[SOLVED] Strange behaviour of Query-by-Example (QBE) fields

Hello,
Jdev 11g TP3.
We have a QBE-Block with input fields.
Follwing steps are performed:
1) Change to find mode: #{bindings.Find.execute}
2) Enter search criteria into the fields
3) execute query: #{bindings.Execute.execute}
4) change to find mode again --> previously entered search criterias are shown --> OK
5) clear the search criteria in all fields
6) execute query again
7) PROBLEM: query is executed with the OLD search criteria again.
If we change into query mode again, the OLD search criterias are shown again
(BUT not for ALL fields).
8) Even wildcards (* or %) doesn't change the situation.
But it's possible to enter a different (but not empty) search criteria into the field.
Details:
=====
For some of the QBE-Fields the search criteria can be cleared but for some of them not.
Differences between this fields:
a) mandatory fields (also part of the key in the view model and NOT NULL in DB) --> criteria can't be cleared
b) non-mandatory fields (also not part of the view model and NULL-able in DB) --> criteria can be cleared
I tried to change one of the mandatory fields to non-mandatory and to be not part of the key in the view model but behaviour doesn't change?
Can anybody confirm this behaviour as bug or is there a mistake on my side?
BR
Peter
Message was edited by:
hofespet

Seems to be solved in TP4.
br
Peter

Similar Messages

  • Strange behaviour after querying a friendly device name

    Hi,
    I have observed some very strange behaviour after querying the friendly name of an audio capturing device. This is my code:
    #include "stdafx.h"
    #include <mmdeviceapi.h>
    #include <Functiondiscoverykeys_devpkey.h>
    int main(int argc, char *argv[]) {
        HRESULT hr = CoInitializeEx(0, COINIT_APARTMENTTHREADED);
        if(FAILED(hr)) throw;
        IMMDeviceEnumerator *pIMMDeviceEnumerator = 0;
        hr = CoCreateInstance(__uuidof(MMDeviceEnumerator), 0, CLSCTX_INPROC_SERVER, __uuidof(IMMDeviceEnumerator), (LPVOID *) &pIMMDeviceEnumerator);
        if(FAILED(hr)) throw;
        IMMDeviceCollection *pIDeviceCollection = 0;
        hr = pIMMDeviceEnumerator->EnumAudioEndpoints(eCapture, DEVICE_STATE_ACTIVE, &pIDeviceCollection);
        if(FAILED(hr)) throw;
        IMMDevice *pIMMDevice = 0;
        hr = pIDeviceCollection->Item(0, &pIMMDevice);
        if(FAILED(hr)) throw;
        IPropertyStore *pIPropertyStore = 0;
        hr = pIMMDevice->OpenPropertyStore(STGM_READ, &pIPropertyStore);
        if(FAILED(hr)) throw;
        PROPVARIANT propVariant;
        PropVariantInit(&propVariant);
        pIPropertyStore->GetValue(PKEY_Device_FriendlyName, &propVariant);
        PropVariantClear(&propVariant);
        pIPropertyStore->Release();
        pIMMDevice->Release();
        pIDeviceCollection->Release();
        pIMMDeviceEnumerator->Release();
        CoUninitialize();
        return -1;
    Please ignore the strange error handling, it's just to make the example code short.
    The problem is the return value (as can be seen in the output window of Visual Studio after running the program or using "echo %errorlevel%" in cmd.exe). Certainly, one would expect to get -1. However, I get 0 instead of -1. What makes this
    really strange is that -1 is properly returned when the line "pIPropertyStore->GetValue..." is commented out. Why does this have any effect on the return value of the process? I have observed this on two different Window 7 machines with different
    soundcard configurations. However, I did another test on a Windows 8.1 machine, where everything worked correctly.
    Can anyone explain what's going on here?
    Thanks,
    Holger

    OK, here's what's going on.
    Everything goes fine until main() exits.
    At that point, mmdevapi.dll's DllMain is called with lpReserved set to a non-NULL value, which is to say, "don't bother to clean up."
    mmdevapi.dll decides to clean up anyway (for shame) and calls into some SetupAPI.dll functions.
    SetupAPI.dll is waiting on a critical section, but at that point ntdll.dll kicks in and says "you know what, the process is shutting down, and this is the last thread, so I can guarantee you this critical section will never be set." ntdll.dll then
    terminates the process, and main()'s exit code is lost.
    Matthew van Eerde

  • SOLVED:Strange behaviour:request shows results in Answers not in dashboard

    Hi all,
    I've come across somethings very strange, which I know has to do with the lowest level of detail of my organisation hierarchy (and the keys used at that level), but I cant understand why its doing the following:
    I have a table in a dashboard that can receive 'is prompted' filter values from a prompt on the same dashboard. When i created this request in answers, for test reasons i gave it values for a "centre" (lowest level of the organization hierarchy), and it displays resultos for my metrics.
    Now I place this request on a dashboard. I have removed my manual test filters, and left the four organization columns of the request as "is prompted", the lowest of these being the centre value. Now in the dashboard, i decide to enter a value directly for the centre dropdown (same value as i previously tested) in my prompt...i get no results found!!!
    The lowest level of this prompt has a key made of two atributes (center and center code), this allows me to "drill down" from the centre to the centre code, although centre code is at the same level. If i remove the drilldown capability, the problem is solved...im not quite sure why.
    Message was edited by:
    zjac

    The problem was with the actual data. WHen analyzing the query that was being done, i realised that one of the keys (the centre name) was using a text field that has spaces after the name of the centre (god knows why), so using a trim resolved the issue.
    I guess the other configuration was using the centre id which was not encountering the above problem.

  • [Solved]Strange behaviour of copy in File Managers (pcmanfm, nautlius)

    Hi, I'm having the weirdest problem I just can't get solved:
    With my default file manager (pcmanfm) I can't copy any files! Everytime I try it gives me the following error:
    "Cannot copy a folder into its sub folder"
    The same action from cli with cp command works without any problems.
    Moving a file with pcmanfm works without any problems though.
    I also tried nautilus and xfe. With thouse file managers I can copy file, but only if I copy them twice. So the first time I select a file and copy it, right click somewhere, the "paste" function is grayed out. If I copy the file again, I can paste...
    Any thoughts? It's driving me crazy.
    FYI: I'm running compiz standalone and installed this system just 2 days ago.
    Edit: It also seems to work with pcmanfm if I copy the file TWICE.
    Last edited by justl (2011-08-26 16:53:18)

    Ok so I kept digging and trying some stuff and it seems the problem is being caused by parcellite.
    If I quit parcellite copying files works again how it should.
    So I guess this is a parcellite bug? The strange thing is I run parcellite also on my other boxes and never had any problems so I'm not sure where this problem suddenly comes from..

