App Browser versus Project Browser

[J. Landman Gay]

On 10/7/2015 1:22 PM, Mark Waddingham wrote:
> Far more useful would be constructive criticism of both the Project
> Browser and the Application Browser. It does seem a little 'silly' to
> maintain two things which serve essentially the same purpose - so Ali's
> idea is perhaps the best way forward - what is it that is good and bad
> about both and is it possible to design something which everybody would
> be happy with?

The issues would probably become clear if you open, say, 10 large 
stacks, each with 50 cards or more, containing dozens of controls per 
card. Since my primary project for the last 2 years uses that setup, I 
haven't been able to use the Project Browser because it isn't practical.

1. The hierarchical organization of the App Browser (AB) is 
indispensable and is the main reason I stay with it. I can see at a 
glance how to drill down to the single object I am looking for and how 
objects are organized on each card by group and layer order. It is by 
far the fastest way to understand how a set of stacks is internally 
structured. The long, scrolling list in the Project Browser (PB) can't 
display the structure as clearly because it is all linear. Multiple 
cards with many objects will run off the top and bottom of the PB window 
and you can't see the overall organization.

2. It is difficult in the PB to quickly find a specific object. If you 
want to know the name of an object on some other card, you have to 
collapse the current card, scroll through 50 cards to find the one 
you're looking for (and if you didn't collapse those already, the 
scrolling is interminable,) expand it, scroll through the objects to 
find the one you want (note the name because it's going to be a long 
trip to find it again,) collapse that card, scroll (forever) again to 
find the card you started with, expand it, find the original object 
again, and continue. In AB, I can just look at the left-hand pane and 
see the name of the target card, click it, note the name of the object, 
then click back where I was. If the AB is sized tall enough to hold 50 
lines of text, I don't have to do much scrolling at all. If I do need to 
scroll, it's minimal because at least 25-30 cards are always visible at 
once.

2. In the AB I can click on any header to view the organization in many 
ways, and I have a choice of which columns I want to display. If I want 
to work only with images, or fields, I can bunch them together in the 
list by type and they are quickly accessible while still allowing me to 
see the other objects on the card. I frequently require info on layering 
order, one click and I have that. I use the ID column extensively. In PB 
I have to type in a filter string to isolate by object type, and then I 
can no longer see any other objects, so if I need some other info I have 
to remove the filter, find what I want, then reinstate the original 
filter. PB does not offer a way to identify an object ID at all, as far 
as I can see, and I need that all the time. (But you could turn off 
those distracting ID tooltips for sure.)

3. Visually, the PB is too cluttered to be quickly scanned. The 
checkmarks in the AB are more useful. In the AB is very easy to see, for 
example, which objects are invisible by simply looking for "gaps" in the 
checkmark column. In the PB I have to examine each object individually 
because the visual difference between the enabled and disabled "eye" 
image is not distinct enough, and even if it were, there's that 
scrolling issue again to see all the objects. Also, there is no single 
column to scan -- the lock icon is interspersed so you have to mentally 
learn to skip over every other icon.

4. I have turned off thumbnails in the PB because with hundreds of 
objects or more, the time required for it to constantly update is (or at 
least, was) unacceptable. Even without thumbnails, it performs much 
slower than the AB. There is also the issue of visual clutter (see 
below) which is main reason I turned off thumbnails on day one. 
Thumbnails also double the amount of scrolling you have to do to find 
things.

5. In the PB there is no clear delineation between cards and substacks. 
Both are left-aligned at the same visual depth. In the AB, all stacks 
are in the left pane, with substacks indented under their mainstack. 
Also, in the PB, the stack you are inspecting scrolls off the top of the 
window, so you are never sure which stack owns the cards that are 
currently displayed. This is a big issue in my project, because all the 
stacks are clones of each other and cards have the same names (usually 
just IDs.) In the AB I can immediately see which stack owns the card 
because the card is highlighted in the left-pane list under its 
easily-viewable owner. Even if I have to scroll to see the stack name, 
the card I'm working with remains selected and its objects remain visible.

6. The icons at the bottom of the PB are so tiny on my screen that they 
are difficult to recognize (and my eyesight isn't great anyway.) I have 
to use the tooltips. That takes too long, so I just open the property 
inspector or use the menu items instead. I suppose with some use I'd 
memorize what each icon does, but the other issues have prevented me 
from becoming familiar enough with it.

That's just what I remember from the few days I tried to work with it. 
I'm not convinced that the current design can accomodate my work style 
unless it can at least be revised to show a columnar view rather than a 
linear one. What I would have preferred is an update for the few 
glitches in the AB (mainly it doesn't always refresh automatically, and 
those blinking tooltips are positively aggressive) and give it a new 
coat of paint if you think it looks too dated. Its plain text layout 
with clear checkmarks is much easier for me to work with. I do like how 
you can change layering order by dragging in the PB, that would be a 
nice addition to the AB.

-- 
Jacqueline Landman Gay         |     jacque at hyperactivesw.com
HyperActive Software           |     http://www.hyperactivesw.com