mozilla memory hog - OS2

This is a discussion on mozilla memory hog - OS2 ; Hi: Until a few months ago, I never had a sys147 in my life. Yesterday, after another sys147, and resulting re-boot, I deceived to investigate the cause, as I thought maybe my 3.0 system wasn't allocating memory correctly, or needed ...

+ Reply to Thread
Page 1 of 2 1 2 LastLast
Results 1 to 20 of 24

Thread: mozilla memory hog

  1. mozilla memory hog

    Hi:

    Until a few months ago, I never had a sys147 in my
    life. Yesterday, after another sys147, and resulting
    re-boot, I deceived to investigate the cause, as I
    thought maybe my 3.0 system wasn't allocating memory
    correctly, or needed high memory support. I know I had
    way more stability with only 8 MB RAM and Netscape
    4.61.

    Well, with Theseus, I saw that SeaMonkey had
    hundreds of megabytes in shared and private RAM. I
    then found some excellent VOICE articles detailing the
    problem across all OS/2 versions, and the culprit as
    Mozilla, as it won't deallocate memory. The only
    "solution" was closing Mozilla frequently, before it
    gobbled up all the shared memory space and crashed
    OS/2.

    I checked a little further, and this major problem
    exists on all OS platforms with Mozilla, the developers
    know, and have tried unsuccessfully to address the
    issue (the suggested config parameters are useless),
    but it appears to be an internal political struggle to
    retain application "speed". They say it might be fixed
    in this new Firefox 3.0 release, but why is the whole
    world subject to this memory hogging crap just to
    maintain some appearances?

    later,
    lin Baden


    -----------------------------------------------------------
    Let Ignorance and Intolerance Rule!

    Baden Kudrenecky
    baden@baden.nu
    http://baden.nu/
    -----------------------------------------------------------



  2. Re: mozilla memory hog

    On Fri, 20 Jun 2008 17:35:45 UTC, baden@baden.nu (Baden Kudrenecky)
    wrote:

    > Hi:
    >
    > Until a few months ago, I never had a sys147 in my
    > life. Yesterday, after another sys147, and resulting
    > re-boot, I deceived to investigate the cause, as I
    > thought maybe my 3.0 system wasn't allocating memory
    > correctly, or needed high memory support. I know I had
    > way more stability with only 8 MB RAM and Netscape
    > 4.61.
    >
    > Well, with Theseus, I saw that SeaMonkey had
    > hundreds of megabytes in shared and private RAM. I
    > then found some excellent VOICE articles detailing the
    > problem across all OS/2 versions, and the culprit as
    > Mozilla, as it won't deallocate memory. The only
    > "solution" was closing Mozilla frequently, before it
    > gobbled up all the shared memory space and crashed
    > OS/2.
    >
    > I checked a little further, and this major problem
    > exists on all OS platforms with Mozilla, the developers
    > know, and have tried unsuccessfully to address the
    > issue (the suggested config parameters are useless),
    > but it appears to be an internal political struggle to
    > retain application "speed". They say it might be fixed
    > in this new Firefox 3.0 release, but why is the whole
    > world subject to this memory hogging crap just to
    > maintain some appearances?
    >
    > later,
    > lin Baden

    Baden,
    If I may ask, why are you still running Warp 3? I have Warp Connect on
    an old PS/2 to play with and have every other version of OS/2 Version
    3, blue and red. Just curious, I still like the Taskbar better than
    the WarpCenter.

  3. Re: mozilla memory hog

    On Fri, 20 Jun 2008 20:42:47 UTC in comp.os.os2.apps, "Count Floyd"
    wrote:

    > I still like the Taskbar better than
    > the WarpCenter.


    You can turn Warpcenter off and turn the old launchpad back on from config.sys
    very easily

    SET AUTOSTART=PROGRAMS,TASKLIST,FOLDERS,launchpad

    --
    Trevor Hemsley, Brighton, UK
    Trevor dot Hemsley at ntlworld dot com

  4. Re: mozilla memory hog

    On 20.06.08 23:01, Trevor Hemsley wrote:

    > On Fri, 20 Jun 2008 20:42:47 UTC in comp.os.os2.apps, "Count Floyd"
    > wrote:
    >
    >> I still like the Taskbar better than the WarpCenter.

