Xojo Conferences
XDCMay2019MiamiUSA

is it possible just LOADING the MBS plugins causes the IDE to leak memory? (Real Studio network user group Mailinglist archive)

Back to the thread list
Previous thread: Icon
Next thread: [ANN] Colorado REALbasic Summit


Re: Lag ProgressBar?   -   Garth Hjelte
  is it possible just LOADING the MBS plugins causes the IDE to leak memory?   -   Jann Gobble
   Re: is it possible just LOADING the MBS plugins causes the IDE to leak memory?   -   Jann Gobble
   Re: is it possible just LOADING the MBS plugins causes the IDE to leak memory?   -   Michael Diehr

is it possible just LOADING the MBS plugins causes the IDE to leak memory?
Date: 03.08.09 16:10 (Mon, 3 Aug 2009 08:10:50 -0700)
From: Jann Gobble
I do not know how to better say this so like any other issue this is
gonna require some back and forth with the list if anyone does not
understand.

My environment: MacBook Pro running Leopard 10.5.7 under RB 2009r3
with MBS plugins 9.4 and 4GB RAM.

I am (it seems) leaking memory all over the place when running in the
IDE -- or that 2009r3 is just a MAJOR memory hog when loading and
using the MBS plugins pack.

Is it possible RB2009R3 is handling MBS plugins differently than it
used to? I say used to, but I really don't know since I have never
used MBS under anything except 2009R3.

Here are the symptoms:

RB2009r3 takes around 2 minutes to load on a 2.53 gHz MBP running with
4Gig of ram under 10.5.7. Most of that time is reading/checking the
MBS plugins as I use no other plugins.

Once compiled, my app uses MBSCurlPlugin, MBSE2Plugin,
MBSRegistrationPlugin, MBSZipFilePlugin (for the MBS plugins) however,
the compiled app under Win and OS X leaks no memory! (or at least
leaks none that I can tell). I am using the RB SQL engine in my app,
Browser Plugin, the REGEX engine and the appearance pack and internet
encodings (those are all the dlls that that are built when using the
windows version so it may belp diagnose the issue)

The only place I *ever* have issues is in the IDE! I frequently get
the spinning beachball of death (once every day or so) and the IDE
alone takes up around 450 MB of ram (1.39GB Virtual). When compiling
the app (Command-B) That RSIZE climbs to 1.39GB and the Virtual climbs
later to around 2.5GB. If i just hit command-r and run the app inside
the debug system of the ide, the RSIZE climbs to 930 and VM climbs to
1.9GB.

Could it have nothing to do with all 29000 functions the plugin is
declaring at startup? It may be something only to do with the way
2009r3 handles compiling the app!

I am trying to make this clear, however, i will be working in the app
for anywhere from hours to a day or so and the system slows to a
CRAWL...then I get the splnning beachball of death -- then crash
reporter comes up! I would assume this would happen more frequently
if i did not have 4GB to play with.

any ideas?

TIA

Jann Gobble

_______________________________________________
Unsubscribe or switch delivery mode:
<http://www.realsoftware.com/support/listmanager/>

Search the archives:
<http://support.realsoftware.com/listarchives/lists.html>

Re: is it possible just LOADING the MBS plugins causes the IDE to leak memory?
Date: 03.08.09 18:06 (Mon, 3 Aug 2009 10:06:12 -0700)
From: Jann Gobble
They are aware of this and have not fixed it? This concerns me if RB
is supposed to be a cross platform extensible language!

I will be making a phone call today.

There should be release notes that state this!

Jann


On Aug 3, 2009, at 9:34 AM, Michael Diehr wrote:

> Yep - long standing bug that REAL is aware of. My hunch is that
> they don't use the MBS plugins regularly, so what is an enormous
> leak to us, to them seems trivial?
>
> In my usage, the IDE leaks about 500MB per platform when I first
> precompile the plugins, and about 50MB per run after that. The
> worst thing you can do is delete the RB cache then build for 2
> platforms -- leaks a *ton* of memory.
>
> I've noticed that in 2009R3, I can now get up to about 2.5GB before
> the IDE starts malfunctioning. In prior versions of the IDE it would
> die around 1.8GB. However, it starts getting slower & slower.
>
> I'm in the habit of restarting the IDE about every 10-20 runs or so...
>
> I also removed all MBS plugins that I'm not using from the plugins
> folder -- this seems to help a little bit.
>
> On Aug 3, 2009, at 8:10 AM, Jann Gobble wrote:
>
>> I do not know how to better say this so like any other issue this
>> is gonna require some back and forth with the list if anyone does
>> not understand.
>>
>> My environment: MacBook Pro running Leopard 10.5.7 under RB 2009r3
>> with MBS plugins 9.4 and 4GB RAM.
>>
>> I am (it seems) leaking memory all over the place when running in
>> the IDE -- or that 2009r3 is just a MAJOR memory hog when loading
>> and using the MBS plugins pack.
>>
>> Is it possible RB2009R3 is handling MBS plugins differently than it
>> used to? I say used to, but I really don't know since I have never
>> used MBS under anything except 2009R3.
>>
>> Here are the symptoms:
>>
>> RB2009r3 takes around 2 minutes to load on a 2.53 gHz MBP running
>> with 4Gig of ram under 10.5.7. Most of that time is reading/
>> checking the MBS plugins as I use no other plugins.
>>
>> Once compiled, my app uses MBSCurlPlugin, MBSE2Plugin,
>> MBSRegistrationPlugin, MBSZipFilePlugin (for the MBS plugins)
>> however, the compiled app under Win and OS X leaks no memory! (or
>> at least leaks none that I can tell). I am using the RB SQL engine
>> in my app, Browser Plugin, the REGEX engine and the appearance pack
>> and internet encodings (those are all the dlls that that are built
>> when using the windows version so it may belp diagnose the issue)
>>
>> The only place I *ever* have issues is in the IDE! I frequently
>> get the spinning beachball of death (once every day or so) and the
>> IDE alone takes up around 450 MB of ram (1.39GB Virtual). When
>> compiling the app (Command-B) That RSIZE climbs to 1.39GB and the
>> Virtual climbs later to around 2.5GB. If i just hit command-r and
>> run the app inside the debug system of the ide, the RSIZE climbs to
>> 930 and VM climbs to 1.9GB.
>>
>> Could it have nothing to do with all 29000 functions the plugin is
>> declaring at startup? It may be something only to do with the way
>> 2009r3 handles compiling the app!
>>
>> I am trying to make this clear, however, i will be working in the
>> app for anywhere from hours to a day or so and the system slows to
>> a CRAWL...then I get the splnning beachball of death -- then crash
>> reporter comes up! I would assume this would happen more
>> frequently if i did not have 4GB to play with.
>>
>> any ideas?
>>
>> TIA
>>
>> Jann Gobble
>>
>> _______________________________________________
>> Unsubscribe or switch delivery mode:
>> <http://www.realsoftware.com/support/listmanager/>
>>
>> Search the archives:
>> <http://support.realsoftware.com/listarchives/lists.html>
> _______________________________________________
> Unsubscribe or switch delivery mode:
> <http://www.realsoftware.com/support/listmanager/>
> Search the archives:
> <http://support.realsoftware.com/listarchives/lists.html>

