Marco Mastria replied to the topic 'inditelescope.cpp left open files' in the forum. 4 months ago

I believe that this works for me:

diff --git a/libindi/libs/indibase/inditelescope.cpp b/libindi/libs/indibase/inditelescope.cpp
index 25e149bb..12b05bb4 100644
--- a/libindi/libs/indibase/inditelescope.cpp
+++ b/libindi/libs/indibase/inditelescope.cpp
@@ -1912,6 +1912,8 @@ const char *Telescope::LoadParkXML()
delXMLEle(ParkdataXmlRoot);

ParkdataXmlRoot = readXMLFile(fp, lp, errmsg);
+ fclose(fp);
+

delLilXML(lp);
if (!ParkdataXmlRoot)
@@ -2038,6 +2040,7 @@ bool Telescope::PurgeParkData()
delXMLEle(ParkdataXmlRoot);

ParkdataXmlRoot = readXMLFile(fp, lp, errmsg);
+ fclose(fp);

delLilXML(lp);
if (!ParkdataXmlRoot)
@@ -2555,6 +2558,7 @@ bool Telescope::LoadScopeConfig()
char ErrMsg[512];

RootXmlNode = readXMLFile(FilePtr, XmlHandle, ErrMsg);
+ fclose(FilePtr);
delLilXML(XmlHandle);
XmlHandle = nullptr;
if (!RootXmlNode)
@@ -2694,6 +2698,7 @@ bool Telescope::HasDefaultScopeConfig()
char ErrMsg[512];

RootXmlNode = readXMLFile(FilePtr, XmlHandle, ErrMsg);
+ fclose(FilePtr);
delLilXML(XmlHandle);
XmlHandle = nullptr;
if (!RootXmlNode)

Read More...

Marco Mastria created a new topic ' inditelescope.cpp left open files' in the forum. 4 months ago

I believe that this may affect all telescopes, as it is a base telescope class.

I'm pooling data from Indi each 10 seconds calling ind_getprop (without any parameters) on raspberry pi.
Each time ind_getprop runs, I noted that open files grows by 2. Few time later, I receive messages on Ekos that there's too many files open.

Using lsof, I could verify that the driver indi_ieq_telescope has 1024 open files: 3 pipes, 1 ttyUSB, sometimes few ParkData.xml and more then 1000 ScopeConfig.xml.

$ lsof -p $(ps -e|grep indi_ieq|awk '{print $1}')|egrep -v mem\|rtd\|txt\|cwd\|FD

Verifing inditelescope.cpp in GitHub I could check that there's certainly some places that the file ScopeConfig.xml is left open, but I'm not sure about ParkData.xml, I believe that this second file is ok.

Maybe need some fclose(FilePtr); in few places on inditelescope.cpp to solve this issue.

Thanks,
Mastria.

Read More...

Marco Mastria replied to the topic 'ieqprolegacydriver.h: No such file or directory' in the forum. 4 months ago

All building ok again. Thanks!

Read More...

Marco Mastria created a new topic ' ieqprolegacydriver.h: No such file or directory' in the forum. 5 months ago

I believe that someone forgot to commit the header file to github:

[ 64%] Building CXX object CMakeFiles/indi_ieqlegacy_telescope.dir/drivers/telescope/ieqprolegacydriver.cpp.o
/root/indi/libindi/drivers/telescope/ieqprolegacydriver.cpp:21:32: fatal error: ieqprolegacydriver.h: No such file or directory
#include "ieqprolegacydriver.h"
^
compilation terminated.

I verified on GitHub site and the file is not there, maybe it's only on another branch.

Mastria.

Read More...

Marco Mastria replied to the topic 'New INDI Atik driver - Feedback requested' in the forum. 1 year ago

Some inteface tests.

When first open focuser tab with Atik, the subframe is disabled "but" with a check-mark (I think this check is disconsidered when disabled, but I don't know). Then I selected the guider camera, that enable the box. It comes "selected", I unselect and switch back to Atik. Then Atik show the box disabled but unselected.



I didn't find the "preview" mode that need to be set in INDI Control Panel. It should be there?



The last picture is from Ekos, nothing to set there too.



I think that you are talking about an "internal" set that need to be checked on driver, not on visual interface (maybe I misunderstood).
I saw people talking about "mutex", is it something that I can "configure" ou it's a code change?
And the last thing, Jpaana says that 383L+ shows frame size as 3362x2537, in my case, same camera, 3354x2529 (firmware 801).

Thanks
Mastria.

Read More...

Marco Mastria replied to the topic 'New INDI Atik driver - Feedback requested' in the forum. 1 year ago

Mine is 383L+ too.
I've notice the same behavior with sub-framing not allowed while testing auto-focus.
I believe that donwload time get slower too, but maybe it's only a "felling", I don't have numbers to compare.

Read More...

Marco Mastria replied to the topic 'New INDI Atik driver - Feedback requested' in the forum. 1 year ago

psjshep,

Yes, it's not on 3rdparty building script.
I managed this (on raspbian) with following (as root, adapt with 'sudo' if you need):

ps: I cloned git repository in my home directory.

