Dojo has a basic issue, amongst other things... with its dojo.dnd.move class. The issue which you can see here:
http://archive.dojotoolkit.org/nightly/dojotoolkit/dojo/tests/dnd/test_parent_constraints.html
Is that when you click on a node, and start dragging, the node itself jumps. It actually moves its position. This is extremely problematic and I was wondering if anyone has seen this happen before.
I am creating an application that requires the moving of nodes but it needs to be precise and thus can't have the initial jump.
Any help would be greatly appreciated.
This is caused by the following highlighted code in Mover.js:
https://github.com/dojo/dojo/blob/master/dnd/Mover.js#L91-92
The odd thing is, based on the comments, it seems like this code is causing the very problem it aims to prevent.
That said, if your body has padding: 0 applied to its style, this code shouldn't affect you. (You can test it on that test page by running document.body.style.padding = "0" in the console before dragging.)
You might want to enter a ticket on the Dojo bug tracker at http://bugs.dojotoolkit.org (or maybe search and see if one has already been entered for it).
Related
I'm trying to display multiple columns with a wxTreeListCtrl. I'm using wxWidgets 3.0.5 and developing with GTK, although I'm intending to use this code in OSX and Windows.
The screenshot below shows what I'm seeing (the application is a file browser, in case that's not obvious). The first column, Filename, is always sized to the width of the tree view. The other two columns are pushed off the right; you can see the scroll bar.
If I manually resize the window, the two columns pop into view, which makes me feel that the tree view hasn't been laid out properly, but the right-hand column's manually set width is always ignored and as I continue to resize the window it gradually gets larger and larger.
I've tried manually setting and getting the column widths, and it appears that the tree view control is mostly ignoring the numbers I give it --- if I call SetColumnWidth(...) and then GetColumnWidth(...) immediately afterwards, I get either 0 or an incorrect number.
The actual code is trivial --- this is all generated with wxformbuilder, so I'd expect it to work:
browserTree = new wxTreeListCtrl( browsePanel, wxID_ANY, wxDefaultPosition, wxDefaultSize, wxTL_MULTIPLE|wxBORDER_THEME );
browserTree->AppendColumn( wxT("Filename"), wxCOL_WIDTH_DEFAULT, wxALIGN_LEFT, 0 );
browserTree->AppendColumn( wxT("Size"), 100, wxALIGN_RIGHT, 0 );
browserTree->AppendColumn( wxT("Mode"), 100, wxALIGN_LEFT, 0 );
(I should add that this behaviour manifests in the wxformbuilder GUI viewer, too.)
The main thing which makes me suspicious is that the widget hierarchy is pretty complex, so it's possible one of the parents is doing something to upset the tree view, but I've tried moving the tree view into a new box sizer in a new frame and exactly the same thing happens there. But there's a screenshot of the hierarchy below in case that helps.
I have found some very old bugs describing this behaviour (e.g. this one: https://github.com/wxWidgets/wxWidgets/issues/17476) but they're all really old and have been fixed anyway. If this were a real bug I'd expect to see a lot more complaints, as it makes the widget essentially unusable; therefore, this must be something I'm doing wrong, probably something really basic.
But what?
Welp, five minutes later I have an answer. I'm going to post this here in case someone else has the same problem, even though it makes me look really stupid.
Short version: it is that five-year old bug I found. It's just that the version of wxWidgets on Debian, which I'm developing on, is so old it hasn't picked up the bugfix.
The good news is that wxWidgets 3.2, which does have the bugfix, is being packaged right now (last post August 28 2022: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=919903). But I don't think that'll help me much, so I'll need to find a workaround. At least now I know what's going on.
In order to refresh a part of the grid, i.e., when font or alignment changes, I was using the following approach:
wxRect rect1=CellToRect(TopLeft);
wxRect rect2=CellToRect(BottomRight);
wxRect r(rect1.GetTopLeft(), rect2.GetBottomRight());
RefreshRect(r);
This was refreshing only a part of the intended block and was not working correctly.
From the suggestions of intellisense I came across RefreshBlock function and it works correctly. I searched the docs and have not found any information on it. I wonder if it is not recommended to use RefreshBlock for some reason? What does RefreshBlock do, does it refresh a block (as the name suggests) or is it equivalent to Refresh?
I am using wxWidgets 3.2 on Win10.
Thanks in advance.
The function RefreshBlock() is indeed the best way to do what you want and it was only undocumented by mistake, i.e. we simply forgot to do it. I've added documentation for it only now, so it will only get included in 3.2.1, but you can still use it in your code, the function itself is available since 3.1.3.
