Re: chopping and changing ...
- From: Maciej Stachowiak <mjs noisehavoc org>
- To: Michael Meeks <michael ximian com>
- Cc: Maciej Stachowiak <mjs noisehavoc org>, Jeff Waugh <jdub perkypants org>, gnome-hackers gnome org, GNOME Desktop List <desktop-devel-list gnome org>
- Subject: Re: chopping and changing ...
- Date: Tue, 3 Sep 2002 15:03:01 -0700
On 03Sep2002 11:05AM (+0100), Michael Meeks wrote:
> Hi Maciej,
>
> On Mon, 2002-09-02 at 19:06, Maciej Stachowiak wrote:
> > Apps can be, and indeed have been, obsoleted by a new better app in
> > the same category, although there may be a period when both must be
> > supported. Examples of past, current and future transitions of this
> > sort include
>
> None of the examples you give happened in a corner; and I'm well aware
> of them. And yes - traditionally Linux distributions have been able to
> simply chop and change at a moment's notice.
>
> However - my thesis, which was clearly not stated explicitely enough is
> that creating productized software is far more expensive than simply
> changing the emphasis between several broken alternatives.
>
> To whit; the documentation alone is as expensive as the code - if
> properly translated, similarly making things accessible can impose quite
> a cost; then the better known costs of i18n, multi-head, and so on also
> increase the cost.
>
> Thus 'just' switching to a different app needs to be based on an
> extremely informed decision. Furthermore, if we have several million
> deployed seats, there has to be a perfectly smooth upgrade path.
>
> All of this miltates against switching to a new application, and in
> favour of maintaining the old.
I think you raise good points. I agree that it is usually foolish to
throw out working code and embark on a from-scratch rewrite; improving
existing code incrementally is almost always the better choice. But,
it is not impossible to replace an app, and on rare occasions it is
the right thing to do.
Indeed, this happens in commercial software all the time. Many newer
apps that are in theory "just a new version" have actually been
majorly rewritten on the inside. Sometimes this is a flaming disaster
(Netscape 6) and sometimes it is a success (Windows XP).
I like Jeff's proposal in his response to you for balancing these
kinds of issues.
- Maciej
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]