Many of my websites have a need for a special kind of input type.  This 
arises when the user can pick no more than one of a list of things, but 
picking none is also acceptable.  The type=radio button and type=select 
options both have a severe drawback; that being, when one item is selected, 
the user cannot get back to "no items selected" state.  It is impossible to 
"unpush" a choice.  The frustrated user has to reset the entire screen to 
get out of this box.  It is truly a moment of carelessness, a lifetime of 
regret scenario.

I noodled around and now have a radio button which can be singularly cleared 
with a double click.  This is obviously an important discovery, alert the 
Nobel Committee!, but, for the life of me, I don't know why this particular 
code works, and I will be most embarrassed when they present me with the 
award.  In particular, why is it a double click that does the deed, rather 
than a single click?  If you can explain it, I will split the medal with 
you.  If you can improve on it, you get the neck ribbon, as well.

Bruce Anderson

(The test site to which the form is submitted returns the values of all vars 
in the form.  Don't be fooled.  HOLD_RADIO is not the name of radio button. 
It is a var used to make a status comparison.  "radio_button" is the name of 
the selectable variable.) 

Hi Bruce

Did not work for me using ThunderBird on SuSE Linux. Might only work on
MS stuff probably because its broken.

Jim

Bruce Anderson wrote:
> Many of my websites have a need for a special kind of input type.  This 
> arises when the user can pick no more than one of a list of things, but 
> picking none is also acceptable.  The type=radio button and type=select 
> options both have a severe drawback; that being, when one item is selected, 
> the user cannot get back to "no items selected" state.  It is impossible to 
> "unpush" a choice.  The frustrated user has to reset the entire screen to 
> get out of this box.  It is truly a moment of carelessness, a lifetime of 
> regret scenario.
> 
> I noodled around and now have a radio button which can be singularly cleared 
> with a double click.  This is obviously an important discovery, alert the 
> Nobel Committee!, but, for the life of me, I don't know why this particular 
> code works, and I will be most embarrassed when they present me with the 
> award.  In particular, why is it a double click that does the deed, rather 
> than a single click?  If you can explain it, I will split the medal with 
> you.  If you can improve on it, you get the neck ribbon, as well.
> 
> Bruce Anderson
> 
> (The test site to which the form is submitted returns the values of all vars 
> in the form.  Don't be fooled.  HOLD_RADIO is not the name of radio button. 
> It is a var used to make a status comparison.  "radio_button" is the name of 
> the selectable variable.) 
> 
> 
> 
> Radio button invention...double click on a selected button to clear it.
> Choices A - D exhibit normal radio button behavior, plus you can clear a
> selection without choosing another value by double clicking on the button.
> Choice E is a standard radio button, which does not have the total clear
> method.
> 
> A
> B
> C
> D
> E
>

Bruce,

>I noodled around and now have a radio button which can be singularly cleared 
>with a double click.  This is obviously an important discovery, alert the 
>Nobel Committee!, but, for the life of me, I don't know why this particular 
>code works, and I will be most embarrassed when they present me with the 
>award.  In particular, why is it a double click that does the deed, rather 
>than a single click?  If you can explain it, I will split the medal with 
>you.  If you can improve on it, you get the neck ribbon, as well.
I looked at it, and I understand how it works.
This works only in IE on windows! not in any other browser, so it is
not a very good solution.
I will look into it later on more deeply, and will come up with a
other solution. But give me a few day's!

