seaside performance in Squeak

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
18 messages Options
Reply | Threaded
Open this post in threaded view
|

seaside performance in Squeak

johnmci
James is busy blogging on seaside squeak, VW, rubyonrails performance  
and I'll note that a bit of Squeak GC tuning can
improve squeak seaside thruput by 100%

http://www.cincomsmalltalk.com/blog/blogView?showComments=true&printTitle=More_Seaside_Testing&entry=3372921925

Beware of course that a VM change and fix is required. If you are  
using the windows or the Unix VM you must ensure you
have a VM that was built with VMMaker-tpr.63, see http://bugs.squeak.org/view.php?id=6667

Otherwise the squeak GC tuning setting will not work.

--
=
=
=
========================================================================
John M. McIntosh <[hidden email]>
Corporate Smalltalk Consulting Ltd.  http://www.smalltalkconsulting.com
=
=
=
========================================================================


_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside
Reply | Threaded
Open this post in threaded view
|

Re: seaside performance in Squeak

Chun, Sungjin
Hi,

I think this is very great news and can I get the source or binary of VM(windows)
for this? Or should I do this by myself?

Thanks in advance.

----- Original Message -----
   From: John M McIntosh <[hidden email]>
   To: Seaside - general discussion <[hidden email]>
   Sent: 07-11-20 05:52:49
   Subject: [Seaside] seaside performance in Squeak

  James is busy blogging on seaside squeak, VW, rubyonrails performance  
and I'll note that a bit of Squeak GC tuning can
improve squeak seaside thruput by 100%

http://www.cincomsmalltalk.com/blog/blogView?showComments=true&printTitle=More_Seaside_Testing&entry=3372921925

Beware of course that a VM change and fix is required. If you are  
using the windows or the Unix VM you must ensure you
have a VM that was built with VMMaker-tpr.63, see http://bugs.squeak.org/view.php?id=6667

Otherwise the squeak GC tuning setting will not work.

--
=
=
=
========================================================================
John M. McIntosh <[hidden email]>
Corporate Smalltalk Consulting Ltd.  http://www.smalltalkconsulting.com
=
=
=
========================================================================


_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside


_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside
Reply | Threaded
Open this post in threaded view
|

Re: seaside performance in Squeak

Chun, Sungjin
In reply to this post by johnmci
Hi,

I think this is very great news and can I get the source or binary of VM(windows)
for this? Or should I do this by myself?

Thanks in advance.

----- Original Message -----
   From: John M McIntosh <[hidden email]>
   To: Seaside - general discussion <[hidden email]>
   Sent: 07-11-20 05:52:49
   Subject: [Seaside] seaside performance in Squeak

  James is busy blogging on seaside squeak, VW, rubyonrails performance  
and I'll note that a bit of Squeak GC tuning can
improve squeak seaside thruput by 100%

http://www.cincomsmalltalk.com/blog/blogView?showComments=true&printTitle=More_Seaside_Testing&entry=3372921925

Beware of course that a VM change and fix is required. If you are  
using the windows or the Unix VM you must ensure you
have a VM that was built with VMMaker-tpr.63, see http://bugs.squeak.org/view.php?id=6667

Otherwise the squeak GC tuning setting will not work.

--
=
=
=
========================================================================
John M. McIntosh <[hidden email]>
Corporate Smalltalk Consulting Ltd.  http://www.smalltalkconsulting.com
=
=
=
========================================================================


_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside


_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside
Reply | Threaded
Open this post in threaded view
|

Re: seaside performance in Squeak

johnmci
In reply to this post by johnmci
http://www.squeakvm.org/win32/

you could talk to who ever built the current windows VM you have and  
see if it was built with VMMaker-tpr.63
also see
http://wiki.squeak.org/squeak/2105

A VMMaker image to build a VM must have the fixed code as detailed in
http://bugs.squeak.org/view.php?id=6667