  • Strange behaviour with query after adding rownum

    Hi,
    We have query which is similar to the one below
    select * from (inner query) where rownum <=25.
    The inner query return only one record but when the entire query is executed we get 2 records and the the second record does not have any data. If we execute the query without the where clause then the query returns only one records. There are many queries similar to the above one and all works fine. The same query works in in all the environment except production. We are using oracle 10g. It seems like a bug. Can somebody help in resolving this issue.
    Thanks,
    Manohar Puttaswamy

    Due to security policies we are not supposed to post the execution plan and the queries. I am extremely sorry for that. We have some workarounds. We have a distinct in the inner query. If we remove distinct and add group by in the query the query is working fine. And also we have an exists clause in the inner query if we replace the exists clause with an in clause then then also the query is executing fine. The main problem i am facing is i am not able to create a testcase.

  • Solved - Strange behaviour in NetBeans 5.5 (and SE 6)

    I am using NB 5.5 and I created a new JInternalFrame and choose Free Design then it created the Layout like following (works perfectly this way):
    javax.swing.GroupLayout caixaOcorrenciaLayout = new javax.swing.GroupLayout(caixaOcorrencia);
            caixaOcorrencia.setLayout(caixaOcorrenciaLayout);
            caixaOcorrenciaLayout.setHorizontalGroup(
                caixaOcorrenciaLayout.createParallelGroup(javax.swing.GroupLayout.Alignment.LEADING)
                .addGroup(javax.swing.GroupLayout.Alignment.TRAILING, caixaOcorrenciaLayout.createSequentialGroup()
                    .addGap(12, 12, 12)
                    .addGroup(caixaOcorrenciaLayout.createParallelGroup(javax.swing.GroupLayout.Alignment.TRAILING, false)
                        .addComponent(jLabel4, javax.swing.GroupLayout.Alignment.LEADING, javax.swing.GroupLayout.DEFAULT_SIZE, javax.swing.GroupLayout.DEFAULT_SIZE, Short.MAX_VALUE)
                        .addComponent(jLabel6, javax.swing.GroupLayout.Alignment.LEADING, javax.swing.GroupLayout.DEFAULT_SIZE, javax.swing.GroupLayout.DEFAULT_SIZE, Short.MAX_VALUE)
                        .addComponent(jLabel2, javax.swing.GroupLayout.Alignment.LEADING, javax.swing.GroupLayout.DEFAULT_SIZE, 86, Short.MAX_VALUE)
                        .addComponent(jLabel8, javax.swing.GroupLayout.Alignment.LEADING))
                    .addGap(16, 16, 16)
                    .addGroup(caixaOcorrenciaLayout.createParallelGroup(javax.swing.GroupLayout.Alignment.TRAILING)
                        .addComponent(cbOcorrencia, javax.swing.GroupLayout.Alignment.LEADING, 0, 461, Short.MAX_VALUE)
                        .addGroup(javax.swing.GroupLayout.Alignment.LEADING, caixaOcorrenciaLayout.createSequentialGroup()
                            .addComponent(campoNumeroNota, javax.swing.GroupLayout.PREFERRED_SIZE, 102, javax.swing.GroupLayout.PREFERRED_SIZE)
                            .addPreferredGap(javax.swing.LayoutStyle.ComponentPlacement.RELATED, 89, Short.MAX_VALUE)
                            .addComponent(jLabel3)
                            .addPreferredGap(javax.swing.LayoutStyle.ComponentPlacement.RELATED)
                            .addComponent(campoCNPJ, javax.swing.GroupLayout.PREFERRED_SIZE, 150, javax.swing.GroupLayout.PREFERRED_SIZE))
                        .addGroup(javax.swing.GroupLayout.Alignment.LEADING, caixaOcorrenciaLayout.createSequentialGroup()
                            .addPreferredGap(javax.swing.LayoutStyle.ComponentPlacement.RELATED)
                            .addGroup(caixaOcorrenciaLayout.createParallelGroup(javax.swing.GroupLayout.Alignment.LEADING)
                                .addComponent(campoTextoLivre, javax.swing.GroupLayout.Alignment.TRAILING, javax.swing.GroupLayout.DEFAULT_SIZE, 461, Short.MAX_VALUE)
                                .addGroup(caixaOcorrenciaLayout.createSequentialGroup()
                                    .addComponent(calData, javax.swing.GroupLayout.PREFERRED_SIZE, 121, javax.swing.GroupLayout.PREFERRED_SIZE)
                                    .addPreferredGap(javax.swing.LayoutStyle.ComponentPlacement.RELATED)
                                    .addComponent(jLabel5)
                                    .addPreferredGap(javax.swing.LayoutStyle.ComponentPlacement.RELATED)
                                    .addComponent(campoHora, javax.swing.GroupLayout.PREFERRED_SIZE, 65, javax.swing.GroupLayout.PREFERRED_SIZE)
                                    .addPreferredGap(javax.swing.LayoutStyle.ComponentPlacement.RELATED)
                                    .addComponent(jLabel7)
                                    .addPreferredGap(javax.swing.LayoutStyle.ComponentPlacement.RELATED)
                                    .addComponent(cbObservacao, javax.swing.GroupLayout.PREFERRED_SIZE, 155, javax.swing.GroupLayout.PREFERRED_SIZE)))))
                    .addGap(12, 12, 12))
                .addGroup(javax.swing.GroupLayout.Alignment.TRAILING, caixaOcorrenciaLayout.createSequentialGroup()
                    .addContainerGap(421, Short.MAX_VALUE)
                    .addComponent(btnConfirmar, javax.swing.GroupLayout.PREFERRED_SIZE, 155, javax.swing.GroupLayout.PREFERRED_SIZE)
                    .addContainerGap())
            );But when I added a JPanel to a JTabbedPane (in an old JInternalFrame I had already) and choose Free Design, it created the Layout as following:
    org.jdesktop.layout.GroupLayout painelOcorrenciaLayout = new org.jdesktop.layout.GroupLayout(painelOcorrencia);
            painelOcorrencia.setLayout(painelOcorrenciaLayout);
            painelOcorrenciaLayout.setHorizontalGroup(
                painelOcorrenciaLayout.createParallelGroup(org.jdesktop.layout.GroupLayout.LEADING)
                .add(painelOcorrenciaLayout.createSequentialGroup()
                    .add(20, 20, 20)
                    .add(jLabel42, org.jdesktop.layout.GroupLayout.PREFERRED_SIZE, 110, org.jdesktop.layout.GroupLayout.PREFERRED_SIZE)
                    .add(jComboBox1, org.jdesktop.layout.GroupLayout.PREFERRED_SIZE, 300, org.jdesktop.layout.GroupLayout.PREFERRED_SIZE))
            );I dont want to use JDesktop, I want to use the Swing package, but what can I do since NB does that (the codification) by it self?
    Thanks
    F.
    Message was edited by:
    Franzisk

