[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-look
Subject:    Re: "Next Link" icon.
From:       Sean Pecor <sean () digitalspinner ! com>
Date:       2002-04-30 3:23:04
[Download RAW message or body]

On Monday 29 April 2002 19:23, Dave Leigh wrote:
> Nobody liked ALL the features in StarOffice 5.2. Sun did a necessary and
> courageous thing by dropping web browsing, e-mail, and that damnable
> desktop from the next release. This is a classic case of trimming fat and
> dropping features to literally save the product. (Judging from OpenOffice
> they did a great job, IMHO). 

This is a great example of trimming fat. I don't think it represents a good 
product going bad via bloat since I'd argue I've yet to see a good product in  
StarOffice. 

> WinZip and PKZip both "went Elvis" with
> overfeatured releases this time 'round. If you like what they USED to be on
> Windows, then you're better off with FilZip. 

I guess I'd ask what you mean by overfeatured. I don't see particularly 
useless functionality in the latest Pkware products. Looking at 
Downloads.com, Winzip was downloaded 522,000 times last week, while Filzip 
was downloaded 552 times. It looks like about 99.9% might prefer what Winzip 
does NOW as opposed to what it did five years ago. 

> PaintShopPro went overboard a couple of releases ago... 

I disagree. I feel PaintShop Pro has been a continually evolving and improving 
product. Other similar products that have since improved and have excelled 
(in my opinion):

1. Adobe Photoshop. I've used Photoshop since v2.5 and it has evolved over the 
years, including a multitude of optional plugins and features. You can argue 
that many 3rd party plugin packs are available but still the core 
functionality has continually increased to the benefit of most Photoshop 
users. I use Gimp as much as humaly possible, but some things just get done 
faster and better in Photoshop.

2. Newtek Lightwave. Simply one of the best 3D modeling and animation tools 
ever. It also has continually improved to include a multitude of optional 
plugins and features in its core product. Many of it's core feature additions 
were those originally found in 3rd party tools, and later integrated into the 
default product. 

By comparison, here are some other graphic design tools that were good, basic 
and simple. But eventually they were outpaced by competing tools that offered 
more core functionality:

1. PC Paint.
2. Autodesk Animator.
3. Deluxe Paint.

> so did RealPlayer. 

Hrm. RealPlayer has never accomplished it's core functionality (dealing with 
streaming content) very well, so I don't see this as a good product going bad 
via feature bloat. It's more of a bad product getting worse because it's 
owners are starving for cash.

> Fortunately some great alternatives are showing up on Linux, and in KDE.
> But Steven's rant is on target, IMHO. Core functionality should be core.
> Basic, simple, and extensible. On top of this you should have a bunch of
> basic, simple, extensions.

I agree in part. In many cases these basic, simple, extensions should be part 
of the KDE project or they just won't be developed or completed. The act of 
inclusion in the KDE project means that more programmers will want to 
contribute. And for the most part, any 3rd party add-on will be ignored by 
the masses that may benefit from it. 

Overall I think that an API to create basic, simple extensions is a good 
thing. I also think that many extensions are best integrated into the core 
package. I think we're well past the notion of KDE being a "window manager". 
It's an entire desktop environment that is gaining ground in part because so 
many options and features are available in the core product.

More and more I think that the whole Unix Philosophy (TM) of creating simple 
extensions via piped applications just can't be applied to a desktop 
environment's application interface. There simply isn't a realistic 
translation. From a programmer's perspective, console tools can interact (via 
piping) without being developed with the same API. Furthermore, they each 
perform a specific task that is useful without also being used in tandem with 
other applications. With KDE "extensions" we're talking about extensions that 
are useless without a KDE toolbar to attach to. With KDE "extensions" we're 
talking about libraries that all have to be compiled with the same compiler 
and with the same API. If we upgrade a single console tool to a new major 
version, it will not break scripts that use it in tandem with other console 
tools (via piping). With KDE "extensions", upgrading KDE to v4.0 would break 
all my v3.0 extensions. The average user would use a major distribution, and 
the major distribution wouldn't necessarily distribute individual 3rd party 
extensions, so these extensions would largely be unobtainable by the average 
user. 

Luckily (for me anyway ;)) KDE continues a progressive functionality policy 
which favors the integration of new usable functionality into the core, at 
least for the forseeable future.

Sean.

-- 
Digital Spinner, Inc.
Cutting edge web design and applications development.
sean@digitalspinner.com
http://www.digitalspinner.com
Phone: 802.948.2020
Fax: 802.948.2749

[prev in list] [next in list] [prev in thread] [next in thread]