Otherwise the tuning hints will not work


On Nov 19, 2007, at 5:53 PM, [hidden email] wrote:

> Hi,
>
> I think this is very great news and can I get the source or binary  
> of VM(windows)
> for this? Or should I do this by myself?
>
> Thanks in advance.
>
> ----- Original Message -----
>   From: John M McIntosh <[hidden email]>
>   To: Seaside - general discussion  
> <[hidden email]>
>   Sent: 07-11-20 05:52:49
>   Subject: [Seaside] seaside performance in Squeak
>
>  James is busy blogging on seaside squeak, VW, rubyonrails performance
> and I'll note that a bit of Squeak GC tuning can
> improve squeak seaside thruput by 100%
>
> http://www.cincomsmalltalk.com/blog/blogView?showComments=true&printTitle=More_Seaside_Testing&entry=3372921925

--
=
=
=
========================================================================
John M. McIntosh <[hidden email]>
Corporate Smalltalk Consulting Ltd.  http://www.smalltalkconsulting.com
=
=
=
========================================================================


_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside
Reply | Threaded
Open this post in threaded view
|

Re: seaside performance in Squeak

Chun, Sungjin
In reply to this post by johnmci
Hi,

I've modified source code manually by referring http://bugs.squeak.org/view.php?id=6667,
but I'm not sure whether modifying referred part is OK. It seems that the code built without
problem, and run correctly up to now :-)

Should I regenerated all VM code instead of doing above?

Thanks in advance.

----- Original Message -----
   From: John M McIntosh <[hidden email]>
   To: Seaside - general discussion <[hidden email]>
   Sent: 07-11-20 15:22:27
   Subject: Re: [Seaside] seaside performance in Squeak

  http://www.squeakvm.org/win32/

you could talk to who ever built the current windows VM you have and  
see if it was built with VMMaker-tpr.63
also see
http://wiki.squeak.org/squeak/2105

A VMMaker image to build a VM must have the fixed code as detailed in
http://bugs.squeak.org/view.php?id=6667

Otherwise the tuning hints will not work


On Nov 19, 2007, at 5:53 PM, [hidden email] wrote:

> Hi,
>
> I think this is very great news and can I get the source or binary  
> of VM(windows)
> for this? Or should I do this by myself?
>
> Thanks in advance.
>
> ----- Original Message -----
>   From: John M McIntosh <[hidden email]>
>   To: Seaside - general discussion  
> <[hidden email]>
>   Sent: 07-11-20 05:52:49
>   Subject: [Seaside] seaside performance in Squeak
>
>  James is busy blogging on seaside squeak, VW, rubyonrails performance
> and I'll note that a bit of Squeak GC tuning can
> improve squeak seaside thruput by 100%
>
> http://www.cincomsmalltalk.com/blog/blogView?showComments=true&printTitle=More_Seaside_Testing&entry=3372921925
--
=
=
=
========================================================================
John M. McIntosh <[hidden email]>
Corporate Smalltalk Consulting Ltd.  http://www.smalltalkconsulting.com
=
=
=
========================================================================


_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside


_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside
Reply | Threaded
Open this post in threaded view
|

Re: seaside performance in Squeak

Chun, Sungjin
In reply to this post by johnmci
Hi,

I've modified source code manually by referring http://bugs.squeak.org/view.php?id=6667,
but I'm not sure whether modifying referred part is OK. It seems that the code built without
problem, and run correctly up to now :-)

Should I regenerated all VM code instead of doing above?

Thanks in advance.