    I dont want to use JDesktop, I want to use the Swing
    package, but what can I do since NB does that (the
    codification) by it self?I assume your question is that you don't want to use the org.jdesktop.layout.GroupLayout class, but the new Java6 javax.swing.GroupLayout. To switch between the two, do the following:
    1. Open the respective form.
    2. In the projects pane, open the sub-items of the form class (the small "+" before the entry).
    3. Click on the "Form" element (for example, if you have an "AboutDialog.java", there are the "AboutDialog" and the "Form AboutDialog" elements there.)
    4. In the properties of the "Form", switch the "Layout Generation Style" to "Standard Java 6 Code".
    This one had me quite a bit; I've had the converse problem, wanting to make an application compatible with Java 5. The thing is, it doesn't change anything for already-generated forms if you change this setting in the global options... and you can't change it manually, as you said, because it's in the generated code.

  • [SOLVED] Strange behaviour of XFce after pressing certain keys

    I noticed that when I press 'Ctrl' and hold it for a while all my currently opened windows will close and I am no longer able to access the Applications Menu on my panel until reboot.
    I also noticed, that when I press Ctrl+Alt+Delete combination my xfce-session is 'terminated' and I am in console environment (with a login prompt).
    Now, I don't know if this type of behaviour is normal, but I do not like it at all.
    I know Ctrl+Alt+Delete should lock the screen, but why I am in console then?
    And why after pressing 'Ctrl' all my currently opened Windows are gone? (or at least not visible and not minimized)
    Any ideas?
    Thanks,
    P.S. I have slim set up to auto-login my user account, but I do not think this is relevant.
    EDIT: Ok, fixed - I realized that I pressed some keys when I was changing the battery on my laptop and did not lock the screen :-) [Stupid me].
    Last edited by MatejLach (2011-10-21 13:35:20)

    swiergot wrote:(if memory serves)
    Well, it doesn't. To make things worse, my sight is not good either. The devices are present in /dev and I did change configuration.
    It turned out to be a permissions problem. I was doing some cleaning in /etc, getting rid of *.pacnew and *.pacsave and I apparently overwrote /etc/group so my user account lost its group assignments. The errors thrown by mplayer were a little confusing because they mentioned "no such file", while in reports from other users with similar problems I could clearly see "permission denied". Only when running mplayer with "-ao oss" it correctly reported "permission denied" on /dev/dsp.
    KDE must be simply assigning permissions when it detects a new device, perhaps it has something to do with ConsoleKit. It would also explain why you have to actually be in Konsole when modprobing because it wouldn't work in a text console even with KDE running in the background.
    Anyway, sorry for bothering. Have a nice day, or good night, depending on your timezone
    Last edited by swiergot (2009-11-22 08:38:27)

  • [SOLVED] Strange behaviour of gnome-terminal and xterm

    Hello,
    I've been on Archlinux for 2 days, and I have a strange problem with gnome-terminal and xterm. When I write a too long command on a line, the text continues from the beginning of this line erasing the text, instead of going to a new line. Do you have any idea of how to fix that?
    Thanks a lot!
    Last edited by crotte (2008-04-13 21:59:34)

    Seems to be in a fresh install, too. I didn't change any setting of bash and no PS1.
    I still can't reproduce it... I hate those errors. :-(
    Edit:
    Fixed by using zsh now.
    Last edited by Misery (2008-05-16 08:48:44)

  • [SOLVED] Strange behaviour of audio driver

    Hi,
    I'm having a very strange problem with sound on my Asus laptop. It started yesterday when I rebooted the machine after five days of running. During that time I didn't upgrade any important package nor did I change any configuration (if memory serves).
    So yesterday when I logged into KDE I got a message saying that audio devices were gone. When I tried mplayer to see if it's a KDE thing, it complained about non-existing sound card and it was actually true as there were no sound related files under /dev, even though all audio modules were loaded. I removed all of them and modprobed snd-hda-intel. KDE immediately detected audio devices and sound was back everywhere.
    Now the funny part. I rebooted again and before logging into KDE, I checked the sound in a console, no luck. I repeated what I did previously to bring the sound back, that is rmmod&modprobe. But this time it didn't work. Then I logged into KDE and tried again, guess what, sound was back!
    So what is it that KDE is doing? At the end exactly the same modules are loaded, the only difference is that hd-audio0 process is not running when the driver is loaded without KDE.
    Any ideas?
    Last edited by swiergot (2009-11-22 08:37:43)

    swiergot wrote:(if memory serves)
    Well, it doesn't. To make things worse, my sight is not good either. The devices are present in /dev and I did change configuration.
    It turned out to be a permissions problem. I was doing some cleaning in /etc, getting rid of *.pacnew and *.pacsave and I apparently overwrote /etc/group so my user account lost its group assignments. The errors thrown by mplayer were a little confusing because they mentioned "no such file", while in reports from other users with similar problems I could clearly see "permission denied". Only when running mplayer with "-ao oss" it correctly reported "permission denied" on /dev/dsp.
    KDE must be simply assigning permissions when it detects a new device, perhaps it has something to do with ConsoleKit. It would also explain why you have to actually be in Konsole when modprobing because it wouldn't work in a text console even with KDE running in the background.
    Anyway, sorry for bothering. Have a nice day, or good night, depending on your timezone
    Last edited by swiergot (2009-11-22 08:38:27)

  • Strange Behaviour of Query. Any Inputs?

