[H-GEN] Is GnuCash ready for prime-time? Was: RFC: SCO

[Greg Black]

On 2003-06-17, Jason Parker-Burlingham wrote:
> Imagine being able to see something misbehave and *immediately*
> connecting a debugger and having the sources *right there* to refer
> to!  I'm pretty sure it'd be possible to do this with a BSD---Greg?
> Any opinions?

Well, it would be possible to do this with any system that you
chose to build entirely from sources -- but few people, other
than developers, would go that route.  It's too time-consuming
and requires levels of skill and persistence that most people
probably don't have.  In my case, I *never* build entire systems
from source unless somebody is paying me to do that.

I would not want general systems built with debugging symbols
and all the other space- and time-wasting flimflam that would
make this kind of thing work.

> In the particular case of leafnode, because I had to fetch, compile,
> and install the software to try to track down the problem, my window
> closed and to this day I don't know what was *really* going on.

If you really wanted to know, you could have setup two boxes
with their clocks set back to the time when the error happened
and repeated the test ;-)  After all, time-sensitive bugs are
common and this is basically the only way to track them.

My approach to software that crashes is simple enough.  If it's
repeatable (or if it repeats), then I assess its place in my
environment.  If it's an unimportant piece of junk (such as the
various browsers I use which all crash more or less at random)
and serves no critical purpose, I either abandon it altogether
if there's an alternative that doesn't crash, or I put up with
it.  If it's something that I need to have working (such as an
editor), I switch to something that is not broken if there's
such a thing that I can find.  Failing that, I'll consider
writing a replacement from scratch.  If that seems too big or
too boring, I'll do some very simple work to see if it's a bug
that I can replicate enough to make it worth putting in a very
small amount of time chasing.

Sometimes, that effort will get me interested in the software
and I'll join the team, at least as an official tester, and help
to fix the stuff I find.  I've been on some free software teams
for over ten years, in that capacity.  Other times, I'll be so
put off by the quality of the code that I'll decide not to touch
the thing ever again.

When I am seriously testing code, whether it's mine or somebody
else's, I always work with all the debugging gloop in place; but
I also make a point of doing parallel testing with a version
that is built the way it should be distributed -- sometimes the
presence of debugging code masks bugs and they still need to be
tracked down.

Finally, I should say that I *loathe* debugging and much prefer
to write clean and correct code in the first place ...

Greg

-- 
Greg Black <gjb at gbch.net> <http://www.gbch.net/gjb.html>
"If you're not living life on the edge, you're taking up
too much space."
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 249 bytes
Desc: not available
URL: <http://lists.humbug.org.au/pipermail/general/attachments/20030618/228aa6bd/attachment.sig>