Always Highlight Most Relevant Quick Drop Item (even if not an exact string match)

[JackDunaway]

When typing into the Quick Drop search, unless the query string is an exact match to a result, nothing is highlighted in the results list, which throws the following error when trying to drop an item:

  "Mer err" is typed, then *click*, produces:  

It would be nice if Quick Drop always highlighted the top match in order to always have a droppable item.

This would make Quick Drop behave like the Win7 Start Menu search - it always highlights the top match, so that pressing "Enter" always has something to invoke:

[JackDunaway]

In the course of typing this idea, I found that hitting the "DOWN" arrow after typing will highlight the first item. This is nice! However, it's a bit of a workflow interruption since it's the left hand on the keyboard trying to hit a key on the opposite of the keyboard.

***EDIT: the TAB key also performs this function! That's an acceptable workaround for the meantime, though perhaps it would still be nice to have this work like the Win7 Start Menu.

Spoiler

If nothing else, the excercise of typing this idea has helped me find a suitable workaround; hopefully this can help someone in the future by stumbling across this post 

[tst]

I really dislike the way QD behaves. I use my own version where the item to be dropped is always the item highlighted in the list and the string is only used for filtering the list. This also means that you don't lose your search terms if you navigate the list.

An example of this type of UI can be seen here (and yes, you should also vote for the idea that example came from).

[alejandro5042]

I love this idea. I imagine it would be really easy to implement. No wonder I thought quick drop was a bit awkward at times.

[Wart]

I normally run into the situation where I know what I want to drop, but quick drop can occasionally be slow to open.  So I'd end up with something like "rge errors" in the QD box.  This idea would fix that issue as well.

[Darin.K]

As I have probably mentioned before, I have recreated QD. I can not say how similar mine is to Darren's version under the hood, but I'll bet it is close. The added bonus is that I can prototype and test drive these suggestions. In this case I will say two things. (1) I like it. (2) It was easy to do, I simply change the listbox to select 1 item instead of 0, reset the listbox value to 0 instead of -1 when the value in the combo box is changed and use the listbox string value when the combo box value gives an error in the Invoke Node. Darren could have this changed in less time then it took me to type this.

[Jarrod_S.]

If anyone is familiar with QuickSilver for Mac, it does a very similar thing, and it is FANTASTIC! You don't have to worry about pesky typos. It's the software's job to figure out what you meant.

[JackDunaway]

---

@jarrod S. wrote:

If anyone is familiar with QuickSilver for Mac, it does a very similar thing, and it is FANTASTIC! You don't have to worry about pesky typos. It's the software's job to figure out what you meant.

---

I use Spotlight, Apple's native "replacement" for Quicksilver, and you're hinting upon another request for Quickdrop; implement machine learning to know which items are dropped most often, and rank them higher in the results list. Spotlight does this, and after just a little learning, it will favor oft-picked items over ne'er-picker.

[tst]

>...implement machine learning to know which items are dropped most often, and rank them higher in the results list.

It's interesting you should mention that. When QD first came out, the two major issues I had with it was that it used auto-complete instead of simply serving as a filter for the list (i.e. essentially this idea) and that it didn't have some kind of preference algorithm (the simplistic one I suggested and implemented was exactly the "often used" one you suggest with some tweaks), but NI wasn't particularly impressed by that at the time. 

I know that QD did actually get some preference sorting algorithm for the list (I think it was in 2009), but I don't know what its algorithm is. Maybe if this is popular enough I will get my wishes 5 years later (although it hasn't exactly been troubling me, since I have my own version, which does work).

[Darren]

I have a prototype for this feature request (minus the "machine learning") that you can try in LabVIEW 2012.  See here and let me know what you think.

[JackDunaway]

---

@Darren wrote:

I have a prototype for this feature request (minus the "machine learning") that you can try in LabVIEW 2012.  See here and let me know what you think.

---

Dude... using this, already and it works *perfectly*. Will give feedback over in that Community thread after using a few weeks. (If you don't hear from me, that's just tacit approval of the changes  )