    I have a query as follows:
    select t1.col1 from table1 t1 where t1.col2=(
    select t2.col1 from table2 t2 where t2.col2=(
    select t3.col1 from table3 t3 where t3.col2=somevalue));
    When executed, it gives me the expected results. But when I execute the subquery alone as
    select t2.col1 from table2 t2 where t2.col2=(
    select t3.col1 from table3 t3 where t3.col2=somevalue);
    it gives me the error as invalid column.
    How does the main query give me the correct results when the subquery gives an error? Can anyone explain this?
    Thanks in advance?

    Hello,
    Please write your actual query here, as I created such type of query and it is working fine here.
    Adi

  • [Solved] Strange behaviour when searhing files in Gnome 3.8 overview

    Searching for "spot" finds the file "The Ink Spots - Maybe.mp3" but searching for "maybe" does not. Any ideas?
    Last edited by discon (2013-06-13 07:51:38)

    OK, I think I got it. "maybe" is a stop word and, apparently, file names are treated as *text* so "maybe" is ignored.
    $ tracker-search spots
    Results:
      file:///home/.../The%20Ink%20Spots%20-%20Maybe.mp3
      The Ink Spots - Maybe.mp3
    $ tracker-search maybe
    Search term 'maybe' is a stop word.
    Stop words are common words which may be ignored during the indexing process.
    Results:
    $

  • Clearing VO dynamic Query-By-Example ViewCriteria

    Looking for some help with JDev 11.1.1.4.0, ADF BC + ADF Faces RC please.
    At it's simplest we're looking how to clear the query-by-example (QBE) filters on a VO without using the table UI component and not accidentally clearing other view criteria on the VO. To explain our use case first let me define the characteristics of our app. We have:
    a) A BTF based on fragments:
    a.1) The transaction/data control scope = Use Existing Transaction if Possible/Shared
    a.2) It contains a single fragment
    a.3) The fragment contains a table based on an ADF BC View Object (VO)
    a.4) The VO has a view criteria defined at design time, implemented via the Application Module (AM)
    a.5) The table allows the user to filter the data
    a.6) The BTF does not include any task flow returns or parent actions
    b) A consuming BTF based on pages:
    b.1) The transaction/data control scope = Always Begin New Transaction/Isolated
    b.2) It contains 2 pages with navigation rules to move between them (back/forward)
    b.3) The fragment from "a" is included on both pages in "b"
    b.4) When navigation occurs between the pages, it is invoked via buttons in this BTF, not the original BTF via parent actions or task flow returns. This aligns with point a.6. As such when navigations occurs between the 2 pages, the region/task flow binding instance of "a" in the region of the current page of "b" is prematurely terminated. The implication is the instance of "a" doesn't get a chance to do any processing before it is closed.
    Note that "a" and "b" are built in isolation of each other. "b" has no inherit knowledge about the components of "a" nor the ADF BC components. The implication is "b" cannot attempt to grab the table UI component, nor manipulate the ADF BC components as it doesn't know anything about the BTFs dropped on its pages.
    If the user opens the 1st page in "b" and enters a query-by-example criteria on the table from the region containing the 1st of instance of "a", then the user navigates to the 2nd page, the filter criteria are still set on the 2nd instance of "a". This is expected behaviour as "a" has data control scope = shared implying when the QBE criteria are set on page 1, they are written down to the VO, which when the user navigates to page 2, as it is sharing the same VO the QBE criteria are there to reuse/share on page 2.
    In our scenario what we want to do:
    1) Is in navigating between the 2 pages, we wish the query-by-example filter criteria to be cleared/reset, such that navigating between pages the filter criteria are empty/fresh
    2) However in navigating between the pages we must not lose the design time view criteria (and it's bind variable value) that has already been applied
    3) Maintain shared data control scope so we minimize the amount of database connections
    The common solution to clearing the QBE filter is as per Timo Hahn's following blog entry: http://tompeez.wordpress.com/2011/10/29/jdev-how-to-reset-a-filter-on-an-aftable/
    Yet that solution assumes the code has access to the table UI component. As we need the QBE criteria reset on navigation caused by "b", and such navigation causes a premature termination of "a" there's no chance for "a" to grab the table component and clear the filter criteria before "b" navigates away. The logical solution to this is on entering "a" again, to programatically reset the QBE criteria as part of the initial steps in the "a" BTF. But on entering "a" using Timo's solution, the method call doesn't have access to table UI component as the fragment isn't yet rendered, making this solution not viable.
    Another solution we've tried is calling applyViewCriteria(null) then reapplying our view criteria from a VO client interface method, but it appears we aren't clearing the QBE criteria. This worries me a bit because it looks like you can't discern the QBE view criteria from the normal view criteria unless you go through the table UI component methods.
    So how do we clear the QBE filter criteria programatically, without clearing (or at least restoring) the design time view criteria, and without relying on the table UI component to clear the QBE criteria?
    Phew, I hope that makes sense? Thanks for your help in advance.
    CM.
    PS. There is an assumption in the title of my post that table Query By Example criteria are stored as ADF BC View Object view criteria. Possibly it's a separate mechanism so please be careful this doesn't lead you astray in answering my post.

    Chris,
    I don't have time to follow this through right now, but I'm on to something.
    I was able to affect the filter via the binding container (at least it's not the UI). See this code:
      public void a(ActionEvent e)
       String s = "#{bindings.EmployeesView1Query.queryDescriptor}";
       Object o = JSFUtils.resolveExpression(s);
       FilterableQueryDescriptor f = (FilterableQueryDescriptor) o;
       f.removeCriterion(f.getCurrentCriterion());
      }Now, it doesn't do exactly what you want (it actually somehow makes no rows show up in the table), but I believe it's moving in the right direction.
    Am leaving for the office (5:40 am), so I'll pass this over to you to have a gander.
    John

  • Query By Example and security issues

    HI,
    I have started looking at security issues in our ADF application.
    Is the default implementation of Query By Example (QBE) on a table safe from Cross Site Scripting and SQL Injection?
    In other words, can a user enter some value in a QBE input field that can either:
    - execute a malicious script (CSS)
    Or
    - somehow change the underlying will change the SQL query
    I am more worried about SQL Injection as QBE takes input from a web user, and makes a corresponding SQL query to the database.
    Are there any ways to prevent any of these?
    Thanks

    Timo thanks for your answer.
    So far I am confident on the following (based on responses and other reading):
    1) default implementation of Query By Example (QBE) (e.g. search fields) is "safe /safer" from/on SQL injection issues.
    2) User entered data  via non QBE fields (I assume this is "For other input text you" Timo mentions) should by checked against special characters (> < etc) to "prevent " cross side scripting.
    However, should I do 2) for QBE filters on alphanumeric columns (default implementation) ? I can do it, but if I do it I would loose some searching functionality
    as >, < are valid wildcard characters.
    Thanks

