{"canonical_url":"https://engineer.company/glossary/engineering-best-practices/","content_markdown":"# Engineering best practices\n\n## What it is\n\nEngineering best practices are ways of working with enough evidence behind them to serve as\nthe normal choice in a stated situation. Regularly restoring a backup in a separate\nenvironment is a good practice when a team needs to know whether it can recover. A written\nrecovery plan, a successful backup job and a measured restore answer different questions. The\npractice earns its place because the restore shows what data can actually be opened, how long\nit takes and where someone still has to improvise.\n\nA practice can be recommended by a research group, an engineering organization, a project or a\nstandards body. The recommendation usually assumes certain conditions, such as a team able to\ndeploy independently or someone able to review a change. A team can adopt it as advice, make\nit an explicit rule or show how it meets a contractual obligation. Those are distinct\ncommitments. An outside recommendation becomes mandatory for a team only when an applicable\nobligation or the team's own agreed rule makes it so.\n\nNIST's Secure Software Development Framework, version 1.1, organizes secure development\npractices around high-level outcomes that can be added to a software development lifecycle. It\npresents tasks and examples of ways to implement them. This distinction is useful outside\nsecurity too. Producing reliable evidence that source changes were reviewed is an outcome.\nRequiring a particular button in a particular hosting service is one possible implementation.\nWhere the outcome is explicit, another team can decide whether its review process supplies\nequivalent evidence and preserves the required protections.\n\nThe title “best practice” is also used formally. The IETF publishes a Best Current Practice\nseries, with a process described in RFC 2026. RFC 8174 clarifies when the capitalized words\ndescribed by its companion RFC carry special meaning in a specification. In another context, a\nteam may use “best practice” to mean established advice. Read who issued it, where it applies\nand what it asks before treating the phrase itself as authority.\n\n## Why it matters\n\nA sound default saves people from settling the same question on every change. A new developer\nsees the expected evidence before opening a review. Operators get comparable information from\none release to another, and teams sharing a service can agree on which actions warrant\nnotification. The saved time depends on the practice being clear enough to apply. “Review\ncarefully” leaves people to invent different criteria. Asking someone to demonstrate that a\ncustomer cannot be charged more than once gives the review something observable to agree on.\n\nDefaults also make exceptions visible. Consider a team that ordinarily releases small changes\nfrequently. A device firmware update may require a field technician to isolate a machine,\napply the update and verify the device at the site. The small-change principle can still\ninfluence the design, test sets and rollout order. The frequency of the physical release\nchanges because the conditions of recovery are different. Stating the reason helps people keep\nthe part that still reduces risk.\n\nEvidence is also local. DORA describes trunk-based development in terms of integrating small\nchanges frequently and identifies capabilities and measures teams can use to examine delivery.\nThat is a practical lead for a team with long-lived branches, slow integration and recurrent\nmerge work. The team still needs to observe its own integration time, rejected changes and\nrecovery behavior. A smaller branch count alone may simply show that the team stopped counting\nsome branches, or began combining more work behind each merge.\n\n## What it looks like in practice\n\nImagine a team that maintains an internal reporting tool. Pull requests must have a second\nreviewer and the shared test suite must pass. This rule usually catches omissions while\nchanges are cheap to correct. When a monthly report is late, the usual test suite spends\nseveral hours replaying fixtures that cover an unrelated export path. The proposed fix changes\nonly the report's filter and is reversible before the report is distributed.\n\nThe team first checks its actual obligations. If approval before release is required, a faster\nroute still needs authorized approval. It then names the relevant risk. The urgent fix must\ninclude the correct records, keep the previously agreed grouping and avoid exposing rows from\nother business units. The author runs focused tests with known input and compares the\ngenerated report against expected rows. A separate reviewer inspects the diff and checks those\nresults. Someone is named to decide whether this evidence is sufficient, and the previous\nreport is kept ready in case the comparison fails. The full test suite runs afterward and will\ninform whether the team wants a faster permanent check.\n\nThe short exception record states the normal rule, why the case qualifies, which protections\nstill apply and when the exception expires. It records the actual elapsed time and outcome. If\nthe team needs the same exception again the following month, it can move the report-specific\nassurance into its regular workflow. Before changing the default it checks that the faster\nroute can detect a wrong date filter and a deliberately excluded business unit. A check that\nfails on those seeded errors supplies stronger evidence than a timer alone.\n\nEngineer’s case on\n[a commit gate of 22 linters with no warning tier](/portfolio/built-a-22-linter-commit-gate/)\ndescribes making a project's agreed checks enforceable at commit time. A different\n[accessibility case](/portfolio/the-accessibility-faults-your-checker-misses/) describes\ndefects encountered despite automatic checks. The two examples expose different questions. A\ngate can enforce a rule consistently while the underlying test still covers only part of what\nusers experience.\n\n## Where it goes wrong\n\n**A checklist becomes the goal.** A process marked complete can still produce the failure it\nwas supposed to prevent. Choose one sample where the practice should clearly make a difference\nand observe it. Check that the reviewer can reconstruct the change, that the restore opens\nusable data or that the release actually reaches the intended users.\n\n**A tool is mistaken for the practice.** Buying a review platform does little if nobody has\nenough time or context to disagree with a change. Write down the outcome that the tool is\nsupposed to support. Then inspect a real example with the people who use it and find out\nwhether the promised information is available in time.\n\n**A local exception becomes a quiet default.** A workable emergency path tends to get reused.\nGive exceptions a stated scope, an accountable person, a compensating measure and a review\npoint. Repetition is useful evidence that the normal route deserves attention. Improve it\nopenly when that evidence holds.\n\n**A team's evidence is copied without its context.** A rule can work under different\ndeployment, staffing or product constraints. Start with the problem described by the source\nand compare it with yours. Try a bounded change, decide in advance what improvement and harm\nwould look like, then keep the practice when the observations support it.\n\n## Further reading\n\n- [NIST Secure Software Development Framework, version 1.1](https://csrc.nist.gov/pubs/sp/800/218/final)\n  — the framework's practices, tasks and examples of implementation.\n- [RFC 2026, section 5](https://www.rfc-editor.org/rfc/rfc2026.html#section-5) — what Best\n  Current Practice means within the IETF's standards process.\n- [RFC 8174](https://www.rfc-editor.org/rfc/rfc8174.html) — when capitalized requirement words\n  have their defined meaning in an IETF document.\n- [DORA on trunk-based development](https://dora.dev/capabilities/trunk-based-development/) —\n  small changes, frequent integration and ways to assess the practice.\n\n\u003chttps://engineer.company/glossary/engineering-best-practices/\u003e\n","date_modified":"2026-11-19T06:00:51Z","description":"How software teams choose proven practices, distinguish guidance from obligations and use evidence to assess an exception.","formats":{"gemtext":"https://engineer.company/glossary/engineering-best-practices/index.gmi","gopher":"https://engineer.company/glossary/engineering-best-practices/gophermap","html":"https://engineer.company/glossary/engineering-best-practices/","json":"https://engineer.company/glossary/engineering-best-practices/index.json","markdown":"https://engineer.company/glossary/engineering-best-practices/index.md","pdf":"https://engineer.company/glossary/engineering-best-practices/index.pdf"},"kind":"page","language":"en","title":"What are engineering best practices?","url":"https://engineer.company/glossary/engineering-best-practices/","version":1}