Re: [Opensim-dev] Ini file(s) loading

2009-03-20 Thread Dr Scofield
Melanie wrote:
 And that brings us full circle to my proposal.
 
 There is much sense in a set of files that are used as defaults, and 
 then overriding settings in specific configuration files.

+1 on that proposal.

DrS/dirk

-- 
dr dirk husemann  virtual worlds research  ibm zurich research lab
SL: dr scofield  drscofi...@xyzzyxyzzy.net  http://xyzzyxyzzy.net/
RL: h...@zurich.ibm.com - +41 44 724 8573 - http://www.zurich.ibm.com/~hud/
___
Opensim-dev mailing list
Opensim-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/opensim-dev


Re: [Opensim-dev] Ini file(s) loading

2009-03-11 Thread Dahlia Trimble
Is there any reason why we wouldn't want to update the defaults in the code
to sane values?

On Fri, Mar 6, 2009 at 1:57 AM, Melanie mela...@t-data.com wrote:

 We did that because the hardcoded defaults won't work anymore. They
 are different fromt he OpenSim.ini.example values.
 However, my proposal provides a sane mechanism to provide external
 defaults, while not requiring user action at all.

 Melanie

 Frisby, Adam wrote:
  I did wonder why we started forcing users to have an opensim.ini. The
 previous 'use defaults' made more sense to me.
 
  Adam
 
  -Original Message-
  From: opensim-dev-boun...@lists.berlios.de [mailto:opensim-dev-
  boun...@lists.berlios.de] On Behalf Of Jeff Ames
  Sent: Thursday, 5 March 2009 7:20 PM
  To: opensim-dev@lists.berlios.de
  Subject: Re: [Opensim-dev] Ini file(s) loading
 
  Melanie wrote:
   read [the config directory] first
   then read the inimaster
   then read the inifile
 
  If I understand this correctly, the config/*.ini files would be
  essentially read-only, and all local changes would be made to the
  inimaster or OpenSim.ini.  But then OpenSim.ini is not broken up, and
  it may be confusing to users why there are two sets of config files.
 
  Is this just due to OpenSim's current behavior of requiring an .ini
  file to be present?  Currently the default values for all settings
  exist in the code itself and in the .ini.example file (itself an
  unfortunate duplication, but that's another topic).  Instead of
  requiring that an .ini be present, we could simply use the default
  values in the code if there is no .ini.  This would also have the
  pleasant side effect of matching the behavior when an empty .ini file
  is present.
 
  Then we could break up and move OpenSim.ini.example entirely to
  config/*.ini.example files, and when the user wants to change a value,
  create foo.ini based on foo.ini.example (copying the whole file if
  they want everything, or only adding the options they want to
  explicitly set).
 
  Then I guees the load order would be:
   - read inimaster (if present)
   - read config/*.ini (if present)
   - use defaults in code for anything not set
 
  I think this would also avoid the merging problem, if users only add
  options they're explicitly setting to the *.ini files.  It would also
  remove the annoyance of having to copy the .ini.example file over
  every time on a new install.
 
  Jeff
  ___
  Opensim-dev mailing list
  Opensim-dev@lists.berlios.de
  https://lists.berlios.de/mailman/listinfo/opensim-dev
  ___
  Opensim-dev mailing list
  Opensim-dev@lists.berlios.de
  https://lists.berlios.de/mailman/listinfo/opensim-dev
 
 
 ___
 Opensim-dev mailing list
 Opensim-dev@lists.berlios.de
 https://lists.berlios.de/mailman/listinfo/opensim-dev

___
Opensim-dev mailing list
Opensim-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/opensim-dev


Re: [Opensim-dev] Ini file(s) loading

2009-03-11 Thread Brianna
Sane values are desperately needed. I deal with many new users every week, in 
game and on forums. Most are x-pat SL content creators that are on top of that 
side but get hyper confused with an OpenSim binary. Many will tough it out and 
get an instance started but too many give up in frustration. 

Lately my improved technique is to have them fetch the OpenSim.ini from OSGrid 
dl, that at least gets past the first hurdles sanely. Or if standalone, use 
that .ini with very minimal edit.

I suggest anyone here open ini.example and look at it from a new user semi 
technical viewpoint.

  - Original Message - 
  From: Dahlia Trimble 
  To: opensim-dev@lists.berlios.de 
  Sent: Tuesday, March 10, 2009 11:21 PM
  Subject: Re: [Opensim-dev] Ini file(s) loading


  Is there any reason why we wouldn't want to update the defaults in the code 
to sane values?


  On Fri, Mar 6, 2009 at 1:57 AM, Melanie mela...@t-data.com wrote:

We did that because the hardcoded defaults won't work anymore. They
are different fromt he OpenSim.ini.example values.
However, my proposal provides a sane mechanism to provide external
defaults, while not requiring user action at all.

Melanie


Frisby, Adam wrote:
 I did wonder why we started forcing users to have an opensim.ini. The 
previous 'use defaults' made more sense to me.

