Please start the program, reproduce the bug, and while the combo is doing whatever it isn't supposed to do click "pause" in the debugger. Then extract the stack trace(s) and upload them for us to inspect. It's impossible to tell what's going on without any information about what's being executed (i.e. no source, no stack).
Every time I click on the drop-down box "Combo Box", I start a thread.
Why do you do that?
Ah okay, I understand now what the problem is. I have no idea why you get that. My advice is to try it as follows:
The precompiled packages that come from Qt Company
In a clean virtual machine. Also one that's not localized.
Finally if 1 and 2 fail, try with a localized virtual machine
If it's present in case 1, 2 or 3, then I'd suggest filing a bug report on the tracker.
@SGaist I am not using Labview's serial port directly because of my code contain many background functionalities. It will take time to implement all in Labview. We want to just give readable data and few User interface to Labview. So user can directly use readable value and interface in Labview.
@SGaist Well, I think no (if I understand your question right), because all paths to the compiler was setted right. Currently I'm trying to do a clean sysroot on my PC x86_64 (just a Qt+glib+gstreamer) with needed libraries (one question with pkg-config and Qt no resolved for me, yet) for the cross-compilation of target apps. As result I will produce a bash-script for a building an arm-sandbox to cross-compile from x86-64 machine to the ARM (specially imx6) and publish it here.
Also thank you for your leading questions (because really currently available in the Ubuntu repos version of compiler is deprecated, I think)!
In addition to what @CKurdu said, you can't do much from one process to another process. At the most you can monitor the second program from first program using the QProcess finished() signal and do something in the first program. Not very clear about your objective as well.
What I have noticed is that drawText(..) for the first time is slow. After that it is very fast. As work-around, you can try like follows. Add one dummy method drawMe() like this. Call this method once before you call show() method. So now drawText(..) inside paintEvent(..) will be fast. It hardly takes anytime. So performance issue is only at startup. After that it is nothing/negligible.
@sierdzio uh really?
hadn‘t yet the time to download it. I was hoping for magical fixes of a couple of issues I had
I have a release date next week and I kind of wanted to use the new lte 😭
Try it out, maybe it'll work for you. I didn't yet have the time to check and report the issues, but it is mighty annoying that working apps suddenly have all their layouts completely broken - and no warnings or errors on the console. "Dear developer, guess what we've changed" kind of thing.
I think I tried it back in Beta1 stage and the app worked, so maybe it will be fixed soon. As far as I know, Qt 5.12.1 already has 200+ bug fixes prepared (if you wonder why on earth would they release 5.12.0 when they already have hundreds of patches ready... I guess they rushed it so they can release during Qt conference).
class MYDLL_EXPORT atgCategoriesManager: public QObject
This seems to be the most elegant way if you are working with CMake: the GENERATE_EXPORT_HEADER macro will generate that mydll_export.h file that keeps all that Microsoftish blah-blah, so your real code stays nice, clean and readable.
@dheerendrahttps://github.com/penk/qml-livereload - this is what I try to achieve: live QML reload on file change. QQuickView works here without window recreation - just changed content, ApplicationWindow and QQuickWindow forces to recreate Window on every file change which I try to avoid.
This is where I'm proposing to split the AppWIndows and Content separately.
Not clear how?
You change the content qml. It will update your AppWindow as well. In the QML what you have posted let me know what you would like to achieve with example.
I mention above: edit file and see results in real time on preview application like qmlscene.
The size of the download seems to be alright. However, I am not sure what targets are in those offline installers,since i never use them. I doubt that those are usable for an update of an existing installation. You probably have to do a new installation.
I was having the impression that you have to do a complete new install. When you have already an existing installation based on an online installer, I would suggest to wait a bit until the major run to update is over. As explained before, I am simply waiting a couple of days because probably the download resources are at the edge at the moment.
With the online installer you shall be able to pick a specific repository for update, The only repository link I found was for older versions.
As I said earlier, very difficult to get the status ethernet using UDP. UDP is fire & forget protocol. You need to build your own mechanism to monitor the status. It could be simple TCP connection to see if the connection alive with remote server. If not you need to get the interfaces status periodically through QNetworkInterface & check if the interface is up. You can use QNetworkInterface::flags() & QNetworkInterface::IsUp to check the status.
i was able to resolve issue, with the use of multipointtoucharea treating mouse click as single touch event and setting minimum and maximum touch points to zero to discard all mouse and touch events in disabled background.
i used this piece of code in place of mouse area:
@sierdzio Since I plan to release all sources under LGPL and/or apache I understand it should not be a problem. Anyone will be able to install the app without using the binary installer, I just want to offer it to inexperienced users to ease installation.
Yeah I tried Mouse Area problem , and worked for me. Now trying on finding successful completion of color animation.
Since I am a new user to the forum I can reply only once in an hour
Thanks for the help
A NULL value in a field that is a foreign key can be considered a design flaw of the database.
When I started to answer your question, that was my initial position (except I wouldn't say "flaw"). When I have used FKs I was never allowed NULL, and I thought that was actually required. However I Googled and plenty said many databases support it (e.g. MySQL) and it was regarded as "acceptable".
I never did think QSqlRelationalTableModel offered much over QSqlTableModel. It's really just a glorified combobox (model being the FK table) interface in the view. Exactly as 2009 says, a custom delegate could do the job. Just a shame that Qt's doesn't offer NULL handling option for what you want.