Greetings,



Any suckers^h^h^h^h^h^h^hbeta testers out there mind giving jalv2 2.5

beta a test drive?



<http://casadeyork.com/jalv2/archive/beta/jalv2.zip>

<http://casadeyork.com/jalv2/archive/beta/jalv2src.zip>



There aren't a lot of new features, but I rewrote the variable

allocator. Preliminary testing shows a 30 - 50% decrease in data

usage. Complex expressions no longer cause the data size to explode.



This code should have all of the released bug fixes also.



New features:



* BIT variables are all allocated out of a global bitbucket (formerly

each procedure had its own bitbucket). If the procedure is re-entrant,

or called through a function pointer, the bitbucket is local.



* The data saved on entry to an interrupt is the minimum necessary

(all of _pic_state is no longer automatically saved)



* PRAGMA FRAME

   Used in a function or procedure, allocates all of the variables in a

contiguous area (aka frame) to minimize bank set commands.



* PRAGMA SPEED

   Generate faster, but sometimes larger code. Currently only affects

left and right shift.



* PRAGMA SIZE

   Generate smaller, but somtimes slower code. Currently only affects

left and right shift.



Let me know if it works, I'd like to start using 2.5 instead of

maintaining the 2.4 base if possible. Please email bugs directly to me.



Thanks



--kyle

Greetings,



Apparently memory management isn't as forgiving in Windows as on

Linux, and I missed an initialization causing a core dump. A fixed

version has been posted for the six of you who've already downloaded it.



Sorry

--kyle

Greetings,



Solved one more problem -- the inline variable one. So...if you're

inclined please give it another go.



<http://www.casadeyork.com/jalv2/archive/beta/jalv2.zip>



--kyle

Hi Kyle. Sorry about the wrong information about JalV2 2.4g.

2.4g is good. It only gives trouble if you try to pragma inline a procedure

that calls another procedure.



In  format.jal if you insert a pragma inline in the _write_digit procedure

compiling with 2.4g will increase program code and reduce the hardware stack

by 1.



For version 2.5 programs appear to work but I cannot the lcd display to

work. Does that make sense? I mean flashing led flashes, digital outputs

turn on and off but the lcd display is blank so I can't verify the analog

measurements are correct.



On Tue, Jun 10, 2008 at 7:41 AM, zmafoox <kyle@...> wrote:



>   Greetings,

>

> Solved one more problem -- the inline variable one. So...if you're

> inclined please give it another go.

>

> <http://www.casadeyork.com/jalv2/archive/beta/jalv2.zip>

>

> --kyle

>

>

>





[Non-text portions of this message have been removed]

Kyle this compiles in 2.4g



Code area: 1063 of 4096 used

Data area: 49 of 240 used

Software stack available: 80 bytes

Hardware stack depth 3 of 8



and 2.5

Code area: 1061 of 4096 used

Data area: 44 of 0 used

Software stack available: 0 bytes

Hardware stack depth 3 of 8



but while the hex file in 2.4g flashes its led and displays12.8 on the lcd

the hex file in 2.5 just sits there doing nothing



--

-----------------------------------------------------------------------------

-- demo JAL v2: LCD_voltage

-- This demo measures an analog voltage and display the value on the LCD.

-- This demo also shows the powerful format routines,

-- which can send a formatted value to any output device.

-- The demo has become rather complex, because we want the demo the work on

-- all kind of PICs, include the 8-pin 12F675.

-- Because in case of 12F675 we've only 6 IO-pins,

-- and the LCD control needs at least 6 pins,

-- we're going to multiplex

--    the datacmd line of the LCD

--    with the analog input

-- In the test setting, a 10k potmeter was connected between GND and +5V,

-- and the slider of the potmeter was connected through a 5k resistor to the

-- analog input. Due to the load of the LCD, the lowest voltage on the ADC

-- will be approximately 0.5 Volt.

--

-----------------------------------------------------------------------------





--

-----------------------------------------------------------------------------

-- include hardware definitions

-- select the appropriate devive library

--

-----------------------------------------------------------------------------



include 16f685

pragma target clock     8_000_000

  pragma target fuses       0x33c4





  include pic_general

  include delay_any_mc

  include format





--

-----------------------------------------------------------------------------





--

-----------------------------------------------------------------------------

-- variable definition and initialisation

--

-----------------------------------------------------------------------------

-- create some space space to store the ADC value

var word ADC_value = 0

var word AD_value

var byte AX[2] at AD_value

var  word  volt

  var  word resistavg

  var bit my_LED1 is pin_c3

-- and set the io-pin as an output

pin_c3_direction = output



-- we need to save the AD settings,

-- to be able to turn it on again

var byte old_ADCON1 = ADCON1







-- define the IO-pins of the LCD

-- and set the correct IO-direction

var volatile bit HD44780_Enable_pin is pin_c6

var volatile bit HD44780_Enable_direction is pin_c6_direction

var volatile bit HD44780_DataCmd_pin is pin_c7

var volatile bit HD44780_DataCmd_direction is pin_c7_direction

var volatile bit HD44780_b4_pin is pin_B4

var volatile bit HD44780_b4_direction is pin_B4_direction

var volatile bit HD44780_b5_pin is pin_B5

var volatile bit HD44780_b5_direction is pin_B5_direction

var volatile bit HD44780_b6_pin is pin_B6

var volatile bit HD44780_b6_direction is pin_B6_direction

var volatile bit HD44780_b7_pin is pin_B7

var volatile bit HD44780_b7_direction is pin_B7_direction



-- define the library constants

-- in this case we set no multiplexing (although we multiplex)

-- because direction is set correct manually

;const HD44780_multiplexed_output       = false

;const LCD_number_of_display_lines      = 2

;const LCD_number_of_display_characters = 16

-- inlcude stringtables and the LCD library

; include HD44780_4_defs

  include lcd_44780



  procedure getvolt is

          ADCON0 = 0b10000001    ;  ' Right justify the result an0  volt

;   delay_10us(5)

;  ADCON0 = 0b10000011

    GO = true                          -- start conversion

   while GO loop end loop

   ;Repeat

;'DelayUS 10

;Until ADCON0 = 0b10000001

;   delay_10us(10)

AX[0] = ADRESL

AX[1] = ADRESH

;Ar(0)=ADRESL

;Ar(1)=ADRESH



end procedure





-- turn cursor off

LCD_cursor = off



-- write text to first line of the display

-- here the string table routine is used

LCD_cursor_pos = 0x00

;LCD_display_string(0,0)