 Adam

 -Original Message-
 From: opensim-dev-boun...@lists.berlios.de [mailto:opensim-dev-
 boun...@lists.berlios.de] On Behalf Of Jeff Ames
 Sent: Thursday, 5 March 2009 7:20 PM
 To: opensim-dev@lists.berlios.de
 Subject: Re: [Opensim-dev] Ini file(s) loading

 Melanie wrote:
  read [the config directory] first
  then read the inimaster
  then read the inifile

 If I understand this correctly, the config/*.ini files would be
 essentially read-only, and all local changes would be made to the
 inimaster or OpenSim.ini.  But then OpenSim.ini is not broken up, and
 it may be confusing to users why there are two sets of config files.

 Is this just due to OpenSim's current behavior of requiring an .ini
 file to be present?  Currently the default values for all settings
 exist in the code itself and in the .ini.example file (itself an
 unfortunate duplication, but that's another topic).  Instead of
 requiring that an .ini be present, we could simply use the default
 values in the code if there is no .ini.  This would also have the
 pleasant side effect of matching the behavior when an empty .ini file
 is present.

 Then we could break up and move OpenSim.ini.example entirely to
 config/*.ini.example files, and when the user wants to change a value,
 create foo.ini based on foo.ini.example (copying the whole file if
 they want everything, or only adding the options they want to
 explicitly set).

 Then I guees the load order would be:
  - read inimaster (if present)
  - read config/*.ini (if present)
  - use defaults in code for anything not set

 I think this would also avoid the merging problem, if users only add
 options they're explicitly setting to the *.ini files.  It would also
 remove the annoyance of having to copy the .ini.example file over
 every time on a new install.

 Jeff
 ___
 Opensim-dev mailing list
 Opensim-dev@lists.berlios.de
 https://lists.berlios.de/mailman/listinfo/opensim-dev
 ___
 Opensim-dev mailing list
 Opensim-dev@lists.berlios.de
 https://lists.berlios.de/mailman/listinfo/opensim-dev


___
Opensim-dev mailing list
Opensim-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/opensim-dev





--


  ___
  Opensim-dev mailing list
  Opensim-dev@lists.berlios.de
  https://lists.berlios.de/mailman/listinfo/opensim-dev
___
Opensim-dev mailing list
Opensim-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/opensim-dev


Re: [Opensim-dev] Ini file(s) loading

2009-03-10 Thread Justin Clark-Casey
Jeff Ames wrote:
 One nice thing about merge conflicts is that it tells you right away
 if a variable you've changed has been modified (e.g., the variable
 name changed), so you can correct it immediately instead of wondering
 why feature X isn't working any more.  But I guess there'd be a lot of
 false alarms, if comments near the variable or the variable's default
 value change in SVN.

Yeah, it's a nice idea but I think those false alarms are the problem.  The 
first time that a conflict happens the 
user's installation will probably stop working due to conflict markers in the 
.ini file.  I have to predict that people 
will change these .ini files directly without reading documentation first.

Another solution I would propose is to name these files *.ini.example in SVN 
(or *.ini-dist a la PHP) and go back to 
using built-in defaults that match these files.  I know this goes against what 
I said before about duplicating settings, 
but I couldn't think up a good solution that uses live .ini files and meets my 
particular concerns.  If a user wants to 
override the built in defaults then they copy *.ini.dist to *.ini or add to 
OpenSim.ini.

At least this way, sophisticated grid operators tracking SVN might be happy 
since they no longer need to merge *.ini 
files.  And individual users using OpenSim in standalone or hypergrid modes 
will also be happy since it will be clear 
how to override built-in defaults and they won't suffer merge conflicts if they 
are tracking SVN.  I still think that 
overriding settings only through OpenSim.ini is too complicated compared to a 
simple copy and edit that keeps the 
setting in a single place.

This solution would require that we have more discipline than before in keeping 
*.ini.dist files aligned with built-in 
settings.

 
 One final idea: instead of config/*.ini.example as I originally
 suggested, have a config/ directory (empty) and config.example/*.ini.
 Then users can copy config.example/foo.ini to config/foo.ini and make
 changes there.  Essentially the same idea, but copying files instead
 of renaming them.  Just trying to avoid having to have users keep
 their modifications in a monolithic .ini file.
 

 Jeff
 ___
 Opensim-dev mailing list
 Opensim-dev@lists.berlios.de
 https://lists.berlios.de/mailman/listinfo/opensim-dev
 


-- 
justincc
Justin Clark-Casey
http://justincc.wordpress.com
___
Opensim-dev mailing list
Opensim-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/opensim-dev


Re: [Opensim-dev] Ini file(s) loading

2009-03-10 Thread Melanie
Built-in settings were pretty much abolished because it was found 
that it's not possible/feasible to keep them sane.
I think that using an additional OpenSim.ini file is not too steep a 
learning curve, and that we should not design to the lowest denominator.

A config override hierarchy is, IMHO, just the ticket. Soon as I 
find some time in the chaos of moving house, I wanted to change all 
three options to accept either a single file, a directory, or a URI. 
That provides the greatest flexibility and would allow them to copy 
config/*.ini to local/*.ini, or somesuch, and edit them there.

Melanie

Justin Clark-Casey wrote:
 Jeff Ames wrote:
 One nice thing about merge conflicts is that it tells you right away
 if a variable you've changed has been modified (e.g., the variable
 name changed), so you can correct it immediately instead of wondering
 why feature X isn't working any more.  But I guess there'd be a lot of
 false alarms, if comments near the variable or the variable's default
 value change in SVN.
 
