BorlandTalk.com Forum Index BorlandTalk.com
Borland discussion newsgroups
 
Archives   FAQFAQ   SearchSearch   MemberlistMemberlist   UsergroupsUsergroups   RegisterRegister 
 ProfileProfile   Log in to check your private messagesLog in to check your private messages   Log inLog in 

Is software reliability important?
Goto page 1, 2  Next
 
Post new topic   Reply to topic    BorlandTalk.com Forum Index -> Delphi Non-Technical
View previous topic :: View next topic  
Author Message
Chris Burrows
Guest





PostPosted: Sat Aug 16, 2003 1:54 pm    Post subject: Is software reliability important? Reply with quote



Recent discussions in these groups on the quality and reliability of the
current version of Delphi and the lack of bug-fixes include views implying
that we don't need to worry to much about bugs as users expect them anyway.
This attitude needs to be challenged. For a start, a thought-provoking
interview with Niklaus Wirth on the subject of software reliability can be
found at:

http://www.eptacom.net/pubblicazioni/pub_eng/wirth.html

The following are pertinent extracts

"Good enough software" is rarely good enough"

"In our profession, precision and perfection are not a dispensible luxury,
but a simple necessity."

"Software development is technical activity conducted by human beings. It is
no secret that human beings suffer from imperfection, limited reliability,
and impatience - among other things. Add to it that they have become
demanding, which leads to the request for rapid, high performance in return
for the requested high salaries. Work under constant time pressure, however,
results in unsatisfactory, faulty products. Generally, the hope is that
corrections will not only be easy, because software is immaterial, but that
the custormers will be willing to share the cost."

"I am convinced that there is a need for high quality software, and the time
will come when it will be recognized that it is worth investing effort in
its development and in using a careful, structured approach based on safe,
structured languages. For the time being, however, this is practicable only
for some "niche products". The niche will expand when a growing number of
people will voice their dissatisfaction with poor software."

Chris Burrows
CFB Software
http://www.cfbsoftware.com





Back to top
Ingvar Nilsen
Guest





PostPosted: Sat Aug 16, 2003 2:43 pm    Post subject: Re: Is software reliability important? Reply with quote



Chris Burrows wrote:

Quote:
"I am convinced that there is a need for high quality software, and the time
will come when it will be recognized that it is worth investing effort in
its development and in using a careful, structured approach based on safe,
structured languages.

Yeah, fine, time to dust off some old compilers Smile
Wonder if he knows what a "structured language" is..

--
Ingvar Nilsen


Back to top
Chris Burrows
Guest





PostPosted: Sat Aug 16, 2003 2:53 pm    Post subject: Re: Is software reliability important? Reply with quote



"Martin Lafferty" <martinl (AT) prel (DOT) co.uk> wrote

Quote:

"However, we know that it is better if each basic concept is represented
by
a single, designated language construct. Not only does this reduce the
effort of learning, but it reduces the volume of the language's
description
and thereby the possibilities of inconsistency and of misunderstanding. "

This has always seemed completely true to me, and yet languages become
every
more complex with ever more ways of expressing the same basic concepts.
When
I read Larry Wall's PERL book I got fed up with the constant mantra
"There's
more than one way to do it" because to me, this was an admission of
failure.


Exactly. I had my first experience with PERL earlier this year and choked
when I saw that statement.

Chris Burrows
CFB Software
http://www.cfbsoftware.com






Back to top
Andrew Gabb
Guest





PostPosted: Sat Aug 16, 2003 3:59 pm    Post subject: Re: Is software reliability important? Reply with quote

Chris Burrows wrote:
Quote:
Recent discussions in these groups on the quality and reliability of the
current version of Delphi and the lack of bug-fixes include views implying
that we don't need to worry to much about bugs as users expect them anyway.
This attitude needs to be challenged. For a start, a thought-provoking
interview with Niklaus Wirth on the subject of software reliability can be
found at:

Wirth was an academic, albeit a relatively smart one back in the 70s
(some were and are abysmally stupid). As he progressed, he moved
further and further away from the real world. So I certainly
wouldn't start there.

It's impossible to develop any large complex system with zero
defects. So it all becomes a tradeoff of what your users expect and
can live with, and how quickly and cost-effectively you can deliver
it. Microsoft couldn't possibly meet their marketing-driven
objectives with 6-sigma quality, and they've proved that they don't
need to.

And reliability is only one quality factor. What about usability for
example? Or what the software *does* when an defect strikes
(survivability?).

