I think it could be done without too much confusion by using preprocessor directives. That way, you don't actually mess with the code; you only change the symbolic constants you wish to affect. For example, if you don't use CW or CAT, then

#define CWINTERFACE ? ??????? false??????? // Set this to true if you want to use CW

#define CATINTERFACE????????? false??????? // Set this to true if you wish to use a CAT interface

and so on. The burden of getting everything right then falls on the programmer, not the user. The user would, however, be responsible for their own code additions/deletions.

As to how much space this saves will depend upon the programs kept/deleted and what you mean by "very much program space". For some users, 50 bytes might be enough to add something they want at the cost of deleting some existing feature. The answer to that's really a try-it-and-see answer.

Jack, W8TEE

On Monday, May 14, 2018, 12:10:53 PM EDT, Tim Gorman <tgorman2@...> wrote:

This is where problems will begin to occur.

Are you going to add in functionality to modify ubitx_menu.ino

on the fly? Or are you going to break up ubitx_menu.ino into a lot of

little pieces that can be included/excluded at compile time? How do you

tie including/excluding menu items to controlling the actual compile of

functions in the code?

If you just inhibit access to the alignment software after it is used it

doesn't lower the amount of memory used for the program unless you do

so with a recompile and reload. Is that really what we want the user to

have to do?

Does deleting a mode actually save very much program space? Most

functions are common between modes, you still have to transition

between receive and transmit, only some variables change value.

Does doing all this actually make it harder for the experimenter to

modify the code because it makes it more difficult to lay out all the

interactions between the code that might be affected? I know it's hard

enough for me already to trace through all the code when I want to

change something. Making it more difficult is not what I would want to

see.

tim ab0wr

On Mon, 14 May 2018 14:18:46 +0000 (UTC)

"Jack Purdum via Groups.Io" <jjpurdum=[email protected]> wrote:

> This is a good idea and might make it easier to peel off those

> features not desired. The source already lends itself to this

> approach. I have not studied it closely enough to know whether

> specific files (e.g., ubitx_keyer.ino, ubitx_cat.ino) can be taken

> out of the compile chain as it currently stands.

>

> Jack, W8TEE

>

>

>? ? On Monday, May 14, 2018, 10:05:13 AM EDT, K9HZ

> <bill@...> wrote:

>

> In fact, to expand on this… I think Ian should consider block defines

> for sections of the code that people want or don’t want.? He has

> already done this in selecting the type of display (2 line or 4

> line).? Some of the things to block define might be:

>

>? ?

>

> 1.?????? CAT

>

> 2.?????? WSPR

>

> 3.?????? Alignment ?(if you’ve done it once, why do it again?)

>

> 4.?????? CW? (some people only use SSB)

>

> 5.?????? SSB? (some people only use CW).

>

>? ?

>

> I’m sure there are others.? This way there is plenty of room if you

> want to experiment.? Just shut stuff off.? In fact, there could be an

> experiment define that turns on or shuts off your experiments too.

>

>? ?

>

> Just a thought…

>

>? ?

>

>? ?

>

> Dr. William J. Schmidt - K9HZ J68HZ 8P6HK ZF2HZ PJ4/K9HZ VP5/K9HZ

> PJ2/K9HZ

>

>? ?

>

> Owner - Operator

>

> Big Signal Ranch – K9ZC

>

> Staunton, Illinois

>

>? ?

>

> Owner – Operator

>

> Villa Grand Piton – J68HZ

>

> Soufriere, St. Lucia W.I.

>

> Rent it:

>

> Like us on Facebook!

>

>? ?

>

>