 Yeah, it's a nice idea but I think those false alarms are the problem.  The 
 first time that a conflict happens the 
 user's installation will probably stop working due to conflict markers in the 
 .ini file.  I have to predict that people 
 will change these .ini files directly without reading documentation first.
 
 Another solution I would propose is to name these files *.ini.example in SVN 
 (or *.ini-dist a la PHP) and go back to 
 using built-in defaults that match these files.  I know this goes against 
 what I said before about duplicating settings, 
 but I couldn't think up a good solution that uses live .ini files and meets 
 my particular concerns.  If a user wants to 
 override the built in defaults then they copy *.ini.dist to *.ini or add to 
 OpenSim.ini.
 
 At least this way, sophisticated grid operators tracking SVN might be happy 
 since they no longer need to merge *.ini 
 files.  And individual users using OpenSim in standalone or hypergrid modes 
 will also be happy since it will be clear 
 how to override built-in defaults and they won't suffer merge conflicts if 
 they are tracking SVN.  I still think that 
 overriding settings only through OpenSim.ini is too complicated compared to a 
 simple copy and edit that keeps the 
 setting in a single place.
 
 This solution would require that we have more discipline than before in 
 keeping *.ini.dist files aligned with built-in 
 settings.
 
 
 One final idea: instead of config/*.ini.example as I originally
 suggested, have a config/ directory (empty) and config.example/*.ini.
 Then users can copy config.example/foo.ini to config/foo.ini and make
 changes there.  Essentially the same idea, but copying files instead
 of renaming them.  Just trying to avoid having to have users keep
 their modifications in a monolithic .ini file.
 
 
 Jeff
 ___
 Opensim-dev mailing list
 Opensim-dev@lists.berlios.de
 https://lists.berlios.de/mailman/listinfo/opensim-dev
 
 
 
___
Opensim-dev mailing list
Opensim-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/opensim-dev


Re: [Opensim-dev] Ini file(s) loading

2009-03-10 Thread krtaylor
Actually, I thought it supported my point very well.  ;-)

I was not suggesting that the split ini in config be the only config 
option - I still like to override idea. But by default, users would edit 
the config/*ini, this does not stop a multi-region grid operator from 
syncing 20 override ini files to the bin of their regions.  My point was 
that there just should not be 2 places by default, one with default 
values and one with the overriding used values - that will just severely 
confuse the situation.

Kurt

Melanie wrote:
 And that brings us full circle to my proposal.

 There is much sense in a set of files that are used as defaults, and 
 then overriding settings in specific configuration files.

 Melanie

 Sean Hennessee wrote:
   
 I am one of those people that run trunk and update fairly often on 9 
 different 
 OpenSim servers. I suspect I always will be. Having to manually merge two 
 huge 
 ini files, on 9 different servers, is a major pain. It would be even worse 
 if I 
 had to merge 6 or 8 different medium sized ini files on each of my 9 
 servers. 
 Where they are located is not important to me, but I really would like to 
 have 
 an SVN new set of ini files, (or one ini file), with all the latest changes 
 that 
 is read first, and a way for me to have a small ini file that I can use for 
 my 
 specific settings which will override the default ini settings. That would 
 make 
 my life much much easier.

 Peace,
 Sean

 Jeff Ames wrote:
 
 Just splitting up the file and putting it in config, with all the
 comments, would help. And for the SVN update - that is a problem
 with any modifications to the tree, not just config. That's what
 backups are for. I guess a simple backup/diff tool for config would
 be a quick solution...
 
 
 Just have config/*.ini and let conflicts happen as they come?  I kinda
 agree... If you're updating some application from version 1.0 to 2.0,
 you expect to have to merge your config changes.  I imagine part of
 the merge problem is just that people are running trunk and updating
 all the time rather than using the (too infrequent?) tagged releases.

 I'm +1 on just using config/*.ini directly.

 Jeff
 ___
 Opensim-dev mailing list
 Opensim-dev@lists.berlios.de
 https://lists.berlios.de/mailman/listinfo/opensim-dev

   
   
 -- 

 Sean Hennessee
 mailto:s...@uci.edu
 http://www.nacs.uci.edu/~sean
 Central Computing Support
 Network  Academic Computing Services
 UC Irvine
 (949)824-8225 Office
 (949)293-5224 Cell


 ... . .- -. /   . -. -. . ... ... . .



 

 ___
 Opensim-dev mailing list
 Opensim-dev@lists.berlios.de
 https://lists.berlios.de/mailman/listinfo/opensim-dev
 
 ___
 Opensim-dev mailing list
 Opensim-dev@lists.berlios.de
 https://lists.berlios.de/mailman/listinfo/opensim-dev
   

-- 
Kurt Taylor (Kurt Stringer)


___
Opensim-dev mailing list
Opensim-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/opensim-dev


Re: [Opensim-dev] Ini file(s) loading

2009-03-09 Thread Jeff Ames
 Just splitting up the file and putting it in config, with all the
 comments, would help. And for the SVN update - that is a problem
 with any modifications to the tree, not just config. That's what
 backups are for. I guess a simple backup/diff tool for config would
 be a quick solution...

