Home > Archive > Software Testing > May 2007 > Re: Improving the quality of software/software quality - Measurement, Standards, Deve
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
| Author |
Re: Improving the quality of software/software quality - Measurement, Standards, Deve
|
|
| Vladimir Trushkin 2007-05-17, 4:29 am |
| On May 16, 9:46 pm, Game <dal...@yahoo.co.uk> wrote:
> Can quality of software be improved by any of the following:
>
> 1) Standards
> 2) Measurement
> 3) Development models (eg. spiral, waterfall, incremental to name a
> few)
> 4) Formal methods
> 5) Testing
>
> Personally I believe they do to some extent as they are all essential
> to SQA but would like any open-ended views.
>
> Also believe testing to a lesser extent than the others because it can
> come after a certain stage of a project has been completed, therefore
> cannot always be improved.
This is too general a question yet [with some assumptions in mind] I
will try to answer.
1). Standards limit the things making them more similar. In some cases
standards contribute quality by specifying more safe and reliable
constructions for coding. Coding standards also contribute the quality
indirectly making code review easier.
2). Measurements are vital for an organization that set quality goals.
The most popular metrics for the product quality are customer-oriented
and represent both validity of the concept and reliability in use.
Those of you who managed quality programs know that what we actually
gain is not always what the perceived benefit of an improvement
program was. Measuring the results allows seeing in which direction
the organization moves, toward the better quality or opposite.
3). Development models are just the way people write the code. Clear
process steps and knowing what to expect from different teams in the
organization and when is much better than a complete chaos when
problems appear at the very end and it takes heroic efforts and much
of overtime to get the project back on track. In my opinion, an
organization can produce a better quality than a chaos. Programming
under high pressure or even in the late night was never resulting in a
good quality. On my memory it was always terrible.
4). Formal methods -- I am not sure I understand what you mean.
5). Testing is well known to be a quality control, a measurement tool.
Yet, the defects it finds are getting fixed, thus providing a better
quality of a product. You have mentioned it correctly that it is
better to think it is too late to find defects.
----
Best Wishes,
Vladimir
| |
|
|
|
|
|