What makes a program robust

Learn more. What defines robust code? Ask Question. Asked 10 years, 2 months ago. Active 2 years, 6 months ago. Viewed 37k times.

So what is the actual definition of "robust code"? Improve this question. Lotus Notes. Lotus Notes Lotus Notes 1 1 gold badge 5 5 silver badges 7 7 bronze badges. This would only hold in a not-strongly-typed language.

In a strongly typed language a variable of type boolean not some integer posing as a boolean , can only be true or false, there is no third option I do not know of any languages where both x and! Show 5 more comments. Active Oldest Votes. Here are a few links that discuss what it means to be robust in terms of software: Robust Programming Robust Definition Robustness, the forgotten code quality.

How to write robust code If you think there is one universally agreed upon definition of "robust" here, good luck. Improve this answer. Community Bot 1. If this was a nullable boolean both Java and c would throw so null should be checked first. There doesn't seem to be any universally agreed upon definition of what a cat or a dog are.

Add a comment. Option A. Option B is trivial to test. Option A kills me Option B gives me an even chance of survival. Fuhrmanator 1, 8 8 silver badges 18 18 bronze badges. They are usually trivally easy to detect once you think of them" - but when you think of them, they're no longer unexpected. There is some question as to if your code if var! I can easily think of a bool that is neither true or false, but it's still unexpected.

In case you say a bool cannot be anything else, if I check whether a letter is a digit character and then convert it to its integer value, I can easily think of that integer value being less than 0 or greater than 9, but it's still unexpected. Null Booleans are supported in Java and C , and have a real-world application.

Consider a database containing a list of people. After a while you decide you need a gender isMale field. Null means "never asked so don't know"; true means male and false means female. OK, trans-gender omitted for simplicity Enumerations are better - can be extended when the need arises in your e.

Asexual, Hermaphrodite, "Refused to Answer" come to mind. Show 2 more comments. I dont get it in spacecrafts is dead code forbidden? I agree that the professor's code example is silly, but not as silly as mine.

David Andersson David Andersson 1 1 silver badge 2 2 bronze badges. The last if certainly be triggered, it wouldn't require really that much effort. Any experienced C programmer have seen values suddenly change. Of course logically, in a controlled single-threaded environment, this should never happen. In real life, the code inside the if will eventually happen.

If there's nothing useful you can do inside that if, then don't code it! I had a funny experience during a particular software development where I raised an exception with curse words in case something impossible happened This should not happen, but may, as a result of a previous error. In university case, you could even augment it by adopting a Design By Contract strategy: Establish invariants for classes eg, size is the number of items in the data list Establish pre-conditions and post-conditions for each function e.

Matthieu M. But the professor specifically said that it was Java, and specifically did NOT say what the type of var is. Well-written, helpful messages are crucial to a quality user experience. Poorly written messages result in low product satisfaction and are a leading cause of avoidable technical support costs. Exception handling needs to be made manageable across applications and for that, it is very important to standardize them into a common framework.

A standardized exception handling framework not only makes applications easier to maintain and debug but also results in cleaner and simpler code. Some of the ways in which it can be done can be.

Sometimes the best response to a serious runtime error is to just release all acquired resources and just abort the program. Let the user rerun the program with proper inputs.

It is really not worth the effort put in. Too much of exceptions creates problems of its own. If you start checking data passed as parameters in every conceivable way and in every conceivable place, your program will be fat and slow.

What is worse is that the additional code added for creating exceptions adds complexity to the program. The Cosmos HackAtom is here! Get feral when you answer to the greatest interview in history Share your philosophy.

A Bhat on September 10th reads. Good code is short, simple, and robust — the challenge is figuring out how to get there. Used imprudently they can wreak havoc on your code and make it impossible to follow and maintain. As a consequence, results can change between runs, even when provided with the same data and parameters. By its nature, this randomness renders strict reproducibility and, therefore, debugging more difficult.

If even the small test set 9 produces different results for each run, new users may not be able to tell whether or not the software is working properly. When comparing results between versions or after changing parameters, even small differences can confuse or muddy the comparison.