Just have config/*.ini and let conflicts happen as they come?  I kinda
agree... If you're updating some application from version 1.0 to 2.0,
you expect to have to merge your config changes.  I imagine part of
the merge problem is just that people are running trunk and updating
all the time rather than using the (too infrequent?) tagged releases.

I'm +1 on just using config/*.ini directly.

Jeff
___
Opensim-dev mailing list
Opensim-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/opensim-dev


Re: [Opensim-dev] Ini file(s) loading

2009-03-09 Thread Tommi Laukkanen
How about splitting to normal ini file and advanced ini file? In the
normal ini file you would have those parameters which 95% of the
people change and rest can be stuffed to advanced ini file.

- Tommi
___
Opensim-dev mailing list
Opensim-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/opensim-dev


Re: [Opensim-dev] Ini file(s) loading

2009-03-09 Thread Sean Hennessee




I am one of those people that run trunk and update fairly often on 9
different OpenSim servers. I suspect I always will be. Having to
manually merge two huge ini files, on 9 different servers, is a major
pain. It would be even worse if I had to merge 6 or 8 different medium
sized ini files on each of my 9 servers. Where they are located is not
important to me, but I really would like to have an SVN new set of ini
files, (or one ini file), with all the latest changes that is read
first, and a way for me to have a small ini file that I can use for my
specific settings which will override the default ini settings. That
would make my life much much easier.

Peace,
Sean

Jeff Ames wrote:

  
Just splitting up the file and putting it in config, with all the
comments, would help. And for the SVN update - that is a problem
with any modifications to the tree, not just config. That's what
backups are for. I guess a simple backup/diff tool for config would
be a quick solution...

  
  
Just have config/*.ini and let conflicts happen as they come?  I kinda
agree... If you're updating some application from version 1.0 to 2.0,
you expect to have to merge your config changes.  I imagine part of
the merge problem is just that people are running trunk and updating
all the time rather than using the (too infrequent?) tagged releases.

I'm +1 on just using config/*.ini directly.

Jeff
___
Opensim-dev mailing list
Opensim-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/opensim-dev

  


-- 

Sean Hennessee
mailto:s...@uci.edu
http://www.nacs.uci.edu/~sean
Central Computing Support
Network  Academic Computing Services
UC Irvine
(949)824-8225 Office
(949)293-5224 Cell


... . .- -. /   . -. -. . ... ... . .


___
Opensim-dev mailing list
Opensim-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/opensim-dev


Re: [Opensim-dev] Ini file(s) loading

2009-03-09 Thread Sean Hennessee
+1 Melanie! ... ... whoa, deja vu!  :)
~Sean

Melanie wrote:
 And that brings us full circle to my proposal.

 There is much sense in a set of files that are used as defaults, and 
 then overriding settings in specific configuration files.

 Melanie

 Sean Hennessee wrote:
   
 I am one of those people that run trunk and update fairly often on 9 
 different 
 OpenSim servers. I suspect I always will be. Having to manually merge two 
 huge 
 ini files, on 9 different servers, is a major pain. It would be even worse 
 if I 
 had to merge 6 or 8 different medium sized ini files on each of my 9 
 servers. 
 Where they are located is not important to me, but I really would like to 
 have 
 an SVN new set of ini files, (or one ini file), with all the latest changes 
 that 
 is read first, and a way for me to have a small ini file that I can use for 
 my 
 specific settings which will override the default ini settings. That would 
 make 
 my life much much easier.

 Peace,
 Sean

 Jeff Ames wrote:
 
 Just splitting up the file and putting it in config, with all the
 comments, would help. And for the SVN update - that is a problem
 with any modifications to the tree, not just config. That's what
 backups are for. I guess a simple backup/diff tool for config would
 be a quick solution...
 
 
 Just have config/*.ini and let conflicts happen as they come?  I kinda
 agree... If you're updating some application from version 1.0 to 2.0,
 you expect to have to merge your config changes.  I imagine part of
 the merge problem is just that people are running trunk and updating
 all the time rather than using the (too infrequent?) tagged releases.

 I'm +1 on just using config/*.ini directly.

 Jeff
   

-- 

Sean Hennessee
Central Computing Support
Network  Academic Computing Services
UC Irvine


... . .- -. /   . -. -. . ... ... . .

___
Opensim-dev mailing list
Opensim-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/opensim-dev


Re: [Opensim-dev] Ini file(s) loading

2009-03-09 Thread Jeff Ames
One nice thing about merge conflicts is that it tells you right away
if a variable you've changed has been modified (e.g., the variable
name changed), so you can correct it immediately instead of wondering
why feature X isn't working any more.  But I guess there'd be a lot of
false alarms, if comments near the variable or the variable's default
value change in SVN.

One final idea: instead of config/*.ini.example as I originally
suggested, have a config/ directory (empty) and config.example/*.ini.
Then users can copy config.example/foo.ini to config/foo.ini and make
changes there.  Essentially the same idea, but copying files instead
of renaming them.  Just trying to avoid having to have users keep
their modifications in a monolithic .ini file.