Regards
Sander Elias

-------------------------------------------------------------
xbHCL (http://www.xbHCL.com) the xbase HTML command layer
PBIH  (http://pbih.eu)       Polar Bear International Hosting
-------------------------------------------------------------
also a member off the XXP (http://www.xxp.nl)

Bruce,

I had the need for a small distraction, so I looked at your problem.

see the attached HTML file. you can check and uncheck any button with
a single click, works in both IE and Firefox (did not test other
browsers, but it should work there too.)
it's a simple case of checking the state when the mouse button comes
'up', and reset the button if it has set before.
you have to use the mouseup in stead off the onclick because the
button is setted before the evaluation of onclick!

hope you can use this.
Regards
Sander Elias

-------------------------------------------------------------
xbHCL (http://www.xbHCL.com) the xbase HTML command layer
PBIH  (http://pbih.eu)       Polar Bear International Hosting
-------------------------------------------------------------
also a member off the XXP (http://www.xxp.nl)

oops, forgot the attachment!
Regards
Sander Elias

-------------------------------------------------------------
xbHCL (http://www.xbHCL.com) the xbase HTML command layer
PBIH  (http://pbih.eu)       Polar Bear International Hosting
-------------------------------------------------------------
also a member off the XXP (http://www.xxp.nl)

Still does not work in Linux with FireFox

Jim

Sander Elias wrote:
> oops, forgot the attachment!
> Regards
> Sander Elias
> 
> -------------------------------------------------------------
> xbHCL (http://www.xbHCL.com) the xbase HTML command layer
> PBIH  (http://pbih.eu)       Polar Bear International Hosting
> -------------------------------------------------------------
> also a member off the XXP (http://www.xxp.nl)
> 
> 
> ------------------------------------------------------------------------
> 
> choiche A
> choiche B
> choiche C
> choiche D
> choiche E
>

Sorry was still trying in Thunderbird It works fine in FireFox and
Konqueror

Jim


Sander Elias wrote:
> oops, forgot the attachment!
> Regards
> Sander Elias
> 
> -------------------------------------------------------------
> xbHCL (http://www.xbHCL.com) the xbase HTML command layer
> PBIH  (http://pbih.eu)       Polar Bear International Hosting
> -------------------------------------------------------------
> also a member off the XXP (http://www.xxp.nl)
> 
> 
> ------------------------------------------------------------------------
> 
> choiche A
> choiche B
> choiche C
> choiche D
> choiche E
>

Jim,

>Sorry was still trying in Thunderbird It works fine in FireFox and
>Konqueror
Thanks for the confirmation!

Regards
Sander Elias

-------------------------------------------------------------
xbHCL (http://www.xbHCL.com) the xbase HTML command layer
PBIH  (http://pbih.eu)       Polar Bear International Hosting
-------------------------------------------------------------
also a member off the XXP (http://www.xxp.nl)

Dear Sander Elias

Only a comment

I am using Thunterbird as news reader, and it show the attachment as 
part of the message, but If I click on the button A and click again on 
the button A, it doesnt check and unchech the button


Best Regards
Osvaldo Ramirez



> Bruce,
> 
> I had the need for a small distraction, so I looked at your problem.
> 
> see the attached HTML file. you can check and uncheck any button with
> a single click, works in both IE and Firefox (did not test other
> browsers, but it should work there too.)
> it's a simple case of checking the state when the mouse button comes
> 'up', and reset the button if it has set before.
> you have to use the mouseup in stead off the onclick because the
> button is setted before the evaluation of onclick!
> 
> hope you can use this.
> Regards
> Sander Elias
> 
> -------------------------------------------------------------
> xbHCL (http://www.xbHCL.com) the xbase HTML command layer
> PBIH  (http://pbih.eu)       Polar Bear International Hosting
> -------------------------------------------------------------
> also a member off the XXP (http://www.xxp.nl)
De

Osvaldo,

>I am using Thunterbird as news reader, and it show the attachment as 
>part of the message, but If I click on the button A and click again on 
>the button A, it doesnt check and unchech the button
Well, that makes sense!
as it is a news/mail reader, it should not execute ANY java script.
although there is very little JS in the sample, it will only run in a
'browser' that has JS enabled!
Regards
Sander Elias

-------------------------------------------------------------
xbHCL (http://www.xbHCL.com) the xbase HTML command layer
PBIH  (http://pbih.eu)       Polar Bear International Hosting
-------------------------------------------------------------
also a member off the XXP (http://www.xxp.nl)

I will inform Stockholm to send the prize to you.  Thank you, sir, for the 
insight.