  • [SOLVED] French Canadian keyboard strange behaviour

    Hello!
    I just upgraded to Gnome 2.30 in [testing] and my keyboard has a strange behaviour. With Gnome 2.28, my keyboard layout was set to Canadian French (legacy), and still is, though some of the keys do not behave the same. For instance, the accents "`", "¸" and ^ do not behave as usual, but rather give me some letter with that accent on it, rather than letting me choose the letter on which put the accent.
    Moreover, the SHIFT+3 combination gives me a # instead of a /, but not in all software. Google Chrome has the good behaviour, but gnome-terminal gives me a #.
    Do you understand any of this?
    I'll post the packages I have upgraded.
    [2010-04-05 01:09] upgraded openssl (0.9.8n-1 -> 1.0.0-2)
    [2010-04-05 01:09] upgraded python (2.6.5-2 -> 2.6.5-3)
    [2010-04-05 01:09] upgraded pygobject (2.20.0-1 -> 2.21.1-1)
    [2010-04-05 01:09] upgraded pygtk (2.16.0-2 -> 2.17.0-1)
    [2010-04-05 01:09] upgraded gnome-menus (2.28.0.1-1 -> 2.30.0-1)
    [2010-04-05 01:09] upgraded libsoup (2.28.2-1 -> 2.30.0-1)
    [2010-04-05 01:09] upgraded libldap (2.4.21-1 -> 2.4.21-2)
    [2010-04-05 01:09] upgraded libidl2 (0.8.13-1 -> 0.8.14-1)
    [2010-04-05 01:09] upgraded orbit2 (2.14.17-1 -> 2.14.18-1)
    [2010-04-05 01:09] upgraded gconf (2.28.0-1 -> 2.28.1-1)
    [2010-04-05 01:09] upgraded gnome-keyring (2.28.2-1 -> 2.30.0-1)
    [2010-04-05 01:09] upgraded heimdal (1.3.1-3 -> 1.3.2-1)
    [2010-04-05 01:09] upgraded libsoup-gnome (2.28.2-1 -> 2.30.0-1)
    [2010-04-05 01:10] upgraded gnome-icon-theme (2.28.0-1 -> 2.30.0-1)
    [2010-04-05 01:10] upgraded libgweather (2.28.0-1 -> 2.30.0-1)
    [2010-04-05 01:10] upgraded evolution-data-server (2.28.3.1-1 -> 2.30.0-2)
    [2010-04-05 01:10] upgraded gnome-desktop (2.28.2-1 -> 2.30.0-1)
    [2010-04-05 01:10] upgraded libwnck (2.28.0-1 -> 2.30.0-1)
    [2010-04-05 01:10] upgraded libunique (1.1.6-1 -> 1.1.6-2)
    [2010-04-05 01:10] upgraded gnome-disk-utility (2.28.1-1 -> 2.30.1-1)
    [2010-04-05 01:11] upgraded libarchive (2.8.3-1 -> 2.8.3-3)
    [2010-04-05 01:11] upgraded smbclient (3.5.1-1 -> 3.5.1-2)
    [2010-04-05 01:11] upgraded gvfs (1.4.3-1 -> 1.6.0-1)
    [2010-04-05 01:11] upgraded gnome-vfs (2.24.2-2 -> 2.24.3-2)
    [2010-04-05 01:11] upgraded libbonobo (2.24.2-1 -> 2.24.3-1)
    [2010-04-05 01:11] upgraded libgnome (2.28.0-2 -> 2.30.0-1)
    [2010-04-05 01:11] upgraded libart-lgpl (2.3.20-1 -> 2.3.21-1)
    [2010-04-05 01:11] upgraded libgnomecanvas (2.26.0-1 -> 2.30.1-1)
    [2010-04-05 01:11] upgraded libbonoboui (2.24.2-1 -> 2.24.3-1)
    [2010-04-05 01:11] upgraded librsvg (2.26.0-2 -> 2.26.2-1)
    [2010-04-05 01:11] upgraded gnome-panel (2.28.0-2 -> 2.30.0-1)
    [2010-04-05 01:11] upgraded libgnomeui (2.24.2-1 -> 2.24.3-1)
    [2010-04-05 01:11] upgraded gnome-python (2.28.0-1 -> 2.28.1-1)
    [2010-04-05 01:11] upgraded alacarte (0.12.4-2 -> 0.13.1-1)
    [2010-04-05 01:11] upgraded apache (2.2.15-1 -> 2.2.15-2)
    [2010-04-05 01:11] upgraded at-spi (1.28.1-1 -> 1.30.0-1)
    [2010-04-05 01:11] upgraded totem-plparser (2.28.2-1 -> 2.30.0-1)
    [2010-04-05 01:12] upgraded brasero (2.28.3-2 -> 2.30.0-1)
    [2010-04-05 01:12] upgraded icu (4.2.1-1 -> 4.4-1)
    [2010-04-05 01:12] upgraded brltty (4.1-2 -> 4.1-3)
    [2010-04-05 01:12] upgraded libgtop (2.28.0-1 -> 2.28.1-1)
    [2010-04-05 01:12] upgraded bug-buddy (2.28.0-1 -> 2.30.0-1)
    [2010-04-05 01:12] upgraded cheese (2.28.1-1 -> 2.30.0-1)
    [2010-04-05 01:12] upgraded clutter (1.0.10-1 -> 1.2.4-1)
    [2010-04-05 01:12] upgraded clutter-gtk (0.10.2-1 -> 0.10.2-2)
    [2010-04-05 01:12] upgraded libxklavier (4.0-1 -> 5.0-1)
    [2010-04-05 01:12] upgraded libgnomekbd (2.28.2-1 -> 2.30.0-1)
    [2010-04-05 01:12] upgraded gnome-settings-daemon (2.28.2-1 -> 2.30.0-1)
    [2010-04-05 01:12] upgraded zenity (2.28.0-1 -> 2.30.0-1)
    [2010-04-05 01:13] upgraded metacity (2.28.1-1 -> 2.30.0-2)
    [2010-04-05 01:13] upgraded gnome-control-center (2.28.1-1 -> 2.30.0-1)
    [2010-04-05 01:13] upgraded compiz-decorator-gtk (0.8.4-2 -> 0.8.4-3)
    [2010-04-05 01:13] upgraded libcups (1.4.2-5 -> 1.4.3-2)
    [2010-04-05 01:13] upgraded openslp (1.2.1-2 -> 1.2.1-3)
    [2010-04-05 01:13] upgraded cups (1.4.2-5 -> 1.4.3-2)
    [2010-04-05 01:13] upgraded curl (7.20.0-1 -> 7.20.0-2)
    [2010-04-05 01:13] upgraded cvs (1.11.23-4 -> 1.11.23-5)
    [2010-04-05 01:13] upgraded postgresql-libs (8.4.3-1 -> 8.4.3-2)
    [2010-04-05 01:13] upgraded cyrus-sasl-plugins (2.1.23-1 -> 2.1.23-2)
    [2010-04-05 01:13] upgraded cyrus-sasl (2.1.23-3 -> 2.1.23-4)