Jeff
___
Opensim-dev mailing list
Opensim-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/opensim-dev


Re: [Opensim-dev] Ini file(s) loading

2009-03-06 Thread Melanie
We did that because the hardcoded defaults won't work anymore. They 
are different fromt he OpenSim.ini.example values.
However, my proposal provides a sane mechanism to provide external 
defaults, while not requiring user action at all.

Melanie

Frisby, Adam wrote:
 I did wonder why we started forcing users to have an opensim.ini. The 
 previous 'use defaults' made more sense to me.
 
 Adam
 
 -Original Message-
 From: opensim-dev-boun...@lists.berlios.de [mailto:opensim-dev-
 boun...@lists.berlios.de] On Behalf Of Jeff Ames
 Sent: Thursday, 5 March 2009 7:20 PM
 To: opensim-dev@lists.berlios.de
 Subject: Re: [Opensim-dev] Ini file(s) loading

 Melanie wrote:
  read [the config directory] first
  then read the inimaster
  then read the inifile

 If I understand this correctly, the config/*.ini files would be
 essentially read-only, and all local changes would be made to the
 inimaster or OpenSim.ini.  But then OpenSim.ini is not broken up, and
 it may be confusing to users why there are two sets of config files.

 Is this just due to OpenSim's current behavior of requiring an .ini
 file to be present?  Currently the default values for all settings
 exist in the code itself and in the .ini.example file (itself an
 unfortunate duplication, but that's another topic).  Instead of
 requiring that an .ini be present, we could simply use the default
 values in the code if there is no .ini.  This would also have the
 pleasant side effect of matching the behavior when an empty .ini file
 is present.

 Then we could break up and move OpenSim.ini.example entirely to
 config/*.ini.example files, and when the user wants to change a value,
 create foo.ini based on foo.ini.example (copying the whole file if
 they want everything, or only adding the options they want to
 explicitly set).

 Then I guees the load order would be:
  - read inimaster (if present)
  - read config/*.ini (if present)
  - use defaults in code for anything not set

 I think this would also avoid the merging problem, if users only add
 options they're explicitly setting to the *.ini files.  It would also
 remove the annoyance of having to copy the .ini.example file over
 every time on a new install.

 Jeff
 ___
 Opensim-dev mailing list
 Opensim-dev@lists.berlios.de
 https://lists.berlios.de/mailman/listinfo/opensim-dev
 ___
 Opensim-dev mailing list
 Opensim-dev@lists.berlios.de
 https://lists.berlios.de/mailman/listinfo/opensim-dev
 
 
___
Opensim-dev mailing list
Opensim-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/opensim-dev


Re: [Opensim-dev] Ini file(s) loading

2009-03-06 Thread Sacha Magne
+1 for me. it will ease all the maintenance and lot of redundnat questions
related to the configuration we got often in #opensim


On Fri, Mar 6, 2009 at 9:57 AM, Melanie mela...@t-data.com wrote:

 We did that because the hardcoded defaults won't work anymore. They
 are different fromt he OpenSim.ini.example values.
 However, my proposal provides a sane mechanism to provide external
 defaults, while not requiring user action at all.

 Melanie

 Frisby, Adam wrote:
  I did wonder why we started forcing users to have an opensim.ini. The
 previous 'use defaults' made more sense to me.
 
  Adam
 
  -Original Message-
  From: opensim-dev-boun...@lists.berlios.de [mailto:opensim-dev-
  boun...@lists.berlios.de] On Behalf Of Jeff Ames
  Sent: Thursday, 5 March 2009 7:20 PM
  To: opensim-dev@lists.berlios.de
  Subject: Re: [Opensim-dev] Ini file(s) loading
 
  Melanie wrote:
   read [the config directory] first
   then read the inimaster
   then read the inifile
 
  If I understand this correctly, the config/*.ini files would be
  essentially read-only, and all local changes would be made to the
  inimaster or OpenSim.ini.  But then OpenSim.ini is not broken up, and
  it may be confusing to users why there are two sets of config files.
 
  Is this just due to OpenSim's current behavior of requiring an .ini
  file to be present?  Currently the default values for all settings
  exist in the code itself and in the .ini.example file (itself an
  unfortunate duplication, but that's another topic).  Instead of
  requiring that an .ini be present, we could simply use the default
  values in the code if there is no .ini.  This would also have the
  pleasant side effect of matching the behavior when an empty .ini file
  is present.
 
  Then we could break up and move OpenSim.ini.example entirely to
  config/*.ini.example files, and when the user wants to change a value,
  create foo.ini based on foo.ini.example (copying the whole file if
  they want everything, or only adding the options they want to
  explicitly set).
 
  Then I guees the load order would be:
   - read inimaster (if present)
   - read config/*.ini (if present)
   - use defaults in code for anything not set
 
  I think this would also avoid the merging problem, if users only add
  options they're explicitly setting to the *.ini files.  It would also
  remove the annoyance of having to copy the .ini.example file over
  every time on a new install.
 
