Hanania Salzer,
>I'm responding only to you, not to the group, because my comment is not
>related to he original question. My note is a pure, industry oriented,
>software developer's observation. (However, I don't mind if you find it
>suitable to respond to the group.)
Your observations are probably the two most important factors in
developers decision making. I think it would be suitable for a response
to the group.
>One of the considerations a developer must have is testing. Even if the
>developer wants to limit his or her testing to the minimum, they would have
>to test at least every change and addition. An absolute minimum is testing
>every statement. An industry minimum is testing every branch.
[deleted material]
In some industries the minimum is testing every branch. In most it
is not even every statement :-(. But this is a separate issue.
The problem with testing is that it takes a long time and is very expensive
Just running tests and checking the results can take days, if not weeks.
My experience is that lots of testing goes into the initial development.
But in maintenance mode management keeps trimming costs and testing
is an easy target.
Moving onto your next point, before we consider them together.
[deleted material]
> An example of such standard is
>the maximum number of nesting in an IF statement. Once a standard is in
>place, the developers' decision is not dependent only on selecting an
>approach promising less labor, but also selecting one that complies with the
>standard.
There are certainly guidelines that talk about nesting constructs. Leaving
aside the issue of whether they are of benefit, lets take this case as an
example.
Lets say a developer is modifying code and finds that some nesting limit
specified in the guidelines is exceeded. Given the resources they would
reorganised the code. But it is not just a matter of editing the sources,
what
about rerunning all those tests to make sure nothing has been broken in the
process?
Maintenance people are well aware of the issues. If they make a big change
they may have to run a large test base, this could be expensive. If they make
the minimum changes necessary to add the new functionality and ignore the
guideline violation they may not have to run the tests.
Payne et al say that when faced with conflicting choices that have strong
emotional attachment (which there is likely to be in these cases), people
ignore these choices and switch the criteria used to make the decision to
other issues.
It would make for an interesting experiment to give subjects a program that
needs to be modified. But whose modification involves selecting between
options that developers have strong emotional attachment to. Would they
base their choice on the real issues, or would they compare some secondary
issues when making their decision? I suspect that many developers would
avoid the emotionally charged issues.
In the testing/guideline case a common refrain is "management made me do it".
Management is probably completely unaware of the choice made on its behalf.
derek
--
Derek M Jones tel: +44 (0) 1252
520 667
Knowledge Software Ltd mailto:[EMAIL PROTECTED]
Applications Standards Conformance Testing http://www.knosof.co.uk
- Automatic footer for [EMAIL PROTECTED] ----------------------------------
To unsubscribe from this list, mail [EMAIL PROTECTED] unsubscribe discuss
To join the announcements list, mail [EMAIL PROTECTED] subscribe announce
To receive a help file, mail [EMAIL PROTECTED] help
This list is archived at http://www.mail-archive.com/discuss%40ppig.org/
If you have any problems or questions, please mail [EMAIL PROTECTED]
