home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.software-eng
- Path: sparky!uunet!caen!jxm
- From: jxm@engin.umich.edu (John Murray)
- Subject: Re: Just how sophisticated should we assume progrr & her/his tools to be?
- Message-ID: <KYV=-5+@engin.umich.edu>
- Date: Fri, 18 Dec 92 20:06:39 EST
- Organization: University of Michigan Engineering, Ann Arbor
- References: <1992Dec16.004059.15927@news.arc.nasa.gov>
- Nntp-Posting-Host: schroedinger.engin.umich.edu
- Lines: 42
-
- In article <1992Dec16.004059.15927@news.arc.nasa.gov> lehman@george.arc.nasa.gov (John Lehman -- GDP) writes:
- >
- >>> [torturous discussion about the use of 'i' and 'j' as loop indices omitted]
- >
- >>I'm glad someone else uses the 'ii' technique. Means that there's some
- >>code I may have to maintain someday that doesn't use 'i' <gag>.
- >
- >So, there were at least 3 of us who were using "ii". I'm not
- >quite sure how sophisticated -- e.g., \<i\> -- I want to be;
- >but lately I've been using "i" because it looks elegant (the i of
- >the beholder), being the shortest possible name -- one less keystroke,
- >times # variables times # modules times # programmers. But I'm
- >sentimental about "ii".
-
- 'ii' and 'jj' are neat substitutes which I hadn't come across before.
- However, the above discussion still provokes some thoughts (flames?):
-
- - Compared to the overall time required to design, write, test, debug,
- document, AND MAINTAIN software, the need to save individual keystroke
- time is minimal. Let's get real here - we're not (I hope) dealing with
- Taylorist efficiency of huge sweat-shops full of keypunch operators
- in this discussion!
-
- - If a loop count is used SO extensively (or a loop is so long) that
- an editor's search function has to be used to locate all instances of
- it, then it's more than just a humble loop count. There's probably some
- real (virtual?) world entity or concept it represents, and consequently
- it should be appropriately named.
-
- - Repeated use of the same variable name leads to unnecessary confusion
- in data dictionaries, variable cross-reference systems, and so on. Many
- instances of database mix-ups between people with the same names or
- birthdays are well-documented; why should we so arrogant as to think
- that variable names are somehow exempt from this phenomenon?
-
- - The assumption that those who have to maintain one's carefully crafted
- code will use (or even have access to) the vi editor, or any other
- given tool, is flawed. Unfortunately, it smacks of elitism and perhaps
- a professional desire to keep out the 'ill-informed'. [For precedents,
- look at the legal and medical professions.]
-
- John Murray
-