    >
    > You can turn Warpcenter off and turn the old launchpad back on from
    > config.sys very easily
    >
    > SET AUTOSTART=PROGRAMS,TASKLIST,FOLDERS,launchpad


    Bad advice, IMO. The more that inis grow, the more problems arise and
    it's always better to start WPS objects like the LaunchPad delayed,
    after the desktop was populated and after maybe some other tasks
    completed with startup. So please use XWP's startup folder and set the
    timings to get the contained objects start safety.

    (I have to admit that the LaunchPad is much less critical in getting
    it started up than the XCenter.)

    --
    Andreas Schnellbacher

  5. Re: mozilla memory hog

    On 06/20/08 10:35 am, Baden Kudrenecky wrote:
    > Hi:
    >
    > Until a few months ago, I never had a sys147 in my
    > life. Yesterday, after another sys147, and resulting
    > re-boot, I deceived to investigate the cause, as I
    > thought maybe my 3.0 system wasn't allocating memory
    > correctly, or needed high memory support. I know I had
    > way more stability with only 8 MB RAM and Netscape
    > 4.61.
    >
    > Well, with Theseus, I saw that SeaMonkey had
    > hundreds of megabytes in shared and private RAM. I
    > then found some excellent VOICE articles detailing the
    > problem across all OS/2 versions, and the culprit as
    > Mozilla, as it won't deallocate memory. The only
    > "solution" was closing Mozilla frequently, before it
    > gobbled up all the shared memory space and crashed
    > OS/2.
    >
    > I checked a little further, and this major problem
    > exists on all OS platforms with Mozilla, the developers
    > know, and have tried unsuccessfully to address the
    > issue (the suggested config parameters are useless),
    > but it appears to be an internal political struggle to
    > retain application "speed". They say it might be fixed
    > in this new Firefox 3.0 release, but why is the whole
    > world subject to this memory hogging crap just to
    > maintain some appearances?
    >
    > later,
    > lin Baden
    >


    There has been quite a few memory fixes, both cross platform and OS/2
    specific fixes. You could try Firefox 3.0 or one of the Seamonkey trunk
    builds and see if it helps.
    Unluckily I'm not even sure if they will run on Warp v3 so if you do try
    it would be interesting if you report results
    Dave


  6. Re: mozilla memory hog

    In , "Count Floyd" writes:
    >On Fri, 20 Jun 2008 17:35:45 UTC, baden@baden.nu (Baden Kudrenecky)
    >wrote:
    >
    >Baden,
    >If I may ask, why are you still running Warp 3? I have Warp Connect on


    That's like asking someone why they still use their
    Maytag washer. {:-) I have never had sufficient reason
    to change my OS, and I really disliked the UI changes
    in 4.0, including the icons, fonts, and general look.
    Somehow, IBM thought it prudent to emulate the most
    ridiculous UI 'feature' ever thrown on the ignorant
    masses, namely the crap "Startup" cascading menu
    structure.

    I have MCP on my drive, and even though I rarely
    access it, its "Desktop" is corrupted, and I don't know
    how to fix it, short of re-installing, which I cannot
    do, as my CDROM drive is dead. With 3.0, there is not
    much I never knew about maintaining it, but I have
    forgotten a lot.

    >an old PS/2 to play with and have every other version of OS/2 Version
    >3, blue and red. Just curious, I still like the Taskbar better than
    >the WarpCenter.


    Yeah, the LaunchPad with npswps is ideal. For me,
    3.0 was the ultimate in Desktop UI development, and it
    maximised functionality and intuitiveness. Everything
    since then has seemed to bastardise the original
    philosophy, design and implementation. The Taskbar is
    an abomination. It's something like having your house
    painter patch up a Picasso.

    If I do change my primary OS, it will be into an all
    64 bit architecture to avoid these memory management
    problems, and I still haven't seen anything that
    impresses me.

    have a great day!
    lin Baden

    -----------------------------------------------------------
    Let Ignorance and Intolerance Rule!