-- write text to second line of the display

-- here the constant string method is used

-- define the constant string

const byte line_voltage[] = "Voltage =       "

-- define a loop counter, will be used as array index

var byte i

-- position the LCD cursor at the start

LCD_cursor_pos = 0x40

-- now loop for all characters in the string

for count(line_voltage) using i loop

   -- send each string element to the LCD

   -- (LCD_char has a auto-increment cursor function)

   LCD_char = line_voltage[i]

end loop

--

-----------------------------------------------------------------------------





--

-----------------------------------------------------------------------------

-- start of the main program

--

-----------------------------------------------------------------------------

forever loop

   -- turn the AD converter on,

   -- in general we could use the procedure ADC_on,

   -- but as the LCD enable pin is also one of the AD-channels,

   -- this IO-pin may not be changed,

   -- so we turn on the ADC manual

   ADCON1 = old_ADCON1

   -- and we only make channel 1 available

;  HD44780_DataCmd_direction = input



   -- this loop needs a general delay,

   -- because the LCD can't change too fast

   -- and this is the optimal place,

   -- because now the AD input pin has enough settling time

   delay_100ms(1)



   -- read the ADC as a 10 bit value

   -- (it doesn't matter if the ADC is set to low or high resolution)

  ; ADC_value = ADC_read(1)

    resistavg = 0

  For 64 loop ;k = 0 To 63 ;must be 64 times

  getvolt ;RAW = ADIn 1' ad_raw = ADIn 0

  resistavg = resistavg + AD_value

  end loop

volt =resistavg/119

   -- turn the ADC off

;  ADC_off

   -- and set A1 again to digital output

   HD44780_DataCmd_direction = output



   -- position the cursor for the voltage value

   LCD_cursor_pos = 0x4A



   -- here the user can select one of the format routines

   -- decimal formatting, total length = 5, 2 decimals

   format_word_dec(LCD_char,volt,5,1)

   -- hexadecimal formatting

   ;format_word_hex(LCD_char,ADC_value)

     my_LED1 = on

   -- wait for half a period

_usec_delay(1000000);  delay_100ms( half_blink_period )

   my_LED1 = off

_usec_delay(1000000);  delay_100ms( half_blink_period )

end loop

-- ---------------------------------





On Tue, Jun 10, 2008 at 7:47 PM, hyt <hoyungh@...> wrote:



> Hi Kyle. Sorry about the wrong information about JalV2 2.4g.

> 2.4g is good. It only gives trouble if you try to pragma inline a procedure

> that calls another procedure.

>

> In  format.jal if you insert a pragma inline in the _write_digit procedure

> compiling with 2.4g will increase program code and reduce the hardware stack

> by 1.

>

> For version 2.5 programs appear to work but I cannot the lcd display to

> work. Does that make sense? I mean flashing led flashes, digital outputs

> turn on and off but the lcd display is blank so I can't verify the analog

> measurements are correct.

>

> On Tue, Jun 10, 2008 at 7:41 AM, zmafoox <kyle@...> wrote:

>

>>   Greetings,

>>

>> Solved one more problem -- the inline variable one. So...if you're

>> inclined please give it another go.

>>

>> <http://www.casadeyork.com/jalv2/archive/beta/jalv2.zip>

>>

>> --kyle

>>

>>

>>

>

>





[Non-text portions of this message have been removed]

Greetings,



Excellent, thank you! This is exactly what I need to track down bugs &

I really appreciate your effort coming up with a minimal program. I'll

work on it later & let you know.



--- In jallist@yahoogroups.com, hyt <hoyungh@...> wrote:

>

> Kyle this compiles in 2.4g

>

> Code area: 1063 of 4096 used

> Data area: 49 of 240 used

> Software stack available: 80 bytes

> Hardware stack depth 3 of 8

>

> and 2.5

> Code area: 1061 of 4096 used

> Data area: 44 of 0 used

> Software stack available: 0 bytes

> Hardware stack depth 3 of 8

>

> but while the hex file in 2.4g flashes its led and displays12.8 on

the lcd

> the hex file in 2.5 just sits there doing nothing

>

> --

>

-----------------------------------------------------------------------------

> -- demo JAL v2: LCD_voltage

> -- This demo measures an analog voltage and display the value on the

LCD.

> -- This demo also shows the powerful format routines,

> -- which can send a formatted value to any output device.

> -- The demo has become rather complex, because we want the demo the

work on

> -- all kind of PICs, include the 8-pin 12F675.

> -- Because in case of 12F675 we've only 6 IO-pins,

> -- and the LCD control needs at least 6 pins,

> -- we're going to multiplex

> --    the datacmd line of the LCD

> --    with the analog input

> -- In the test setting, a 10k potmeter was connected between GND and

+5V,

> -- and the slider of the potmeter was connected through a 5k

resistor to the

> -- analog input. Due to the load of the LCD, the lowest voltage on

the ADC

> -- will be approximately 0.5 Volt.

> --

>

-----------------------------------------------------------------------------

>

>

> --

>

-----------------------------------------------------------------------------

> -- include hardware definitions

> -- select the appropriate devive library

> --

>

-----------------------------------------------------------------------------

>

> include 16f685

> pragma target clock     8_000_000

>  pragma target fuses       0x33c4

>

>

>  include pic_general

>  include delay_any_mc

>  include format

>

>

> --

>

-----------------------------------------------------------------------------

>

>

> --

>

-----------------------------------------------------------------------------

> -- variable definition and initialisation

> --

>

-----------------------------------------------------------------------------

> -- create some space space to store the ADC value

> var word ADC_value = 0

> var word AD_value

> var byte AX[2] at AD_value

> var  word  volt

>  var  word resistavg

>  var bit my_LED1 is pin_c3

> -- and set the io-pin as an output

> pin_c3_direction = output

>

> -- we need to save the AD settings,

> -- to be able to turn it on again

> var byte old_ADCON1 = ADCON1

>

>

>

> -- define the IO-pins of the LCD

> -- and set the correct IO-direction

> var volatile bit HD44780_Enable_pin is pin_c6

> var volatile bit HD44780_Enable_direction is pin_c6_direction

> var volatile bit HD44780_DataCmd_pin is pin_c7

> var volatile bit HD44780_DataCmd_direction is pin_c7_direction

> var volatile bit HD44780_b4_pin is pin_B4

> var volatile bit HD44780_b4_direction is pin_B4_direction

> var volatile bit HD44780_b5_pin is pin_B5

> var volatile bit HD44780_b5_direction is pin_B5_direction

> var volatile bit HD44780_b6_pin is pin_B6

> var volatile bit HD44780_b6_direction is pin_B6_direction

> var volatile bit HD44780_b7_pin is pin_B7

> var volatile bit HD44780_b7_direction is pin_B7_direction

>

> -- define the library constants

> -- in this case we set no multiplexing (although we multiplex)

