with a few thousand people working on it.

I've played GTA since I was a kid and never missed a release. When GTA V came out, I was at uni. Now I'm a working parent with a bit of grey hair, and I was one of the people who waited years for this one. Thirteen years feels different when you spend them on the waiting side.

Somewhere right now, a project manager is reading the headlines, opening a slide deck and moving a go-live date for the fourth time. The difference is that nobody pre-ordered their release, and nobody is refreshing a download bar at midnight for it.

I modernise legacy systems for a living. I've never shipped a game, but I've seen enough "two-year" software modernisation projects to recognise a well-run big project when I see one. And Rockstar just did almost everything big IT projects are famous for not doing.

---

## Three release dates and nobody got fired

GTA 6 was announced for 2025. Then it moved to May 2026. Then to today.

Now picture the same thing in a typical IT project. The first delay gets a workshop to agree on a new date. The second delay gets blamed on the supplier, who is replaced. The third delay gets a new name, "strategic realignment", and the project starts again from zero.

Rockstar moved the date twice, said they needed more time to reach the quality people expect, and went back to work. The internet complained for a weekend, and everyone still bought it.

They could afford to say "not yet" because the date was never the product. The product was the product. Revolutionary concept, I know.

---

## The real hero is GTA V

Here's the part nobody talks about. GTA V came out in 2013, sold more than 200 million copies, and was released again on two new console generations. GTA Online got updates for thirteen years. The old game kept paying for the new one, the whole time.

That's not luck. That's the most important pattern in legacy modernisation, and most companies get it backwards.

When a company wants to replace a system that has run its business for decades, the first instinct is to treat the old system as the enemy. Freeze it. Stop investing. Put everyone on the new one. Switch the old one off on a Monday, ideally after a long weekend.

Rockstar did the opposite. They kept the old system alive, earning and improving, while the new one took as long as it needed. In our world that's called the strangler pattern: keep the old thing running, replace it piece by piece, and switch off parts only when the new ones have proven they do the same job.

Your legacy system is GTA V. It's old, it's a bit ugly, everyone has strong opinions about it, and it's paying everybody's salary, including the people writing slides about killing it.

---

## How a rewrite dies

I've seen this one up close. A company grows very fast, and its whole business runs on one old internal system that was never designed for a company that size.

So it's held together with hacks. Every new requirement gets a workaround, and every workaround gets another workaround. Everybody hates it, and it still runs the entire company, every single day.

The decision is made: replace it. Reasonable so far.

Then come the mistakes, in the classic order.

**Impatience.** The plan is to rebuild everything on a new stack at once, with a date that makes everyone nervous except the people who chose it.

**A wish list.** Why just replace the old system when you can also build all the features people have wanted for years? Same budget, same date, of course.

**Failures stacking up.** Each one is "a small delay". Together they are the project.

**The knowledge walks out.** The pressure grows, and the people who understand the old system best start leaving.

Costs keep climbing until the project is cancelled. The old hacked-together system outlives its replacement.

Nobody in that story is stupid. Every decision makes sense in a meeting. Together they make a textbook big-bang failure.

---

## Unfinished software always looks terrible

In 2022, someone hacked Rockstar and leaked about ninety videos of early GTA 6 footage. Half the internet decided the game looked bad. Placeholder graphics, debug text everywhere, characters walking through walls.

Game developers everywhere responded by posting screenshots of their own unfinished games, which looked just as bad, because that's what software looks like in the middle.

Every legacy replacement goes through this moment. Six months in, someone demos the new system to management. It's slower than the old one and missing half the features. Panic. The project gets "re-scoped", someone gets replaced, and the pressure goes up exactly when the team needs calm.

Rockstar ignored the leak and kept building. If you judge a rewrite on its worst demo instead of on what it does in production, you'll cancel every good project at exactly the moment it looks worst.

---

## "But we're not Rockstar"

Correct. You don't have GTA V's money, a few thousand developers or fans who'll wait years and say thank you.

But the lessons that matter here are free:

**Keep the old system earning.** Don't freeze it and starve it while the new one is built. It's your GTA Online. It pays for everything.

**Replace first, improve later.** A migration and a wish list in the same project is two projects with one budget.

**Replace piece by piece.** Move one part at a time, run old and new side by side, and switch off the old part only when the new one gives the same results on real data.

**Protect the people who know the system.** They are the documentation. When they leave, the project doesn't slow down. It goes blind.

**Let quality move the date, not the other way round.** A date set before anyone understands the old system isn't a plan. It's a wish with a Gantt chart.

---

## Wasted

Every GTA player knows the screen. Things are going well, you're driving fast, you're very confident, and then the screen goes grey and one word appears: WASTED.

That's what a big-bang rewrite looks like from the inside. Years of driving fast and feeling confident, and then the meeting where someone quietly says "let's stop".

Rockstar took thirteen years and shipped. A lot of rewrites take less and ship nothing. So before you set the date for replacing the system that has run your business since before some of your developers were born, ask yourself one question:

Would your company have cancelled GTA 6?

Cheers,
**Salim**