    Baden Kudrenecky
    baden@baden.nu
    http://baden.nu/
    -----------------------------------------------------------



  7. Re: mozilla memory hog

    Hi Dave:

    I just tried running Firefox 3.0 on OS/2 3.0, and it
    does not draw anything at all inside the main window.
    I don't know if this is a bug, as it is not on a
    supported OS.

    thanks,
    lin Baden


    In , Dave Yeo writes:
    >On 06/20/08 10:35 am, Baden Kudrenecky wrote:
    >> Hi:
    >>
    >> Until a few months ago, I never had a sys147 in my
    >> life. Yesterday, after another sys147, and resulting
    >> re-boot, I deceived to investigate the cause, as I
    >> thought maybe my 3.0 system wasn't allocating memory
    >> correctly, or needed high memory support. I know I had
    >> way more stability with only 8 MB RAM and Netscape
    >> 4.61.
    >>
    >> Well, with Theseus, I saw that SeaMonkey had
    >> hundreds of megabytes in shared and private RAM. I
    >> then found some excellent VOICE articles detailing the
    >> problem across all OS/2 versions, and the culprit as
    >> Mozilla, as it won't deallocate memory. The only
    >> "solution" was closing Mozilla frequently, before it
    >> gobbled up all the shared memory space and crashed
    >> OS/2.
    >>
    >> I checked a little further, and this major problem
    >> exists on all OS platforms with Mozilla, the developers
    >> know, and have tried unsuccessfully to address the
    >> issue (the suggested config parameters are useless),
    >> but it appears to be an internal political struggle to
    >> retain application "speed". They say it might be fixed
    >> in this new Firefox 3.0 release, but why is the whole
    >> world subject to this memory hogging crap just to
    >> maintain some appearances?
    >>
    >> later,
    >> lin Baden
    >>

    >
    >There has been quite a few memory fixes, both cross platform and OS/2
    >specific fixes. You could try Firefox 3.0 or one of the Seamonkey trunk
    >builds and see if it helps.
    >Unluckily I'm not even sure if they will run on Warp v3 so if you do try
    >it would be interesting if you report results
    >Dave
    >




    -----------------------------------------------------------
    Let Ignorance and Intolerance Rule!

    Baden Kudrenecky
    baden@baden.nu
    http://baden.nu/
    -----------------------------------------------------------



  8. Re: mozilla memory hog

    On Fri, 20 Jun 2008 17:35:45 UTC, baden@baden.nu (Baden Kudrenecky) wrote:

    > Until a few months ago, I never had a sys147 in my
    > life. Yesterday, after another sys147, and resulting
    > re-boot, I deceived to investigate the cause, as I
    > thought maybe my 3.0 system wasn't allocating memory
    > correctly, or needed high memory support.


    High memory support is a (the?) primary reason for running
    a 4.5 kernel. Without it, some latter-day apps will be
    functionally unavailable to you - as you've discovered.

    [from a later post]
    > I have MCP on my drive, and even though I rarely
    > access it, its "Desktop" is corrupted, and I don't know
    > how to fix it, short of re-installing, which I cannot
    > do, as my CDROM drive is dead.


    A dead CD drive isn't a very compelling reason. How
    much would a replacement cost? $25? If that's too
    much, check out a Mom-and-Pop computer store that
    sells used systems. You can probably find one in
    their junk bin for $5.


    --
    == == almost usable email address: Rich AT E-vertise.Com == ==
    __________________________________________________ _________________
    |
    | DragText v3.9 with NLS support
    Rich Walsh | A Distinctly Different Desktop Enhancement
    Ft Myers, FL | http://e-vertise.com/dragtext/
    __________________________________________________ _________________


  9. Re: mozilla memory hog

    On 06/21/08 12:06 pm, Baden Kudrenecky wrote:

    > Somehow, IBM thought it prudent to emulate the most
    > ridiculous UI 'feature' ever thrown on the ignorant
    > masses, namely the crap "Startup" cascading menu
    > structure.


