Software design and development methods?

Software design and development methods?

[mjsilva]

This isn't really an Ada question, but I think I'm likely to get more
good info and less nonsense asking it here than in almost any other
group.  My question is how does an individual, or a small company with
limited resources, improve the software design and development process
beyond an ad-hoc collection of self-discovered "methodologies".  What,
for example, are the -right- ways to do a requirements analysis, create
a functional specification and come up with a program framework?  What
is the right way to create a test plan?

The goal is to come up with a consistent way to develop "better"
(including but not limited to more reliable) software, including
embedded applications, quicker.  These applications are not in the
safety-critical arena, BTW.

Any and all guidance is greatly appreciated!

Mike
mjsilva@jps.net


Sent via Deja.com http://www.deja.com/
Before you buy.

Re: Software design and development methods?

[Jerry Petrey]

mjsilva@my-deja.com wrote:
> 
> This isn't really an Ada question, but I think I'm likely to get more
> good info and less nonsense asking it here than in almost any other
> group.  My question is how does an individual, or a small company with
> limited resources, improve the software design and development process
> beyond an ad-hoc collection of self-discovered "methodologies".  What,
> for example, are the -right- ways to do a requirements analysis, create
> a functional specification and come up with a program framework?  What
> is the right way to create a test plan?
> 
> The goal is to come up with a consistent way to develop "better"
> (including but not limited to more reliable) software, including
> embedded applications, quicker.  These applications are not in the
> safety-critical arena, BTW.
> 
> Any and all guidance is greatly appreciated!
> 
> Mike
> mjsilva@jps.net


From what I've seen, most 'big' companies haven't figured this out yet,
although, on paper, they claim to have.

In any case, a good start would be reading the book "The Capability
Maturity Model : Guidelines for Improving the Software Process (Sei
Series in Software Engineering)", ISBN: 0201546647 .

Also look at "Software Process Improvement : Practical Guidelines for
Business Success 
(Sei Series in Software Engineering)", ISBN: 020117782X , "Managing the
Software Process", ISBN: 0201180952 ,
"A Discipline for Software Engineering (Sei Series in Software
Engineering)",  ISBN: 0201546108 and
"Managing Risk: Methods for Software Systems Development (SEI Series in
Software Engineering)", ISBN: 0201255928 .

Hope this helps.

Jerry

-- 
-----------------------------------------------------------------------------
-- Jerry Petrey                                                
-- Senior Principal Systems Engineer - Member Team Ada & Team Forth
-- Raytheon Missile Systems 
-- NOTE: please remove NOSPAM in email address to reply                  
-----------------------------------------------------------------------------

Re: Software design and development methods?

[Brian Rogoff]

On Tue, 10 Oct 2000 mjsilva@my-deja.com wrote:
> This isn't really an Ada question, but I think I'm likely to get more
> good info and less nonsense asking it here than in almost any other
> group.  My question is how does an individual, or a small company with
> limited resources, improve the software design and development process
> beyond an ad-hoc collection of self-discovered "methodologies".  What,
> for example, are the -right- ways to do a requirements analysis, create
> a functional specification and come up with a program framework?  What
> is the right way to create a test plan?
> 
> The goal is to come up with a consistent way to develop "better"
> (including but not limited to more reliable) software, including
> embedded applications, quicker.  These applications are not in the
> safety-critical arena, BTW.
> 
> Any and all guidance is greatly appreciated!

Have you looked at "Extreme Programming"? There seem to be some Ada
precursors to XP, see 

	http://stsc.hill.af.mil/CrossTalk/1996/jul/manageme.asp?

Speaking of XP, I thought Ed Falis was going to release his Ada unit test
framework. What gives Ed?

-- Brian

Re: Software design and development methods?

[Ehud Lamm]

PSP and TSP are from SEI, and more info can be found on ther site
(http://www.sei.cmu.edu).
Alas, on this topic I found much less info than on the CMM.
I'd be glad to find more in depth sources on the Net.

--
Ehud Lamm   mslamm@mscc.huji.ac.il
http://purl.oclc.org/NET/ehudlamm <==  Me!



Ken Garlington <Ken.Garlington@computer.org> wrote in message
news:ZbSE5.1189$ln6.160997@news.flash.net...
> <mjsilva@my-deja.com> wrote in message news:8rvov4$imr$1@nnrp1.deja.com...
> > This isn't really an Ada question, but I think I'm likely to get more
> > good info and less nonsense asking it here than in almost any other
> > group.  My question is how does an individual, or a small company with
> > limited resources, improve the software design and development process
> > beyond an ad-hoc collection of self-discovered "methodologies".  What,
> > for example, are the -right- ways to do a requirements analysis, create
> > a functional specification and come up with a program framework?  What
> > is the right way to create a test plan?
>
> For an individual, consider the Personal Software Process (PSP). For a
small
> company, consider the Team Software Process (TSP). These won't really
answer
> all your questions, but they will give you an approach to improve the
> answers.
>
>

Re: Software design and development methods?