    [2010-04-05 01:13] upgraded dbus-python (0.83.0-2.1 -> 0.83.1-1)
    [2010-04-05 01:13] upgraded gnome-python-desktop (2.28.0-3 -> 2.30.0-1)
    [2010-04-05 01:13] upgraded deskbar-applet (2.28.0-3 -> 2.30.0-1)
    [2010-04-05 01:13] upgraded dnsutils (9.6.1-2 -> 9.6.1-3)
    [2010-04-05 01:14] upgraded eog (2.28.2-2 -> 2.30.0-1)
    [2010-04-05 01:14] upgraded libwebkit (1.1.15.4-3 -> 1.1.90-2)
    [2010-04-05 01:14] upgraded gobject-introspection (0.6.6-1 -> 0.6.9-1)
    [2010-04-05 01:14] upgraded epiphany (2.28.2-2 -> 2.30.0-1)
    [2010-04-05 01:14] upgraded evince (2.28.2-1 -> 2.30.0-1)
    [2010-04-05 01:14] upgraded gtkhtml (3.28.3-1 -> 3.30.0-1)
    [2010-04-05 01:14] upgraded evolution-exchange (2.28.3-1 -> 2.30.0-2)
    [2010-04-05 01:14] upgraded evolution-webcal (2.28.0-1 -> 2.28.1-1)
    [2010-04-05 01:14] upgraded file-roller (2.28.2-1 -> 2.30.0-1)
    [2010-04-05 01:15] upgraded gcalctool (5.28.2-1 -> 5.30.0-1)
    [2010-04-05 01:15] upgraded gconf-editor (2.28.0-1 -> 2.30.0-1)
    [2010-04-05 01:15] upgraded gnome-session (2.28.0-1 -> 2.30.0-1)
    [2010-04-05 01:15] upgraded gdm (2.28.2-1 -> 2.30.0-1)
    [2010-04-05 01:15] upgraded gtksourceview2 (2.8.2-1 -> 2.10.0-1)
    [2010-04-05 01:15] upgraded pygtksourceview2 (2.8.0-1 -> 2.10.0-1)
    [2010-04-05 01:15] upgraded gedit (2.28.3-1 -> 2.30.0-1)
    [2010-04-05 01:15] upgraded git (1.7.0.3-1 -> 1.7.0.4-1)
    [2010-04-05 01:15] upgraded libsigc++2.0 (2.2.4.2-1 -> 2.2.5-1)
    [2010-04-05 01:15] upgraded glibmm (2.22.1-1 -> 2.24.0-1)
    [2010-04-05 01:15] upgraded gucharmap (2.28.2-1 -> 2.30.0-1)
    [2010-04-05 01:16] upgraded gnome-applets (2.28.0-2 -> 2.30.0-1)
    [2010-04-05 01:16] upgraded gnome-backgrounds (2.28.0-1 -> 2.30.0-1)
    [2010-04-05 01:16] upgraded gnome-bluetooth (2.28.6-1 -> 2.30.0-1)
    [2010-04-05 01:16] upgraded vte (0.22.5-1 -> 0.24.0-2)
    [2010-04-05 01:16] upgraded gnome-desktop-sharp (2.26.0-3 -> 2.26.0-5)
    [2010-04-05 01:16] upgraded gnome-doc-utils (0.18.1-1 -> 0.20.0-1)
    [2010-04-05 01:17] upgraded gnome-games (2.28.2-1 -> 2.30.0-1)
    [2010-04-05 01:17] upgraded gnome-games-extra-data (2.28.0-1 -> 2.30.0-1)
    [2010-04-05 01:17] upgraded gnome-mag (0.15.9-1 -> 0.16.1-1)
    [2010-04-05 01:17] upgraded gnome-media (2.28.5-1 -> 2.30.0-2)
    [2010-04-05 01:17] upgraded gnome-netstatus (2.28.0-1 -> 2.28.1-1)
    [2010-04-05 01:17] upgraded gnome-nettool (2.28.0-1 -> 2.30.0-1)
    [2010-04-05 01:17] upgraded gnome-power-manager (2.28.3-1 -> 2.30.0-1)
    [2010-04-05 01:17] upgraded gnome-screensaver (2.28.3-1 -> 2.30.0-1)
    [2010-04-05 01:17] upgraded gtkmm (2.18.2-1 -> 2.20.0-1)
    [2010-04-05 01:17] upgraded gnome-system-monitor (2.28.0-1 -> 2.28.1-1)
    [2010-04-05 01:18] upgraded system-tools-backends (2.8.3-1 -> 2.10.0-1)
    [2010-04-05 01:18] upgraded liboobs (2.22.2-1 -> 2.30.0-1)
    [2010-04-05 01:18] upgraded nautilus (2.28.4-1 -> 2.30.0-1)
    [2010-04-05 01:18] upgraded gnome-system-tools (2.28.2-1 -> 2.30.0-2)
    [2010-04-05 01:18] upgraded gnome-terminal (2.28.2-1 -> 2.30.0-1)
    [2010-04-05 01:18] upgraded gtk-engines (2.18.5-1 -> 2.20.0-1)
    [2010-04-05 01:18] upgraded gnome-themes (2.28.1-1 -> 2.30.0-1)
    [2010-04-05 01:18] upgraded gnome-user-share (2.28.2-1 -> 2.30.0-1)
    [2010-04-05 01:18] upgraded gnome-utils (2.28.3-1 -> 2.30.0-1)
    [2010-04-05 01:19] upgraded xulrunner (1.9.2.2-2 -> 1.9.2.3-1)
    [2010-04-05 01:19] upgraded yelp (2.28.1-2 -> 2.30.0-1)
    [2010-04-05 01:19] upgraded gnome2-user-docs (2.28.2-1 -> 2.30.0-1)
    [2010-04-05 01:19] upgraded gok (2.28.1-1 -> 2.30.0-1)
    [2010-04-05 01:19] upgraded grep (2.6.2-1 -> 2.6.3-1)
    [2010-04-05 01:19] upgraded gstreamer0.10-bad (0.10.18-2 -> 0.10.18-3)
    [2010-04-05 01:19] upgraded neon (0.28.6-2 -> 0.28.6-4)
    [2010-04-05 01:19] upgraded gstreamer0.10-bad-plugins (0.10.18-2 -> 0.10.18-3)
    [2010-04-05 01:19] upgraded hamster-applet (2.28.2-1 -> 2.30.0.1-1)
    [2010-04-05 01:19] upgraded net-snmp (5.5-2 -> 5.5-3)
    [2010-04-05 01:19] upgraded hplip (3.9.12-5 -> 3.10.2-1)
    [2010-04-05 01:20] upgraded imap (2007e-1 -> 2007e-2)
    [2010-04-05 01:20] upgraded jre (6u18-2 -> 6u19-2)
    [2010-04-05 01:21] upgraded kdebase-workspace (4.4.2-1 -> 4.4.2-2)
    [2010-04-05 01:21] upgraded kernel26-firmware (2.6.33.1-1 -> 2.6.33.2-1)
    [2010-04-05 01:24] upgraded kernel26 (2.6.33.1-1 -> 2.6.33.2-1)
    [2010-04-05 01:24] upgraded kernel26-headers (2.6.33.1-1 -> 2.6.33.2-1)
    [2010-04-05 01:24] upgraded lib32-glib2 (2.22.5-1 -> 2.24.0-1)
    [2010-04-05 01:24] upgraded lib32-atk (1.28.0-1 -> 1.30.0-1)
    [2010-04-05 01:24] upgraded lib32-e2fsprogs (1.41.10-1 -> 1.41.11-1)
