Hi list,
Long story short : I am using LM Studio as a local server to send
requests to an LLM model.
I use the shell command in LC 9.6 with curl on MacOS Sequoia 15.1.1.
The request looks like that :
curl http://127.0.0.1:1234/v1/embeddings \
-H "Content-Type: application/json" \
-d '{
s producing the erroneous third line, you may have to test the
range of each line.
Bob S
On Dec 30, 2024, at 8:13 AM, jbv via use-livecode
wrote:
Hi list,
Long story short : I am using LM Studio as a local server to send
requests to an LLM model.
I use the shell command in LC 9.6 with cur
xt"
}'
Le 2024-12-30 13:06, Mark Wieder a écrit :
On 12/30/24 08:13, jbv via use-livecode wrote:
Because of that I can't parse the json and I get an error with
JsonImport.
What am I doing wrong ?
How can I fix it ?
A couple of curl commandline options to try:
1. use the --si
Hi list,
Long story short : I am using LM Studio as a local server to send
requests to an LLM model.
I use the shell command in LC 9.6 with curl on MacOS Sequoia 15.1.1.
The request looks like that :
curl http://127.0.0.1:1234/v1/embeddings \
-H "Content-Type: application/json" \
-d '{
Le 2025-02-03 11:07, Bob Sneidar via use-livecode a écrit :
I heard that the base64encode puts a carriage return at a certain
interval of characters (not sure how many).
Bob S
Yes, that's true, but even when I removed the returns it didn't work.
__
Hi list,
Someone asked me privately, so I thought that maybe others
would be interested in a simple way to interact between LC
and AI models locally on their machine.
The method uses Ollama, which is available for Mac, Win and
Linux : https://ollama.com
1- Download and install Ollama (plenty of
Richard & Mark,
Thank you both for your suggestions.
I actually made some progress, as there were 2 different issues :
1- the syntax of the multimodal json curl must be
{"type" : "image_url", "image_url" : {"url":
"data:image/jpeg;base64,{_BASE64_}"}}
instead of
{"type" : "image_url", "i
I am stuck.
No matter what I try, I always get the same error :
"url" field must be a base64 encoded image
So I'm back to my very first question : how to base64encode
a jpeg file in LC to be used in a regular multimodal json
curl request...
Le 2025-02-03 19:00, Richard Gaskin via use-livecode
Hi Monte,
It works ! Thank you very much !
jbv
Le 2025-02-05 04:35, Monte Goulding via use-livecode a écrit :
Hi jbv
Here is my code for base64 encoding :
put URL ("file:" & "myfile.jpg") into timagedata
Use `binfile:` here instead of `file:` so the engine doesn’t treat the
data as nativ
Hi list,
Is it possible to close a tab of a website displayed
inside a browser widget ?
I am on Mac and I know this can be done using AppleScript,
but is there an LC-only solution ?
Thank you in advance.
jbv
___
use-livecode mailing list
use-livecode@
Hi list,
I need to base64encode a serie of jpeg files to include
the code in a curl request.
What is the best way to do that ?
Do I need to import each file into an image control and
base64encode the "text" of the image ?
Thank you,
jbv
___
use-liveco
Le 2025-02-02 21:10, Richard Gaskin via use-livecode a écrit :
jbv wrote:
I need to base64encode a serie of jpeg files to include
the code in a curl request.
What is the best way to do that ?
Depends. What's on the receiving end, and what options does it
support?
Actually, my question w
For the record, I tried to register at Azure to test their free offer,
but the registration includes a tel.# validation that doesn't work...
So I tried AWS for their free offer, registration went fine, but their
free offer doesn't include GPU usage.
So now I am trying Google Colab.
Regarding Appw
Hi again list,
Has anyone tried to do text embedding from LC on
a CPU-only device ?
Thank you in advance.
jbv
___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription
pre
Hi list,
I have a small LC desktop app that sends requests via curl
to ollama and some AI models. Everything runs fine on my Mac
and now I would like to try to move it to the cloud, but I'm
not sure which platform to choose.
I have read a few blogs like this one :
https://www.shiksha.com/online-c
Le 2025-03-17 09:27, Craig Newman via use-livecode a écrit :
If you place a simple one-liner in “more stuff”, like
“put random(99) into msg”
does that work? Another words, is it something in your own “more stuff”
that is the problem?
Craig,
It stops working.
It seems like messages aren't sen
Hi list,
Is there anything specific to consider/include
about datagrids when building a standalone, like
inclusions ?
When I look at the inclusions list in the
"standalone settings" dialog, I see "Datagrid"
but I thought that selecting "search for required
inclusions when saving the standalone" w
Hi list,
Here's what I am trying to do :
on opencard
send "handler1" to me with messages
end opencard
on handler1
send "handler2" to me in 200 millisec
-- more stuff
end handler1
on handler2
if condition then
send "handler2" to me in 200 millisec
end if
end handler2
The i
Le 2025-03-17 10:01, Alex Tweedly via use-livecode a écrit :
That sounds quite likely. If handler1 is running continuously, then
other messages will not happen.
I'd suggest putting various
wait for 0 seconds with messages
inside the running loops of handler1, in order that pending messages
Hi list,
I have a group containing several fields and 2 images,
and these images are of png format with some transparent
areas.
I need to detect "mouseup" events on these elements with
a "on mouseup" handler in the group script.
