BPI inside

Category: Uncategorized

  • Lachrymologie?

    NULL

  • More illness

    NULL

  • Software-update, second try…

    Next week there will be a second try to update our software at one of our customers systems, which failed last time. My colleagues have fixed the most probable cause of this and I have also improved my software-part for more stability although it didn’t have any stability-problems. But who knows how long this release will be running until the next (hopefully very major) update, I don’t want to have it fail under any circumstance.

    We’re quite sure that this time the release will work as expected and without problems appearing after it.

    Well, for us it just has the look of a "service pack" because currently our software has improved quite much since the last install there and we just made some changes, bugfixes and minor improvements to that old version.
    It’s also already running at the upper third of its capacity because although there are only slightly more clients using it for which it was initially designed, some of these clients use it through a "proxy", so one client generates as much load as several dozens of normal ones.

    Our current software already has many design and capability improvements to handle a greater load and offer greater stability and flexibility on the clients side.

    Hopefully THIS version gets installed the next time and not another update of the old piece. Maintaining the old version is more difficult every time.

  • Cough, cough…

    Headache, fever and hurting arms and legs.

    I’m getting ill in the last two days of my attendance, but I don’t expect something to happen in these last hours. Been in bed the whole day, watching TV/DVD, reading magazines and sleep. But I don’t feel much better now, perhaps the peak of illness will strike this night or tomorrow.

    Damn, I have to get up on Monday no matter if I’m ill or not:

    • I have to pass on the attendance phone
    • I’m getting new shock absorbers onto my car, this should be the last thing to fix from the accident. Hopefully…

  • Taskzilla updated, anybody cares?

    Near the end of the last week, I updated our installation of Taskzilla with the new layout made by the designer.

    Now people have been using it for several days, reported some minor bugfixes to me and everything was fine.

    But today I recognized, that on the edit page of every task the category entries were completely misaligned. The category "Task Type" contained the entries from "Dev-Type" and some more swapping around.

    It was also impossible to add more CC people, if there was at least one entered, because the entry-field was disabled in that case.

    I wonder, why no one has seen this swapping before me. Perhaps this design is that intuitive, but I rather guess, that people are already used to Taskzilla. But why no one has recognized the bug with CC is a miracle for me… Does anyone actually use this feature?

  • Democenter gets deadline?

    Yesterday during a presentation for some of our developers/managers, a demo-center which runs entirely from CD was promised for upcoming Friday…

    Nice that I’m at least informed during a presentation, that I have a target date for my Knoppix-based CD.

    Meanwhile more problems arose, I think I have to remove NTFS-support from the CD for now because it always locks up doing nothing during mount and I don’t have the time to dig into this until Friday.

  • Building demo centers

    For some weeks now, a few of my collegues are working on an "All-in-one Demo Center". This is a special installation of all our solutions combined on one system, to be able to show anybody instantly any of our different products if he is interested and to have a complete reference system running inhouse, which you can put almost everywhere.

    Its currently placed on a notebook which is taken to customers as well be present on the <a href="http://www.cebit.de/”CeBit somewhere around Cisco’s corner.

    The operating system is an installation of Suse Enterprise Linux 9(?), the database backend is running Oracle and the webinterface runs on top of Apache.

    To distribute this solution on other machines, ISOs are created with Mondo Linux, which then can be inserted in almost any other computer, booted and puts the ISOs data on the computers harddisk. Existing data is lost with that solution.

    I was granted to try to make anoter All-In-One center, which boots completely from CD on any computer but without harming the existing data on the disk. Running completely in a ramdisk without even touching the disks in any way.

    I’m basing my installation on a stripped down distribution of Knoppix, which already runs completely from CD. It uses a compressed image as data-source, so it can hold some GB of data on a single CD.

    Initially this seemed to work quite well, the database files (>2GB) compressed down to about 100 MB and the oracle binaries also shrinked a bit. But soon I ran into problems because oracle could not start up in a complete read-only mode. It needed write-access to its database files, lockfiles and so on which was impossible to do on CD.

    So I decided to temporarily mount existing drives and put the needed writable files as well as a swapfile (oracle needs plenty of RAM) there only during the time of demonstration and remove it completely on shutdown again. Seemed cool idea, worked theoretically and found gread acceptance. Practically I had the next problem: the notebooks of our marketing people, where this should run on, have only NTFS partitions and Linux currently has only reasonable readonly-support for NTFS. Luckily I found the Captive project, which grants RW-access to NTFS by including the original WinXP driver into linux.

    My work is to integrate all the pieces and give it a foolproof interface so that even the least experienced marketing person can use it to demonstrate our solutions anywhere.

    Currently, I’m almost finished with preparing the environment for oracle. This means auto-detection of all partitions, their free space and file-system, let the user choose the destination, copy the datafiles, correctly split up and soft-link together again read-only and read-write files and finally start up oracle in such a patched together combination 😉

    Somewhere by next week I should be finished with oracle and also have integrated the administration interface. That means, if no other problems appear.

  • Raising my responsibility level

    NULL

  • throw in depth

    What I meant recently with "Never use throw-declaration in C++" is following:

    int throwing(int i) throw (exception_A)
{
  if(i > 0)
    throw exception_B;
  return i;
}

    When compiling on ie. AIX no error is shown during compiling, but if the exception_B gets ever thrown, it magically converts into an uncatchable unknown_exception and crashes your app. I’m not sure if Linux also makes it uncatchable, but there are also no signs of dangers there during compiling.

    Microsoft’s Visual Studio ships around this elegantly: It just ignores any throw()-declaration everywhere.

    I understand why no compiler enforces this to be correct, because of compatibility with old code and huge catch-statements with bad exception-derivates, but I don’t know why no compiler has even an option to enforce the correctness of throw-declarations. This would save many people much headache when porting software between platforms.

  • Attendance

    NULL