    It's quite different in what it contains, though - it's just quick
    access to the stuff on the desktop, if you have windows covering it.

  10. Re: mozilla memory hog

    Hi Baden

    Baden Kudrenecky wrote:
    >



    ----- snip -----


    > Somehow, IBM thought it prudent to emulate the most
    > ridiculous UI 'feature' ever thrown on the ignorant
    > masses, namely the crap "Startup" cascading menu
    > structure.
    >




    And I find that feature extremely useful for quick access to apps buried
    within Desktop folders...


    Ah well, different (mouse)strokes for different folks :-)


    Pete

  11. Re: mozilla memory hog

    Sir:

    Baden Kudrenecky wrote:
    > In , "Count Floyd" writes:
    >> On Fri, 20 Jun 2008 17:35:45 UTC, baden@baden.nu (Baden Kudrenecky)
    >> wrote:
    >>
    >> Baden,
    >> If I may ask, why are you still running Warp 3? I have Warp Connect on

    >
    > That's like asking someone why they still use their
    > Maytag washer. {:-) I have never had sufficient reason
    > to change my OS, and I really disliked the UI changes
    > in 4.0, including the icons, fonts, and general look.
    > Somehow, IBM thought it prudent to emulate the most
    > ridiculous UI 'feature' ever thrown on the ignorant
    > masses, namely the crap "Startup" cascading menu
    > structure.
    >
    > I have MCP on my drive, and even though I rarely
    > access it, its "Desktop" is corrupted, and I don't know
    > how to fix it, short of re-installing, which I cannot
    > do, as my CDROM drive is dead. With 3.0, there is not
    > much I never knew about maintaining it, but I have
    > forgotten a lot.

    Boot to the command line and run makeini using the two rc files. That
    should clear up most corruption, or enough to get a desktop where you
    can run checkini/c.
    --
    Bill
    Thanks a Million!

  12. Re: mozilla memory hog

    On 21.06.08 21:10, Baden Kudrenecky wrote:
    > I just tried running Firefox 3.0 on OS/2 3.0, and it
    > does not draw anything at all inside the main window.
    > I don't know if this is a bug, as it is not on a
    > supported OS.


    I take it this means that you don't even see the main menu? Then
    that's probably a bug in your graphics driver. SDD or SNAP should
    work.

    >> On 06/20/08 10:35 am, Baden Kudrenecky wrote:
    >>> Well, with Theseus, I saw that SeaMonkey had
    >>> hundreds of megabytes in shared and private RAM. I
    >>> then found some excellent VOICE articles detailing the
    >>> problem across all OS/2 versions, and the culprit as
    >>> Mozilla, as it won't deallocate memory.


    That is not correct. As I have written fairly often already, it
    _does_ deallocate memory. It's just that the used memory is
    maintained by the C library and if that memory has fragmented, the
    OS and other applications will not be able to make use of it. Search
    for "memory fragmentation" or see
    http://blog.pavlov.net/2007/11/10/memory-fragmentation/
    for more details.

    For Firefox 3 (can also be seen in the SeaMonkey 2.0a1pre nightly
    builds) I worked around that a bit by allocating large parts of the
    memory using OS/2 functions directly. If you e.g. close a tab or a
    window with FF 3 you will immediately see that it gives back some
    memory (so while it may be a bit heavier at first, in the long term
    it will use less RAM than FF 2).

    Peter.

  13. Re: mozilla memory hog

    Hi Rick:

    In , "Rich Walsh" writes:
    >On Fri, 20 Jun 2008 17:35:45 UTC, baden@baden.nu (Baden Kudrenecky) wrote:
    >
    >> Until a few months ago, I never had a sys147 in my
    >> life. Yesterday, after another sys147, and resulting
    >> re-boot, I deceived to investigate the cause, as I
    >> thought maybe my 3.0 system wasn't allocating memory
    >> correctly, or needed high memory support.

    >
    >High memory support is a (the?) primary reason for running
    >a 4.5 kernel. Without it, some latter-day apps will be
    >functionally unavailable to you - as you've discovered.