It seems from the source code that, depending on the location of its parameters, RefreshBlock refreshes any of the following:
corner grid
frozen cols grid
frozen rows grid
main grid
Since the area I wanted to refresh was on the main grid the following approach works (the idea is similar to RefreshBlock's approach):
auto GridWnd = CellToGridWindow(TL);
wxRect rect = BlockToDeviceRect(TL, BR, GridWnd);
GetGridWindow()->RefreshRect(rect);
Now everything is refreshed correctly.
Notes:
If only RefreshRect(rect) is called, things will NOT work as expected.
Little experiment showed that BlockToDeviceRect(TL, BR) also works, therefore eliminating the need for auto GridWnd = CellToGridWindow(TL);
Every now and then my selenium tests randomly fail with an "element X not found" error message. I do a simple select by ID, eg.
click('sideBar_queryButton')
When I use the dom inspector, the element is there, so I wonder why selenium doesn't find it sometimes. When I run the same test again, it works or it breaks again, there seems to be no reliable way of reproducing it. Even tests which are there for ages seem to randomly break and then magically work again. Inserting a few sleep statements sometimes helps but not reliably. So I wonder if I'm using it incorrectly. Has anyone else had these problems with selenium and if so, how did you fix them?
Edit: I found it to be much more reliable to put some test markers in my pages and wait for them to appear. If you use asynchronous operations which might create race conditions in your tests, inserting a test marker into your html after you finish the operation worked pretty well for me. E.g.
$('<div>').addClass("testMarker").append("OpXYZFinished").appendTo($('#content'));
That way, you can do a simple "waitForTextPresent" to see if things worked out and its much more reliable than guessing the browser's loading state. The testmarker class needs to be formatted in a way that it is not visible to the user (e.g. font color == background color).
Thanks for all your comments. After some deeper digging on the net and
in our tests I finally found combining these statements instead of a
simple waitForPageToLoad to be the cure for our issues:
waitForPageToLoad('')
// wait until all ajax activity has ceased. That check's jQuery's $.active
waitForCondition('selenium.browserbot.getUserWindow().$.active == 0', 5000)
// wait a second for all JS to properly initialize
pause(1000)
There is still a pause in there which is somewhat ugly, but it does
the trick.
I was started to work on my first sample of the dijit.Tree control.
When it didn't work, I added a couple of dijit.form.Textbox'es to the screen, and they didn't work either, even though I have them working fine on a similar form.
Interesting enough, when I killed FireFox and restarted it, I was got a clear message in FireBug console on what was wrong with my tree control, and got it fixed.
But now I'm rather puzzled while the simple dijits like NumberTextBox that I added are not working.
My Samples:
1) http://3wcloud-com-provisioning-qa.appspot.com/testDijitDate - this works great
(except the initial date value not showing - I have a separate question open on that)
2) http://3wcloud-com-provisioning-qa.appspot.com/testDijitTree
I've done several "diff"s on the two files, and I kind find the difference that would cause one to work and one not to work.
Thanks,
Neal
You should put your dojo.require statements immediately in a script block, not wrapped in the onload callback. Also, your input tags should be HTML style without the XHTML style slash (you didn't declare XHTML in your document) though I doubt that's the problem. See if either of those things help.
After setting up a table model in Qt 4.4 like this:
QSqlTableModel *sqlmodel = new QSqlTableModel();
sqlmodel->setTable("Names");
sqlmodel->setEditStrategy(QSqlTableModel::OnFieldChange);
sqlmodel->select();
sqlmodel->removeColumn(0);
tableView->setModel(sqlmodel);
tableView->show();
the content is displayed properly, but editing is not possible, error:
QSqlQuery::value: not positioned on a valid record
I can confirm that the bug exists exactly as you report it, in Qt 4.5.1, AND that the documentation, e.g. here, still gives a wrong example (i.e. one including the removeColumn call).
As a work-around I've tried to write a slot connected to the beforeUpdate signal, with the idea of checking what's wrong with the QSqlRecord that's about to be updated in the DB and possibly fixing it, but I can't get that to work -- any calls to methods of that record parameter are crashing my toy-app with a BusError.
So I've given up on that idea and switched to what's no doubt the right way to do it (visibility should be determined by the view, not by the model, right?-): lose the removeColumn and in lieu of it call tableView->setColumnHidden(0, true) instead. This way the IDs are hidden and everything works.
So I think we can confirm there's a documentation error and open an issue about it in the Qt tracker, so it can be fixed in the next round of docs, right?
It seems that the cause of this was in line
sqlmodel->removeColumn(0);
After commenting it out, everything work perfectly.
Thus, I'll have to find another way not to show ID's in the table ;-)
EDIT
I've said "it seems", because in the example from "Foundations of Qt development" Johan Thelin also removed the first column. So, it would be nice if someone else also tries this and reports results.
I use Qt 4.6.1 in PyQt and the problem is still here. Removing "removeColumn(0)" solves the issue.