Hello out there,
I am venturing into new territory and want to use RBC. One thing that
strikes me immediately is the handling of the time resp. x axis: All
examples I saw so far label it with the number of samples.
This strikes me as very odd. I am used to seeing scales like some
seconds or some hundreds of milli seconds - but with several thousand
data points.
Surely I am missing something here.
Any link to an example or other form of clarification will be greatly
appreciated.
Helmut

Helmut Giese <hgiese@ratiosoft.com> wrote:
> Hello out there,
> I am venturing into new territory and want to use RBC. One thing that
> strikes me immediately is the handling of the time resp. x axis: All
> examples I saw so far label it with the number of samples.
> This strikes me as very odd. I am used to seeing scales like some
> seconds or some hundreds of milli seconds - but with several thousand
> data points.
> Surely I am missing something here.
> Any link to an example or other form of clarification will be greatly
> appreciated.
> Helmut

I only find some "Royal Bank of ..." in Google. What RBC do you
have in mind?

Andreas Leitgeb <avl@logic.at> schrieb:

>Helmut Giese <hgiese@ratiosoft.com> wrote:
>> Hello out there,
>> I am venturing into new territory and want to use RBC. One thing that
>> strikes me immediately is the handling of the time resp. x axis: All
>> examples I saw so far label it with the number of samples.
>> This strikes me as very odd. I am used to seeing scales like some
>> seconds or some hundreds of milli seconds - but with several thousand
>> data points.
>> Surely I am missing something here.
>> Any link to an example or other form of clarification will be greatly
>> appreciated.
>> Helmut
>
>I only find some "Royal Bank of ..." in Google. What RBC do you
>have in mind?

My fault, I assumed it was common knowledge. RBC is a (the?) successor
to BLT - the great graphcical package from a George Howlett.
As far as I recall
- he worked on it all by himself
- never reacted to any attempt to contact him and
- eventually quit.
I believe BLT never made it into 8.5 - but certainly not into 8.6.
So some nice people took over and released at least parts of it as
'RBC'.
Helmut

On 3/24/24 06:24, Helmut Giese wrote:
> Hello out there,
> I am venturing into new territory and want to use RBC. One thing that
> strikes me immediately is the handling of the time resp. x axis: All
> examples I saw so far label it with the number of samples.
> This strikes me as very odd. I am used to seeing scales like some
> seconds or some hundreds of milli seconds - but with several thousand
> data points.
> Surely I am missing something here.
> Any link to an example or other form of clarification will be greatly
> appreciated.
> Helmut

I've not worked with RBC, but BLT used to be able to handle times on the
X-axis.

I think Ron Fox may have used RBC.

>> I am venturing into new territory and want to use RBC. One thing that
>> strikes me immediately is the handling of the time resp. x axis: All
>> examples I saw so far label it with the number of samples.
>> This strikes me as very odd. I am used to seeing scales like some
>> seconds or some hundreds of milli seconds - but with several thousand
>> data points.
>> Surely I am missing something here.
>> Any link to an example or other form of clarification will be greatly

Some fragments from an old trending program using rbc::graph imported into the ::t2k:: namespace.
A lot of the program is accessing data from a custom networked trend database (t2k is Trend2000)
that is no longer used anywhere that I know of. However it does show how times can be formatted for display on the x axis of an rbc/blt graph.

proc trend {win} {
global t2k

set g $win.g
destroy $g
::t2k::graph $g -background darkSlateGray -foreground lightgray \
-plotbackground black -plotrelief sunken
$g axis configure x -command {formatTimeTick} -color lightgray
$g legend configure -hide yes
$g pen configure activeLine -symbol "" -linewidth 3
$g grid on
pack $g -side left -expand yes -fill both
set t2k(win) $win
set t2k(graph) $g
set t2k(day,first) [date2Day [today]]
set t2k(day,last) $t2k(day,first)
t2k_ZoomStackTime $g
t2k_Crosshairs $g
}

# format time ticks as time and date

proc formatTimeTick {graph sec} {
if {[expr {$sec < 100000}]} {
return [format {%11.8g} $sec]
}
set ss [expr int($sec)]
set sf [expr {int(1000*($sec - $ss) + 0.5)}]
if {0 == $sf} { set sf ""} {set sf .[format %03d $sf]}
return "[clock format $ss -format {%T}]$sf\n[clock format $ss -format {%x}]"
}

t2k_ZoomStackTime and t2k_Crosshairs are from a modified library/graph.tcl from the rbc source.
and are not needed for a basic trend graph.

formatTimeTick displays date and time for large times, and fractions of seconds for short times.

Dave B

Hello Dave,
>Some fragments from an old trending program using rbc::graph imported into the ::t2k:: namespace.
>A lot of the program is accessing data from a custom networked trend database (t2k is Trend2000)
>that is no longer used anywhere that I know of. However it does show how times can be formatted for display on the x axis of an rbc/blt graph.
>
thanks a lot for the example. I didn't quite get it to work but I
beleive I have an idea now how to handle my "problem".
Best regards
Helmut

On 3/24/2024 6:04 AM, Helmut Giese wrote:
> Andreas Leitgeb <avl@logic.at> schrieb:
>
>> Helmut Giese <hgiese@ratiosoft.com> wrote:
>>> Hello out there,
>>> I am venturing into new territory and want to use RBC. One thing that
>>> strikes me immediately is the handling of the time resp. x axis: All
>>> examples I saw so far label it with the number of samples.
>>> This strikes me as very odd. I am used to seeing scales like some
>>> seconds or some hundreds of milli seconds - but with several thousand
>>> data points.
>>> Surely I am missing something here.
>>> Any link to an example or other form of clarification will be greatly
>>> appreciated.
>>> Helmut
>>
>> I only find some "Royal Bank of ..." in Google. What RBC do you
>> have in mind?
>
> My fault, I assumed it was common knowledge. RBC is a (the?) successor
> to BLT - the great graphcical package from a George Howlett.
> As far as I recall
> - he worked on it all by himself
> - never reacted to any attempt to contact him and
> - eventually quit.
> I believe BLT never made it into 8.5 - but certainly not into 8.6.
> So some nice people took over and released at least parts of it as
> 'RBC'.
> Helmut

There's a subset of the original BLT available for 8.6. I found it from a post on the wiki on page https://wiki.tcl-lang.org/page/BLT down near the bottom about getting it from an androwish distribution file. I then also found a 64 bit version as I wrote on that page.

Hi et99
>
>There's a subset of the original BLT available for 8.6. I found it from a post on the wiki on page https://wiki.tcl-lang.org/page/BLT down near the bottom about getting it from an androwish distribution file. I then also found a 64 bit version as I wrote on that page.
that's cool. Many thanks for this link
Helmut