At 08:28 PM 4/9/2007, you wrote:

>Ditto for software. If you want software fast, expect low quality.

Not necessarily. That may be a good analysis for the state of RAD 
software now, but the one doesn't necessarily lead to the other.

Besides, the point being made isn't about speed so much as it is 
showstopper bugs getting into the releases. I mind slowness of 
startup of RB only when RB crashes on me continually and I have to 
restart it all the time.

>How does one measure the BQ "Bug Quotient" of a piece of software
>before it's released?

The company tests it! That's the way it's always been done. Although 
these days companies use public betas to help matters, and I don't 
have a problem with that. I think RS's system is a good one. The 
problem, though, is that too many showstoppers make it in the 
releases. That's what this thread is about. Ultimately that's RS's 
responsibility, not the public who have been invited to beta.

>Although steps have been taken in the field of 'software 
>correctness', I think early pioneers caught on when they realized 
>that there is NO way to conclusively prove that a software program 
>doesn't have any bugs.

That isn't good history. The "pioneers" had to debug their work 
conclusively. They didn't have an easy way to update software, so a 
bad release could be absolutely fatal. And sometimes it was.

 From what I read consistently, software in the 70-80-90's was 
tested-tested-tested and was expected to be as close to bug-free as 
possible. I never have read anything where any firm threw up their 
hands and said "software just doesn't work sometimes".

Remember the classic challenge of the Donald Knuth in TeX? He said 
that he'd pay $20 for every person that found a bug in his program, 
and he intended to double it every year the program existed.

Read "Writing Solid Code" (Microsoft Press) by Steve Maguire. The 
illusion that programs with any significant bugs should exist is 
shattered in his Introduction.

I don't think there's any doubt that companies have used the modern 
faster delivery systems to speed releases and back off of software 
quality. I hope that's not a off-the-cuff remark, but the temptation 
is incredible, and I see it given into numerous times, in my company 
and also in most that I work for.

But again, the point here is "show-stoppers that make it in releases".

All I know is that I use a 6-8 year old copy of Visual Basic and it 
crashes very seldom. I can honestly say it has no show-stoppers. I 
never ever have any inkling to even upgrade to VB6 for the reason of 
productivity. Not so with RB. RB isn't filled with them, they are 
very few, but they are upfront enough to cause serious concern. 
They've wasted several hours of my time already.

If they were user-error or had workarounds I'd be the first to use 
them and not complain (constructively, hopefully). But not these.

>Let's lighten up, people. Hopefully, now that RS is using RB to
>build RB, they can clear up a lot of the doodlypoop in 2007r2.

I actually would rather that they used something more low level. 
Using RB to do RB makes for a nice happy ending, but it makes for a 
rough ride in the meantime.

But speaking very pragmatically, when it gets to the point it makes 
absolutely no difference to me at all how RS makes RB. Certain 
programs are work programs, like a sound editor, a DAW (digital audio 
workstation) a word processor, a graphics program, or a IDE like RB. 
They are the foundation of peoples livelihood. They need to be 
rock-solid. Your band saw may be touchy, but not your screwdriver.

Features are nice, but in stuff like this the program has to be 
rock-solid and give the user confidence that it'll be up and running 
for 8 hours straight. RB is that type of program. And that's why it 
causes a stink when the "rock-solidness" is compromised, even in the 
smallest degree.

So, whatever it takes is fine with me, whether cow-dung is used in 
the memory-manager or it's leftover sandwiches. Reasonable speed for 
common functions, no crashing, and all basic functions work as advertised.

* * * * * * * * * * * * * * * * * * * * * * * * * * *
| Garth Hjelte                                      |
| Customer Service Representative, President        |
| Chicken Systems, Inc, Rubber Chicken Software Co. |
| 714 5th Street SE                                 |
| Willmar, MN 56201 USA                             |
|                                                   |
| 800-8-PRO-EPS    Toll Free Order Line (US Only)   |
| 320-235-9798     Tech Support, Sampler Questions  |
|                  International Line               |
| 360-838-7689     Fax                              |
| Product Sales:   [EMAIL PROTECTED]             |
| Product Support: [EMAIL PROTECTED]           |
| Sampler Q+A:     [EMAIL PROTECTED]                |
| Web Page:        www.chickensys.com               |
* * * * * * * * * * * * * * * * * * * * * * * * * * *

_______________________________________________
Unsubscribe or switch delivery mode:
<http://www.realsoftware.com/support/listmanager/>

Search the archives:
<http://support.realsoftware.com/listarchives/lists.html>

Reply via email to