----- Original Message -----
   From: John M McIntosh <[hidden email]>
   To: Seaside - general discussion <[hidden email]>
   Sent: 07-11-20 15:22:27
   Subject: Re: [Seaside] seaside performance in Squeak

  http://www.squeakvm.org/win32/

you could talk to who ever built the current windows VM you have and  
see if it was built with VMMaker-tpr.63
also see
http://wiki.squeak.org/squeak/2105

A VMMaker image to build a VM must have the fixed code as detailed in
http://bugs.squeak.org/view.php?id=6667

Otherwise the tuning hints will not work


On Nov 19, 2007, at 5:53 PM, [hidden email] wrote:

> Hi,
>
> I think this is very great news and can I get the source or binary  
> of VM(windows)
> for this? Or should I do this by myself?
>
> Thanks in advance.
>
> ----- Original Message -----
>   From: John M McIntosh <[hidden email]>
>   To: Seaside - general discussion  
> <[hidden email]>
>   Sent: 07-11-20 05:52:49
>   Subject: [Seaside] seaside performance in Squeak
>
>  James is busy blogging on seaside squeak, VW, rubyonrails performance
> and I'll note that a bit of Squeak GC tuning can
> improve squeak seaside thruput by 100%
>
> http://www.cincomsmalltalk.com/blog/blogView?showComments=true&printTitle=More_Seaside_Testing&entry=3372921925
--
=
=
=
========================================================================
John M. McIntosh <[hidden email]>
Corporate Smalltalk Consulting Ltd.  http://www.smalltalkconsulting.com
=
=
=
========================================================================


_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside


_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside
Reply | Threaded
Open this post in threaded view
|

Re: seaside performance in Squeak

johnmci
In reply to this post by johnmci
If your C code in interp.c looks like

  if ((((longAt(foo->freeBlock)) & AllButTypeMask) < foo-
  >growHeadroom) && (foo->gcBiasToGrow > 0)) {
             /* begin biasToGrow */
             growSize = (((sqInt) (foo->growHeadroom * 3) >> 1)) -  
((longAt(foo-
  >freeBlock)) & AllButTypeMask);
             /* begin growObjectMemory: */
             foo->statGrowMemory += 1;
             limit = sqGrowMemoryBy(foo->memoryLimit, growSize);
             if (!(limit == foo->memoryLimit)) {
                 foo->memoryLimit = limit - 24;
                 initializeMemoryFirstFree(foo->freeBlock);
             }
             weDidGrow = 1;
         }


then you have the fix needed to use

Smalltalk setGCBiasToGrowGCLimit: 64*1024*1024.
Smalltalk setGCBiasToGrow: 1.

The 64*1024*1024.  value is chosen based on a guess. The proper value  
for your application can only be chosen by testing and observing how  
your server image responds to some particular test suite, or to  
production loading, whichever is less risky or mandated by your  
organization.
The other values

> SmalltalkImage current  vmParameterAt: 5 put: 8000.  "do an  
> incremental GC after this many allocations"
> SmalltalkImage current  vmParameterAt: 6 put: 4000.  "tenure when  
> more  than this many objects survive the GC"

are just double the defaults but could be adjusted higher depending on  
the acceptable delay for an incremental GC, also the "tenure when  
more  than this many objects survive the GC" might not need to be 50%  
of the allocation value it could be some other percentage of the total.
These values below could also be changed, but one would need to  
collect some GC statistical data and see if they require changing.
> Smalltalk vmParameterAt: 25 put: 24*1024*1024. "grow headroom"
> Smalltalk vmParameterAt: 24 put: 48*1024*1024.  "shrink threshold"


My point was the current default values were set in 1996 or so, and  
are NOT valid for a seaside server, plus the setGCBiasToGrow fixes an  
interesting performance issue with how the default grow logic deals  
with incremental GCs versus growth concerns.
On Nov 19, 2007, at 11:08 PM, [hidden email] wrote:

