I've just started to work with iBird PRO. I'm impressed and
excited, but a few issues jump out at me immediately, within the first
hour of use. I speak from the point of view of a birder experienced
with using more
traditional reference materials and of an experienced user of
software. It seems to me this view is relevant to your "PRO" product.
Before
giving a long list of perceived weaknesses, I should preface by saying
that I'm not writing a review -- if I were, I would dwell on the many
many things this software seems to do so well. I've taken considerable
time here to detail the following issues because I think this software
is worth careful attention and high expectations. Obviously a lot of
each has already gone into this software.
The main issue
is a cluster of awkward/missing features related to searching/limiting
and presenting lists of species. I'll describe an imaginary user
session. The "browse" view by family gives a list of families, and you
can drill down. Great, this is what every experienced birder wants.
But that's not how the application starts up (it starts by first name,
the least useful of the three choices), and there's no way to correct
the default. That's one easy user setting, or else remember the
selected view between launches. Then I want to limit the 914 birds to
those in my geographical region (alas, the tradition is by state,
whether Texas- or Rhode Island-sized, but this isn't iBird's fault). I
go to search, select my state, hit "view" and I get a list. I'm bumped
back to first-name order, so select family order again. But I don't
get a list of families with drill-down, it's a list of 290 birds with a
headline introducing each family group. Not useful for 290 birds.
Conceptually, it makes no sense to me to have one concept (icon,
program function, mode of presentation) for search-results and another
for browsing. Just have one concept, call it browsing, and then filter
it with criteria such as region and physical characteristics. Call it
browsing and filtering because there's already a "search" function, the
search field at the top in both "browse" and "search" modes, with
magnifying glass and "keyword or /Latin or &BandCode" prompt. That
one's useful, and should be the only feature called "search." Back to
filtering: for some criteria, region in particular (but why not allow
any/all criteria?), the user will want to be able to select persistence
over application launches, while having non-persistent criteria reset
at re-launch (or of course at explicit request to reset).
Now
to some particular filter criteria. I notice that selecting my US
state in "Location" is a different operation from selecting my
US-state-plus-month(s). Hunh? I can speculate that the history of the
program's development led to separate filter criteria, but from user
perspective this makes no sense. Region and month should be separate,
multi-selectable criteria. Since the state-month list gives only
states, I assume you don't (yet) have monthly data for the non-state
locations (e.g., Aleutians, Atlantic Coast). So when one of those
locations is selected, gray-out the months, and vice versa, and explain
the behavior in the docs. There are probably other solutions but the
present overlapping is not a good situation.
Song pattern:
The first species I tried with this attribute in mind was Chipping
Sparrow, and I really don't understand (yet) why you say it is
"falling" not "flat." I'm sure that with further experience with this
attribute, I will develop an intuition for how you have applied it, but
it would facilitate learning (and probably always be of interest) if
the "play vocalization" screen listed the attribute value for song
pattern and also song-type (e.g., Rattle-Guttural -- which you
incidentally misspell "gutteral"). More confusion: when I limit to
"warble-trill" I get American Woodcock on the list. On audio screen
for Woodcock, Common Nighthawk is (very appropriately) mentioned as a
similar-sounding species. But Common Nighthawk isn't on the list
limited to "warble-trill"!?
This leads to observation that if
CONI wasn't on the limited list with AMWO, CONI shouldn't have been
mentioned on AMWO's audio screen as a sound-alike species. I confirmed
this by limiting to Massachusetts and seeing that Band-tailed Pigeon,
which doesn't occur in MA, is nevertheless listed as both a sound-alike
and look-alike species to Mourning Dove. This is definitely a problem
for the user using iBird as an aid to identification; the very useful
sound/look-alike lists are misleading and so not as useful as they
could be. On the other hand, I can think of scenarios where a user
using iBird as a learning tool would want this information. A program
setting doesn't seem like the best solution, because it's too clunky to
switch, and I can think of situations where I want to switch while look
at one species a/c. So I suggest a toggle on those two screens
(look-alike and sound-alike) whether to apply or not apply the current
filter to the list of lookalikes.