  Jeff
  ___
  Opensim-dev mailing list
  Opensim-dev@lists.berlios.de
  https://lists.berlios.de/mailman/listinfo/opensim-dev
  ___
  Opensim-dev mailing list
  Opensim-dev@lists.berlios.de
  https://lists.berlios.de/mailman/listinfo/opensim-dev
 
 
 ___
 Opensim-dev mailing list
 Opensim-dev@lists.berlios.de
 https://lists.berlios.de/mailman/listinfo/opensim-dev




-- 

http://K-grid.com
Just be cause it's Kool
___
Opensim-dev mailing list
Opensim-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/opensim-dev


Re: [Opensim-dev] Ini file(s) loading

2009-03-06 Thread Chris Hart
I like it - no more merging, defaults split up between files for
manageability, no more hunt the ini in the mammoth bin directory.
Defaults are readable - if you got rid of the example file or the need
for an ini at all you'd have to dig through code to figure out what the
defaults were. You still get plenty of control over options via the
master (if you use a master), and then individual differences in local
instances. 

+1

-Original Message-
From: opensim-dev-boun...@lists.berlios.de
[mailto:opensim-dev-boun...@lists.berlios.de] On Behalf Of Melanie
Sent: 06 March 2009 08:56
To: opensim-dev@lists.berlios.de
Subject: Re: [Opensim-dev] Ini file(s) loading

Hi,


Jeff Ames wrote:
 Melanie wrote:
 read [the config directory] first
 then read the inimaster
 then read the inifile

This is because the config/* files would hold the defaults, they 
must be read first.

[...]

 
 Then we could break up and move OpenSim.ini.example entirely to
 config/*.ini.example files, and when the user wants to change a value,
 create foo.ini based on foo.ini.example (copying the whole file if
 they want everything, or only adding the options they want to
 explicitly set).
 

I was suggesting to get away from *.example files. Instead, I am 
proposing to break up OpenSim.ini into *.ini files that provide the 
defaults. The in-code defaults are a requirement of Nini, but they 
are not functional, just ignore that they're there. The requirement 
for an OpenSim.ini file in code needs to be dropped, of course.

That is why the config directory needs to be read first - it 
provides the defaults! Inimaster would be rendered useless by the 
load order you're proposing!

 Then I guees the load order would be:
  - read inimaster (if present)
  - read config/*.ini (if present)
  - use defaults in code for anything not set
 

See above, this won't serve the users' needs. In your scenario, if 
the config/*.example changes, they'd still need to merge.

What I propose, to reiterate, is this:

Break up Opensim.ini.example into config/*.ini. Uncomment all the 
options so they serve as defaults.
Remove the explicit requirement for an ini file.
This way, sitewide changes from the provided defaults go to 
inimaster, and machine/instance settings go to opensim.ini.
The files in config/ can remain unchanged and can be updated on 
every svn update, without manual merging.

Of course, many other use cases are possible, including 
modifications to config/*.ini if desired, or omitting either 
configuration source.

Melanie
___
Opensim-dev mailing list
Opensim-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/opensim-dev

No virus found in this incoming message.
Checked by AVG - www.avg.com 
Version: 8.0.237 / Virus Database: 270.11.5/1977 - Release Date:
03/05/09 19:32:00
___
Opensim-dev mailing list
Opensim-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/opensim-dev


Re: [Opensim-dev] Ini file(s) loading

2009-03-06 Thread Justin Clark-Casey
Melanie wrote:
 Hi,
 
 
 Jeff Ames wrote:
 Melanie wrote:
 read [the config directory] first
 then read the inimaster
 then read the inifile
 
 This is because the config/* files would hold the defaults, they 
 must be read first.
 
 [...]
 
 Then we could break up and move OpenSim.ini.example entirely to
 config/*.ini.example files, and when the user wants to change a value,
 create foo.ini based on foo.ini.example (copying the whole file if
 they want everything, or only adding the options they want to
 explicitly set).

 
 I was suggesting to get away from *.example files. Instead, I am 
 proposing to break up OpenSim.ini into *.ini files that provide the 
 defaults. The in-code defaults are a requirement of Nini, but they 
 are not functional, just ignore that they're there. The requirement 
 for an OpenSim.ini file in code needs to be dropped, of course.

If/when we split up OpenSim,ini then we do largely need to get away from 
.example files - renaming lots of 
config/*.ini.example files is obviously a no go.

If we get rid of OpenSim.ini entirely, then when a new user wants to change 
their storage/network settings (which I 
suspect pretty much everyone does sooner or later), they will either

(a)  Follow some instructions to copy config/storage.ini to OpenSim.ini and 
then tweak it or
(b)  Change config/storage.ini directly

Even if we encouraged (a), I think that realistically people will just end up 
doing (b).  If config/storage.ini ends up 
in SVN, then when it changes we will wreck their configuration.  I think that 
overriding default storage settings would 
also be confusing to users.

So I would argue that we should retain mandatory 
OpenSim.ini/OpenSim.ini.example just for storage and other absolutely 
basic settings (e.g. gridmode).  This would also mean that many users wouldn't 
need to worry at all about the contents 
of config/

I acknowledge this may be a pita for those supplying configuration purely over 
the network, though maybe there could be 
some extra setting that switches off a mandatory OpenSim.ini in these 
situations.

 
 That is why the config directory needs to be read first - it 
 provides the defaults! Inimaster would be rendered useless by the 
 load order you're proposing!
 
