Linked by Thom Holwerda on Thu 13th Oct 2011 21:33 UTC, submitted by mahmudinashar
Ubuntu, Kubuntu, Xubuntu Ah, it's that time of the year again. We already had this up on the sidebar, but I figured we'd turn it into a proper front page item - mostly because I want to discuss the move by the Ubuntu team to no longer install GNOME 2 as the 'classic' desktop option - which pretty much ends any and all involvement for me with Ubuntu (KDE 4 here I come). There's more to this than just that, of course, so those of you who do like Unity still have enough reason to upgrade.
Thread beginning with comment 493237
To view parent comment, click here.
To read all comments associated with this story, please click here.
RE[5]: Sigh...
by lemur2 on Mon 17th Oct 2011 09:52 UTC in reply to "RE[4]: Sigh..."
lemur2
Member since:
2007-02-17

Do you believe that kernel drivers and hardware have something to do with the crash dialog crashing?


I don't know, I've never seen it crash on KDE4.

Presumably something caused a crash to bring up a crash dialog in the first place. Whatever bug that was, it could easily explain a subsequent crash of the crash dialog as well.

Reply Parent Score: 2

RE[6]: Sigh...
by saynte on Mon 17th Oct 2011 10:34 in reply to "RE[5]: Sigh..."
saynte Member since:
2007-12-10

That is possible, of course, that the crash triggered some fundamental bug somewhere lower in the stack. However, usually the blame should work top-down, not bottom-up.

If kMail happens to crash, do you immediately file a bug against the kernel, or let Intel/AMD they have a bug in their CPU? Of course not, you first blame kMail, then if the maintainer can show that it's somewhere in a library he/she passes down the stack so it can be fixed.

Thom (quite likely) found a few bugs in KDE, denying this won't help anything.

Reply Parent Score: 1

RE[7]: Sigh...
by lemur2 on Mon 17th Oct 2011 11:17 in reply to "RE[6]: Sigh..."
lemur2 Member since:
2007-02-17

That is possible, of course, that the crash triggered some fundamental bug somewhere lower in the stack. However, usually the blame should work top-down, not bottom-up.

If kMail happens to crash, do you immediately file a bug against the kernel, or let Intel/AMD they have a bug in their CPU? Of course not, you first blame kMail, then if the maintainer can show that it's somewhere in a library he/she passes down the stack so it can be fixed.

Thom (quite likely) found a few bugs in KDE, denying this won't help anything.


Au contraire, it is clear that no such bug exists on any of my systems, which are running the same version of KDE. None of my systems exhibit anything like the behaviour which Thom describes.

All of my well-behaved, bug-free systems and Thom's system have KDE (and KDE applications) in common.

The only thing that sets Thom's system apart is "somewhere lower in the stack".

This observation is not proof, but nevertheless it is a strong indication that the fault on Thom's system does indeed lie "somewhere lower in the stack". Certainly, given all the described symptoms (and lack thereof on other systems), that is the first place I would look.

Reply Parent Score: 2

RE[6]: Sigh...
by aaronb on Mon 17th Oct 2011 20:02 in reply to "RE[5]: Sigh..."
aaronb Member since:
2005-07-06

You are reading a techie website yet you seem to have little understanding of what could make software unstable.

The problem with trying to force your opinions on others is additional frustration as now not only do they have to work through issues with KDE they also have to put up with people repeatedly saying "Well it works fine here so you must be wrong".

At best you could encourage people who are having issues to raise a bug report and if many people chime in with evidence of the problem it will probably be confirmed and then fixed.

At worst (what you are currently doing) is just denying the possibility that although something works for you it may not work for others.

For example, it is not a good idea for people who do not have type 1 diabetes to inject insulin just before or after meals, but injecting insulin can work great for people who do have type 1 diabetes.

Reply Parent Score: 2

RE[7]: Sigh...
by lemur2 on Mon 17th Oct 2011 22:17 in reply to "RE[6]: Sigh..."
lemur2 Member since:
2007-02-17

You are reading a techie website yet you seem to have little understanding of what could make software unstable.


Au contraire I have many years experience at debugging and fixing software, even on real-time systems, highly complex systems and embedded systems (where diagnostic tools are often minimal, and bugs are very hard to track down).

