A couple of simple ergonomic adjustments

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
108 messages Options
1 ... 3456
Reply | Threaded
Open this post in threaded view
|

Re: Ergonomics - Rant

Simon Kirk
stephane ducasse <stephane.ducasse <at> free.fr> writes:
>
> Could you make a precise list of the possible improvements of  
> ecompletion?
>

Sure - I was planning to do that. As soon as I'm back doing some
Squeak in the next few days I'll start cataloguing things as I think of them.

Cheers,
S


Reply | Threaded
Open this post in threaded view
|

RE: Ergonomics - Rant

J J-6
In reply to this post by tblanchard



> Date: Mon, 16 Jul 2007 08:47:32 -0500

> From: [hidden email]
> To: [hidden email]
> Subject: Re: Ergonomics - Rant
>
> Damien, I agree with Todd on this one.
>
> While it is certainly a defensible position to say that the method
> should disappear because after editing, it doesn't match the browser's
> criteria, but I think it is less useful.
>
> That is, I think of the senders browser as being a filter on open not
> a continuously updating filter. Once open, I want to edit my code. I
> don't want it to keep on filtering.

This is part of the whole MVC paradigm, no?  When you change a method some updates will get generated and the window will know to check this method that changed, it does fit anymore so it's dropped.  I personally have never had this problem, because knowing this was MVC I always open a class browser on that method (it's either double click it or a right click menu, don't remember) if I want to change the code not to send the browsed message anymore *but* I still want to look at it after.


Missed the show?  Watch videos of the Live Earth Concert on MSN. See them now!

Reply | Threaded
Open this post in threaded view
|

RE: Ergonomics - Rant

J J-6
In reply to this post by tblanchard
Now I'm confused.  This has been the state of affairs since ST-80 or earlier no?  This isn't a morphic thing, the event stuff is in Object itself.




> From: [hidden email]

> To: [hidden email]
> Date: Mon, 16 Jul 2007 15:55:44 +0100
> Subject: RE: Ergonomics - Rant
>
> Is that new for 3.10, doesn't seem to happened in my 3.9...
>
> -----Original Message-----
> From: [hidden email]
> [mailto:[hidden email]] On Behalf Of Damien
> Cassou
> Sent: 16 July 2007 3:47 pm
> To: The general-purpose Squeak developers list
> Subject: Re: Ergonomics - Rant
>
>
> 2007/7/16, Gary Chambers <[hidden email]>:
> > Indeed, would be nice if it monitored new senders of the method
> > selector too!
>
>
> This is already done: the sender browser is informed when a new sender
> appears.
>
>
> > Though, as I know, that would be rather cpu heavy.
>
> This does not consume more. The notification system informs the browser when
> a method is changed or when a new method is created.
>
> > Best to treat it as a snapshot at the time it was asked for.
>
> --
> Damien Cassou
>
>


Don't get caught with egg on your face.    Play Chicktionary! 

Reply | Threaded
Open this post in threaded view
|

RE: Ergonomics - Rant

Schwab,Wilhelm K
In reply to this post by tblanchard
Jason,

There is room for debate on what it should do, but this one sounds like
a preference to me.  Options would be to drop, keep, or launch a new
browser on the method.

You want ergonomics?  We should move the Browse-it command to just under
inspect; I have to go to the second menu quite often.  With my current
project, I found a modal way to select a folder, but not files, other
than a chain of menus that (pro) show all hidden files, (con) cannot
hide them, force the user to resort to a more... option, and do not
allow sorting.  It screams for (sorry) modal dialog.  I created a really
cheesy looking "tool bar" composed to ordinary buttons because that was
pretty much the only option.  I have to "balance" the mouse to keep the
input focus in the method name field of the method finder; great tool,
pathetic interface.  I could go on.  The IDE, and sadly the basic pieces
available to give to a would-be user, are 25 years out of date.  How's
that for rant? :)

Bill



===========================
J J <azreal1977@...> azreal1977@...  wrote:

> Date: Mon, 16 Jul 2007 08:47:32 -0500
> From: david.mitchell@...
> To: squeak-dev@...
> Subject: Re: Ergonomics - Rant
>
> Damien, I agree with Todd on this one.
>
> While it is certainly a defensible position to say that the method
> should disappear because after editing, it doesn't match the browser's
> criteria, but I think it is less useful.
>
> That is, I think of the senders browser as being a filter on open not
> a continuously updating filter. Once open, I want to edit my code. I
> don't want it to keep on filtering.

This is part of the whole MVC paradigm, no?  When you change a method
some updates will get generated and the window will know to check this
method that changed, it does fit anymore so it's dropped.  I personally
have never had this problem, because knowing this was MVC I always open
a class browser on that method (it's either double click it or a right
click menu, don't remember) if I want to change the code not to send the
browsed message anymore *but* I still want to look at it after.