> Hi,
>
> I've modified source code manually by referring http://bugs.squeak.org/view.php?id=6667 
> ,
> but I'm not sure whether modifying referred part is OK. It seems  
> that the code built without
> problem, and run correctly up to now :-)
>
> Should I regenerated all VM code instead of doing above?
>
> Thanks in advance.
>
> ----- Original Message -----
>   From: John M McIntosh <[hidden email]>
>   To: Seaside - general discussion  
> <[hidden email]>
>   Sent: 07-11-20 15:22:27
>   Subject: Re: [Seaside] seaside performance in Squeak
>
>  http://www.squeakvm.org/win32/
>
> you could talk to who ever built the current windows VM you have and
> see if it was built with VMMaker-tpr.63
> also see
> http://wiki.sq 
> Vhay5vcmcvc3F1ZWFrLzIxMDUNCg0KQSBWTU1ha2VyIGltYWdlIHRvIGJ1aWxkIGEgVk0g
> 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
> lDQo
> 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
> u

--
=
=
=
========================================================================
John M. McIntosh <[hidden email]>
Corporate Smalltalk Consulting Ltd.  http://www.smalltalkconsulting.com
=
=
=
========================================================================


_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside
Reply | Threaded
Open this post in threaded view
|

Re: seaside performance in Squeak

Adrian Lienhard
In reply to this post by johnmci
Has anybody experience with GC tuning for Seaside applications?

Adrian


On Nov 20, 2007, at 7:22 AM, John M McIntosh wrote:

> http://www.squeakvm.org/win32/
>
> you could talk to who ever built the current windows VM you have and  
> see if it was built with VMMaker-tpr.63
> also see
> http://wiki.squeak.org/squeak/2105
>
> A VMMaker image to build a VM must have the fixed code as detailed in
> http://bugs.squeak.org/view.php?id=6667
>
> Otherwise the tuning hints will not work
>
>
>
> On Nov 19, 2007, at 5:53 PM, [hidden email] wrote:
>
>> Hi,
>>
>> I think this is very great news and can I get the source or binary  
>> of VM(windows)
>> for this? Or should I do this by myself?
>>
>> Thanks in advance.
>>
>> ----- Original Message -----
>>  From: John M McIntosh <[hidden email]>
>>  To: Seaside - general discussion  
>> <[hidden email]>
>>  Sent: 07-11-20 05:52:49
>>  Subject: [Seaside] seaside performance in Squeak
>>
>> James is busy blogging on seaside squeak, VW, rubyonrails performance
>> and I'll note that a bit of Squeak GC tuning can
>> improve squeak seaside thruput by 100%
>>
>> http://www.cincomsmalltalk.com/blog/blogView?showComments=true&printTitle=More_Seaside_Testing&entry=3372921925
>
> --
> =
> =
> =
> =
> =
> ======================================================================
> John M. McIntosh <[hidden email]>
> Corporate Smalltalk Consulting Ltd.  http://
> www.smalltalkconsulting.com
> =
> =
> =
> =
> =
> ======================================================================
>
>
> _______________________________________________
> seaside mailing list
> [hidden email]
> http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside

_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside
Reply | Threaded
Open this post in threaded view
|

Re: seaside performance in Squeak

David T. Lewis
In reply to this post by johnmci
On Mon, Nov 19, 2007 at 12:52:49PM -0800, John M McIntosh wrote:
> you must ensure you
> have a VM that was built with VMMaker-tpr.63, see
> http://bugs.squeak.org/view.php?id=6667

And VMMaker-tpr.63 is located where?

:-/

_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside
Reply | Threaded
Open this post in threaded view
|

Re: seaside performance in Squeak

Alexander Lazarevic'
David T. Lewis schrieb:
> And VMMaker-tpr.63 is located where?

Found it?

Alex
_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside
Reply | Threaded
Open this post in threaded view
|

Re: seaside performance in Squeak

timrowledge

On 23-Nov-07, at 12:09 AM, Alexander Lazarevic' wrote:

> VMMaker-tpr.63


Well, *I* have a copy but I'm not sure how anyone else would since I  
don't usually publish numbered versions.

tim
--
tim Rowledge; [hidden email]; http://www.rowledge.org/tim
Why do we want intelligent terminals when there are so many stupid  
users?