# build libatik

cd ~/indi/3rdparty/
mkdir build_libatik
cd build_libatik
cmake -DCMAKE_INSTALL_PREFIX=/usr . ../libatik
make
make install
# build driver

cd ~/indi/3rdparty/indi-atik/
mkdir build 
cd build 
cmake -DCMAKE_INSTALL_PREFIX=/usr .. 
make 
make install


Read More...

Marco Mastria replied to the topic 'Drivers dont load saved and/or default parameters at startup' in the forum. 1 year ago

The Vantage driver have parameters that can be configured. I saved the configuration and restarted the server, and none of them was loaded. After I clicked on button "Load", my configured parameters are loaded correctly:

I don't have a GPS dongle, so I setted location data. These data aren't loaded automaticly, only after click:



This other tab are very important! The default parameters need to be changed. For example, when "forecast" is 1 it's not a "lock" condition to me, but my rolloff-root don't let me unpark (weather warning). It should warning only when condition is 3 (in my situation):


Below, the status with default parameters, and the status after a clicke "Load" button.


By the way, I copied the Vantage_config.xml to Vantage_config.xml.default, so, they are equals with my configured parameters. None of them are loaded automatically.

Siimilar situations occurs with other drivers. As these settings are saved on driver's xml file, I don't undertood when you said "Vantage driver for example doesn't have any settings that it saves for itself". I'm wrong expecting this? So I realy need to use my python script to load the correct configuration? It's not a real problem to me, but the "general" users will not have the expertise to write code to just "click" the load button.

If you need I can verify this situation on others drivers. At the moment, I'm using 16 drivers, when possible I'll check all of them:

indi_armadillo_focus
indi_astrometry
indi_atik_ccd
indi_ieq_telescope
indi_ioptronv3_telescope
indi_meta_weather
indi_qhy_ccd
indi_rtlsdr_detector
indi_script_dome
indi_sx_wheel
indi_v4l2_ccd
indi_vantage_weather
indi_watchdog
indi_watcher_weather
indi_webcam_ccd
indi_wiringpi_gpio

Thanks,
Mastria.

Read More...

Marco Mastria replied to the topic 'Distributing the sequence of captures' in the forum. 1 year ago

Yes, I saw this today, and understood, I've never used the scheduler, I believe that I need to "previously save" a sequence on CCD tab and configure this in schedule tab after that. I was thinking on something simpler, straight on CCD tab where you configure the sequence, maybe only a box that I can say that I want this N times.

I'll give a try on schedule when weather permits.

Thanks.

Read More...

Marco Mastria created a new topic ' Drivers dont load saved and/or default parameters at startup' in the forum. 1 year ago

I've been using several indi drivers, and noted that some of them don't load saved parameters on startup.

I wrote a small python program that run just after drivers startup to "click" the LOAD button. After this, everything works ok.

I believe there's other drivers with this same situation, but at the moment I identified this 3:

- indi_vantage_weather.py
- indi_watcher_weather.py
- indi_wiringpi_gpio.py

I could not locate in code "when" the driver load the saved or default configuration.

Thanks,
Mastria

Read More...

Marco Mastria created a new topic ' Distributing the sequence of captures' in the forum. 1 year ago

In other softwares, i.e., ATIK Artemis, MaximDL, etc., there's an option to distribute the the captures sequence.
For example, if I want to get 10 frames from RGBL and D(ark), I can get this like Ekos do (1, 2, 3...10 R, then 1,2,...10 G, and so on), or I can do something like R1, G1, B1, D1, R2, G2,... Artemis permits, for example, to configure a sequence (for example, 1xL, 2xR, 2xG, 4xHA, anyway) and there's a parameter that I request that this sequence should be repeated N times. This is good a aproach, very versatile.

There's advantages and disvantages, as filters moves on each capture, maybe need for refocus, etc, but the main advantage in my opinion is that I can configure a bunch of frames to take, and leave all night long, so, at the end, I'll discard several (clouds, etc), but I'll end with aproximatelly the same number of frames of each filter/dark. In the normal Ekos sequence, sometimes I get 20 frames from R and G and the wheather changes and I can't take B for example.

This option exists on Ekos and I didn't locate that yet? (I know, I can create a large sequencing repeating each capture several times, but it's a kind of "dumb way").

If not, it could be a great improvment in Ekos in my opnion.
Thanks,
Mastria.

Read More...

Marco Mastria replied to the topic 'Can anybody recommend USB GPS dongle for RPi/StellarMate?' in the forum. 1 year ago

Not exactly about a dongle, but...
If my mount have a GPS (iOptron CEM60/120), can I configure the GPS source (on other drivers) to my mount? This works?
For exemple, in indi_watcher_weather or indi_vantage_weather (and others), on GPS property (snoop devices), there's a way to get the data from Mount's GPS?
Could it be just put the mount id in this field? If not, maybe this can be on a "wish list".

Read More...

Marco Mastria created a new topic ' last changes on dome_script.cpp inserted a new bug' in the forum. 1 year ago

Last change on dome_script.cpp removed the use of tmpnam leaving name string with invalid value - this is the name of temp file where SCRIPT_STATUS script will save the state.

I believe that this same situation will occur with telescope_script.cpp.

Attached the log with invalid file name reported.

Read More...