> -- because direction is set correct manually

> ;const HD44780_multiplexed_output       = false

> ;const LCD_number_of_display_lines      = 2

> ;const LCD_number_of_display_characters = 16

> -- inlcude stringtables and the LCD library

> ; include HD44780_4_defs

>  include lcd_44780

>

>  procedure getvolt is

>          ADCON0 = 0b10000001    ;  ' Right justify the result an0  volt

> ;   delay_10us(5)

> ;  ADCON0 = 0b10000011

>    GO = true                          -- start conversion

>   while GO loop end loop

>   ;Repeat

> ;'DelayUS 10

> ;Until ADCON0 = 0b10000001

> ;   delay_10us(10)

> AX[0] = ADRESL

> AX[1] = ADRESH

> ;Ar(0)=ADRESL

> ;Ar(1)=ADRESH

>

> end procedure

>

>

> -- turn cursor off

> LCD_cursor = off

>

> -- write text to first line of the display

> -- here the string table routine is used

> LCD_cursor_pos = 0x00

> ;LCD_display_string(0,0)

>

> -- write text to second line of the display

> -- here the constant string method is used

> -- define the constant string

> const byte line_voltage[] = "Voltage =       "

> -- define a loop counter, will be used as array index

> var byte i

> -- position the LCD cursor at the start

> LCD_cursor_pos = 0x40

> -- now loop for all characters in the string

> for count(line_voltage) using i loop

>   -- send each string element to the LCD

>   -- (LCD_char has a auto-increment cursor function)

>   LCD_char = line_voltage[i]

> end loop

> --

>

-----------------------------------------------------------------------------

>

>

> --

>

-----------------------------------------------------------------------------

> -- start of the main program

> --

>

-----------------------------------------------------------------------------

> forever loop

>   -- turn the AD converter on,

>   -- in general we could use the procedure ADC_on,

>   -- but as the LCD enable pin is also one of the AD-channels,

>   -- this IO-pin may not be changed,

>   -- so we turn on the ADC manual

>   ADCON1 = old_ADCON1

>   -- and we only make channel 1 available

> ;  HD44780_DataCmd_direction = input

>

>   -- this loop needs a general delay,

>   -- because the LCD can't change too fast

>   -- and this is the optimal place,

>   -- because now the AD input pin has enough settling time

>   delay_100ms(1)

>

>   -- read the ADC as a 10 bit value

>   -- (it doesn't matter if the ADC is set to low or high resolution)

>  ; ADC_value = ADC_read(1)

>    resistavg = 0

>  For 64 loop ;k = 0 To 63 ;must be 64 times

>  getvolt ;RAW = ADIn 1' ad_raw = ADIn 0

>  resistavg = resistavg + AD_value

>  end loop

> volt =resistavg/119

>   -- turn the ADC off

> ;  ADC_off

>   -- and set A1 again to digital output

>   HD44780_DataCmd_direction = output

>

>   -- position the cursor for the voltage value

>   LCD_cursor_pos = 0x4A

>

>   -- here the user can select one of the format routines

>   -- decimal formatting, total length = 5, 2 decimals

>   format_word_dec(LCD_char,volt,5,1)

>   -- hexadecimal formatting

>   ;format_word_hex(LCD_char,ADC_value)

>     my_LED1 = on

>   -- wait for half a period

> _usec_delay(1000000);  delay_100ms( half_blink_period )

>   my_LED1 = off

> _usec_delay(1000000);  delay_100ms( half_blink_period )

> end loop

> -- ---------------------------------

>

>

> On Tue, Jun 10, 2008 at 7:47 PM, hyt <hoyungh@...> wrote:

>

> > Hi Kyle. Sorry about the wrong information about JalV2 2.4g.

> > 2.4g is good. It only gives trouble if you try to pragma inline a

procedure

> > that calls another procedure.

> >

> > In  format.jal if you insert a pragma inline in the _write_digit

procedure

> > compiling with 2.4g will increase program code and reduce the

hardware stack

> > by 1.

> >

> > For version 2.5 programs appear to work but I cannot the lcd

display to

> > work. Does that make sense? I mean flashing led flashes, digital

outputs

> > turn on and off but the lcd display is blank so I can't verify the

analog

> > measurements are correct.

> >

> > On Tue, Jun 10, 2008 at 7:41 AM, zmafoox <kyle@...> wrote:

> >

> >>   Greetings,

> >>

> >> Solved one more problem -- the inline variable one. So...if you're

> >> inclined please give it another go.

> >>

> >> <http://www.casadeyork.com/jalv2/archive/beta/jalv2.zip>

> >>

> >> --kyle

> >>

> >>

> >>

> >

> >

>

>

> [Non-text portions of this message have been removed]

>

Greetings,



I made a rather bone-headed mistake on the last beta. How it past

sanity is beyond me. Ah well, a new one is posted for the adventurous:



<http://www.casadeyork.com/jalv2/archive/beta/jalv2.zip>

<http://www.casadeyork.com/jalv2/archive/beta/jalv2src.zip>



--kyle

Good, I was trying to figure out what I had done wrong!



Gordon Williams



----- Original Message -----

From: "zmafoox" <kyle@...>

To: <jallist@yahoogroups.com>

Sent: Saturday, June 07, 2008 6:17 PM

Subject: [jallist] Re: jalv2 2.5 beta





Greetings,



Apparently memory management isn't as forgiving in Windows as on

Linux, and I missed an initialization causing a core dump. A fixed

version has been posted for the six of you who've already downloaded it.



Sorry

--kyle

Kyle,



I haven't checked out if everything works properly by these are the changes

in memory size:



2.4g ----> 2.5alpha

4477       4725   program

321         144    data



So there has been an increase in program size - I haven't played with any of

the switches.



Good decrease in data.



