> ## Content Index
> Fetch the complete content index at: https://blog.manugames.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# Foundry Gave Up Its Release Date to Thirteen Forum Comments
- URL: https://blog.manugames.com/foundry-update-5-delay-experimental-branch-loaders/
- Published: 2026-11-17T15:00:00.000Z
- Updated: 2026-11-17T15:00:00.000Z
- Description: Foundry had November 19, the same day as GTA 6, and gave it up. The test branch did not find a bug, it found a price.
- Author: Seth Tideward
- Tags: The Deep Dive, Patch Notes Roundup

## The Deep Dive

On Friday, Foundry's developer gave up a release date and told you exactly why. "I originally planned to release it on November 19th, because I found it pretty funny to do it on the same day as GTA6\. However we have received a lot of constructive feedback from the experimental branch and I want to spend some extra time reacting to it." November 19 is the date Rockstar Games has on its own site for Grand Theft Auto VI. Update 5 now has no day at all, only "the plan is still this year."

Foundry is the first-person voxel factory builder from Channel 3 Entertainment, published by Paradox Interactive, in Early Access on Steam since May 2, 2024 at $29.99\. Update 5 went onto a public test branch on November 3, carrying blueprints, new loaders, conveyor lifts, small cargo ships and powered train tracks, with one stated purpose: "the goal is to gather as much feedback as possible." Ten days later the date was gone, and what the branch had found was not a bug.

### It found a price, not a defect

The new loaders are the change at the centre of this. Foundry used to have several Loader items, one per length, placed like any other building. Update 5 replaces all of them with a single Loader you drag from a start point to an end point, with the drag direction setting item flow. In July, DevBlog #107 listed the four goals: transfer machine to machine without a belt, transfer belt to belt, "Reduce the amount of unique items," and "Improve user-experience and reduce the amount of clicks needed."

Read that list and the redesign is unambiguously better: it does more, carries less clutter, and is fewer clicks per loader. Players agreed in July, and two weeks later the developer said so, "I'm glad that the feedback was very positive." Then people got their hands on it. On November 4, day two of the branch, Cyfagy wrote: "Before, I could walk past the machine and just go click, click, click... now I have to stop every single time, hold the click, pull, and then let go." Same day, insane\_Alex: "The time and number of clicks required to add new loaders have increased significantly." Same day, onetoxiczombie, with the line that makes the whole thread credible: "This isnt just me saying 'older is better' more customization is nice."

All three said in the same breath that they liked the loaders. That is the finding. The new system is fewer clicks per loader and more clicks per row of loaders, and a factory player does not place one loader, they place forty in a walk-past. A drag is a better primitive and a worse reflex. You cannot count clicks in a video, which is why July's round could not catch it: that was feedback on a description, and the description was correct.

So a public test branch is not a bug hunt. Bugs get found by anyone eventually, including by the people who paid full price on release day. What only a branch can measure is the cost of a change that is right on paper, in the one action a player performs ten thousand times. Foundry paid a release date for that reading, so the date was the cheaper of the two.

### The fix is the expensive answer, and it undoes a goal

Friday's blog is not a revert and not a shrug: "we are going to explore ways of having both systems. One of the things we'll try is the option to have fixed sized loaders you can select from the radial menu, with the drag mode being the new default." That is nearly word for word what insane\_Alex asked for on day two. It is the right call, and the one that costs the most, permanently.

Look back at goal three, "Reduce the amount of unique items." Shipping both systems partly unwinds it, and leaves two placement paths to build, document, teach and keep working in a game with no mouse cursor, which the developer admits is the hard part: "Sometimes things are really hard to implement in first person view."

Newsletters love to print "developer listens to community" and stop there. The honest version is that listening changed the design document, not just the ship date, and the permanent maintenance is the real payment. It is the right bill to pay, because players adapt to features and do not adapt to a hand movement they make four hundred times an hour.

### Thirteen comments is a different instrument, not a small sample

The thread I keep quoting had thirteen comments when I read it on Monday. Thirteen, on a game two and a half years into Early Access with a publisher behind it, and it moved a release date. The cleanest trace in it: on November 4 Hesulan asked for the input and output loaders to be colour coded again, "maybe as a settings option or something. That made it a lot easier to see at a glance while running around building stuff." The November 9 branch changelog contains the line "Added Loader colors, indicating direction." Five days, one comment, one changelog entry.

The blog also shows the half nobody praises, where a developer declines. Blueprint feedback split ("Some of you said it's a bit overwhelming, while others said it's missing the even more complex features for erkle's mod"), and his answer is to keep the built-in system accessible and let the community mod carry repeat mode, rather than average two groups into a feature that serves neither.

### What makes a feedback round affordable

Cadence. Update 3 hit its test branch on December 16, 2025 and went live thirty-six days later; Update 4, Rails and Reactors, shipped April 28\. A studio on that schedule spends one window reacting, not a year of silence, which is why a studio with one annual release cannot buy this: its feedback lands inside a frozen build, and by the time next year's patch ships the regression has become the game. The sentence that makes the post work is the one nobody will quote: "we're still in Early Access. Going forward we'll have more rounds of quality of life improvements with every future update." Rounds, plural, is the product. The date never was.

### What I am taking into my own game

Disclosure, since it belongs here: I build Tideward, a premium idle RPG for Apple hardware, so I am a developer writing about another developer's schedule, not a neutral party.

An idle RPG cannot run this experiment the same way. The loop is metered in days of wall-clock time, so two weeks of hands-on testing says little about how a trade feels in week six. What I have instead is an open beta and a Discord where keepers report, and the reports that no design note would have produced are exactly the Foundry kind: 1.1.4 fixed the Sailing Orders menu that had lost its place, because kotori.piyo put hands on it and I had not.

The other thing I am taking is a rule I already made and will now defend harder. Every Toolbelt pattern in Tideward is a sidegrade rather than an upgrade, more XP for a slower round, that kind of trade, so a new tool never retires the one you were using. This is the failure mode that rule exists for: the moment a new tool replaces an old one, you have told everyone their muscle memory was worth nothing, and you find out from strangers whether that was true only after it ships.

Foundry had a funny release date and a strictly better loader. It kept the loader, dropped the joke, and will carry two placement systems for the rest of the game's life because a handful of players' hands disagreed with four correct design goals. That is a studio reading its own production line properly. The bottleneck was never the feature, it was the reflex, and reflexes are only measurable in public.

[DevBlog #116, the timeline and the feedback](https://store.steampowered.com/news/app/983870/view/1849192144187628?ref=blog.manugames.com) | [The Update 5 experimental branch post](https://steamcommunity.com/app/983870/discussions/0/571553842245712383/?ref=blog.manugames.com) | [The branch feedback thread](https://steamcommunity.com/app/983870/discussions/0/571553842245712490/?ref=blog.manugames.com) | [DevBlog #107, the loader redesign goals](https://store.steampowered.com/news/app/983870/view/1837955055356922?ref=blog.manugames.com) | [Foundry on Steam](https://store.steampowered.com/app/983870/?ref=blog.manugames.com)

P.S. Tideward is up for pre-order on the App Store for iPhone, iPad, Mac, Apple TV and Apple Vision Pro. Free to download, one $9.99 unlock, launching February 11, 2027\. [Pre-order it now](https://apps.apple.com/us/app/tideward/id6765781814?ref=blog.manugames.com).

---

*The Assembly Line is a free twice-weekly newsletter on factory builders and idle games. Read on the web at* [*blog.manugames.com*](https://blog.manugames.com/)*. Brought to you by Manu Games, makers of Tideward, a premium idle RPG for iPhone, iPad & Mac.*