It all depends on your market and how they might react to quality
problems. I can't afford to write 6-sigma software either - my
clients won't pay for it. So it becomes a matter of compromise -
fortunately my clients have been somewhat tempered by Microsoft and
the unreliability of software on the web in general.

I'm about to swap from Netscape to Outlook/IE (dammit!) because
Netscape/Mozilla has become too fragile (*my* assessment) and has
dropped functionality that I seriously need (and which Outlook has).
The future also looks bleak with open-source, so I vote with my
feet. I trade off the swap with the embuggerance it might cause me.

My final analogy is something I call Terminal Hotel Disease (or
'Ainslee Disease' named after a specific hotel in Canberra), which
is a common problem world-wide. I used to travel a lot and stay in
3* hotels/motels. Quite often I'd find that the rooms in a hotel
were getting crummier and crummier as the management cut back on
maintenance and cleaning. Eventually they went broke and sold out at
a loss. 'Health centres' start selling life memberships cheap (and
you know how much *they* are likely to be worth). Travel agencies
sell cutprice travel with big discounts for payments long in advance.

Going down.

Andrew
--
Andrew Gabb
email: [email]agabb (AT) tpgi (DOT) com.au[/email] Adelaide, South Australia
phone: +61 8 8342-1021, fax: +61 8 8269-3280
-----


Back to top
Jim Cooper
Guest





PostPosted: Sat Aug 16, 2003 4:51 pm    Post subject: Re: Is software reliability important? Reply with quote


Quote:
Recent discussions in these groups on the quality and reliability of the
current version of Delphi and the lack of bug-fixes include views implying
that we don't need to worry to much about bugs as users expect them anyway.

I think that's an implication too far Smile I doubt you'd find anyone
suggesting fewer bugs would be better. However it is certainly
unrealistic to expect zero bugs in any complex piece of software.
Correctness proofs are just not viable for large pieces of software, and
to my knowledge have only ever been done in certain restricted
circumstances.

How many and which bugs are "allowable" in a released application is a
debate with little point, IMO, as experience and opinions vary too much.
Eg I have zero problems with D7 bugs, whilst others apparently find it
completely unusable (with all shades in between, of course).

Quote:
For a start, a thought-provoking interview with Niklaus Wirth

I found some of the opinions a little naive, but that could be because
it's an old interview now.

Quote:
"Good enough software" is rarely good enough"

Demonstrably false. You are running an operating system right now that
is "good enough" (and I can say that without even knowing which one it
is). And there is at least one other dev tool we both use that falls
into that category :-)

Quote:
"In our profession, precision and perfection are not a dispensible luxury,
but a simple necessity."

If he was talking about medicine then he might have a better point Smile
Perfection is a noble goal, but one we rarely, if ever, achieve (I'm
obviously not talking about myself here <vbg>).

Cheers,
Jim Cooper

____________________________________________

Jim Cooper [email]jcooper (AT) tabdee (DOT) ltd.uk[/email]
Tabdee Ltd http://www.tabdee.ltd.uk

TurboSync - Connecting Delphi with your Palm
____________________________________________

Back to top
Alessandro Federici
Guest





PostPosted: Sat Aug 16, 2003 5:18 pm    Post subject: Re: Is software reliability important? Reply with quote

"Chris Burrows" <info (AT) cfbsoftware (DOT) com.au> wrote

Quote:
Recent discussions in these groups on the quality and reliability of the
current version of Delphi and the lack of bug-fixes include views implying
that we don't need to worry to much about bugs as users expect them
anyway.


Right, but the issue is that even if they might expect them and be tolerant
with them for a week or two, it doesn't mean they shouldn't get patched in
the current version.

Quote:
"In our profession, precision and perfection are not a dispensible luxury,
but a simple necessity."


That actually depends on what you consider more important: deliver quality
or make money. They are both important of course, but how you prioritize
them has immediate results in the bug-fixing process.

Quote:
[..] Work under constant time pressure, however,
results in unsatisfactory, faulty products. [..]

<speculation>
Which is where it all boils down to IMO, expecially in the RAD division at
Borland, where the same team has to deal with 4 products (or any close
variant to this).
</speculation>

Quote:
For the time being, however, this is practicable only
for some "niche products". The niche will expand when a growing number of
people will voice their dissatisfaction with poor software."

I hope so Wink)) but let me add something to this: why is it that people find
ok to blame Borland/Microsoft/whatever for their bugs but never mention they
could have found them by just doing a little research before (i.e. using the
downloadable trials)? Isn't the fault on both sides? There's plenty of
options in our market. Why is the sloppiness of customers starting to write
code before thinking ok, and Borland/Microsoft/whatever bug's aren't?
Just wondering...