Regards,



Gordon Williams



----- Original Message -----

From: "zmafoox" <kyle@...>

To: <jallist@yahoogroups.com>

Sent: Saturday, June 07, 2008 6:17 PM

Subject: [jallist] Re: jalv2 2.5 beta





Greetings,



Apparently memory management isn't as forgiving in Windows as on

Linux, and I missed an initialization causing a core dump. A fixed

version has been posted for the six of you who've already downloaded it.



Sorry

--kyle

Just downloaded 2.5 to try



with 2.4g  got



Code area: 1230 of 3840 used

Data area: 96 of 255 used

Software stack available: 80 bytes

Hardware stack depth 5 of 8



with 2.5 got (20080607 version)



Code area: 1172 of 3840 used

Data area: 56 of 0 used

Software stack available: 0 bytes

Hardware stack depth 5 of 8



  program did not work. am using 16f685 with lcd display



On Sun, Jun 8, 2008 at 8:48 AM, Gordon Williams <g_will@...> wrote:



>   Kyle,

>

> I haven't checked out if everything works properly by these are the changes

> in memory size:

>

> 2.4g ----> 2.5alpha

> 4477 4725 program

> 321 144 data

>

> So there has been an increase in program size - I haven't played with any

> of

> the switches.

>

> Good decrease in data.

>

> Regards,

>

>

> Gordon Williams

>

> ----- Original Message -----

> From: "zmafoox" <kyle@... <kyle%40videovideo.com>>

> To: <jallist@yahoogroups.com <jallist%40yahoogroups.com>>

> Sent: Saturday, June 07, 2008 6:17 PM

> Subject: [jallist] Re: jalv2 2.5 beta

>

> Greetings,

>

> Apparently memory management isn't as forgiving in Windows as on

> Linux, and I missed an initialization causing a core dump. A fixed

> version has been posted for the six of you who've already downloaded it.

>

> Sorry

> --kyle

>

>

>





[Non-text portions of this message have been removed]

Hi Kyle,



Tried 2.5 beta. Program compiled OK but would not run. No display on

2 line LCD.



With 2.4d:  Code 3323 out of 8192 on a pic 16f877

             Data 91 of 368



With 2.5 beta: Code 3434 , Data 41

Back to 2.4d for the moment!



Cheers



Colin



--- In jallist@yahoogroups.com, "zmafoox" <kyle@...> wrote:

>

> Greetings,

>

> Any suckers^h^h^h^h^h^h^hbeta testers out there mind giving jalv2

2.5

> beta a test drive?

>

> <http://casadeyork.com/jalv2/archive/beta/jalv2.zip>

> <http://casadeyork.com/jalv2/archive/beta/jalv2src.zip>

>

> There aren't a lot of new features, but I rewrote the variable

> allocator. Preliminary testing shows a 30 - 50% decrease in data

> usage. Complex expressions no longer cause the data size to explode.

>

> This code should have all of the released bug fixes also.

>

> New features:

>

> * BIT variables are all allocated out of a global bitbucket

(formerly

> each procedure had its own bitbucket). If the procedure is re-

entrant,

> or called through a function pointer, the bitbucket is local.

>

> * The data saved on entry to an interrupt is the minimum necessary

> (all of _pic_state is no longer automatically saved)

>

> * PRAGMA FRAME

>   Used in a function or procedure, allocates all of the variables

in a

> contiguous area (aka frame) to minimize bank set commands.

>

> * PRAGMA SPEED

>   Generate faster, but sometimes larger code. Currently only affects

> left and right shift.

>

> * PRAGMA SIZE

>   Generate smaller, but somtimes slower code. Currently only affects

> left and right shift.

>

> Let me know if it works, I'd like to start using 2.5 instead of

> maintaining the 2.4 base if possible. Please email bugs directly to

me.

>

> Thanks

>

> --kyle

>

sorry Kyle.  Even blink led does not work with version 2.5



include 16f685

pragma target clock    4_000_000

  pragma target fuses       0x33c4





--

-----------------------------------------------------------------------------

-- variable definition and initialisation

--

-----------------------------------------------------------------------------

-- define the io-pin as a logical name

;var volatile bit my_LED1 is pin_A0

var bit my_LED1 is pin_c3

-- and set the io-pin as an output

pin_c3_direction = output



-- define the blink period in units of [100 msec]

const byte half_blink_period = 1

--

-----------------------------------------------------------------------------





--

-----------------------------------------------------------------------------

-- start of the main program

--

-----------------------------------------------------------------------------

forever loop

   my_LED1 = on

   -- wait for half a period

_usec_delay(1000000);  delay_100ms( half_blink_period )

   my_LED1 = off

_usec_delay(1000000);  delay_100ms( half_blink_period )

end loop



On Sun, Jun 8, 2008 at 4:57 PM, colin.tallis <colin.tallis@...> wrote:



>   Hi Kyle,

>

> Tried 2.5 beta. Program compiled OK but would not run. No display on

> 2 line LCD.

>

> With 2.4d: Code 3323 out of 8192 on a pic 16f877

> Data 91 of 368

>

> With 2.5 beta: Code 3434 , Data 41

> Back to 2.4d for the moment!

>

> Cheers

>

> Colin

>

>

> --- In jallist@yahoogroups.com <jallist%40yahoogroups.com>, "zmafoox"

> <kyle@...> wrote:

> >

> > Greetings,

> >

> > Any suckers^h^h^h^h^h^h^hbeta testers out there mind giving jalv2

> 2.5

> > beta a test drive?

> >

> > <http://casadeyork.com/jalv2/archive/beta/jalv2.zip>

> > <http://casadeyork.com/jalv2/archive/beta/jalv2src.zip>

> >

> > There aren't a lot of new features, but I rewrote the variable

> > allocator. Preliminary testing shows a 30 - 50% decrease in data

> > usage. Complex expressions no longer cause the data size to explode.

> >

> > This code should have all of the released bug fixes also.

> >

> > New features:

> >

> > * BIT variables are all allocated out of a global bitbucket

> (formerly

> > each procedure had its own bitbucket). If the procedure is re-

> entrant,

> > or called through a function pointer, the bitbucket is local.

> >

> > * The data saved on entry to an interrupt is the minimum necessary

> > (all of _pic_state is no longer automatically saved)

> >

> > * PRAGMA FRAME

> > Used in a function or procedure, allocates all of the variables

> in a

> > contiguous area (aka frame) to minimize bank set commands.

