Revisiting Turnover-Induced Knowledge Loss in Software Projects
Bibliographic record
Abstract
In large software projects, tacit knowledge of the system is threatened by developer turnover. When a developer leaves the project, their knowledge may be lost if the other developers do not understand the design decisions made by the leaving developer. Understanding the source code written by leaving developers thus becomes a burden for their successors. In a previous paper, Rigby et al. reported on a case study of turnover-induced knowledge loss in two large projects, Chromium and a project at Avaya, using risk evaluation methods usually applied to financial systems. They found that the two projects were susceptible to large knowledge losses that are more than three times the average loss. We report on a replication of their study on the Chromium project, as well as seven other large and medium-sized open source projects. We also extended theirwork by studying two variations of the knowledge loss metric, as well as the location and persistence of abandoned files. We found that all projects had a similar knowledge loss probability distribution, but extreme knowledge loss can be more severe than those originally discovered in Chromium and the project at Avaya. We also found that, in the systems under study, abandoned files often remained in the system for long periods.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.025 | 0.221 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.007 | 0.005 |
| Science and technology studies | 0.002 | 0.004 |
| Scholarly communication | 0.004 | 0.010 |
| Open science | 0.003 | 0.006 |
| Research integrity | 0.002 | 0.004 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".