--
Regards,
Alessandro Federici

RemObjects Software, Inc.
http://www.remobjects.com



Back to top
Ingvar Nilsen
Guest





PostPosted: Sat Aug 16, 2003 5:32 pm    Post subject: Re: Is software reliability important? Reply with quote

Alessandro Federici wrote:

Quote:
That actually depends on what you consider more important: deliver
quality or make money. They are both important of course, but how
you prioritize them has immediate results in the bug-fixing process.
This is an important point, very important...!!!!


I can imagine various views, attitudes regarding this and imagine tough
negotiations in certain companies <g> about this.
At the end of the day, lack of fixes and an attempt to "milk" customers
will backfire, "at the end of the day".

--
Ingvar Nilsen


Back to top
Alessandro Federici
Guest





PostPosted: Sat Aug 16, 2003 5:47 pm    Post subject: Re: Is software reliability important? Reply with quote

"Ingvar Nilsen" <telcontr@[Remove-This-Part]online.no> wrote


Quote:
At the end of the day, lack of fixes and an attempt to "milk" customers
will backfire, "at the end of the day".

That's what I've always been saying when arguing that saying "Dx outsold
Dx-1" is really shortsighted Wink but hey, what do I know? <G>



Back to top
Kristofer Skaug
Guest





PostPosted: Sat Aug 16, 2003 9:35 pm    Post subject: Re: Is software reliability important? Reply with quote

"Alessandro Federici" wrote
Quote:
Right, but the issue is that even if they might expect them and be
tolerant with them for a week or two, it doesn't mean they shouldn't
get patched in the current version.

BINGO!

Quote:
That actually depends on what you consider more important: deliver quality
or make money. They are both important of course, but how you prioritize
them has immediate results in the bug-fixing process.

There's another aspect to this, which I think we all know from our own
professional practice:
Knowingly shipping defective software (and subsequently not being allowed
time to fix it) has a negative influence on morale on the part of the
developers. From DeMarco&Lister, "Peopleware":

<quote>
[M]anagers tend to think of quality as just another attribute of the
product, something that may be supplied in varying degrees according to the
needs of the marketplace (...). The builders [on the other hand] tend to
impose quality standards of their own. The minimum that will satisfy them is
more or less the best quality they have achieved in the past. This is
invariably a higher standard than what the market requires and is willing to
pay for. (...)
</quote>

When the profitability of each release is always the highest goal,
motivation and quality consciousness within an organisation will diminish
over time, it becomes a negative spiral. It is only to be hoped that this is
not a manifest problem at Borland.

Quote:
ok to blame Borland/Microsoft/whatever for their bugs but never mention
they
could have found them by just doing a little research before (i.e. using
the
downloadable trials)?

I actually did this. Downloaded D7 Architect trial and ran my core library
test suites on it.
This resulted in the detection of a quite embarrassing compiler bug, new in
D7, which I reported in QC#2258.
At that time (September 2002), I was 100% convinced that Borland would
scramble to fix it in a patch within a month or two, max.
And I silently thought that I'd buy D7, just to be "up to date" (and in
spite of the relative lack of new features), as soon as the patch was out
and verified. It never came, so I didn't buy D7.

Kristofer



Back to top
Ixtlan
Guest





PostPosted: Sat Aug 16, 2003 10:39 pm    Post subject: Re: Is software reliability important? Reply with quote


"Andrew Gabb" <agabb (AT) tpgi (DOT) com.au> wrote


Quote:
The future also looks bleak with open-source, so I vote with my
feet. I trade off the swap with the embuggerance it might cause me.

Where did you get this impression from, that the future looks bleak with
open-source????

Just wondering why.

As far as I'm concerned, 5 years ago I shared your bleak outlook on OS, but
today, nearly all my software development is based on OS, and to me the
future of OS looks very bright indeed!

Ixtlan



Back to top
Dennis Landi
Guest





PostPosted: Sun Aug 17, 2003 4:57 am    Post subject: Re: Is software reliability important? Reply with quote


"Jim Cooper" <jcooper (AT) tabdee (DOT) ltd.uk> wrote


Quote:
"In our profession, precision and perfection are not a dispensible
luxury,
but a simple necessity."

If he was talking about medicine then he might have a better point Smile
Perfection is a noble goal, but one we rarely, if ever, achieve (I'm
obviously not talking about myself here <vbg>).


I agree completely. Most of us strive for zero defects. But lets face it,
most of use are not writing software for NASA's space shuttle systems...
Much software has an intentionally short-life span. And even for more
substantial systems that are expensive to build and intended to pay for
itself over time (up-time), critical bugs can be identified and (drum roll,
please) FIXED.