    The current problem, as I see it, is not that I have
    apps that need high memory support, but that certain
    applications are hogging all the shared area which is
    within the first 512 MiB. It is frustrating to see the
    sys147 when there are still 100s MiB RAM free. I read
    on VOICE that the situation is pervasive across all
    OS/2 platforms, so high memory support wouldn't help
    anything.

    Just to check, on Saturday, I booted up MCP-2 (4.5),
    to see how Firefox 3.0 would handle memory
    deallocation, and the system ran only 1:46 hours before
    it hung, and on reboot, chkdsk had a field day with my
    full 30 GB data partition, which I ended up
    reformatting. So much for testing, so I will stay with
    3.0, as that's the exact kind of **** I don't desire,
    and why I stay with something that is reliable and
    functional.

    My 3.0 runs fast and faultlessly on HPFS386, and I
    don't know if the MCP disc error was due to the HPFS on
    that partition, or possibly a blind coincidence, but
    why risk it? Since I had some reservations about the
    disc integrity, I ran a program to write all '1' then
    '0' across the whole drive, twice, and HPFS386 found no
    errors, so now, when another chkdsk /f:3 completes, I
    can start restoring the data to the partition. For
    some reason, "format /fs:hpfs /l" took forever, so I
    stopped it after six hours when it had only formatted
    5%.

    >[from a later post]
    >> I have MCP on my drive, and even though I rarely
    >> access it, its "Desktop" is corrupted, and I don't know
    >> how to fix it, short of re-installing, which I cannot
    >> do, as my CDROM drive is dead.

    >
    >A dead CD drive isn't a very compelling reason. How
    >much would a replacement cost? $25? If that's too
    >much, check out a Mom-and-Pop computer store that
    >sells used systems. You can probably find one in
    >their junk bin for $5.


    I bought a replacement CDROM drive, and it had the
    same problem, so I guessed there was something awry on
    the ThinkPad motherboard. Strange . . . After this
    latest episode, my incentive to fix or use the MCP
    partition has been somewhat diminished. {:-)

    have fun!
    in Baden


    -----------------------------------------------------------
    Let Ignorance and Intolerance Rule!

    Baden Kudrenecky
    baden@baden.nu
    http://baden.nu/
    -----------------------------------------------------------



  14. Re: mozilla memory hog

    In <485e8df5$0$7079$4c368faf@roadrunner.com>, "William L. Hartzell" writes:
    >Sir:
    >
    >Baden Kudrenecky wrote:
    >> In , "Count Floyd" writes:
    >>> On Fri, 20 Jun 2008 17:35:45 UTC, baden@baden.nu (Baden Kudrenecky)
    >>> wrote:


    >> I have MCP on my drive, and even though I rarely
    >> access it, its "Desktop" is corrupted, and I don't know
    >> how to fix it, short of re-installing, which I cannot
    >> do, as my CDROM drive is dead. With 3.0, there is not
    >> much I never knew about maintaining it, but I have
    >> forgotten a lot.

    >Boot to the command line and run makeini using the two rc files. That
    >should clear up most corruption, or enough to get a desktop where you
    >can run checkini/c.


    I think I did that, but IIRC, in 4.0, unlike 3.0,
    you cannot repair INI files, only construct a new set.

    Anyway, without "SET DESKTOP=F:\DESKTOP", the
    Desktop will not boot up, and there are still many
    problems, and many applications will not install.

    have fun!
    lin Baden


    -----------------------------------------------------------
    Let Ignorance and Intolerance Rule!

    Baden Kudrenecky
    baden@baden.nu
    http://baden.nu/
    -----------------------------------------------------------



  15. Re: mozilla memory hog

    Hi Peter:

    In <6c7iv7F3dqsdnU1@mid.individual.net>, Peter Weilbacher writes:
    >On 21.06.08 21:10, Baden Kudrenecky wrote:
    >> I just tried running Firefox 3.0 on OS/2 3.0, and it
    >> does not draw anything at all inside the main window.
    >> I don't know if this is a bug, as it is not on a
    >> supported OS.

    >
    >I take it this means that you don't even see the main menu? Then
    >that's probably a bug in your graphics driver. SDD or SNAP should
    >work.


