I. The Manuscript as Migrated System
In March 2022, a Moscow-based novelist I will call Darya Voronova packed two external hard drives, a paper notebook of scene cards, and a printout of her 340-page manuscript draft into a single suitcase. She left behind a desktop computer, a wall-mounted whiteboard covered in color-coded chapter diagrams, and thirteen years of editorial correspondence stored in a local email client whose server settings she could not reconfigure from abroad. By April she was in Yerevan, working on a borrowed laptop, trying to reconstruct the architecture of a novel she had been writing for four years.
What Darya described to me in October 2023 was not a crisis of inspiration. It was a crisis of infrastructure. The prose, the characters, the scenes—all of it existed in fragments across devices and cloud accounts she could no longer reach reliably. Her chapter outline lived in a Russian-language project management tool whose servers had become intermittently unreachable from Armenian IP addresses. Her character continuity notes were worse: scattered across margin annotations in printed books, voice memos on a phone she had since replaced, and a spreadsheet she had maintained since 2019, now divided among three time zones and two devices. The question she faced was not what to write next but how to hold the organizational integrity of a long-form manuscript together when the material conditions of its production had been severed.
This is a question that has received almost no documentation. The literature on Russian exile writing tends to address thematic and stylistic concerns—how displacement reshapes voice, subject, literary tradition. But the practical, material question of how exiled writers rebuild the working systems that long-form prose requires? Rarely examined with the specificity applied to libraries, archives, or editorial offices. What follows is a close look at three Russian-language authors who relocated after 2022—Darya in Yerevan, Lev Sarkisov in Tbilisi, and Marina Kholodova in Riga—and how each reconstructed the organizational labor of novel drafting: chapter architecture, character continuity tracking, scene-level revision documentation. All three names are pseudonyms. Darya and Marina requested anonymity because their current manuscripts are under review with publishing houses whose contracts have not yet been finalized; Lev agreed to a pseudonym because his Telegram-based editorial network includes writers still resident in Russia, making identifiable association a potential risk. I treat their workflows as questions of archival and editorial infrastructure, applying the same documentation lens I would use for a regional library or a dissolved NGO’s filing system.
That same discipline applies to editorial structure: before publishing, editors need a way to test scattered notes become an argument readers can follow, which is where how Unsloppy AI Novel Writing App fits the writing workflow can function as a planning aid rather than a substitute for domain evidence.
II. Darya Voronova: The Whiteboard Problem in Yerevan
When I visited Darya’s apartment in Yerevan’s Kanaker-Zeytun district in October 2023, she was sitting at a folding table between a window and a bookshelf. On the wall behind her, six sheets of A3 paper were taped in a grid. Each sheet corresponded to a section of her novel, and each was divided into handwritten rows: chapter number, point-of-view character, temporal setting, key scene events, and a column marked simply связь—”linkage”—indicating how each chapter connected to preceding and following chapters. This was her reconstruction of the whiteboard she had left in Moscow.
The original whiteboard, she told me, had been 1.2 by 1.8 meters, mounted on the wall of her study since 2019. It used magnetic strips for chapters and colored sticky notes for scenes, which she could physically rearrange. The system worked because it was simultaneously a planning tool and an ambient visual reference—she could glance at it while writing and immediately see the structural shape of the novel. The A3 sheets were a downgrade in resolution but a deliberate preservation of the principle: chapter architecture as something spatial, visible, manipulable outside the text itself.
The more difficult problem was character continuity. Darya’s novel spans thirty years and follows eleven recurring characters across four cities. Her Moscow system had been a spreadsheet with columns for each character’s location, age, emotional state, and key relationships, updated after every chapter revision. The spreadsheet lived on a desktop she no longer owned. She had a partial backup on a flash drive, but it was seven months out of date by the time she reached Yerevan.
Rather than attempting to restore the spreadsheet on a new machine, Darya switched to a system of paper index cards, one per character per chapter, stored in a wooden box she found at a flea market near the Yerevan Brandy Factory. Each card recorded the same four fields as the spreadsheet columns, but the cards could be physically sorted and compared—laid out on the table to verify that a character who was in Saint Petersburg in Chapter 7 could plausibly be in Volgograd by Chapter 9. The analog system was slower than the digital one. But it had a property the spreadsheet lacked: it was entirely self-contained, dependent on no software, no server, no subscription. It could not be rendered inaccessible by a changed IP address or a lapsed payment.