    [2010-04-05 01:24] upgraded lib32-openssl (0.9.8n-1 -> 1.0.0-2)
    [2010-04-05 01:24] upgraded lib32-heimdal (1.3.1-3 -> 1.3.2-1)
    [2010-04-05 01:24] upgraded lib32-libcups (1.4.2-5 -> 1.4.3-2)
    [2010-04-05 01:24] upgraded lib32-pango (1.26.2-1 -> 1.28.0-1)
    [2010-04-05 01:24] upgraded lib32-gtk2 (2.18.9-2 -> 2.20.0-2)
    [2010-04-05 01:24] upgraded lib32-libgl (7.7-1 -> 7.7.1-0.1)
    [2010-04-05 01:24] upgraded lib32-libxml2 (2.7.6-2 -> 2.7.7-1)
    [2010-04-05 01:24] upgraded lib32-mesa (7.7-1 -> 7.7.1-0.1)
    [2010-04-05 01:24] upgraded lib32-qt (4.6.2-2 -> 4.6.2-3)
    [2010-04-05 01:24] upgraded libepc (0.3.10-1 -> 0.3.11-1)
    [2010-04-05 01:24] upgraded libfetch (2.30-1 -> 2.30-3)
    [2010-04-05 01:24] upgraded libgail-gnome (1.20.1-1 -> 1.20.2-1)
    [2010-04-05 01:24] upgraded libgdata (0.4.0-1 -> 0.6.4-1)
    [2010-04-05 01:24] upgraded libgnomecups (0.2.3-6 -> 0.2.3-7)
    [2010-04-05 01:24] upgraded libgnomeprint (2.18.6-2 -> 2.18.7-2)
    [2010-04-05 01:24] upgraded libgnomeprintui (2.18.4-1 -> 2.18.5-1)
    [2010-04-05 01:24] upgraded librpcsecgss (0.19-2 -> 0.19-3)
    [2010-04-05 01:24] upgraded libssh (0.4.1-1 -> 0.4.1-3)
    [2010-04-05 01:25] upgraded libxml++ (2.26.0-1 -> 2.30.0-1)
    [2010-04-05 01:25] upgraded links (2.2-3 -> 2.2-4)
    [2010-04-05 01:25] upgraded lynx (2.8.7-1 -> 2.8.7-2)
    [2010-04-05 01:25] upgraded sbcl (1.0.35-1 -> 1.0.37-1)
    [2010-04-05 01:25] upgraded maxima (5.20.1-2 -> 5.20.1-3)
    [2010-04-05 01:25] upgraded mercurial (1.5-1 -> 1.5.1-1)
    [2010-04-05 01:26] upgraded miro (2.5.4-4 -> 3.0-1)
    [2010-04-05 01:26] upgraded mousetweaks (2.28.2-1 -> 2.30.0-1)
    [2010-04-05 01:26] upgraded nasm (2.07-1 -> 2.08.01-1)
    [2010-04-05 01:26] upgraded nautilus-actions (1.12.2-1 -> 2.30.0-1)
    [2010-04-05 01:26] upgraded nautilus-open-terminal (0.18-1 -> 0.18-2)
    [2010-04-05 01:26] upgraded nautilus-sendto (2.28.2-1 -> 2.28.4-1)
    [2010-04-05 01:26] upgraded samba (3.5.1-1 -> 3.5.1-2)
    [2010-04-05 01:26] upgraded nautilus-share (0.7.2-4 -> 0.7.2-5)
    [2010-04-05 01:26] upgraded network-manager-applet (0.8-2 -> 0.8-3)
    [2010-04-05 01:26] upgraded nfs-utils (1.2.2-1 -> 1.2.2-2)
    [2010-04-05 01:26] upgraded nmap (5.21-1 -> 5.21-2)
    [2010-04-05 01:26] upgraded ptlib (2.6.5-1 -> 2.6.5-2)
    [2010-04-05 01:27] upgraded opal (3.6.6-1 -> 3.6.6-2)
    [2010-04-05 01:27] upgraded openntpd (3.9p1-10 -> 3.9p1-11)
    [2010-04-05 01:28] upgraded openoffice-base (3.2.0-1 -> 3.2.0-3)
    [2010-04-05 01:28] upgraded openssh (5.4p1-3 -> 5.4p1-4)
    [2010-04-05 01:28] upgraded openvpn (2.1.1-1 -> 2.1.1-2)
    [2010-04-05 01:28] upgraded orca (2.28.3-1 -> 2.30.0-1)
    [2010-04-05 01:28] upgraded perl-crypt-ssleay (0.57-3 -> 0.57-4)
    [2010-04-05 01:28] upgraded perl-uri (1.52-1 -> 1.54-1)
    [2010-04-05 01:28] upgraded php (5.3.2-4 -> 5.3.2-6)
    [2010-04-05 01:28] upgraded php-apache (5.3.2-4 -> 5.3.2-6)
    [2010-04-05 01:28] upgraded pixman (0.16.6-1 -> 0.18.0-1)
    [2010-04-05 01:28] upgraded postfix (2.7.0-1 -> 2.7.0-2)
    [2010-04-05 01:28] upgraded qt (4.6.2-2 -> 4.6.2-3)
    [2010-04-05 01:28] upgraded qtscriptgenerator (0.1.0-3 -> 0.1.0-4)
    [2010-04-05 01:29] upgraded rhythmbox (0.12.7-1 -> 0.12.8-1)
    [2010-04-05 01:29] upgraded ruby (1.9.1_p378-1 -> 1.9.1_p378-2)
    [2010-04-05 01:29] upgraded seahorse (2.28.1-1 -> 2.30.0-1)
    [2010-04-05 01:29] upgraded seahorse-plugins (2.28.1-1 -> 2.30.0-1)
    [2010-04-05 01:29] upgraded sound-juicer (2.28.1-2 -> 2.28.2-1)
    [2010-04-05 01:29] upgraded subversion (1.6.9-2 -> 1.6.9-4)
    [2010-04-05 01:29] upgraded syslog-ng (3.0.4-3 -> 3.0.4-4)
    [2010-04-05 01:29] upgraded telepathy-glib (0.10.1-1 -> 0.10.2-1)
    [2010-04-05 01:29] upgraded tomboy (1.1.0-1 -> 1.2.0-2)
    [2010-04-05 01:29] upgraded totem (2.28.5-2 -> 2.30.0-1)
    [2010-04-05 01:29] upgraded transmission-gtk (1.92-1 -> 1.92-2)
    [2010-04-05 01:30] upgraded vala (0.7.10-1 -> 0.8.0-1)
    [2010-04-05 01:30] upgraded vinagre (2.28.1-2 -> 2.30.0-1)
    [2010-04-05 01:30] upgraded vino (2.28.1-3 -> 2.28.2-1)
    [2010-04-05 01:30] upgraded virtuoso (6.1.0-1 -> 6.1.0-2)
    [2010-04-05 01:30] upgraded wget (1.12-1 -> 1.12-2)
    [2010-04-05 01:30] upgraded wpa_supplicant (0.6.10-1 -> 0.6.10-2)
    [2010-04-05 01:30] upgraded xorg-server (1.7.6-2 -> 1.7.6-3)
    [2010-04-05 01:30] upgraded xorg-xinit (1.2.0-1 -> 1.2.1-1)
    [2010-04-05 01:30] upgraded xterm (255-1 -> 256-1)
    [2010-04-05 09:21] upgraded fofix-svn (1991-1 -> 2038-1)
    [2010-04-05 12:31] upgraded gloobus-preview-bzr (212-1 -> 214-1)
    Last edited by valandil (2010-04-06 16:52:28)