> >

> > * PRAGMA SPEED

> > Generate faster, but sometimes larger code. Currently only affects

> > left and right shift.

> >

> > * PRAGMA SIZE

> > Generate smaller, but somtimes slower code. Currently only affects

> > left and right shift.

> >

> > Let me know if it works, I'd like to start using 2.5 instead of

> > maintaining the 2.4 base if possible. Please email bugs directly to

> me.

> >

> > Thanks

> >

> > --kyle

> >

>

>

>





[Non-text portions of this message have been removed]

Greetings,



Thanks for this, I was just going to ask someone to send me a

non-working sample! I'll take a look.



--- In jallist@yahoogroups.com, hyt <hoyungh@...> wrote:

>

> sorry Kyle.  Even blink led does not work with version 2.5

>

> include 16f685

> pragma target clock    4_000_000

>  pragma target fuses       0x33c4

>

>

> --

>

-----------------------------------------------------------------------------

> -- variable definition and initialisation

> --

>

-----------------------------------------------------------------------------

> -- define the io-pin as a logical name

> ;var volatile bit my_LED1 is pin_A0

> var bit my_LED1 is pin_c3

> -- and set the io-pin as an output

> pin_c3_direction = output

>

> -- define the blink period in units of [100 msec]

> const byte half_blink_period = 1

> --

>

-----------------------------------------------------------------------------

>

>

> --

>

-----------------------------------------------------------------------------

> -- start of the main program

> --

>

-----------------------------------------------------------------------------

> forever loop

>   my_LED1 = on

>   -- wait for half a period

> _usec_delay(1000000);  delay_100ms( half_blink_period )

>   my_LED1 = off

> _usec_delay(1000000);  delay_100ms( half_blink_period )

> end loop

>

> On Sun, Jun 8, 2008 at 4:57 PM, colin.tallis <colin.tallis@...> wrote:

>

> >   Hi Kyle,

> >

> > Tried 2.5 beta. Program compiled OK but would not run. No display on

> > 2 line LCD.

> >

> > With 2.4d: Code 3323 out of 8192 on a pic 16f877

> > Data 91 of 368

> >

> > With 2.5 beta: Code 3434 , Data 41

> > Back to 2.4d for the moment!

> >

> > Cheers

> >

> > Colin

> >

> >

> > --- In jallist@yahoogroups.com <jallist%40yahoogroups.com>, "zmafoox"

> > <kyle@> wrote:

> > >

> > > Greetings,

> > >

> > > Any suckers^h^h^h^h^h^h^hbeta testers out there mind giving jalv2

> > 2.5

> > > beta a test drive?

> > >

> > > <http://casadeyork.com/jalv2/archive/beta/jalv2.zip>

> > > <http://casadeyork.com/jalv2/archive/beta/jalv2src.zip>

> > >

> > > There aren't a lot of new features, but I rewrote the variable

> > > allocator. Preliminary testing shows a 30 - 50% decrease in data

> > > usage. Complex expressions no longer cause the data size to explode.

> > >

> > > This code should have all of the released bug fixes also.

> > >

> > > New features:

> > >

> > > * BIT variables are all allocated out of a global bitbucket

> > (formerly

> > > each procedure had its own bitbucket). If the procedure is re-

> > entrant,

> > > or called through a function pointer, the bitbucket is local.

> > >

> > > * The data saved on entry to an interrupt is the minimum necessary

> > > (all of _pic_state is no longer automatically saved)

> > >

> > > * PRAGMA FRAME

> > > Used in a function or procedure, allocates all of the variables

> > in a

> > > contiguous area (aka frame) to minimize bank set commands.

> > >

> > > * PRAGMA SPEED

> > > Generate faster, but sometimes larger code. Currently only affects

> > > left and right shift.

> > >

> > > * PRAGMA SIZE

> > > Generate smaller, but somtimes slower code. Currently only affects

> > > left and right shift.

> > >

> > > Let me know if it works, I'd like to start using 2.5 instead of

> > > maintaining the 2.4 base if possible. Please email bugs directly to

> > me.

> > >

> > > Thanks

> > >

> > > --kyle

> > >

> >

> >

> >

>

>

> [Non-text portions of this message have been removed]

>

Thank you for Jalv2.4g.

the 16f685 include file is from http://www.robh.nl/jal/



On Sun, Jun 8, 2008 at 10:54 PM, zmafoox <kyle@...> wrote:



>   Greetings,

>

> Thanks for this, I was just going to ask someone to send me a

> non-working sample! I'll take a look.

>

>

> --- In jallist@yahoogroups.com <jallist%40yahoogroups.com>, hyt <hoyungh@...>

> wrote:

> >

> > sorry Kyle. Even blink led does not work with version 2.5

> >

> > include 16f685

> > pragma target clock 4_000_000

> > pragma target fuses 0x33c4

> >

> >

> > --

> >

> ----------------------------------------------------------

> > -- variable definition and initialisation

> > --

> >

> ----------------------------------------------------------

> > -- define the io-pin as a logical name

> > ;var volatile bit my_LED1 is pin_A0

> > var bit my_LED1 is pin_c3

> > -- and set the io-pin as an output

> > pin_c3_direction = output

> >

> > -- define the blink period in units of [100 msec]

> > const byte half_blink_period = 1

> > --

> >

> ----------------------------------------------------------

> >

> >

> > --

> >

> ----------------------------------------------------------

> > -- start of the main program

> > --

> >

> ----------------------------------------------------------

> > forever loop

> > my_LED1 = on

> > -- wait for half a period

> > _usec_delay(1000000); delay_100ms( half_blink_period )

> > my_LED1 = off

> > _usec_delay(1000000); delay_100ms( half_blink_period )

> > end loop

> >

> > On Sun, Jun 8, 2008 at 4:57 PM, colin.tallis <colin.tallis@...> wrote:

> >

> > > Hi Kyle,

> > >

> > > Tried 2.5 beta. Program compiled OK but would not run. No display on

> > > 2 line LCD.

> > >

> > > With 2.4d: Code 3323 out of 8192 on a pic 16f877

> > > Data 91 of 368

> > >

> > > With 2.5 beta: Code 3434 , Data 41

> > > Back to 2.4d for the moment!

> > >

> > > Cheers

> > >

> > > Colin

> > >

> > >

> > > --- In jallist@yahoogroups.com <jallist%40yahoogroups.com> <jallist%

> 40yahoogroups.com>, "zmafoox"

>

> > > <kyle@> wrote:

> > > >

> > > > Greetings,

> > > >

> > > > Any suckers^h^h^h^h^h^h^hbeta testers out there mind giving jalv2

> > > 2.5

> > > > beta a test drive?