 Then I guees the load order would be:
  - read inimaster (if present)
  - read config/*.ini (if present)
  - use defaults in code for anything not set

 
 See above, this won't serve the users' needs. In your scenario, if 
 the config/*.example changes, they'd still need to merge.
 
 What I propose, to reiterate, is this:
 
 Break up Opensim.ini.example into config/*.ini. Uncomment all the 
 options so they serve as defaults.
 Remove the explicit requirement for an ini file.
 This way, sitewide changes from the provided defaults go to 
 inimaster, and machine/instance settings go to opensim.ini.
 The files in config/ can remain unchanged and can be updated on 
 every svn update, without manual merging.
 
 Of course, many other use cases are possible, including 
 modifications to config/*.ini if desired, or omitting either 
 configuration source.
 
 Melanie
 ___
 Opensim-dev mailing list
 Opensim-dev@lists.berlios.de
 https://lists.berlios.de/mailman/listinfo/opensim-dev
 


-- 
justincc
Justin Clark-Casey
http://justincc.wordpress.com
___
Opensim-dev mailing list
Opensim-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/opensim-dev


Re: [Opensim-dev] Ini file(s) loading

2009-03-06 Thread Justin Clark-Casey
Frisby, Adam wrote:
 I did wonder why we started forcing users to have an opensim.ini. The 
 previous 'use defaults' made more sense to me.

Unfortunately, the hardcoded defaults had become out of alignment with sane 
default settings, causing various degrees of 
user distress.

Any time where settings exist in two places there is a big risk of them getting 
out of sync.  I would argue that modules 
should fail on startup rather than contain hardcoded defaults.  External files 
such as OpenSim.ini.example (or 
config/*.ini) are better in that they provide explanatory comments and make 
available settings user visible.

 
 Adam
 
 -Original Message-
 From: opensim-dev-boun...@lists.berlios.de [mailto:opensim-dev-
 boun...@lists.berlios.de] On Behalf Of Jeff Ames
 Sent: Thursday, 5 March 2009 7:20 PM
 To: opensim-dev@lists.berlios.de
 Subject: Re: [Opensim-dev] Ini file(s) loading

 Melanie wrote:
 read [the config directory] first
 then read the inimaster
 then read the inifile
 If I understand this correctly, the config/*.ini files would be
 essentially read-only, and all local changes would be made to the
 inimaster or OpenSim.ini.  But then OpenSim.ini is not broken up, and
 it may be confusing to users why there are two sets of config files.

 Is this just due to OpenSim's current behavior of requiring an .ini
 file to be present?  Currently the default values for all settings
 exist in the code itself and in the .ini.example file (itself an
 unfortunate duplication, but that's another topic).  Instead of
 requiring that an .ini be present, we could simply use the default
 values in the code if there is no .ini.  This would also have the
 pleasant side effect of matching the behavior when an empty .ini file
 is present.

 Then we could break up and move OpenSim.ini.example entirely to
 config/*.ini.example files, and when the user wants to change a value,
 create foo.ini based on foo.ini.example (copying the whole file if
 they want everything, or only adding the options they want to
 explicitly set).

 Then I guees the load order would be:
  - read inimaster (if present)
  - read config/*.ini (if present)
  - use defaults in code for anything not set

 I think this would also avoid the merging problem, if users only add
 options they're explicitly setting to the *.ini files.  It would also
 remove the annoyance of having to copy the .ini.example file over
 every time on a new install.

 Jeff
 ___
 Opensim-dev mailing list
 Opensim-dev@lists.berlios.de
 https://lists.berlios.de/mailman/listinfo/opensim-dev
 ___
 Opensim-dev mailing list
 Opensim-dev@lists.berlios.de
 https://lists.berlios.de/mailman/listinfo/opensim-dev
 


-- 
justincc
Justin Clark-Casey
http://justincc.wordpress.com
___
Opensim-dev mailing list
Opensim-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/opensim-dev


Re: [Opensim-dev] Ini file(s) loading

2009-03-05 Thread Mic Bowman
that setup would be very helpful.
--mic


On Thu, Mar 5, 2009 at 5:29 AM, Melanie mela...@t-data.com wrote:
 To me, what would make the most sense is this:

 - break up the Opensim.ini.example into chunks, and but them into a
 config directory
 - read this directory first
 - then read the inimaster
 - then read the inifile
 - this means that new opensim.ini options would, in the future,
 appear in these defaults files, which could be left unchanged in
 most cases, therefore adding new options is easy on grid operators,
 no more merging

 Melanie


 Mic Bowman wrote:
 fwiw... we use a shared inimaster as the base for the configuration
 of simulators running on multiple servers. then each simulator has a
 very small opensim.ini, typically just to configure the network ports
 and the storage parameters. that is, the local opensim.ini overwrites
 the shared values in the inimaster. bulk changes are easy (just change
 the inimaster) and customizations are very small.

 --mic


 On Wed, Mar 4, 2009 at 4:14 PM, MW michaelwr...@yahoo.co.uk wrote:
 Well I only inserted the loading from the config folder, inbetween the two
 existing steps. It has always been that the 'masterini' file was loaded
 first (if one was given as a argument) and then opensim.ini (or whatever was
 set with -inifile) was loaded.

 The original idea behind the master being that it could be a master file for
 say a whole grid, while each one had some local changes in the
 bin\opensim.ini.