Wilhelm K. Schwab, Ph.D.
University of Florida
Department of Anesthesiology
PO Box 100254
Gainesville, FL 32610-0254

Email: [hidden email]
Tel: (352) 846-1285
FAX: (352) 392-7029


Reply | Threaded
Open this post in threaded view
|

RE: Ergonomics - Rant

J J-6
In reply to this post by tblanchard
I had this problem too all the time, until I discovered that if instead of putting the paren at the end, you just highlight the whole code that goes in parens and press the left one.  It puts both on around the highlighted code, so now it's gone from an irritation to something I miss when it's not there. :)




> From: [hidden email]

> Date: Tue, 17 Jul 2007 12:32:32 -0700
> To: [hidden email]
> Subject: Re: Ergonomics - Rant
>
> Oh, I'll chime in - it doesn't work the way I do and it drives me nuts.
>
> I often type most of a line, realize I need parens around part of it
> at the end of the line, put in the closing paren, then go back to put
> in the opening one - where it then insists on providing a matching
> closing one. In general, I often end up typing the closing quote/
> paren first, then go back to put in the opening ones.
>
> In other words, I never code left to right - I code like an artist
> builds a picture - starting with a sketch and gradually filling in
> detail jumping around. eCompletion is a huge impediment when you are
> doing this as you spend more time deleting spurious matching quotes
> and parens than you do just typing stuff.
>
> I'm quite willing to agree that I might not be typical - but that's
> how I work - successive approximations of proper code rather than
> left to right perfection.
>
> -Todd Blanchard
>
> On Jul 16, 2007, at 10:58 AM, Damien Cassou wrote:
>
> > 2007/7/16, Gary Chambers <[hidden email]>:
> >> Since we do business devlopment we like to keep things a bit more
> >> stable,
> >> though try to keep up with the main release.
> >> Besides, I find the eCompletion stuff a bit problematic (some of
> >> us use it
> >> though).
> >
> > What is problematic for you might be problematic for others too.
> > What is it?
> >
> > --
> > Damien Cassou
> >
>
>


Don't get caught with egg on your face.    Play Chicktionary! 

Reply | Threaded
Open this post in threaded view
|

Re: Ergonomics - Rant

Chris Muller-3
In reply to this post by Colin Putney
> What drives me nuts about them is that they impede the default
> behavior for typing when you already have a selection - deleting the
> selection and replacing it with what you type. If what you type
> happens to start with $(, you get very different behavior that if it
> starts with something else.

No it doesn't.  You have a selection, you want to replace it with
other text that begins with a parenthesis.  So you type the
parenthesis and the selected text is surrounded by parentheses, but
the text remains selected, so just keep typing, therefore the text is
then replaced with what you type.

Granted, it gives you the closing parenthesis too, but how often do
you want an opening without a closing?

 - Chris

Reply | Threaded
Open this post in threaded view
|

Re: Ergonomics - Rant

Chris Muller-3
In reply to this post by cbeler
> Actually, the problem I have is I don't know how to undo auto
> parens...is there a way to do ?

Use RbFormatter.  After coding any method, I now hit two key-strokes
to save it, the first one formats it, which also remove unnecessary
parenthesis, the second is Alt+s to save it.

Reply | Threaded
Open this post in threaded view
|

Re: Ergonomics - Rant

Colin Putney
In reply to this post by Chris Muller-3

On Jul 26, 2007, at 7:14 PM, Chris Muller wrote:

>> What drives me nuts about them is that they impede the default
>> behavior for typing when you already have a selection - deleting the
>> selection and replacing it with what you type. If what you type
>> happens to start with $(, you get very different behavior that if it
>> starts with something else.
>
> No it doesn't.  You have a selection, you want to replace it with
> other text that begins with a parenthesis.  So you type the
> parenthesis and the selected text is surrounded by parentheses, but
> the text remains selected, so just keep typing, therefore the text is
> then replaced with what you type.
>
> Granted, it gives you the closing parenthesis too, but how often do
> you want an opening without a closing?

Whenever you already have one. That seems to happen to me fairly  
often. The other annoying thing about it is that it breaks the  
behavior of cmd-J. Normally I can select text, type to replace it,  
then hit cmd-J a bunch of times to make similar edits. If the auto-
parenthesis stuff is triggered that doesn't work.

This is silly because we already *have* a command to wrap selected  
text in parentheses. This breakage doesn't even save a keystroke. It  
makes a three-key combination into a two-key combination. Totally  
gratuitous.

Colin

1 ... 3456