And especially when producing results for publications, grants, or diagnoses, any analysis should be absolutely reproducible. Given the size of biological data, it is unreasonable to suggest that random algorithms be removed. However, most programs use a pseudo-random number generator, which uses a starting seed and an equation to approximate random numbers.

Setting the seed to a consistent value can remove randomness between runs. Allow the user to optionally provide the random seed as an input parameter, thus rendering the program deterministic for those cases where it matters. If the seed is set internally e. If setting the seed is not possible, make sure the acceptable tolerance is known and detailed in documentation and in the tests.

There has been extended discussion over the past few years of the sustainability of research software, but this question is meaningless in isolation: any piece of software can be sustained if its users are willing to put in enough effort.

The real equation is the ratio between the skill and effort available and the ease with which software can be installed, understood, used, maintained, and extended. That said, not every coding effort needs to be engineered to last. Exploratory analysis is an iterative process that is developed quickly and revised often [ 4 , 11 ]. However, if a script is dusted off and run three or four times for slightly different purposes, is crucial to a publication or a lab, or is being passed on to someone else, it may be time to make your software more robust.

The authors thank the reviewers who provided feedback on the manuscript before and after submission. This work was partially funded by the Ontario Institute for Cancer Research. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

National Center for Biotechnology Information , U. PLoS Comput Biol. Published online Apr Author information Copyright and License information Disclaimer. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

This article has been cited by other articles in PMC. A checklist summarizing these ten simple rules to apply to your own software. Introduction Scientific software is typically developed and used by a single person, usually a graduate student or postdoc [ 1 ].

More specifically, we mean that: it can be installed on more than one computer with relative ease, it works consistently as advertised, and it can be integrated with other tools. Rule 1: Use version control Version control is essential to sustainable software development [ 17 , 18 ]. Rule 2: Document your code and usage How to write high-quality documentation has been described elsewhere [ 21 ], and so here, we only cover two minimal types: the README and usage.

List required dependencies. We address dependencies in more detail in Rule 5. Provide compilation or installation instructions. List a few example commands to get a user started quickly. List the most commonly used arguments, a description of each, and the default values. State where to find more information.

Rule 3: Make common operations easy to control Being able to change parameters on the fly to determine if and how they change the results is important as your software gains more users since it facilitates exploratory analysis and parameter sweeping.

Rule 4: Version your releases Software evolves over time, with developers adding or removing features as need dictates. Rule 5: Reuse software within reason In the spirit of code reuse and interoperability, developers often want to reuse software written by others.

Rule 6: Rely on build tools and package managers for installation To compile code, deploy applications, and automate other tasks, programmers routinely use build tools like Make, Rake, Maven, Ant, or MS Build.

Rule 9: Include a small test set that can be run to ensure the software is actually working Every package should come with a set of tests for users to run after installation.

Rule Produce identical results when given identical inputs The usage message tells users what the program could do. Conclusion There has been extended discussion over the past few years of the sustainability of research software, but this question is meaningless in isolation: any piece of software can be sustained if its users are willing to put in enough effort.

Supporting information S1 Checklist Robust software checklist. PDF Click here for additional data file. Acknowledgments The authors thank the reviewers who provided feedback on the manuscript before and after submission.

References 1. Toward effective software solutions for big biology. Nature Biotechnology. Baker M. A Survey of the Practice of Computational Science. In: State of the Practice Reports. Lawlor B, Walsh P. Engineering bioinformatics: building reliability, performance and productivity into bioinformatics software.

Tools and data services registry: a community effort to document bioinformatics resources. Nucleic Acids Research. A decade of web server updates at the bioinformatics links directory: — Genome Research. Seemann T. Ten recommendations for creating usable bioinformatics command line software. Nekrutenko A, Taylor J. Next-generation sequencing data interpretation: enhancing reproducibility and accessibility.

Nature ReviewsGenetics. PLoS Computational Biology. Sanders R, Kelly D. Dealing with Risk in Scientific Software Development. Software, IEEE. The Galaxy platform for accessible, reproducible and collaborative biomedical analyses: update. Howe B. Computing in Science Engineering. Bioconductor: open software development for computational biology and bioinformatics.