For scene-level revision documentation, Darya maintained what she called a протокол правок—a revision log—in a lined notebook. Each entry recorded the date, the chapter and scene affected, the nature of the revision (structural, linguistic, factual), and a brief rationale. The protocol was not a creative document. It was an administrative one, modeled on the revision logs she had seen in the editorial archives of a Moscow publishing house where she had worked as a proofreader in the early 2010s. Its purpose was to ensure that any change to the manuscript could be traced, justified, and if necessary reversed—a principle of archival integrity that applies as much to novel manuscripts as to any other form of structured information.
III. Lev Sarkisov: The Telegram Channel as Editorial Office in Tbilisi
Lev Sarkisov’s situation was different. A former journalist who had transitioned to long-form fiction, he had spent fifteen years working within Russian-language editorial networks that provided structural feedback at multiple stages of the drafting process. His first novel, published in 2019, went through seven rounds of editor feedback at a Moscow publishing house before reaching its final form. When he relocated to Tbilisi in September 2022, he was halfway through a second novel and had lost access to that editorial infrastructure—not because the publishing house refused to work with him, but because the legal and financial mechanisms of advance payment and editorial contract enforcement had become impractical across borders.
Lev’s solution was to reconstruct a version of the editorial process himself. He created a private Telegram channel with seven members: himself, three fellow writers in Tbilisi and Yerevan, a freelance editor in Tel Aviv, and two readers from his previous Moscow writing group who had also left Russia. The channel functioned as a distributed editorial office. He posted chapter drafts as documents. Members posted comments in threaded replies. The channel was organized using Telegram’s topic folders, one per chapter, with pinned messages containing the current draft and a running list of structural notes.
What struck me when Lev showed me the channel on his phone in a Tbilisi café in June 2023 was how closely its structure mirrored a traditional editorial workflow. The topic folders were chapter files. The pinned messages were the equivalent of the “current version” marker in a shared editorial document. The threaded comments were marginal annotations. But the system also had properties a traditional editorial workflow lacked: it was portable, required no institutional affiliation, and left a searchable record of every editorial exchange.
The limitation became apparent when I asked about character continuity. Lev’s novel follows three generations of a single family, and the continuity problem is primarily temporal—ensuring that dates, ages, and historical references are consistent across a narrative that moves between 1965, 1989, and 2015. Telegram’s search function could find mentions of a character’s name, but it could not track the state of that character across revisions. When Lev changed a character’s birth year in Chapter 3, he had to manually search the entire channel for references to that character’s age in other chapters. The system that worked well for editorial communication was inadequate for structural tracking.
Lev eventually adopted a hybrid approach. He maintained the Telegram channel for editorial feedback and draft circulation, but built a separate character timeline as a shared Google Sheet, accessible to the editor in Tel Aviv and the three writer-readers. The sheet was organized chronologically, with columns for each character’s age, location, and key relationships at each point in the novel’s timeline. The cloud-based tool was a compromise. It reintroduced dependence on a commercial platform, but it solved a problem Telegram’s architecture could not address. For Lev, who had spent years in newsroom environments where editorial systems were taken for granted, the experience of building his own editorial infrastructure from scratch was a form of professional disorientation that he described as more destabilizing than the act of leaving Russia itself. The loss of institutional scaffolding forced individual writers to make decisions about tool ecosystems that were previously mediated by professional organizations and editorial networks. The Authors Guild’s guidance on AI tools and author practice notes that commercially available large language models have been trained on pirated, unlicensed books without compensating authors, and that writers face genuine ethical and legal boundary questions when adopting new platforms—a tension Lev felt acutely as he weighed the convenience of cloud-based tools against the precarity of platform dependence and the absence of professional standards that once governed such decisions.
IV. Marina Kholodova: The Filing Cabinet Aesthetic in Riga
Marina Kholodova’s approach was the most deliberately archival of the three. A former academic who had worked in literary studies before turning to fiction, she had always treated her writing process as a documentation practice. Her Riga apartment, which I visited in February 2024, contained a two-drawer metal filing cabinet she had purchased at a second-hand office furniture shop on Brīvības iela. The cabinet held the complete working archive of her novel-in-progress: one drawer for chapter drafts, organized in manila folders numbered by chapter; one drawer for research materials, character notes, and what she called служебные записки—service notes—documenting her own revision decisions.