> > > >

> > > > <http://casadeyork.com/jalv2/archive/beta/jalv2.zip>

> > > > <http://casadeyork.com/jalv2/archive/beta/jalv2src.zip>

> > > >

> > > > There aren't a lot of new features, but I rewrote the variable

> > > > allocator. Preliminary testing shows a 30 - 50% decrease in data

> > > > usage. Complex expressions no longer cause the data size to explode.

> > > >

> > > > This code should have all of the released bug fixes also.

> > > >

> > > > New features:

> > > >

> > > > * BIT variables are all allocated out of a global bitbucket

> > > (formerly

> > > > each procedure had its own bitbucket). If the procedure is re-

> > > entrant,

> > > > or called through a function pointer, the bitbucket is local.

> > > >

> > > > * The data saved on entry to an interrupt is the minimum necessary

> > > > (all of _pic_state is no longer automatically saved)

> > > >

> > > > * PRAGMA FRAME

> > > > Used in a function or procedure, allocates all of the variables

> > > in a

> > > > contiguous area (aka frame) to minimize bank set commands.

> > > >

> > > > * PRAGMA SPEED

> > > > Generate faster, but sometimes larger code. Currently only affects

> > > > left and right shift.

> > > >

> > > > * PRAGMA SIZE

> > > > Generate smaller, but somtimes slower code. Currently only affects

> > > > left and right shift.

> > > >

> > > > Let me know if it works, I'd like to start using 2.5 instead of

> > > > maintaining the 2.4 base if possible. Please email bugs directly to

> > > me.

> > > >

> > > > Thanks

> > > >

> > > > --kyle

> > > >

> > >

> > >

> > >

> >

> >

> > [Non-text portions of this message have been removed]

> >

>

>

>





[Non-text portions of this message have been removed]

Greetings,



--- In jallist@yahoogroups.com, hyt <hoyungh@...> wrote:

>

> Thank you for Jalv2.4g.

> the 16f685 include file is from http://www.robh.nl/jal/



First place I went :) I've reproduced the problem -- it has to do with

inline and variables specified to be `at' other variables. I'm working

on it....



>

> On Sun, Jun 8, 2008 at 10:54 PM, zmafoox <kyle@...> wrote:

>

> >   Greetings,

> >

> > Thanks for this, I was just going to ask someone to send me a

> > non-working sample! I'll take a look.

> >

> >

> > --- In jallist@yahoogroups.com <jallist%40yahoogroups.com>, hyt

<hoyungh@>

> > wrote:

> > >

> > > sorry Kyle. Even blink led does not work with version 2.5

> > >

> > > include 16f685

> > > pragma target clock 4_000_000

> > > pragma target fuses 0x33c4

> > >

> > >

> > > --

> > >

> > ----------------------------------------------------------

> > > -- variable definition and initialisation

> > > --

> > >

> > ----------------------------------------------------------

> > > -- define the io-pin as a logical name

> > > ;var volatile bit my_LED1 is pin_A0

> > > var bit my_LED1 is pin_c3

> > > -- and set the io-pin as an output

> > > pin_c3_direction = output

> > >

> > > -- define the blink period in units of [100 msec]

> > > const byte half_blink_period = 1

> > > --

> > >

> > ----------------------------------------------------------

> > >

> > >

> > > --

> > >

> > ----------------------------------------------------------

> > > -- start of the main program

> > > --

> > >

> > ----------------------------------------------------------

> > > forever loop

> > > my_LED1 = on

> > > -- wait for half a period

> > > _usec_delay(1000000); delay_100ms( half_blink_period )

> > > my_LED1 = off

> > > _usec_delay(1000000); delay_100ms( half_blink_period )

> > > end loop

> > >

> > > On Sun, Jun 8, 2008 at 4:57 PM, colin.tallis <colin.tallis@> wrote:

> > >

> > > > Hi Kyle,

> > > >

> > > > Tried 2.5 beta. Program compiled OK but would not run. No

display on

> > > > 2 line LCD.

> > > >

> > > > With 2.4d: Code 3323 out of 8192 on a pic 16f877

> > > > Data 91 of 368

> > > >

> > > > With 2.5 beta: Code 3434 , Data 41

> > > > Back to 2.4d for the moment!

> > > >

> > > > Cheers

> > > >

> > > > Colin

> > > >

> > > >

> > > > --- In jallist@yahoogroups.com <jallist%40yahoogroups.com>

<jallist%

> > 40yahoogroups.com>, "zmafoox"

> >

> > > > <kyle@> wrote:

> > > > >

> > > > > Greetings,

> > > > >

> > > > > Any suckers^h^h^h^h^h^h^hbeta testers out there mind giving

jalv2

> > > > 2.5

> > > > > beta a test drive?

> > > > >

> > > > > <http://casadeyork.com/jalv2/archive/beta/jalv2.zip>

> > > > > <http://casadeyork.com/jalv2/archive/beta/jalv2src.zip>

> > > > >

> > > > > There aren't a lot of new features, but I rewrote the variable

> > > > > allocator. Preliminary testing shows a 30 - 50% decrease in data

> > > > > usage. Complex expressions no longer cause the data size to

explode.

> > > > >

> > > > > This code should have all of the released bug fixes also.

> > > > >

> > > > > New features:

> > > > >

> > > > > * BIT variables are all allocated out of a global bitbucket

> > > > (formerly

> > > > > each procedure had its own bitbucket). If the procedure is re-

> > > > entrant,

> > > > > or called through a function pointer, the bitbucket is local.

> > > > >

> > > > > * The data saved on entry to an interrupt is the minimum

necessary

> > > > > (all of _pic_state is no longer automatically saved)

> > > > >

> > > > > * PRAGMA FRAME

> > > > > Used in a function or procedure, allocates all of the variables

> > > > in a

> > > > > contiguous area (aka frame) to minimize bank set commands.

> > > > >

> > > > > * PRAGMA SPEED

> > > > > Generate faster, but sometimes larger code. Currently only

affects

> > > > > left and right shift.

> > > > >

> > > > > * PRAGMA SIZE

> > > > > Generate smaller, but somtimes slower code. Currently only

affects

> > > > > left and right shift.