Bruce,

>I will inform Stockholm to send the prize to you.  Thank you, sir, for the 
>insight. 

Well, Thanks 
But on a more serious side, this kind off programming is a direct
violation off the w3.org standard. I understand why you would want to
use anything like this, but you should consider using checkboxes in
stead of radio button's 
with a little scripting you can let them behave exactly the same as
radio button, and it would be 'legal' to uncheck all off them.

quote from w3.org html 4.01 specs:
---------------------------------------------------------------------------------------------------------------
radio buttons
    Radio buttons are like checkboxes except that when several share
the same control name, they are mutually exclusive: when one is
switched "on", all others with the same name are switched "off". The
INPUT element is used to create a radio button control.
    If no radio button in a set sharing the same control name is
initially "on", user agent behavior for choosing which control is
initially "on" is undefined. Note. Since existing implementations
handle this case differently, the current specification differs from
RFC 1866 ([RFC1866] section 8.1.2.4), which states:

        At all times, exactly one of the radio buttons in a set is
checked. If none of the <INPUT> elements of a set of radio buttons
specifies `CHECKED', then the user agent must check the first radio
button of the set initially.

    Since user agent behavior differs, authors should ensure that in
each set of radio buttons that one is initially "on".
-----<quote end
>------------------------------------------------------------------------------------------------------

If you want a sample off a radio-check-boxes just drop me a note, and
I will see what I can do!

Regards
Sander Elias

-------------------------------------------------------------
xbHCL (http://www.xbHCL.com) the xbase HTML command layer
PBIH  (http://pbih.eu)       Polar Bear International Hosting
-------------------------------------------------------------
also a member off the XXP (http://www.xxp.nl)

I have been doing that checkbox trick for the last 3 years. And I know about 
the w3.org standard on radio buttons; in fact, I believe being able to set 
the "checked" value is said to be impossible in that same standard.  Luckily 
for me, it ignores its own standard.

The problem I have with the checkbox solution is that it is based in a 
javascript which grows geometrically with the number of checkboxes, and the 
script does not lend itself to being generated algorithmically.  The 
clearance routine requires each checkbox in the group to have a list of all 
the affected vars, which is an approach that shouts "typo error". 
Furthermore, it means that there are multiple variables to be checked to 
answer the single question "What do you want for lunch?" because each 
checkbox has its own variable name.  Consequently, data analysis coding 
becomes more complex.

The radio button solution means that I can generate pages from a database 
because each button contains its own clearing code, and that code is 
identical across all of the buttons.  There is only one variable to examine 
on the analysis side, so that coding is simplier, too.  My goal is to build 
websites that create their pages dynamically.  This approach means my 
customer can enter all of their mutually exclusive choices into a dbf and my 
page builder algorithm will just do it, regardless of how many there are, 
and then my program does the all typing, rather than me.

The original push button radios, on which this control is modelled, were in 
the cars I drove through the early '70s.  On those radios, you could unpush 
a button.  I think this modified input has precedence on its side, and it 
fills a need that exists in real programming world.  Viva la click, click!

Seems to me the easiest way is to simply add a button called NONE. This
would also give the user a clue that it is an option.

Jim

Bruce Anderson wrote:
> I have been doing that checkbox trick for the last 3 years. And I know about 
> the w3.org standard on radio buttons; in fact, I believe being able to set 
> the "checked" value is said to be impossible in that same standard.  Luckily 
> for me, it ignores its own standard.
> 
> The problem I have with the checkbox solution is that it is based in a 
> javascript which grows geometrically with the number of checkboxes, and the 
> script does not lend itself to being generated algorithmically.  The 
> clearance routine requires each checkbox in the group to have a list of all 
> the affected vars, which is an approach that shouts "typo error". 
> Furthermore, it means that there are multiple variables to be checked to 
> answer the single question "What do you want for lunch?" because each 
> checkbox has its own variable name.  Consequently, data analysis coding 
> becomes more complex.
> 
> The radio button solution means that I can generate pages from a database 
> because each button contains its own clearing code, and that code is 
> identical across all of the buttons.  There is only one variable to examine 
> on the analysis side, so that coding is simplier, too.  My goal is to build 
> websites that create their pages dynamically.  This approach means my 
> customer can enter all of their mutually exclusive choices into a dbf and my 
> page builder algorithm will just do it, regardless of how many there are, 
> and then my program does the all typing, rather than me.
> 
> The original push button radios, on which this control is modelled, were in 
> the cars I drove through the early '70s.  On those radios, you could unpush 
> a button.  I think this modified input has precedence on its side, and it 
> fills a need that exists in real programming world.  Viva la click, click! 
> 
>