QA - TESTING, is a fact of life, and part of any serious development life
cycle. QA is predicated upon the acknowledgement that bugs exist, period.
That acknowledgement has become a corner-stone of modern software
development. The question isn't whether or not bugs SHOULD exist. That is
just naive in the extreme. The question is: what mechanism is in place to
catch the bugs as efficiently as one's resources allow. But even after the
system has been deployed and gone live, bugs can still be intelligently
tracked and addressed. (obviously) From a *theory* stand-point this is a
non-issue...

-dennis



Back to top
Ingvar Nilsen
Guest





PostPosted: Sun Aug 17, 2003 7:39 am    Post subject: Re: Is software reliability important? Reply with quote

Kristofer Skaug wrote:

Quote:
When the profitability of each release is always the highest goal,
motivation
and quality consciousness within an organisation will diminish over
time, it becomes a negative spiral. It is only to be hoped that
this is not a manifest problem at Borland.

Well said! I have made myself the same thoghts.
BTW, what you point your finger at is a well known fact in most software
companies.

--
Ingvar Nilsen


Back to top
Chris Burrows
Guest





PostPosted: Mon Aug 18, 2003 2:39 am    Post subject: Re: Is software reliability important? Reply with quote

"Jim Cooper" <jcooper (AT) tabdee (DOT) ltd.uk>

Quote:
I found some of the opinions a little naive, but that could be because
it's an old interview now.


1997? Old?

A more recent, less academic article, discussing the attitudes that concern
me (expectation / tolerance of bugs, lack of interest in fixing bugs, bugs
are only an inconvenience not a real problem, bugs are inevitable etc. etc.)
is:

http://www.cbsnews.com/stories/2003/04/29/tech/main551492.shtml

The last thing I want to happen is for lawyers to get in on the act.
However, I believe it might be only when software manufacturers face the
threat of being sued for negligence or whatever, that they will start to
take software quality issues seriously. It is up to us, as professional
developers, to do what we can to ensure this doesn't need to happen.

Chris Burrows
CFB Software
http://www.cfbsoftware.com




Back to top
Kristofer Skaug
Guest





PostPosted: Mon Aug 18, 2003 7:14 am    Post subject: Re: Is software reliability important? Reply with quote

"Chris Burrows" wrote
Quote:

1997? Old?

well, a whole epidemic of dotcom-quality software production philosophy
has washed over us and gone again in the meantime.

Quote:
It is up to us, as professional developers, to do what
we can to ensure this doesn't need to happen.

True. If there are minimum quality (safety, reliability) standards
imposed on the production of cars (to get "road qualified"), airplanes,
ships, toasters, shower gel bottles and video recorders, why shouldn't
there be a minimum quality standard for software?

In my own workplace we spend far more time testing and certifying the
custom hardware that our software is running on/with, than testing the
software itself ( - which is patchable). Now there's admittedly fairly
limited opportunity for 3-year-olds to get their fingers jammed in our
software, or for airheads to use it as an experimental pet-hairdryer, but
OTOH our software *does* run the risk of damaging upto half a billion
Euro's worth of satellite hardware with one well-timed crash. So... yeah
I've been wondering why oh why

Kristofer



Back to top
John Elrick
Guest





PostPosted: Tue Aug 19, 2003 9:50 pm    Post subject: Re: Is software reliability important? Reply with quote


"Chris Burrows" <info (AT) cfbsoftware (DOT) com.au> wrote

Quote:
Recent discussions in these groups on the quality and reliability of the
current version of Delphi and the lack of bug-fixes include views implying
that we don't need to worry to much about bugs as users expect them
anyway.
This attitude needs to be challenged. For a start, a thought-provoking
interview with Niklaus Wirth on the subject of software reliability can be
found at:

http://www.eptacom.net/pubblicazioni/pub_eng/wirth.html

The following are pertinent extracts

"Good enough software" is rarely good enough"

Wirth writes software...he doesn't buy it.

The average buyer looks for:

The first solution on his/her radar that solves a relevant business problem
and is reasonably stable for a reasonable cost.

That is the reality of the marketplace.



John



Back to top
Display posts from previous:   
Post new topic   Reply to topic    BorlandTalk.com Forum Index -> Delphi Non-Technical All times are GMT
Goto page 1, 2  Next
Page 1 of 2

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum


Powered by phpBB © 2001, 2006 phpBB Group
SEO toolkit © 2004-2006 webmedic.