    I am using SDD:

    SciTech SNAP Graphics for OS/2
    Version 3.1.8, build 505
    Professional Edition

    Graphics
    Screen size : 1400x1050 pixels
    Bit depth : 32 bits
    Plane count : 1 plane
    Number of colors : 16777216 colors
    Palette manager : No
    Raster caps : 0075
    Screen
    Physical size : 283x212 mm
    Resolution : 126x126 dpi
    Manufacturer: ATI
    Chipset: Mobility Radeon 7500
    Memory: 32768 Kb
    DAC: ATI Internal 24 bit DAC
    Clock: ATI Internal Clock
    Driver Revision: 3.1, Build 28
    Driver Build: May 19 2005

    Processor & memory
    Number of CPUs : 1
    CPU family : 6
    CPU model : 9
    CPU stepping : 5
    Page size : 4096 bytes
    Physical RAM : 511 MB
    Virtual address limit: 0 MB
    High memory support : No
    Operating system
    Operating system : OS/2 Warp
    Version : 3.0
    Build : 8.264
    Boot drive : C: HPFS

    >
    >>> On 06/20/08 10:35 am, Baden Kudrenecky wrote:
    >>>> Well, with Theseus, I saw that SeaMonkey had
    >>>> hundreds of megabytes in shared and private RAM. I
    >>>> then found some excellent VOICE articles detailing the
    >>>> problem across all OS/2 versions, and the culprit as
    >>>> Mozilla, as it won't deallocate memory.

    >
    >That is not correct. As I have written fairly often already, it
    >_does_ deallocate memory. It's just that the used memory is
    >maintained by the C library and if that memory has fragmented, the
    >OS and other applications will not be able to make use of it. Search
    >for "memory fragmentation" or see
    > http://blog.pavlov.net/2007/11/10/memory-fragmentation/
    >for more details.


    Last night, I checked this out, and after SeaMonkey
    had used up about 420 MB, I closed all the tabs and
    pages, so that only my home page (no graphics) was
    open, and the memory usage stayed at 420 MB. (200 MB
    private, 220 MB shared). After 15 minutes or so, I
    closed SeaMonkey, and most the memory was released.

    I still see a memory leak on my system from
    somewhere that never used to exist, and three days is
    about the maximum uptime. If Java 1.3.1 crashes
    (frequently) in SeaMonkey, then that memory is never
    deallocated.

    >For Firefox 3 (can also be seen in the SeaMonkey 2.0a1pre nightly
    >builds) I worked around that a bit by allocating large parts of the
    >memory using OS/2 functions directly. If you e.g. close a tab or a
    >window with FF 3 you will immediately see that it gives back some
    >memory (so while it may be a bit heavier at first, in the long term
    >it will use less RAM than FF 2).
    >
    > Peter.


    I may try a SeaMonkey nightly build out later to see
    what transpires.

    thanks a lot,
    lin Baden


    -----------------------------------------------------------
    Let Ignorance and Intolerance Rule!

    Baden Kudrenecky
    baden@baden.nu
    http://baden.nu/
    -----------------------------------------------------------



  16. Re: mozilla memory hog

    On Mon, 23 Jun 2008 22:07:51 UTC in comp.os.os2.apps, baden@baden.nu (Baden
    Kudrenecky) wrote:

    > in 4.0, unlike 3.0,
    > you cannot repair INI files, only construct a new set.


    Not true. It works the same way as it always did, if you delete os2[sys].ini
    then run makeini it creates them from scratch, if you don't delete them then it
    tries to repair them.

    --
    Trevor Hemsley, Brighton, UK
    Trevor dot Hemsley at ntlworld dot com

  17. Re: mozilla memory hog

    In , baden@baden.nu (Baden Kudrenecky) writes:
    >Hi Peter:
    >
    >In <6c7iv7F3dqsdnU1@mid.individual.net>, Peter Weilbacher writes:
    >>On 21.06.08 21:10, Baden Kudrenecky wrote:

    >
    > I may try a SeaMonkey nightly build out later to see
    >what transpires.