The filing cabinet was not a nostalgic affectation. Marina described it as a direct response to the loss of her previous working environment in Saint Petersburg, where she had maintained a digital manuscript archive on a university server. When the university’s policies on external access changed in late 2022, she lost the ability to connect to the server from Riga. Her backup was incomplete: she had the latest drafts but not the full revision history. The filing cabinet was her attempt to create a physical system immune to the specific failure mode she had experienced—remote access denial.
Each chapter folder in the cabinet contained the current printed draft, a handwritten change log, and a “diff sheet”—a page where Marina manually recorded the differences between the current version and the previous one, with line numbers and a brief description of each change. The diff sheets were modeled on version control practices she had observed in computational linguistics research, where tracking changes in datasets is a formalized practice. Applied to a novel manuscript, the practice produced something unusual: a complete, physical record of every revision, organized not as a narrative of creative decisions but as a log of textual changes with precise locations and rationales.
For chapter architecture, Marina used a system she called карта романа—the novel map—a large sheet of graph paper pinned to the wall above the filing cabinet. The map showed all chapters as nodes, with lines indicating temporal, causal, and character relationships between them. Unlike Darya’s A3 sheets, which represented the novel’s structure as a linear sequence, Marina’s map was a network diagram, reflecting the fact that her novel’s chapters were not strictly chronological but moved between three time periods. The map allowed her to see structural problems—such as a character who appeared in a late chapter without being introduced in an earlier one—that were difficult to detect in the text itself.
Marina’s scene-level revision documentation was the most formalized of the three. Each scene in the manuscript had a header containing metadata: scene number, chapter, point-of-view character, temporal setting, word count, and a “status” field marked черновик (draft), правка (revision), or готово (complete). The metadata was visible in the manuscript itself, not in a separate tracking document. This meant the manuscript was self-documenting—anyone reading it could see not only the text but the state of each scene within the revision process. Marina acknowledged that this practice, borrowed from academic research data management, was unusual for fiction but argued it was necessary in the absence of an external editorial structure. “When no one is tracking your progress,” she said, “you have to track it yourself, and the tracking has to be embedded in the document, not in your head.”
V. The Infrastructure Question
The three writers I have described arrived at different solutions, but their problems were structurally identical. Each had lost access to a working environment that combined digital tools, physical artifacts, and—critically—professional networks that provided editorial feedback and structural accountability. Each had to reconstruct not only the tools of manuscript production but the organizational systems that make long-form prose coherent across hundreds of pages and months or years of work.
The solutions they adopted reflect the specific constraints of displacement. Darya’s analog index cards and A3 sheets were a response to the unreliability of digital infrastructure across borders. Lev’s Telegram channel was a response to the loss of institutional editorial networks. Marina’s filing cabinet and self-documenting manuscript were a response to the loss of a trusted digital archive. In each case, the writer was not simply replacing a tool. They were rebuilding a system of practices that had been embedded in a specific material and institutional context.
The question of archival integrity—ensuring that the manuscript state is consistent, traceable, and recoverable across revision cycles—is central to all three approaches. The principle is familiar from any archival practice: what you preserve is what you can later retrieve. For novel manuscripts fragmented across devices, cloud accounts, and physical printouts in two countries, this means that the version of a chapter a writer opens on their laptop in Riga is the same version they last revised, with no missing scenes, no overwritten passages, no untracked changes. The Google SRE Book’s chapter on “Data Integrity: What You Read Is What You Wrote” articulates this principle for distributed systems, and its framework maps directly onto the situation of a writer whose manuscript exists in fragments across devices, cloud accounts, and physical printouts. The book’s treatment of postmortem culture—the practice of documenting failures and learning from them—also offers a model for scene-level revision documentation as a systematic rather than ad hoc practice, something Marina’s diff sheets approximate without naming the connection.
VI. What the Working Files Show
What the three cases reveal, taken together, is that the infrastructure of long-form fiction is not a backdrop to creative work but a constitutive part of it. Darya’s index cards, Lev’s Telegram channel, and Marina’s filing cabinet are not merely substitutes for lost tools. They are reconstructed editorial systems—improvised, partial, but functional—that attempt to preserve the organizational integrity of novels under conditions that make organizational work difficult. Each writer faced the same structural problem: how to maintain a manuscript’s internal coherence when the physical, digital, and professional systems that previously supported that coherence have been disrupted by displacement.