Bruce,

>I have been doing that checkbox trick for the last 3 years. And I know about 
>the w3.org standard on radio buttons; in fact, I believe being able to set 
>the "checked" value is said to be impossible in that same standard.  Luckily 
>for me, it ignores its own standard.
Sorry, but your wrong on this one! it clearly states that the designer
should check one by default, using the 'checked attribute on the
input. I do this on a frequent base!

>The problem I have with the checkbox solution is that it is based in a 
>javascript which grows geometrically with the number of checkboxes, and the 
>script does not lend itself to being generated algorithmically.  The 
>clearance routine requires each checkbox in the group to have a list of all 
>the affected vars, which is an approach that shouts "typo error". 
>Furthermore, it means that there are multiple variables to be checked to 
>answer the single question "What do you want for lunch?" because each 
>checkbox has its own variable name.  Consequently, data analysis coding 
>becomes more complex.
Well, I think I can create a radio-button solution that is almost as
simple as my radio-button one (although the function will be slightly
larger!) I have to try it to be entirely sure, but I think if you name
all your checkboxes equal (as with radio button) you can use the same
algorithm for both the checkbox solution, and the radio-button.
If I had more time available, I would try it out for you.


>The radio button solution means that I can generate pages from a database 
>because each button contains its own clearing code, and that code is 
>identical across all of the buttons.  There is only one variable to examine 
>on the analysis side, so that coding is simplier, too.  My goal is to build 
>websites that create their pages dynamically.  This approach means my 
>customer can enter all of their mutually exclusive choices into a dbf and my 
>page builder algorithm will just do it, regardless of how many there are, 
>and then my program does the all typing, rather than me.
As I stated above,I believe  this would be possible for checkboxes
too!


>The original push button radios, on which this control is modelled, were in 
>the cars I drove through the early '70s.  On those radios, you could unpush 
>a button.  I think this modified input has precedence on its side, and it 
>fills a need that exists in real programming world.  Viva la click, click! 
Well, you have a partial point there, I remember those radio's. the
design was not meant to un push buttons. If one did that, most off the
time the would only produce a loud noise after that! (also the radio
did not 'know' what to do in such a case...)\
the same goes for radio-buttons, you can fiddle around to un push all,
but what's the state after that???

As always', there are arguments to neglect a standard. But I would
rather follow them. In this age people are more and more accustomed to
a certain way off doing things, and breaking with that convention
might not be wise.

Just my 2 cents

Regards
Sander Elias

-------------------------------------------------------------
xbHCL (http://www.xbHCL.com) the xbase HTML command layer
PBIH  (http://pbih.eu)       Polar Bear International Hosting
-------------------------------------------------------------
also a member off the XXP (http://www.xxp.nl)

>>...the same goes for radio-buttons, you can fiddle around to un push all, 
>>but what's the state after that???
oHtml:getVar() will return a NIL.  Assuming that you configure a wrapper 
function around your oHtml:getVar() to set NILs = chr(0) (smart practice in 
a character type only environment), then one radio var handles a one-to-many 
list of exclusive choices, and the data driven page generation routine can 
step through the appropriate dbf spitting out lines of html code.