    I tried out the latest nightly SeaMonkey build on
    3.0, and it does not draw in the main window either. A
    screen shot is available at:

    http://baden.nu/Seamonkey_2.gif

    thanks,
    lin Baden

    -----------------------------------------------------------
    Let Ignorance and Intolerance Rule!

    Baden Kudrenecky
    baden@baden.nu
    http://baden.nu/
    -----------------------------------------------------------



  18. Re: mozilla memory hog

    Hi Trevor:

    You are correct! I cannot remember why I thought
    that.

    thanks
    lin Baden


    In , "Trevor Hemsley" writes:
    >On Mon, 23 Jun 2008 22:07:51 UTC in comp.os.os2.apps, baden@baden.nu (Baden
    >Kudrenecky) wrote:
    >
    >> in 4.0, unlike 3.0,
    >> you cannot repair INI files, only construct a new set.

    >
    >Not true. It works the same way as it always did, if you delete os2[sys].ini
    >then run makeini it creates them from scratch, if you don't delete them then it
    >tries to repair them.
    >
    >--
    >Trevor Hemsley, Brighton, UK
    >Trevor dot Hemsley at ntlworld dot com




    -----------------------------------------------------------
    Let Ignorance and Intolerance Rule!

    Baden Kudrenecky
    baden@baden.nu
    http://baden.nu/
    -----------------------------------------------------------



  19. Re: mozilla memory hog

    On Mon, 23 Jun 2008 22:34:03 UTC, Baden Kudrenecky wrote:

    > Hi Peter:
    >
    > In <6c7iv7F3dqsdnU1@mid.individual.net>, Peter Weilbacher writes:
    > >On 21.06.08 21:10, Baden Kudrenecky wrote:
    > >> I just tried running Firefox 3.0 on OS/2 3.0, and it
    > >> does not draw anything at all inside the main window.
    > >> I don't know if this is a bug, as it is not on a
    > >> supported OS.

    > >
    > >I take it this means that you don't even see the main menu? Then
    > >that's probably a bug in your graphics driver. SDD or SNAP should
    > >work.

    >
    > I am using SDD:
    >
    > SciTech SNAP Graphics for OS/2
    > Version 3.1.8, build 505
    > Professional Edition


    OK, then it's not that. Maybe something else is missing from Warp3 that
    is available in Warp4. But I was surprised to hear that it started at
    all, as I have heard before that the prerequisite libc063.dll isn't
    working on Warp3.

    > >That is not correct. As I have written fairly often already, it
    > >_does_ deallocate memory. It's just that the used memory is
    > >maintained by the C library and if that memory has fragmented, the
    > >OS and other applications will not be able to make use of it. Search
    > >for "memory fragmentation" or see
    > > http://blog.pavlov.net/2007/11/10/memory-fragmentation/
    > >for more details.

    >
    > Last night, I checked this out, and after SeaMonkey
    > had used up about 420 MB, I closed all the tabs and
    > pages, so that only my home page (no graphics) was
    > open, and the memory usage stayed at 420 MB. (200 MB
    > private, 220 MB shared). After 15 minutes or so, I
    > closed SeaMonkey, and most the memory was released.


    Sure, with SeaMonkey 1.x that's what I would expect. Waiting won't help
    memory fragmentation.

    > I still see a memory leak on my system from
    > somewhere that never used to exist, and three days is
    > about the maximum uptime. If Java 1.3.1 crashes
    > (frequently) in SeaMonkey, then that memory is never
    > deallocated.


    Not surprised to hear that. In my experience Java (even the 1.4.x
    variants for OS/2) was always very bad to clean up after itself.
    --
    Greetings, | My From: address is valid as is the version without "spam"
    Peter. | I try to find real messages among the spam once a week

  20. Re: mozilla memory hog

    On Tue, 24 Jun 2008, Baden Kudrenecky wrote:

    > I tried out the latest nightly SeaMonkey build on
    > 3.0, and it does not draw in the main window either. A
    > screen shot is available at:


    What do you guys use to build Seamonkey under OS/2?

+ Reply to Thread
Page 1 of 2 1 2 LastLast