_______________________________________________
Unsubscribe or switch delivery mode:
<http://www.realsoftware.com/support/listmanager/>

Search the archives:
<http://support.realsoftware.com/listarchives/lists.html>

Re: is it possible just LOADING the MBS plugins causes the IDE to leak memory?
Date: 03.08.09 17:34 (Mon, 3 Aug 2009 09:34:41 -0700)
From: Michael Diehr
Yep - long standing bug that REAL is aware of. My hunch is that they
don't use the MBS plugins regularly, so what is an enormous leak to
us, to them seems trivial?

In my usage, the IDE leaks about 500MB per platform when I first
precompile the plugins, and about 50MB per run after that. The worst
thing you can do is delete the RB cache then build for 2 platforms --
leaks a *ton* of memory.

I've noticed that in 2009R3, I can now get up to about 2.5GB before
the IDE starts malfunctioning. In prior versions of the IDE it would
die around 1.8GB. However, it starts getting slower & slower.

I'm in the habit of restarting the IDE about every 10-20 runs or so...

I also removed all MBS plugins that I'm not using from the plugins
folder -- this seems to help a little bit.

On Aug 3, 2009, at 8:10 AM, Jann Gobble wrote:

> I do not know how to better say this so like any other issue this is
> gonna require some back and forth with the list if anyone does not
> understand.
>
> My environment: MacBook Pro running Leopard 10.5.7 under RB 2009r3
> with MBS plugins 9.4 and 4GB RAM.
>
> I am (it seems) leaking memory all over the place when running in
> the IDE -- or that 2009r3 is just a MAJOR memory hog when loading
> and using the MBS plugins pack.
>
> Is it possible RB2009R3 is handling MBS plugins differently than it
> used to? I say used to, but I really don't know since I have never
> used MBS under anything except 2009R3.
>
> Here are the symptoms:
>
> RB2009r3 takes around 2 minutes to load on a 2.53 gHz MBP running
> with 4Gig of ram under 10.5.7. Most of that time is reading/
> checking the MBS plugins as I use no other plugins.
>
> Once compiled, my app uses MBSCurlPlugin, MBSE2Plugin,
> MBSRegistrationPlugin, MBSZipFilePlugin (for the MBS plugins)
> however, the compiled app under Win and OS X leaks no memory! (or at
> least leaks none that I can tell). I am using the RB SQL engine in
> my app, Browser Plugin, the REGEX engine and the appearance pack and
> internet encodings (those are all the dlls that that are built when
> using the windows version so it may belp diagnose the issue)
>
> The only place I *ever* have issues is in the IDE! I frequently get
> the spinning beachball of death (once every day or so) and the IDE
> alone takes up around 450 MB of ram (1.39GB Virtual). When
> compiling the app (Command-B) That RSIZE climbs to 1.39GB and the
> Virtual climbs later to around 2.5GB. If i just hit command-r and
> run the app inside the debug system of the ide, the RSIZE climbs to
> 930 and VM climbs to 1.9GB.
>
> Could it have nothing to do with all 29000 functions the plugin is
> declaring at startup? It may be something only to do with the way
> 2009r3 handles compiling the app!
>
> I am trying to make this clear, however, i will be working in the
> app for anywhere from hours to a day or so and the system slows to a
> CRAWL...then I get the splnning beachball of death -- then crash
> reporter comes up! I would assume this would happen more frequently
> if i did not have 4GB to play with.
>
> any ideas?
>
> TIA
>
> Jann Gobble
>
> _______________________________________________
> Unsubscribe or switch delivery mode:
> <http://www.realsoftware.com/support/listmanager/>
> Search the archives:
> <http://support.realsoftware.com/listarchives/lists.html>

_______________________________________________
Unsubscribe or switch delivery mode:
<http://www.realsoftware.com/support/listmanager/>

Search the archives:
<http://support.realsoftware.com/listarchives/lists.html>