[Jeff Carter]

mjsilva@my-deja.com wrote:
> 
> This isn't really an Ada question, but I think I'm likely to get more
> good info and less nonsense asking it here than in almost any other
> group.  My question is how does an individual, or a small company with
> limited resources, improve the software design and development process
> beyond an ad-hoc collection of self-discovered "methodologies".  What,
> for example, are the -right- ways to do a requirements analysis, create
> a functional specification and come up with a program framework?  What
> is the right way to create a test plan?

There is NO right way to create a functional specification. It is a
truth universally accepted that you Must be Object Oriented, or you're
No Good.

Serially, for an individual or small company, you might want to look
into Personal Process Improvement from SEI.

-- 
Jeff Carter
"Perfidious English mouse-dropping hoarders."
Monty Python & the Holy Grail

Re: Software design and development methods?

[Ken Garlington]

<mjsilva@my-deja.com> wrote in message news:8rvov4$imr$1@nnrp1.deja.com...
> This isn't really an Ada question, but I think I'm likely to get more
> good info and less nonsense asking it here than in almost any other
> group.  My question is how does an individual, or a small company with
> limited resources, improve the software design and development process
> beyond an ad-hoc collection of self-discovered "methodologies".  What,
> for example, are the -right- ways to do a requirements analysis, create
> a functional specification and come up with a program framework?  What
> is the right way to create a test plan?

For an individual, consider the Personal Software Process (PSP). For a small
company, consider the Team Software Process (TSP). These won't really answer
all your questions, but they will give you an approach to improve the
answers.

Re: Software design and development methods?

[mjsilva]

In article <Pine.BSF.4.21.0010111609420.24789-
100000@shell5.ba.best.com>,
  Brian Rogoff <bpr@shell5.ba.best.com> wrote:
> On Tue, 10 Oct 2000 mjsilva@my-deja.com wrote:
> > This isn't really an Ada question, but I think I'm likely to get
more
> > good info and less nonsense asking it here than in almost any other
> > group.  My question is how does an individual, or a small company
with
> > limited resources, improve the software design and development
process
> > beyond an ad-hoc collection of self-discovered "methodologies".
What,
> > for example, are the -right- ways to do a requirements analysis,
create
> > a functional specification and come up with a program framework?
What
> > is the right way to create a test plan?
> >
> > The goal is to come up with a consistent way to develop "better"
> > (including but not limited to more reliable) software, including
> > embedded applications, quicker.  These applications are not in the
> > safety-critical arena, BTW.
> >
> > Any and all guidance is greatly appreciated!
>
> Have you looked at "Extreme Programming"? There seem to be some Ada
> precursors to XP, see
>
> 	http://stsc.hill.af.mil/CrossTalk/1996/jul/manageme.asp?

No I haven't, mostly because of the silly name and the hype (I'm just a
giant inductor in the circuit of what's-hot-today...).  I'll take a
look at this URL, as well as the books and the SEI material.  Thanks to
all for the recommendations.

Mike


Sent via Deja.com http://www.deja.com/
Before you buy.

Re: Software design and development methods?

[Jubilation]

If you need a quick jumpstart, read Software Process Survival Guide by Steve
McConnell. His book "Rapid Development" is a more detailed analysis of
different process aspects.

But you won't find a process from these books. A lot of useful information,
but not a cookbook approach for sure. And if you want to know how much
people disagree about what is the right way, get a book about Cleanroom or
PSP and another about XP.

Kari

--
Please drop the 1 away from my email address if you reply

Re: Software design and development methods?

[Al Christians]

See:

http://www.lyee.co.jp/en/docs/index.htm


This one describes a new set of methods from Japan.  There is a
report of a very successful implementation of a 2 million line system.
Unfortunately, these documents are very hard to understand in
concrete terms, being perhaps obfuscated in translation from 
Japanese.  Any plane-English info on Lyee available?


Al

Re: Software design and development methods?

[Ed Falis]

Brian Rogoff wrote:


> Speaking of XP, I thought Ed Falis was going to release his Ada unit
> test
> framework. What gives Ed?
>
> -- Brian
>
>

"Real soon now".  Just ironing out some details about gettting it
checked in and such.

Sorry about the delayed response - can't write to newsgroups while I'm
on the road.

- Ed