> ...but I think if you name all your checkboxes equal (as with radio 
> button) you can use the same algorithm for both the checkbox solution, and 
> the radio-button.
If you name all the checkboxes the same, then you have to give them return 
values that are not boolean, which is contrary to their design of returning 
a logical TRUE.  You also have to identify which of the several identically 
identified checkboxes you have to set to "not checked" when you check this 
one, a tricky piece of code for me, because identifying them individually by 
name has been eliminated.  The beauty of the radio button is that the 
browser treats it as an array.  If you want to do that with checkboxes, you 
have to make your own array handler.

A lot of my business is in creating and running questionnaires, and then 
analyzing the collected data.  I want to know that a respondent actually 
looked at and answered a question, rather than be fooled into thinking he 
did because I have loaded in default values.  My "prior to submission 
routine" flags those questions which have not been answered to the 
responder.  Hopefully, I can get them to go back and correct their 
oversights.  Those that do come through blank will generate a followup phone 
call from me, and when I start dialing the phone, I start burning money.  I 
agree with you that we should make programs harmonize with how people do 
things, but my experience is that most people are not programmers; they do 
things quite differently from the ways you and I do.  And the next time you 
use a shoe for an impromptu hammer, realize that you have just become one of 
them.

Most respectfully,
Bruce

Bruce;

>>>...the same goes for radio-buttons, you can fiddle around to un push all, 
>>>but what's the state after that???
>oHtml:getVar() will return a NIL.  Assuming that you configure a wrapper 
>function around your oHtml:getVar() to set NILs = chr(0) (smart practice in 
>a character type only environment), then one radio var handles a one-to-many 
>list of exclusive choices, and the data driven page generation routine can 
>step through the appropriate dbf spitting out lines of html code.
Hmm, apparently I wasn't too clear. I meant this as a rhetorical
question  
but if 'none' is a valid choice, you should list it! I know what it
returns program wise (actually, nothing is returned!, that is what
causes html:getvar to give you back a nil, it could not find the
variable in the response, so it gives back a nil!



>> ...but I think if you name all your checkboxes equal (as with radio 
>> button) you can use the same algorithm for both the checkbox solution, and 
>> the radio-button.
>If you name all the checkboxes the same, then you have to give them return 
>values that are not boolean, which is contrary to their design of returning 
>a logical TRUE.  You also have to identify which of the several identically 
>identified checkboxes you have to set to "not checked" when you check this 
>one, a tricky piece of code for me, because identifying them individually by 
>name has been eliminated.  The beauty of the radio button is that the 
>browser treats it as an array.  If you want to do that with checkboxes, you 
>have to make your own array handler.
Well, I think it is easily done using just a few lines off JS, you
don't need a complete array handler, just traversing the DOM is
enough. I will write a sample for you to demonstrate it, but not right
now! leaving for a small trip within 48 hours. 


>A lot of my business is in creating and running questionnaires, and then 
>analyzing the collected data.  I want to know that a respondent actually 
>looked at and answered a question, rather than be fooled into thinking he 
>did because I have loaded in default values.  My "prior to submission 
>routine" flags those questions which have not been answered to the 
>responder.  Hopefully, I can get them to go back and correct their 
>oversights.  Those that do come through blank will generate a followup phone 
>call from me, and when I start dialing the phone, I start burning money.  I 
>agree with you that we should make programs harmonize with how people do 
>things, but my experience is that most people are not programmers; they do 
>things quite differently from the ways you and I do.  And the next time you 
>use a shoe for an impromptu hammer, realize that you have just become one of 
>them.
well Ok, here you have a valid reason to not select a default choice!
I still think you should go for checkboxes. I have stated my opinion
on that already, so I'm not going to repeat them now.

On the other side, you have a point on the shoe story  Damn, I'm one
off them.... 

regards
Sander
Regards
Sander Elias

-------------------------------------------------------------
xbHCL (http://www.xbHCL.com) the xbase HTML command layer
PBIH  (http://pbih.eu)       Polar Bear International Hosting
-------------------------------------------------------------
also a member off the XXP (http://www.xxp.nl)