Build1 distinct publisher3 min readUpdated
A former Jagex developer's account makes the port a question of scope and expectations, not graphics. The audit list is the part that transfers to any long-lived codebase facing a new target.
The Engineer · Build desk

Compiled by The EngineerSomething wrong?How this is made
The expensive item on that audit list is not the hard-coded screen dimension. That one is findable and mechanical. The costly one is the right-click.
RuneScape's desktop model put a context menu behind nearly every object in the world: talk to an NPC, pick up an item, examine it, attack it [9]. Hover states let a player preview information before committing to a click [10]. A mouse hands you two independent channels for free, plus pixel precision. A finger gives you one contact point, no preview, and a contact patch wider than several of the elements it has to hit. So every interaction that used to open a menu becomes a question about what a single tap should default to, and the reasoning behind the original refusal to pick a default is not in a document you can search, because much of the code was written by people who had already left the company, in patterns that predate current convention [5].
The panels repeat the problem at a different scale. Inventory, equipment, prayer, magic and combat stats were all on screen at once on desktop [11]. Hiding most of them is the obvious answer and also the one the audience notices, because the community arrived with firm expectations about how the game should look and feel [13].
The memory figures set the outer bound. Desktop rendering paths assumed a floor of 2GB or more of available RAM, while mobile devices were often limited to 1-2GB in total, shared with the operating system [7]. At the bottom of that range, the desktop floor alone is twice the whole device budget before the OS takes its share [2]. Since both clients sit on one persistent world with one player-driven economy [12], the savings cannot come from carrying less world. They have to come out of how the same world is represented.
Two numbers deserve care. The publisher's headline sells a 20-year-old game, and the article's own inventory cites over 20 years of content from hundreds of developers [17], while the same piece dates the codebase at more than 15 years when the author joined [3] and puts his tenure at 2017 to 2019 [2]. Measured from the 2001 release [1], the codebase was roughly 16 years old when he started [1]. The 20 is the game's age at the time of writing, not the age of the thing being ported. The account is also single-sourced: a former employee, writing afterwards, whose current studio sells the audit-first method it describes and has applied it to remaking Nub [15][16]. The claim that weeks of audit saved months of production is his estimate [8], and Jagex is not represented in the piece.
What transfers is the order of operations rather than the checklist. Settle the scope boundary first, since a parity client and a companion app price out as different projects [18], then write down every service the old platform was quietly providing: fixed screen geometry, pointer precision, hover, memory headroom, and the assumption that you are the only client touching the world [6][4]. Code comes after that. Anything on the list nobody wrote down arrives later as a production incident with a ship date attached [8].
Follow any of these and your For You feed starts watching them — no settings page required.
Ranked by verification strength, evidence, and original report placement.
RuneScape was first released in 2001 and is described as one of the longest-running MMORPGs in history.
The author worked at Jagex from 2017 to 2019 as a Technical Developer on the RuneScape Mobile team.
The game had already accumulated more than 15 years of content, systems, interfaces and player expectations when the author joined the mobile effort.
The mandate was full cross-platform parity: a mobile player needed to be able to do everything a desktop player could, in the same world, at the same time, rather than a simplified mobile version.
The codebase contained code written by developers who had long since left the company, using patterns and conventions that predated modern best practices, and the game's scripting language, RuneScript, was proprietary to Jagex.
Before writing any mobile-specific code the team audited the systems it would need to touch, finding hard-coded screen dimensions embedded deep in UI rendering logic, mouse-specific input assumptions baked into core gameplay mechanics, desktop-specific rendering paths assuming minimum hardware specifications far above any mobile device, and desktop memory usage patterns.
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
Single self-published recollection, no artifacts
All content derives from one dev.to post authored by the studio whose services it promotes. The technical detail is plausible and internally coherent, and specific enough to be actionable (named audit findings, explicit memory figures, an enumerated desktop interaction model), but nothing is externally verifiable: no code, profiling data, build metrics, Jagex confirmation, or second account. One quantitative claim is unbacked entirely, and the article's own age figures disagree between headline and body.
Two self-reported projects, no outcome data
Adoption of the described audit-first practice rests on two self-reported instances from the same voice: the author's 2017-2019 work on the RuneScape Mobile team, and Ocean View Games' Nub remake with Inferna Games in Unity for iOS, Android and Steam. Both are asserted rather than evidenced — no ship dates, store listings, download or retention figures, or third-party confirmation appear — and no other studio is shown using the method.
Modestly overstated by framing, not by substance
The underlying engineering observations are measured in tone and mostly proportionate to what a practitioner could know. The overstatement is in the packaging: a headline age figure larger than the body's own timeline, an unquantified 'audit took weeks, saved months' return presented as a lesson, and a marquee legacy MMORPG used to credential a small studio's current porting services while simultaneously disclaiming any commercial role in that project. The gap is real but narrow — the transferable checklist largely delivers what the framing promises.
Vendor content marketing with explicit disclaimer
The post is published on dev.to under the Ocean View Games account and explicitly pivots from the RuneScape recollection to 'how they now inform the mobile development and porting work we do', naming a current commercial project. That is a direct lead-generation incentive to make legacy porting look both hazardous and tractable-with-help. Mitigating the score somewhat: an unusually clear attribution note stating the work was done at Jagex Games Studio, that RuneScape is a Jagex trademark, and that Ocean View Games was not a contractor or agency partner on the project.
Coherent but unverifiable single-voice account
Confidence is limited by structure, not plausibility. Everything traces to one self-interested publisher with no second source, so the descriptive claims about the codebase, parity mandate and audit findings can be reported but not confirmed, and the one payoff claim cannot be tested at all. What raises confidence above the floor is that the technical content is internally consistent, mechanically sensible for a 2001-era desktop MMORPG moving to touch hardware, and accompanied by an unusually explicit attribution disclosure.
build
One game, two codebases: where parity belongs when you ship native on iOS and Android1 distinct publisher
build
Unity's AI problem is not the prompt: the load-bearing context lives in the prefabs1 distinct publisher
build
Roblox's Prompt to Prod: typing was the easy part, trust is a retrofit on a 20-year codebase1 distinct publisher
build
Mipmaps are three levers, not one checkbox: stability, residency, and meaning1 distinct publisher
Distinct publishers with included, body-backed reporting in this cluster.
dev.to
1 article · August 24, 2026