_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside
Reply | Threaded
Open this post in threaded view
|

Re: seaside performance in Squeak

David T. Lewis
In reply to this post by Alexander Lazarevic'
On Fri, Nov 23, 2007 at 09:09:33AM +0100, Alexander Lazarevic' wrote:
> David T. Lewis schrieb:
> > And VMMaker-tpr.63 is located where?
>
> Found it?

No.

Dave

_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside
Reply | Threaded
Open this post in threaded view
|

Re: seaside performance in Squeak

Adrian Lienhard
In reply to this post by timrowledge
May I ask why VMMaker is not stored in one dedicated SqueakSource  
repository? It should be a matter of minutes to create a project and  
commit the current version to it and then people can commit and merge  
their fixes etc. The current situation with different versions in  
different places is quite obscure and IMHO should be improved for such  
an important part of Squeak.

Adrian


On Nov 23, 2007, at 16:55 , tim Rowledge wrote:

>
> On 23-Nov-07, at 12:09 AM, Alexander Lazarevic' wrote:
>
>> VMMaker-tpr.63
>
>
> Well, *I* have a copy but I'm not sure how anyone else would since I  
> don't usually publish numbered versions.
>
> tim
> --
> tim Rowledge; [hidden email]; http://www.rowledge.org/tim
> Why do we want intelligent terminals when there are so many stupid  
> users?
>
>
> _______________________________________________
> seaside mailing list
> [hidden email]
> http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside

_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside
Reply | Threaded
Open this post in threaded view
|

Re: seaside performance in Squeak

timrowledge

On 23-Nov-07, at 9:00 AM, Adrian Lienhard wrote:

> May I ask why VMMaker is not stored in one dedicated SqueakSource  
> repository?

Because I tried it once and had problems and never got around to  
making another attempt.

> It should be a matter of minutes to create a project and commit the  
> current version to it and then people can commit and merge their  
> fixes etc.
You're kidding, yes? The chances of me letting other people commit  
changes to such an important chunk of code are exactly (no floating  
point comparison to confuse things) zero. If other people want to do  
it for themselves and risk losing important work then they're welcome.  
I won't - couldn't - prevent it, nor would I ever help solve the  
problems.

When you use a tim branded VMMaker you are getting something that has  
had attention paid to
it. It certainly won't be perfect but it will be good.

tim
--
tim Rowledge; [hidden email]; http://www.rowledge.org/tim
Useful random insult:- Cackles a lot, but I ain't seen no eggs yet.


_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside
Reply | Threaded
Open this post in threaded view
|

Re: seaside performance in Squeak

David T. Lewis
On Fri, Nov 23, 2007 at 09:57:53AM -0800, tim Rowledge wrote:

>
> On 23-Nov-07, at 9:00 AM, Adrian Lienhard wrote:
>
> >May I ask why VMMaker is not stored in one dedicated SqueakSource  
> >repository?
>
> Because I tried it once and had problems and never got around to  
> making another attempt.
>
> >It should be a matter of minutes to create a project and commit the  
> >current version to it and then people can commit and merge their  
> >fixes etc.
> You're kidding, yes? The chances of me letting other people commit  
> changes to such an important chunk of code are exactly (no floating  
> point comparison to confuse things) zero.

Hi Tim,

Actually, a read-only copy of the numbered MCZs on SqueakSource
would be very helpful to me for understanding version changes
relative to the Subversion sources.

Assuming that you have on the order of say 100 numbered versions
sitting on a hard drive someplace, I would expect it take be about
60 to 90 minutes of annoyance and tedium to get the whole mess
copied over to SqueakSource. From that point forward it's no real
work to keep the SqueakSource copies up to date with your hard drive.

I started doing this with OSProcess, CommandShell and related plugins
a while back, and once I got past the initial annoyances I've been
happy with the process. So if you can find it in your heart to endure
an hour and a half of tedium and annoyance, I'll be most grateful.