> > > > >

> > > > > Let me know if it works, I'd like to start using 2.5 instead of

> > > > > maintaining the 2.4 base if possible. Please email bugs

directly to

> > > > me.

> > > > >

> > > > > Thanks

> > > > >

> > > > > --kyle

> > > > >

> > > >

> > > >

> > > >

> > >

> > >

> > > [Non-text portions of this message have been removed]

> > >

> >

> >

> >

>

>

> [Non-text portions of this message have been removed]

>

sorry to bring more bad news  Kyle. Is something similar happening with

V2.4g?



Tried Gordon's trick of using pragma inline in the format procedures. With

format time word it actually reduced program size and the program worked but

with format word dec the program appeared to start and then hung up.



On Mon, Jun 9, 2008 at 5:02 AM, zmafoox <kyle@...> wrote:



>   Greetings,

>

>

> --- In jallist@yahoogroups.com <jallist%40yahoogroups.com>, hyt <hoyungh@...>

> wrote:

> >

> > Thank you for Jalv2.4g.

> > the 16f685 include file is from http://www.robh.nl/jal/

>

> First place I went :) I've reproduced the problem -- it has to do with

> inline and variables specified to be `at' other variables. I'm working

> on it....

>

> >

> > On Sun, Jun 8, 2008 at 10:54 PM, zmafoox <kyle@...> wrote:

> >

> > > Greetings,

> > >

> > > Thanks for this, I was just going to ask someone to send me a

> > > non-working sample! I'll take a look.

> > >

> > >

> > > --- In jallist@yahoogroups.com <jallist%40yahoogroups.com> <jallist%

> 40yahoogroups.com>, hyt

>

> <hoyungh@>

> > > wrote:

> > > >

> > > > sorry Kyle. Even blink led does not work with version 2.5

> > > >

> > > > include 16f685

> > > > pragma target clock 4_000_000

> > > > pragma target fuses 0x33c4

> > > >

> > > >

> > > > --

> > > >

> > > ----------------------------------------------------------

> > > > -- variable definition and initialisation

> > > > --

> > > >

> > > ----------------------------------------------------------

> > > > -- define the io-pin as a logical name

> > > > ;var volatile bit my_LED1 is pin_A0

> > > > var bit my_LED1 is pin_c3

> > > > -- and set the io-pin as an output

> > > > pin_c3_direction = output

> > > >

> > > > -- define the blink period in units of [100 msec]

> > > > const byte half_blink_period = 1

> > > > --

> > > >

> > > ----------------------------------------------------------

> > > >

> > > >

> > > > --

> > > >

> > > ----------------------------------------------------------

> > > > -- start of the main program

> > > > --

> > > >

> > > ----------------------------------------------------------

> > > > forever loop

> > > > my_LED1 = on

> > > > -- wait for half a period

> > > > _usec_delay(1000000); delay_100ms( half_blink_period )

> > > > my_LED1 = off

> > > > _usec_delay(1000000); delay_100ms( half_blink_period )

> > > > end loop

> > > >

> > > > On Sun, Jun 8, 2008 at 4:57 PM, colin.tallis <colin.tallis@> wrote:

> > > >

> > > > > Hi Kyle,

> > > > >

> > > > > Tried 2.5 beta. Program compiled OK but would not run. No

> display on

> > > > > 2 line LCD.

> > > > >

> > > > > With 2.4d: Code 3323 out of 8192 on a pic 16f877

> > > > > Data 91 of 368

> > > > >

> > > > > With 2.5 beta: Code 3434 , Data 41

> > > > > Back to 2.4d for the moment!

> > > > >

> > > > > Cheers

> > > > >

> > > > > Colin

> > > > >

> > > > >

> > > > > --- In jallist@yahoogroups.com <jallist%40yahoogroups.com><jallist%

> 40yahoogroups.com>

> <jallist%

> > > 40yahoogroups.com>, "zmafoox"

> > >

> > > > > <kyle@> wrote:

> > > > > >

> > > > > > Greetings,

> > > > > >

> > > > > > Any suckers^h^h^h^h^h^h^hbeta testers out there mind giving

> jalv2

> > > > > 2.5

> > > > > > beta a test drive?

> > > > > >

> > > > > > <http://casadeyork.com/jalv2/archive/beta/jalv2.zip>

> > > > > > <http://casadeyork.com/jalv2/archive/beta/jalv2src.zip>

> > > > > >

> > > > > > There aren't a lot of new features, but I rewrote the variable

> > > > > > allocator. Preliminary testing shows a 30 - 50% decrease in data

> > > > > > usage. Complex expressions no longer cause the data size to

> explode.

> > > > > >

> > > > > > This code should have all of the released bug fixes also.

> > > > > >

> > > > > > New features:

> > > > > >

> > > > > > * BIT variables are all allocated out of a global bitbucket

> > > > > (formerly

> > > > > > each procedure had its own bitbucket). If the procedure is re-

> > > > > entrant,

> > > > > > or called through a function pointer, the bitbucket is local.

> > > > > >

> > > > > > * The data saved on entry to an interrupt is the minimum

> necessary

> > > > > > (all of _pic_state is no longer automatically saved)

> > > > > >

> > > > > > * PRAGMA FRAME

> > > > > > Used in a function or procedure, allocates all of the variables

> > > > > in a

> > > > > > contiguous area (aka frame) to minimize bank set commands.

> > > > > >

> > > > > > * PRAGMA SPEED

> > > > > > Generate faster, but sometimes larger code. Currently only

> affects

> > > > > > left and right shift.

> > > > > >

> > > > > > * PRAGMA SIZE

> > > > > > Generate smaller, but somtimes slower code. Currently only

> affects

> > > > > > left and right shift.

> > > > > >

> > > > > > Let me know if it works, I'd like to start using 2.5 instead of

> > > > > > maintaining the 2.4 base if possible. Please email bugs

> directly to

> > > > > me.

> > > > > >

> > > > > > Thanks

> > > > > >

> > > > > > --kyle

> > > > > >

> > > > >

> > > > >

> > > > >

> > > >

> > > >

> > > > [Non-text portions of this message have been removed]

> > > >

> > >

> > >

> > >

> >

> >

> > [Non-text portions of this message have been removed]

> >

>

>

>





[Non-text portions of this message have been removed]

Kyle,



I've lost track of where you are with the current version of 2.5.  Are you

still working on solving a bug or is the current version bug free (as far as

you know)?