Going forward, the option to display a sonogram of the sound being played would be a terrific feature.
I
have tried to figure out by means of experimentation what the "remember
location" toggle setting does, but I can't. There should be a
"chapter" in the reference help for "settings".
Let's go back and improve the family drill-down. With this version of
the software, drill into a family, and you always get a list of
species, never a list of genuses into which you can drill again. Like
US states, some families are much bigger than others, and this will
certainly be true once we start filtering. For the same reasons that
users want to drill into families, they want to drill into genuses when
the resulting list of species would be large (nearly 60 for Emberizidae
and Parulidae). It would be pretty easy, I think, to program a user
settable threshold value: when a species list will be >N, list
genuses instead. Some heuristic refinement would be necessary to avoid
listing a single genus, and perhaps other edge cases.
Now
miscellaneous issues. The organization of the screen you get when you
hit the "help" icon at the bottom needs some help itself! On this
screen, the fourth colored button at the top is labeled "help". But I
just selected "help," didn't I?? Pressing that button doesn't actually
give much help, just a couple of web links to soft resources like the
forum -- and doesn't mention where the REAL help is, which is the
"reference" section, which on that first screen is linked WAY below the
fold -- below the first screen, below several FAQ answers. Speaking of
FAQs and FAQ answers, there are two links to them (the colored button
at the top and a link at the bottom) with a different set of FAQs and
answers stuck inbetween.(*)
Why are some settings outside the
application and some within? Is this an iPhone issue? From a user's
perspective, it's confusing to have two distinct sets of settings. Two
paths to access a list of settings is fine; if the two lists can't be
made identical, an on-screen reminder of the other list would be
helpful, and can one list at least be a superset of the other?
In
browse function, first-name-order view, enter "crown" in the search
field. 16 species are found with "crown" in the name. Great, this is
correct. Switch to (or search while in) family-order view, and you get
a list of the 7 families containing those 16 species. So far, so
good. But select Columbidae now, and you get the full list of 17
Columbidae species, not the single species (White-crowned Pigeon) with
"crown" in the name, as expected. This seems like a bug, rather than a
design issue.
Documentation bugs: In Reference, under "browse",
heading "first, last, and family buttons": the text here is entirely
incorrect, belongs somewhere else. It starts with "Includes five size
categories". Likewise under the next heading "Reset" it starts talking
about habitat.
Two editor's quibbles: (1) why are all (?) the
phonetic texts enclosed in quotation marks? Quotation marks
distinguish text inside them from text outside them. But there's
never any text outside them -- not in any of the species I've looked at
so far. So they seem to me just visual clutter and they raise a false
question in users as to what they're doing there. (2) Why does every
label for an audio track end with the word "Voice"? E.g., "American
Robin Voice". I can't find any label that doesn't end with "voice" so
it's more visual clutter. It would be great if in the future you had
multiple separate sounds for each bird (with corresponding sound-alike
lists) but they would use the more technical labels "song" "chip note"
"flight" "distress" or "non-vocal" for e.g., dove wings, Woodcock
wings, etc.
Some long-named species have labels for the audio
track that don't fit; need some scrolling thing to solve? (automated or
else user touch)
In the search-by-keyword field, can you modify
the iPhone keyboard that pops up so that the user doesn't have to press
two extra keys to get in and out of the number/punctuation keyboard to
enter '/' and '&'? No species names start with Q, so you could
steal that character's space on the qwerty layout as a kludge, if you
could change the glyph on the "key": / would be shift-& (or vice
versa, but I guess bandcodes used more often than Latin). A better
solution would be to make the prefix unnecessary. I don't know about
the Latin, but there aren't significant conflicts with the banding
codes -- Birder's Diary handles both just fine. Maybe it's an issue of
iPhone processor not strong enough for satisfactory performance?
-- Matt
(*) It is like trying to stop the tide, but I'll make the general point anyway
that with good documentation there are no FAQs; FAQ answers are (or
ought to be) essentially temporary, until documentation is revised or a
transient situation passes. Lists of FAQ answers originated on usenet
discussion groups, which of course didn't have manuals.