When I click on the transparent area of each image,
"mouseup" is det
Hi,
I am doing a lot of things as a hobbyist with LC and IA, either
locally or on a vps with ollama, and also with the OpenAI API.
That includes embedding, vision, syntaxic trees of sentences, etc.
A few days ago, I have posted on this forum the code I use to
interact with ollama via curl reques
Hi list,
I have a datagrid with dgStyle as "table".
I have populated the columns and the dgText.
Now I need to adjust the width of each column by script :
put "100,120,150" into Lw
set the dgColumnWidths of group "DataGrid 1" to Lw
But nothing changes when running the above commands.
What am I
Le 2025-03-06 19:04, Bob Sneidar via use-livecode a écrit :
Your supposition is incorrect. We are not ALL using these gadgets. I
for one never will.
I for one use these tools only for things I am not familiar with, like
python scripts
for instance, with the risk of using pieces of code I don't
Le 2025-03-03 16:05, Bob Sneidar via use-livecode a écrit :
What you are looking for is dgProp [“column widths”] which takes a
comma delimited list.
The API for datagrid properties is here:
https://lessons.livecode.com/m/datagrid/l/7343-data-grid-properties
Bob S
Thank you, that works.
New
Le 2025-03-04 13:07, Klaus major-k via use-livecode a écrit :
...
set the dgProp["sort by column"] of group "your DG here..." to empty
...
:-) I tried :
set the dgProp["sort by column"] of group "your DG here..." to false
Thank you.
jbv
___
use-live
Le 2025-03-04 13:07, Klaus major-k via use-livecode a écrit :
...
set the dgProp["sort by column"] of group "your DG here..." to empty
...
Well, actually it doesn't work as expected : the columns aren't sorted
when filled with data, but users can still click on column headers and
sort the data.
Hi list,
One more question : is there a way to hide the top row
of a datagrid when no column names need to be displayed ?
I tried
set the dgProp["columns"] of group "DataGrid 1" to empty
but it displays "col1. col2. etc" by default.
Thanks you in advance.
jbv
_
: how does auto-detection of
inclusions work ? Does it analyze the structure
of the stacks, or does it scan the scripts for
commands relative to libraries ?
Best,
Le 2025-05-24 09:13, jbv via use-livecode a écrit :
Hi list,
I am trying to update a stack created with LC 9.6.9 using
LiveCode
Hi list,
I am trying to update a stack created with LC 9.6.9 using
LiveCode Community 9.6.3 .
This stack uses various libraries : dialogs, xml, zip, internet,
table, datagrid...
In the IDE everything works as expected.
The problem begins with inclusions when I try to build standalones :
- if I c
Hi list,
I have just finalized an app that will process a lot
of text and images through AI models, with a main loop
running for days (or even weeks) that will send ten of
thousands of API requests to Ollama and/or Llama.cpp,
and feed the returned data into various DBs.
I have made numerous test
Hi list,
Has anyone tried to use a version of LiveCode Community
on a Mac M4 with Sequoia ?
Every installation I tried crashed at startup (9.6.2 fo
instance), although they work fine on iMac intel.
And Rosetta 2 doesn't seem to help...
Am I missing something ?
Thank you in advance.
jbv
Hi Klaus,
This should work :
put "ben=USER&password=PWD&sprach=de" into tData
post urlencode(tData) to URL "http://XXX/X.php";
jbv
Le 2025-07-06 10:17, Klaus major-k via use-livecode a écrit :
Hi friends,
quick question, how to format the data in LC for posting them to a PHP
script lik
Hi list,
When using a browser widget, I can get the whole code of a
page displayed in the browser with :
put the htmlText of widget "Browser" into myHtml
I can also run chunks of javascript inside the browser,
even if the code isn't there, for instance :
do "var T=document.getElementsByCla
Le 2025-06-27 15:47, Roger Guay via use-livecode a écrit :
I believe it was Richmond, but I’ve tried it on my M1 Mac and I
Couldn’t get it to work.
Did it crash at startup ? Did you use Rosetta ? Which Community version
dis you try ?
___
use-live
I think I read on this list or on the forum that someone
used it on M1 or M2 Mac. Perhaps with Rosetta ?
Le 2025-06-27 12:12, Bob Sneidar via use-livecode a écrit :
Did Community ever work on ARM based processors?
Bob S
On Jun 26, 2025, at 10:35 PM, jbv via use-livecode
wrote:
Hi list
Thank you Klaus.
I don't see how I could have missed that...
Le 2025-07-24 11:57, Klaus major-k via use-livecode a écrit :
Hi jbv,
uncheck "Backdrop" in menu "View", that should do the trick. :-)
___
use-livecode mailing list
use-livecode@lists.run
Hi list,
For the past few months at startup of the IDE, no matter
which version of LC I use, the backdrop is set to white
by default. And each time I have to set it to "none" in
the message box.
Is there a way to have the backdrop set to "none" by
default at startup ? I haven't found anything in
Hi list,
Let's say I have a portion of a script like this :
global myVar
put 0 into myVar
-- some code
repeat until myVar > 0
end repeat
-- some code
and the content of global myVar being changed
in a handler of another control :
on mouseUp
global myVar
put 1 into myVar
Hi list,
Years ago I built a desktop app with LC, and the client is still using
it.
Users of the app work in the same company but are scattered in different
cities. Since there were small differences between cities, I found it
easier to build a slightly different version of the app for each cit
201 - 239 of 239 matches
Mail list logo
