|
|
||
Kate Error Messages and Compiling the Kernel—Day 30© Copyright Darrell Anderson. I updated to a more recent Slackware-current containing the 2.4.28 kernel and KDE 3.3.2. I hoped that the changes might resolve the Kate problem of opening in a small window. I also hoped that the updated kernel might resolve some of the kernel error messages that distracted me. Nada. Kate is now worse. Although I did figure out how to set window sizes in KDE (why aren’t window sizes remembered automatically instead of having to configure a bunch of window parameters?), I seem to have traded one nuisance for another. My ~/.xsession-errors file is filled with a mysterious message of: ASSERT: "!m_doc->wrapCursor()" in katerenderer.cpp (626)Sometimes the message is repeated like a wild fire. I did stumble into the KDE bug report web site and apparently this new error is common. I find hard to believe that such an error message could have escaped the KDE developers before releasing version 3.3.2. The message is too repeatable not to have been noticed. So now I face either living with this error message or reverting back to KDE 3.3.1. Aargh! Does anyone know how to run two concurrent KDE sessions without receiving error messages from aRts? Speaking of sound I notice that occasionally KDE takes its time with event sounds. Sometimes there is a noticeable delay before an event sound happens. The problem is inconsistent. Sometimes there is a delay, sometimes not. I did some more surfing and the best I could find was to enable real-time mode and increase the buffer size. I am trying that now. However, every time I press the Apply button when in the Sounds section of the KDE control panel I receive a Seg Fault error. Every time. I finally am realizing that people who use GNU/Linux are beta testers for open source software developers. The world is their playground. I am not saying they produce bad software. No, they do quite well. My concern is that the open source development model creates a never ending moving target. For example, I wish KDE would run without any error messages. Just run. No seg faults. No silly messages about QPixmap. No messages about X Error: BadWindow. No kdecore (KAction) messages every time I open a file using Kate from the Konqueror file manager. I wish programs opened and remembered my window positions without manual fiddling. I wish I could run two concurrent KDE sessions without aRts error messages. Yet, I see that the KDE team is already planning the next major release. How about a stable bug-free release before moving on? Why isn’t KDE developed in a similar manner to kernels—one stable branch and one for development and bleeding edge? Keep releasing patches and bug fixes for the stable branch and don’t render that branch obsolete. I’d love to see a version 3.3.3 or 3.3.4 that is bug free. Don’t get me wrong, I like KDE and I intend to keep using that desktop environment. I just want the irritations fixed. To make room for the newer Slackware I installed to the partitions originally storing VectorLinux. Thus, now I have two versions of Slackware on my system in addition to my old Mandrake 9.2. Mandrake 9.2 is slow but stable and will remain on my system for a long time as that installation provides me a baseline to tweak Slackware. This time I did some things differently. I did not perform a full installation, instead selecting the files I wanted to install. Because of my Slackware inexperience I missed de-selecting a few packages and later had to use removepkg to remove those packages. Now the only window managers on my system are KDE and the absolutely awful twm, which is by default par tof the X package. Do people actually work using twm?—shudder! I also did not press Enter when asked about my keyboard. Instead I selected the keyboard from the selection list. I did this for a simple reason: to force the Slackware installation script to create an /etc/rc.d/rc.keymap file. I’m somewhat lost about power management. My motherboard BIOS is an early generation supporting ACPI. At least, that is what the original manual declares, However, when I enable the APM daemon, those are the kernel messages I see and nothing about ACPI (yes, I enabled the acpid script). However, I have a Seagate hard drive that can be powered down and a DPMS compliant LCD monitor. I’d like to enable those features. The BIOS seems to do all right spinning down the hard drive, but not the monitor. My confusion is over how the GNU/Linux side mixes with the BIOS. Do I need to enable the BIOS settings or is GNU/Linux sufficient? I don’t know how to go about tweaking this. Some people might think that odd that I am so ignorant about power management, but I’ve never owned a laptop and NT4 never fully supported power management. This is new territory for me. I continued my adventures with recompiling the kernel using release 2.4.28. Same problems. ALSA breaks, dep mod errors, and modules don’t compile or install. Discouraging. Finis. |
||