The very first step is to figure out when and where it works, and when and where it doesn't work. Once you can reliably make it fail, the next step is to figure out what is different about when and where it fails compared to when and where it doesn't.

If there are a lot of different systems where it doesn't fail, and a few where it does, then by far the most likely site of the fault is in the places where the systems differ, rather than the places where they are common.

The problem with trying to force your opinions on others is additional frustration as now not only do they have to work through issues with KDE they also have to put up with people repeatedly saying "Well it works fine here so you must be wrong". At best you could encourage people who are having issues to raise a bug report and if many people chime in with evidence of the problem it will probably be confirmed and then fixed. At worst (what you are currently doing) is just denying the possibility that although something works for you it may not work for others. For example, it is not a good idea for people who do not have type 1 diabetes to inject insulin just before or after meals, but injecting insulin can work great for people who do have type 1 diabetes.


Der. Let me teach you how to suck eggs ...

Edited 2011-10-17 22:17 UTC

Reply Parent Score: 2

RE[7]: Sigh...
by lemur2 on Mon 17th Oct 2011 22:41 in reply to "RE[6]: Sigh..."
lemur2 Member since:
2007-02-17

You are reading a techie website yet you seem to have little understanding of what could make software unstable. The problem with trying to force your opinions on others is additional frustration as now not only do they have to work through issues with KDE they also have to put up with people repeatedly saying "Well it works fine here so you must be wrong". At best you could encourage people who are having issues to raise a bug report and if many people chime in with evidence of the problem it will probably be confirmed and then fixed.


Indeed. There is plenty of scope for the bug in Thom's system to be amongst one of these:

http://reports.qa.ubuntu.com/reports/kernel-bugs/reports/_kernel_on...

... most of which, on the face of it, have nothing to do with KDE4.

At worst (what you are currently doing) is just denying the possibility that although something works for you it may not work for others.


We are discussing the most likely location for bugs on Thom's system.

How is this in any way denying that Thom's system is not working?

Edited 2011-10-17 22:42 UTC

Reply Parent Score: 2

RE[7]: Sigh...
by lemur2 on Tue 18th Oct 2011 00:15 in reply to "RE[6]: Sigh..."
lemur2 Member since:
2007-02-17

You are reading a techie website yet you seem to have little understanding of what could make software unstable. The problem with trying to force your opinions on others is additional frustration as now not only do they have to work through issues with KDE they also have to put up with people repeatedly saying "Well it works fine here so you must be wrong". At best you could encourage people who are having issues to raise a bug report and if many people chime in with evidence of the problem it will probably be confirmed and then fixed. At worst (what you are currently doing) is just denying the possibility that although something works for you it may not work for others.


BTW, on my Acer Aspire One 522 netbook, I experienced this bug for Kubuntu Oneiric:

https://bugs.launchpad.net/ubuntu/+source/linux/+bug/818730

I have downloaded the Ubuntu 11.10 install/LiveCD from "http://cdimage.ubuntu.com/releases/oneiric/alpha-2/oneiric-desktop-... and created a USB stick for install (with the option of no extra space for persistent files selected). I then tried to run the LiveCD on an Acer AO522 (netbook with AMD C50 dual core processor).
If I do *not* connect the wired ethernet the netbook freezes when during the boot the system is trying to use the wireless card to access the internet. If I have the wired ethernet connected to the internet, the install *does not* freeze.


Even though this bug freezes the netbook when I try to use the BCM4313 wifi via the KDE4 network manager, I do not blame KDE4 for this.

It truns out that the problem is actually with the unstable Kernel module "atl1c", which is the driver for the wired ethernet. I used the wired ethernet during the install (via the LiveCD, or in the case of my netbook, the LiveUSB), and once I had Oneiric installed, I blacklisted the ethernet driver atl1c because I normally use the wifi.

I will monitor progress of this bug. In time it will get fixed, and I will then un-blacklist the atl1c driver.

What I won't do is go on internet forums and complain loudly that the KDE4 wifi manager is broken.

Edited 2011-10-18 00:18 UTC

Reply Parent Score: 2