Jacqueline McNally schrieb:
Hello
I would like some feedback on the following comment.
"However, while there are plenty of proponents, the open-source movement
does have legitimate detractors. One is Michael Goulde, analyst of
software infrastructureRelevant Products/Services from BMC Software for
Forrester Research. According to him, the coding for open-source
products often provides a built-in drawback. By way of example, he
points to OpenOffice. "The actual code for the first version was
spaghetti," he said. "The code for 2.0 isn't much better. This is going
to be a bear to continue to evolve, and they should probably start from
scratch. It's no accident that in Forrester surveys OpenOffice usage
barely shows up.""
From: http://www.newsfactor.com/story.xhtml?story_id=10200002DYPI
But what code is Goulde looking at to write-off 8 millions lines of code
as spaghetti?
Please can you tell me that if in your opinion the OpenOffice.org is
spaghetti, if not the whole, which parts, and why?
As others have written: if you first look at an unknown project of
several million lines of code, the structure of which you don't know
yet, you'll see only spaghetti.
It also is the case that the code of OOo has evolved over some 15 years,
most of it as closed source. In this time programming techniques,
particularly for C++, progressed a lot.
As a result, you will find areas done at different times and by
differently skilled or trained people that exhibit code of varying quality.
There are some areas that look relatively 'spaghetti'. We know about
them, because they are harder to change and prone to cause regressions.
And we are incrementally cleaning them up. But generally just rewriting
from scratch doesn't make everything better automatically, because you
lose the accumulated bug fixes and special cases that were needed to
make the code work in the real world.
And if you see how OOo evolved from 1.x to 2.0, you have living proof
that our code is not all that bad and unmaintainable.
BTW, most of the spaghetti-ish code is rather old and dates back to the
closed-source days. The code that was newly written since the code was
released on OpenOffice.org generally uses a more modern style and is
easier to maintain. So this proves the opposite of the point of that
researcher.
Even more telling is what happened at mozilla.org where much unhard to
maintain Netscape code that had been developed as closed source was
scrapped to be replaced by better code from the open-source developpers.
It appears that closed-source programmers are more prone to producing
ugly code - under the pressure of schedules set by management or simply
because they think noone else will ever see that code so they needn't be
ashamed.
Ciao, Joerg
--
Joerg Barfurth Sun Microsystems - Desktop - Hamburg
>>>>>>>>>>>>>>>>>> using std::disclaimer <<<<<<<<<<<<<<<<<<<<<<<
Software Engineer [EMAIL PROTECTED]
OpenOffice.org Configuration http://util.openoffice.org
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