Dave

_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside
Reply | Threaded
Open this post in threaded view
|

Re: seaside performance in Squeak

Adrian Lienhard
In reply to this post by timrowledge

On Nov 23, 2007, at 18:57 , tim Rowledge wrote:

>
> On 23-Nov-07, at 9:00 AM, Adrian Lienhard wrote:
>
>> May I ask why VMMaker is not stored in one dedicated SqueakSource  
>> repository?
>
> Because I tried it once and had problems and never got around to  
> making another attempt.
>
>> It should be a matter of minutes to create a project and commit the  
>> current version to it and then people can commit and merge their  
>> fixes etc.
> You're kidding, yes?

No, I'm serious. Of course, you are not going to blindly merge  
everything people commit.  You can skip whatever changes you don't  
like in your main line. This is how Seaside is done, and I think it  
works well. The nice thing is that you don't need to mail mcz or cs  
files around and hence it is more likely that people share changes  
that are worth being looked at.
If you don't like this, you can also make the project read only or  
only writable by the guys you trust. This would allow outsiders to see  
what is going on.

Adrian
_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside
Reply | Threaded
Open this post in threaded view
|

Re: seaside performance in Squeak

Colin Putney

On 23-Nov-07, at 12:32 PM, Adrian Lienhard wrote:

>> Because I tried it once and had problems and never got around to  
>> making another attempt.
>>
>>> It should be a matter of minutes to create a project and commit  
>>> the current version to it and then people can commit and merge  
>>> their fixes etc.
>> You're kidding, yes?
>
> No, I'm serious. Of course, you are not going to blindly merge  
> everything people commit.  You can skip whatever changes you don't  
> like in your main line. This is how Seaside is done, and I think it  
> works well. The nice thing is that you don't need to mail mcz or cs  
> files around and hence it is more likely that people share changes  
> that are worth being looked at.
> If you don't like this, you can also make the project read only or  
> only writable by the guys you trust. This would allow outsiders to  
> see what is going on.

Agreed. It works quite well for OmniBrowser too.  I'm picky about what  
changes I merge into the mainline, but I let anyone save versions to  
my repository. If people commit code that's problematic, that's fine,  
it just doesn't get merged into the trunk. Even so, contributions  
usually get incorporated in one form or another, because the  
visibility makes design decisions easier to discuss.

Colin
_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside
Reply | Threaded
Open this post in threaded view
|

Re: seaside performance in Squeak

timrowledge

On 23-Nov-07, at 8:31 PM, Colin Putney wrote:

>
> On 23-Nov-07, at 12:32 PM, Adrian Lienhard wrote:
>
>>> Because I tried it once and had problems and never got around to  
>>> making another attempt.
>>>
>>>> It should be a matter of minutes to create a project and commit  
>>>> the current version to it and then people can commit and merge  
>>>> their fixes etc.
>>> You're kidding, yes?
>>
>> No, I'm serious. Of course, you are not going to blindly merge  
>> everything people commit.  You can skip whatever changes you don't  
>> like in your main line. This is how Seaside is done, and I think it  
>> works well. The nice thing is that you don't need to mail mcz or cs  
>> files around and hence it is more likely that people share changes  
>> that are worth being looked at.
>> If you don't like this, you can also make the project read only or  
>> only writable by the guys you trust. This would allow outsiders to  
>> see what is going on.
>
> Agreed. It works quite well for OmniBrowser too.  I'm picky about  
> what changes I merge into the mainline, but I let anyone save  
> versions to my repository. If people commit code that's problematic,  
> that's fine, it just doesn't get merged into the trunk.

Hmph. If someone can point me to documentation/discussion of how to do  
this I might consider it. I'm used to a rather linear approach with MC  
so I have no idea really about allowing or managing a main trunk and  
branches.

tim
--
tim Rowledge; [hidden email]; http://www.rowledge.org/tim
Fractured Idiom:- MERCI RIEN - Thanks for nothin'.


_______________________________________________
seaside mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside