Re: Changes I've been thinking of...

[Richard Frith-Macdonald]

On 10 Oct 2009, at 10:03, Michael Thaler wrote:

> Hi,
>
> > I'm one of them. I got interested into GNUstep also because of it  
> > looks.
> > I love windowmaker. It is the only thing I use on free Unices.
> > It is sleek, unobtrusive, professional. The same should be true for
> > GNUstep. GNUstep stuff is generally almost there, but there are  
> > drawing
> > glitches, ugly icons, imperfectly done interfaces which makes it  
> > not on
> > par with openstep. And we should be even better. Improved, more  
> > complete.
>
> I really dislike the term "looks professional". How do you define  
> "looks
> professional"? I think the only sane definition of "looks  
> professional" is that
> something that looks professional is something that is used for  
> professional
> work.

I think the intended meaning is ... well suited to continuous  
productive use (ie at least several hours a day actually doing things).

> By this definition Windows XP / Vista, KDE, Gnome and MacOS X all look
> professional because they are used for professional work.

But that's clearly not the definition that other people are using.

> Why is a user interface more "professional" if all buttons are  
> squared,
> everything is gray and there are no gradients?

I think you are making a 'straw man' argument here...  I can't see  
anyone saying that a small set of specific features is what they are  
concerned about or that what an interface ends up being used for  
defines what it's suitable for.

> In my opinion, to do professional work, it is much more important to  
> have
> professional tools (IDEs etc.) then to have no gradients and square  
> buttons.

I suspect nobody will disagree with that, but it's not relevant to a  
discussion of theming.

> > Also, this concept of "outdated" is really ridiculous. Style has no
> > time. People like Rolex. Waterman. Montblanc. Breguet.
> > People like Vetiver, 4711 Koelnisch Wasser.
> > People like Veuve Cliquot Poinsardin.
> > These items are made as our fathers or our grand-fatehrs could have
> > bought them. Serious people like them because they are masterpieces.
>
> Well, most people I know (scientists, engineers etc.) think that  
> wasting money
> for a Rolex is ridiculous. I don't even think that a Rolex looks  
> good. I
> consider a Rolex a status symbol that is not worth its money because  
> you can
> get better, cheaper, better looking (this is my personal opinion)  
> for less
> money.
>
> Maybe you consider scientists and engineer not serious people, but I  
> do and I
> know lots of serious people (by my definition) that consider all you  
> mentioned
> above as waste of money.
>
> And I don't buy that masterpieces made by our farthers cannot be  
> improved.
> Imagine you build a real masterpiece carriage one hundred years ago  
> when there
> were no cars. How many people would buy a carriage instead of an  
> ordinary car
> today just because the carriage is a real masterpiece and an  
> ordinary car is
> just an ordinary car?

Again, you seem to be missing the point ... which was clearly not that  
a specific design is good value for money or that 'serious' people  
must like a specific design, or that old designs cannot be improved  
upon.

The point was that some designs do not become outdated, and that you  
can't dismiss the opinion of people who like particular old designs as  
being frivolous.
The implication was that the NeXTstep gui design is one of those which  
has not become outdated ... a 'classic' if you like.

I don't even think he was saying that the NeXTstep design can't be  
improved on.
My personal opinion is that I want to *see* an improved design before  
adopting it.

> The NeXTSTEP GUI was designed fifteen years ago when it was  
> basically not
> possible to have round buttons, gradients, transparency, shadows  
> etc. because
> the hardware was not powerful enough for that.

Actually that's only slightly true... NeXTstep was well ahead of its  
time.  The original NeXT machines were very good at display performance.

1. They used display postscript, and the engine for that could support  
true round buttons (ie easily detect a hit inside a circle, or any  
other curved shape for that matter).
2. The display postscript supported gradients
3. NeXTstep certainly supported transparency
I don't know about shadows or 'etc', but while performance would have  
been a consideration, it was certainly not the major factor in the  
design of the NeXT look.  The appearance of NeXTstep was due to a  
certain design aesthetic, not due to technical considerations.
You don't have to take my word for it ... you can ask any user of an  
old-time NeXT machine what the graphics performance was like, and you  
can look at the old NeXT manuals to see what the capabilities of  
display prostscript were.

> But the world moved on and
> today almost nobody wants to have square buttons, no gradients etc.  
> (There are
> certainly people today which would prefer to have a masterpiece  
> carriage
> instead of an ordinary car, but most people that just need something  
> to go to
> work would certainly take the car).

Sounds OK, but you need to quantify 'almost nobody' ... what surveys  
have been done?, how were they conducted? were they big enough to be  
statistically significant?, how did they allow for systematic biases  
etc.  Without facts like that, 'almost nobody' is nothing but an  
personal opinion.

> > Now of course, other people change dresses every few months, have a
> > Swatch, use the latest perfume from Kiko or Pupa or whatever. They
> > drink  some fashion-drink like bacardi breezer.
>
> So what? People did not use cell-phones, GPS receivers, DVD burners,  
> > 1 TB
> hard drives etc. 20 years ago. Now they do. Is that a good thing or  
> a bad?

He's merely saying that some people like 'timeless' styles, while  
other people like following the latest trends.
The clear implication is that it's not legitimate to dismiss an old  
design just because you are a person who likes new ones.

> The gnustep project is about 15 years old, older then both KDE and  
> GNOME. How
> many developers does gnustep have today and how many does KDE (or  
> GNOME) have?
> Why are there so many people working on KDE and GNOME, but almost  
> none working
> on Gnustep?

Greg clearly thinks it's marketing.

In part I agree.  I think marketing is what we most need now, but I  
think there was a historical design issue which caused GNUstep to lose  
out to Gnome and KDE.

Way, way back, it was decided to copy NeXT and implement the gui using  
display postscript (so as to have the same low level drawing APIs),  
but to put that on top of X.
Unfortunately, implementing display postscript was just too huge a  
task (NeXT didn't need to develop it, they bought it from Adobe along  
with the expertise to optimise its performance).  This meant that the  
gui couldn't really progress to a usable state and was held back for  
several years, by which time Gnome and KDE had achieved 'market  
dominance'.

> But if the gnustep projects wants to increase the
> number of users and developers, I think it is absolutely necessary  
> to improve
> / change the look and feel to something more familiar / pleasent for  
> typical
> OSS users.

I think having good themes is great for initial impression (eye  
candy)... and it would be great to have a page with screenshots of  
GNUstep running a variety of different themes to attract different  
groups of people, but I'd go back to your earlier comment:

> In my opinion, to do professional work, it is much more important to  
> have
> professional tools (IDEs etc.) then to have no gradients and square  
> buttons.

What we most need is a small set of really good apps, and sufficient  
interoperability to work well with any non-GNUstep apps we need.

So ...

1. Marketing to get people to give us a look.

2. Eye-candy to draw people in and get them to try things out  
(changing the default theme won't do that ... we need to have a group  
of three or four good themes to appeal to different people)

3. Enough good quality stuff so that people don't try once and then  
give up.

Basically, 1 and 2 are what Greg was saying at the start of this thread.