General and Desktop

[quote author="Andre" date="1317547623"]I'm not so sure about that. I think you can just spin a QEventLoop from any thread, QThread or not. Sure, QThread has it build in, but I don't think there is much magic involved there. Once you have a running eventloop, you can use slots in the thread, right?[/quote] Yes, it should be that way. If a QObject class is created inside a thread that is not a QThread (like the main thread) a pseudo QZhread object is created and it works.

I think that a faster way maybe a single function with a flag. You should only set the negation of the flag inside the slot and fire it every timer trigger. Then the flag indicate the way the slot function should follow.

Sure, it can be done using pure c++. Or you can use all the work already done by the Quick team and create a widget that is internally using QML to do the same with less work. Really, Quick is made for this kind of graphic UI stuff, and Quick 1 can be easily mixed with classic C++.

Personally, I think that you need to re-think that UI. Sounds like hell to use a UI like that. Why don't you put the 200 items in a list view themselves? You do realize that list views can also display checkboxes, don't you. That way, you can at least provide sorting and filtering on the list, so your users can actually find the items they need from it. If you use a model-view approach to begin with, then displaying a list that only contains the items checked in the other list is trivial. You only need a QSortFilterProxyModel to filter on the checkstate role.

This reads like a nice howto. Would you considder putting the above in a page on the wiki?

Sorry, I misread your post. A "cell in the table" usually triggers the concept of something like a [[Doc:QTableWidget]] in Qt world (and not only there). You were referring to a BLOB column of a database table, which make much more sense. In that case: Read the BLOB data into a [[Doc:QByteArray]], feed that as that data storage to a [[Doc:QBuffer]] and feed that to the QuaZip constructor mentioned by Vass.

There is no separate 32bit/64bit toolchain on the Mac. The toolchains differ only in compiler versions (gcc 4.0, 4.2, clang). All of them support 32bit/64bit and PowerPC/Intel object code simultaneously. You will have to set the appropriate combination by command line switches to gcc or in the .pro file for qmake: To build only 32bit targets use: @ CONFIG += x86 ppc CONFIG -= x86_64 ppc64 @ You can link your application against every lib that includes object code for the same architectures. It does not harm if you libs contain more architectures than you need.

The answers above explain a problem I came to ask about, so thank you. However, in the general case of a user-entered pattern I do not see any way of getting around this bad restriction. Suppose the pattern (example from QRegExp documentation) is <b>.*</b>, this cannot succeed when the document is: @ some text or <b>bold text</b> other etc. @ You could examine the user's pattern and change it into: @ <b>.$ ^.</b> @ But that would fail for the case of <b>on one line</b> and also would fail for the case of a target spanning three or more lines. In short I see no work-around for the general case of patterns matching across multiple lines. Comment?

If you don't want to use the checkable property of the button, your only way to achieve this is to keep the status of the animation (e.g., started = true|false) somewhere else, and each time you handle the clicked() signal you manually toggle the animation status. Another way is to provide two different slots, let say animation_start and animation_stop. In each slot you disconnect the slot itself from the signal handling and connect the other one, so each time the button is clicked one of the two slots handles the signal alternatively.

"Qt Bug Tracker":http://bugreports.qt.nokia.com/ has a section dedicated to Qt Creator. Check it out.

I believe the IM/GM guys and the other graphics library folks do know what they are doing. And in doubt, the surely are of more wisdom in that topic than we all together. This discussion is becoming too academic for my taste and I'm out now. Happy bit squeezing to everyone who wants to save some three bytes on his 8 GB workstation...

You can check whether your help collection is prepared correctly by loading it into Qt Assistant. This way you make sure the data you try to load is sane.

no, but you can install the event filter on all childs of a widget; e.g subclass QWidget overrinding method eventFilter() (see "there":http://doc.qt.nokia.com/4.2/eventsandfilters.html;) Create a widget of this new type (in qtdesigner create a qwidget and then use promote to... to change is type) put all widget you want inside it; get the list of widgets with "findChildren()":http://doc.qt.nokia.com/4.2/qobject.html#findChildren iterate over the list and use installEventFilter on each widget. You should be fine.

In Qt namespace documentation see Qt::WidgetAttribute type

you mention a label. Could be the edit has this label as a proxy. from the docs : bq. If first or second has a focus proxy, setTabOrder() correctly substitutes the proxy. also have a look at "setFocusProxy":http://doc.qt.nokia.com/4.7/qwidget.html#setFocusProxy

Any progress? I find it incredible that this bug - which makes the documentation on QSqlRelationalTableModel erroneous - has been marked as being fixed for a "future release" for 7 months!

now i know where is the problem...thx vass :D:D