    Problem solved!
    I remembered I once configured my keyboard with HAL and /etc/hal/fdi/policy/10-keymap.fdi was set to xkb.layout=ca and xkd.variant=fr. I changed the GDM layout to Canada and removed the xkb.variant=fr from the .fdi file and everything was fixed!

  • [solved] AMD Radeon - strange behaviour after mesa-update

    Good afternoon Ladies and Gentlemen,
    This is my first posting, I will try to give as much as detailed information refering to my problem as possible. Please be patient if I ask for solutions which for you all seem easy and logical, I am a Newbie.
    The details: I am running a HP Pavilion 17 " AMD Quadcore laptop, 8gb Ram, Arch Linux 32bit Kernel 3.17.1 XFCE desktop with an AMD Radeon graphic card for about a few weeks now. Everything went fine so far, but last Saturday I did an update via pacman-Syu, the update included a mesa package. All went smooth as ever. Then after a reboot the XFCE desktop had stripes across the laptop screen, the mouse cursor had disappeared, also letters of the icon descriptions and the menus were missing, some strange "hieroglyphics" appeared instead of them. I wasn't able to downgrade the mesa-package as I nearly wasn't able to open the terminal because of the missing mouse cursor, then I managed somehow to get into the terminal and gave in the commands but the writing was incomplete and/or did not work because of these strange and missing letters.
    No problem, I thought to myself, a fresh install will do the work, so I wipped my hard drive using Parted Magic, then I did a fresh Arch install (Netinstall), this time using LXDE instead of XFCE as my desktop environment. When I booted into the login screen first, it looked fine but after logging into the desktop all the stripes, the missing mouse cursor and the strange letters appeared again. So is there any solution to that strange behaviour ? Before the Saturday-update everything worked without any problems for weeks. BTW other OS work without these problems, I tried Macpup just for fun, no messing-up with my Radeon graphics card happened.
    Sorry for my piss poor English I am Austrian. Sorry for eventually causing any inconvenience asking this question. Regards, Fladi.
    Last edited by fladi (2014-11-04 05:01:28)

    fafik1234 wrote:
    I have the same problem, free drivers, xfce4, procesor AMD5350(Kabini), vga radeon R3 (8400)
    downgrade linux kernel to linux-3.16.4
    download: http://seblu.net/a/arm/packages/l/linux … pkg.tar.xz
    pacma -U linux-3.16.4-1-i686.pkg.tar.xz
    edit /etc/pacman.conf
    IgnorePkg   =  linux linux-headers
    screenshot
    linux 3.17 xfce http://pics.tinypic.pl/i/00592/ek3z3ip81ky2.png
    linux 3.17 compiz http://pics.tinypic.pl/i/00592/bddprwn4tey8.png
    linux 3.16 compiz http://pics.tinypic.pl/i/00592/ja5ma8c7rdgl.png
    On a laptop, free drivers, linux 3.17.1-pae, openbox, procesor core i5, vga integrated card - everything OK
    The screenshots you posted match exactly to my problem ! Are you really sure the bug is coming from Kernel 3.17 ? I had this kernel running for about a week without any problems, the problems came when updating mesa.

Maybe you are looking for