I'm currently having problems and I don't want to duplicate your effort

unnecessarily.



Regards,



Gordon Williams



----- Original Message -----

From: "zmafoox" <kyle@...>

To: <jallist@yahoogroups.com>

Sent: Monday, June 09, 2008 7:41 PM

Subject: [jallist] Re: jalv2 2.5 beta





Greetings,



Solved one more problem -- the inline variable one. So...if you're

inclined please give it another go.



<http://www.casadeyork.com/jalv2/archive/beta/jalv2.zip>



--kyle

Kyle,



Seems to be major problems with this version



Compilation in jaledit took 40 sec and then a box came up indicating that it

timed out.  There appears to be a proper asm file though.  Running it

through MPlab indicates that things aren't working.



I'll have another look tomorrow.



Regards,



Gordon Williams



----- Original Message -----

From: "zmafoox" <kyle@...>

To: <jallist@yahoogroups.com>

Sent: Monday, June 16, 2008 6:59 PM

Subject: [jallist] Re: jalv2 2.5 beta





Greetings,



I made a rather bone-headed mistake on the last beta. How it past

sanity is beyond me. Ah well, a new one is posted for the adventurous:



<http://www.casadeyork.com/jalv2/archive/beta/jalv2.zip>

<http://www.casadeyork.com/jalv2/archive/beta/jalv2src.zip>



--kyle

--- In jallist@yahoogroups.com, "Gordon Williams" <g_will@...> wrote:

>

> Kyle,

>

> Seems to be major problems with this version

>

> Compilation in jaledit took 40 sec and then a box came up indicating

that it

> timed out.  There appears to be a proper asm file though.  Running it

> through MPlab indicates that things aren't working.



Depending on the size of the program, and especially the number of

variables, the longer compile time is to be expected. I've not

optimized it yet, and it's rather greedy. If you compile with the

'-debug' flag, look for `interference graph create passes' and

`variable graph edges'. The larger these numbers, the longer the compile.



>

> I'll have another look tomorrow.

>

> Regards,

>

> Gordon Williams

>

> ----- Original Message -----

> From: "zmafoox" <kyle@...>

> To: <jallist@yahoogroups.com>

> Sent: Monday, June 16, 2008 6:59 PM

> Subject: [jallist] Re: jalv2 2.5 beta

>

>

> Greetings,

>

> I made a rather bone-headed mistake on the last beta. How it past

> sanity is beyond me. Ah well, a new one is posted for the adventurous:

>

> <http://www.casadeyork.com/jalv2/archive/beta/jalv2.zip>

> <http://www.casadeyork.com/jalv2/archive/beta/jalv2src.zip>

>

> --kyle

>

Hate to tell you this after all the time you spent on it. I'm not getting

much joy with it either.



With Rob's device files the program appears to cycle repeatedly from the

beginning to the line

forever loop

but this doesn't happen with Bert's device files. Even with bert's device

files although the program appears to work the variable values appear to be

wrong. Since the displayed variables appear unintelligible I can't think of

a way to check them. Strangely ascii text display correctly so I cannot

understand why figures appear  as graphics mixed with numbers.



On Tue, Jun 17, 2008 at 10:20 PM, zmafoox <kyle@...> wrote:



>   --- In jallist@yahoogroups.com <jallist%40yahoogroups.com>, "Gordon

> Williams" <g_will@...> wrote:

> >

> > Kyle,

> >

> > Seems to be major problems with this version

> >

> > Compilation in jaledit took 40 sec and then a box came up indicating

> that it

> > timed out. There appears to be a proper asm file though. Running it

> > through MPlab indicates that things aren't working.

>

> Depending on the size of the program, and especially the number of

> variables, the longer compile time is to be expected. I've not

> optimized it yet, and it's rather greedy. If you compile with the

> '-debug' flag, look for `interference graph create passes' and

> `variable graph edges'. The larger these numbers, the longer the compile.

>

> >

> > I'll have another look tomorrow.

> >

> > Regards,

> >

> > Gordon Williams

> >

> > ----- Original Message -----

> > From: "zmafoox" <kyle@...>

> > To: <jallist@yahoogroups.com <jallist%40yahoogroups.com>>

> > Sent: Monday, June 16, 2008 6:59 PM

> > Subject: [jallist] Re: jalv2 2.5 beta

> >

> >

> > Greetings,

> >

> > I made a rather bone-headed mistake on the last beta. How it past

> > sanity is beyond me. Ah well, a new one is posted for the adventurous:

> >

> > <http://www.casadeyork.com/jalv2/archive/beta/jalv2.zip>

> > <http://www.casadeyork.com/jalv2/archive/beta/jalv2src.zip>

> >

> > --kyle

> >

>

>

>





[Non-text portions of this message have been removed]

Usually compilation would take only 3-4 seconds with the 2.4 versions and

earlier 2.5 versions.



I will send you some info separately.



Regards,



Gordon Williams



----- Original Message -----

From: "zmafoox" <kyle@...>

To: <jallist@yahoogroups.com>

Sent: Tuesday, June 17, 2008 10:20 AM

Subject: [jallist] Re: jalv2 2.5 beta





--- In jallist@yahoogroups.com, "Gordon Williams" <g_will@...> wrote:

>

> Kyle,

>

> Seems to be major problems with this version

>

> Compilation in jaledit took 40 sec and then a box came up indicating

that it

> timed out.  There appears to be a proper asm file though.  Running it

> through MPlab indicates that things aren't working.



Depending on the size of the program, and especially the number of

variables, the longer compile time is to be expected. I've not

optimized it yet, and it's rather greedy. If you compile with the

'-debug' flag, look for `interference graph create passes' and

`variable graph edges'. The larger these numbers, the longer the compile.



>

> I'll have another look tomorrow.

>

> Regards,

>

> Gordon Williams

>

> ----- Original Message -----

> From: "zmafoox" <kyle@...>

> To: <jallist@yahoogroups.com>

> Sent: Monday, June 16, 2008 6:59 PM

> Subject: [jallist] Re: jalv2 2.5 beta

>

>

> Greetings,

>

> I made a rather bone-headed mistake on the last beta. How it past

> sanity is beyond me. Ah well, a new one is posted for the adventurous:

>

> <http://www.casadeyork.com/jalv2/archive/beta/jalv2.zip>

> <http://www.casadeyork.com/jalv2/archive/beta/jalv2src.zip>

>

> --kyle

>