> Hi all !
> This is a long post, replying to many previous posts, and adding some parts
> from IRC chats, and some even from discussions with Gimp developers.
> ...

I have a degree of sympathy with this post although it seems to go to 
the other extreme.  Backwards compatibility *is* overrated.  Autotools 
is a great example of what happens if you don't trim the crud out of 
your software.  No doubt, I can compile Gimp on every Unix platform ever 
conceived between now and 1972 but I still have to learn about 4 
different languages before I can begin to start understanding how it 
works, let alone deciphering the scripts.  Call me an idiot if you like, 
but it shouldn't be easier to learn the finer points of C++ than the 
tool that simply builds it.

On the other hand, we have KDE 4.  Plenty of improvements, no doubt, but 
no users left to appreciate them.  :)

As for customisability?  I think that it's probably underestimated. Take 
the example of me spending half an hour or more on google this evening 
trying to work out how to enable menu tear-offs in Gnome.  As far as I 
can tell, the feature just isn't there any more.  Luckily for me, canvas 
right-click still provides the feature.  I needed it, btw, because I 
wanted to add several guides to my project, one after the other.  Rather 
than having to navigate all the way to that menu item every time, it's 
much easier to just have it available on the screen.

That's not the only use of this feature, btw.  Another good use is for 
learning keyboard shortcuts.  No need to hover the mouse over an icon 
for half a second; just glance at the menu.

Like I said, I don't know what's happened to this (really nice) feature 
but I did find a clue in some Java/GTK SDK documentation which states 
that usability studies have concluded that menu tear-offs are a bad idea 
and should be avoided.  Oh dear, I thought.  Someone's been conducting 
usability studies.

Not that I have anything particularly against UI studies, but the method 
had better be sound, the assumptions had better be correct, and the 
results had better be applied appropriately.  If the conclusion comes as 
a surprise, re-examine the experiment...carefully.

Anyway, getting back to the Gimp, I'd be willing to bet real money that 
whatever ideas you have about a typical Gimp user are probably wrong.  
By all means, design for whatever you think the common use case is 
likely to be but remember that Gimp is (to borrow a programming term) a 
low-level IMP.  That makes it even more likely that usage patterns from 
one user to the next will be radically different.  If you don't enable 
Gimp's users to do the things that you can't think of, you'll just have 
non-plussed Gimpers.  If you take away those things?  Well, good luck 
with all the hate. :)

---AV & Spam Filtering by M+Guardian - Risk Free Email (TM)---

Gimp-developer mailing list

Reply via email to