PDA

View Full Version : patch to enable index numbers to be entered directlywhen selecting items from a list



Christopher Key
2006-04-09, 04:46
This patch modifies the behaviour when a user selects items in a list with
the digit keys. For 10 items or less, the functionality is unchanged, i.e.
1-9 maps to 1-9 and 0 maps to 10. For 11 or more items, the user can enter
a number directly, e.g. the input sequence 1,4,2 would select indexes
1,14,142 in turn. The timeout before new key presses start a new number is
the same as when using numbers to enter letters. If a user enters a number
beyond the end of the list, they are taken to the last item in the list. To
maintain similarity to the scheme for 10 or less items, if a number is
entered starting with 0, a preceding 1 is implicit, e.g. the sequence 0,2
would yield index 102.



Chris Key

dean
2006-04-09, 07:18
I've been meaning to do this for a long time. Thanks, Chris!

I'm out of town and off svn right now, can somebody merge and test
for me?

-dean

On Apr 9, 2006, at 4:46 AM, Christopher Key wrote:

> This patch modifies the behaviour when a user selects items in a
> list with the digit keys. For 10 items or less, the functionality
> is unchanged, i.e. 1-9 maps to 1-9 and 0 maps to 10. For 11 or
> more items, the user can enter a number directly, e.g. the input
> sequence 1,4,2 would select indexes 1,14,142 in turn. The timeout
> before new key presses start a new number is the same as when using
> numbers to enter letters. If a user enters a number beyond the end
> of the list, they are taken to the last item in the list. To
> maintain similarity to the scheme for 10 or less items, if a number
> is entered starting with 0, a preceding 1 is implicit, e.g. the
> sequence 0,2 would yield index 102.
>
>
> Chris Key
>
> <EnterIndexNo.patch>
>

kdf
2006-04-09, 11:13
The patch is coming up with bad line endings for me. Can you repost
with unix line endings and perhaps gzipped (in case its the attaching
that has got it confused)?

the timeout should probably make use of the existing character entry
time, "displaytexttimeout", for some level of consistency. users can
expect something to happen based on that timing. I would think most
never change it, but anyone who does happen to set a high number might
be surprised by a 1 second reaction, especially one that enters a new
mode.
-k
On 9-Apr-06, at 4:46 AM, Christopher Key wrote:

> This patch modifies the behaviour when a user selects items in a list
> with the digit keys. For 10 items or less, the functionality is
> unchanged, i.e. 1-9 maps to 1-9 and 0 maps to 10. For 11 or more
> items, the user can enter a number directly, e.g. the input sequence
> 1,4,2 would select indexes 1,14,142 in turn. The timeout before new
> key presses start a new number is the same as when using numbers to
> enter letters. If a user enters a number beyond the end of the list,
> they are taken to the last item in the list. To maintain similarity
> to the scheme for 10 or less items, if a number is entered starting
> with 0, a preceding 1 is implicit, e.g. the sequence 0,2 would yield
> index 102.
>
> Chris Key
> <EnterIndexNo.patch>