 But I agree that it can be confusing, especially with the chosen names. So
 maybe we do need to rethink that.

 --- On Wed, 4/3/09, Sean Hennessee s...@uci.edu wrote:

 From: Sean Hennessee s...@uci.edu
 Subject: Re: [Opensim-dev] Ini file(s) loading
 To: michaelwr...@yahoo.co.uk, opensim-dev@lists.berlios.de
 Date: Wednesday, 4 March, 2009, 10:37 PM

 Yay!! +1

 Updating ini files is often a pain. This is an excellent step in the right
 direction.

 In this scenario, my suggestion/preference would be to have the bin/config
 directory contain all the SVN default values, (any number of files), with NO
 bin/opensim.ini file present by default. Then I could create a
 bin/opensim.ini file as a very small set of overriding values specific to my
 grids. The inimaster really wouldn't be used since it would always get
 overwritten.

 Below are my suggestions instead of using this parsing order.

 Looking at the three steps, I'm not sure I understand the logic of the order
 chosen. The 'inimaster' sounds like it should be the master and not ever
 be overwritten, i.e. it should be parsed last. To keep things at least more
 familiar to what they are now, I would prefer that the bin/opensim.ini file
 be the same one it is now, the default one that comes with the SVN, and
 parsed first. The bin/config directory would then be the second set of ini's
 parsed to allow my multiple ini files to override the standard default.
 Then, if one specifically identifies an inimaster, that should override
 everything else, especially since you have to specify it on the command
 line. My line of thinking is that if you are specifically saying use this
 inimaster on the command line that would mean you really want to use it
 only, as in a testing scenario without having to worry about default/other
 overrides. IIRC other linux/server type programs, like sendmail, behave
 similar to this.

 It might also be nice to have an include type of statement in the ini
 file/url allowing you to reference files/urls without having them in a
 special location as well as specifying the order you want them read instead
 of having to make them alphabetical for order.

 Anyway, my OS$2. And thanks for working on this!

 Peace,
 Sean

 MW wrote:

 Last week, I added the ability for opensim to search a folder for ini files
 and load (and merge together) all of those files.

 By default it will look for the folder 'bin\config' and search that for .ini
 files. The folder it searches can be changed by using the command line
 argument -inidirectory=path.

 So now our ini loading is a three step process:

 Step 1: If the -inimaster=filepath or url command line argument was used,
 load that file/url.

 Step 2: Then check if the config directory (default 'bin\config' or one set
 by -inidirectory=path)  exists and search that folder for ini files and
 load them. They are merged on top of the inimaster file. So any duplicate
 entries will be overwrote.

 Step 3: Load the inifile, which by default is 'bin\opensim.ini' but can be
 changed by using the -inifile=filepath or url command line argument. This
 again is merged on top of all previous settings so will overwrite the older
 (from previous steps) settings.

 Now the question is how we should set things up in SVN as the default. It
 would be nice to split up the opensim.ini, into multiple ini files in the
 config directory. And then just have a small opensim.ini that contains the
 values likely to change ofter, like network settings etc.

 But because

Re: [Opensim-dev] Ini file(s) loading

2009-03-05 Thread Frisby, Adam
I did wonder why we started forcing users to have an opensim.ini. The previous 
'use defaults' made more sense to me.

Adam

 -Original Message-
 From: opensim-dev-boun...@lists.berlios.de [mailto:opensim-dev-
 boun...@lists.berlios.de] On Behalf Of Jeff Ames
 Sent: Thursday, 5 March 2009 7:20 PM
 To: opensim-dev@lists.berlios.de
 Subject: Re: [Opensim-dev] Ini file(s) loading

 Melanie wrote:
  read [the config directory] first
  then read the inimaster
  then read the inifile

 If I understand this correctly, the config/*.ini files would be
 essentially read-only, and all local changes would be made to the
 inimaster or OpenSim.ini.  But then OpenSim.ini is not broken up, and
 it may be confusing to users why there are two sets of config files.

 Is this just due to OpenSim's current behavior of requiring an .ini
 file to be present?  Currently the default values for all settings
 exist in the code itself and in the .ini.example file (itself an
 unfortunate duplication, but that's another topic).  Instead of
 requiring that an .ini be present, we could simply use the default
 values in the code if there is no .ini.  This would also have the
 pleasant side effect of matching the behavior when an empty .ini file
 is present.

 Then we could break up and move OpenSim.ini.example entirely to
 config/*.ini.example files, and when the user wants to change a value,
 create foo.ini based on foo.ini.example (copying the whole file if
 they want everything, or only adding the options they want to
 explicitly set).

 Then I guees the load order would be:
  - read inimaster (if present)
  - read config/*.ini (if present)
  - use defaults in code for anything not set

 I think this would also avoid the merging problem, if users only add
 options they're explicitly setting to the *.ini files.  It would also
 remove the annoyance of having to copy the .ini.example file over
 every time on a new install.

 Jeff
 ___
 Opensim-dev mailing list
 Opensim-dev@lists.berlios.de
 https://lists.berlios.de/mailman/listinfo/opensim-dev
___
Opensim-dev mailing list
Opensim-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/opensim-dev


Re: [Opensim-dev] Ini file(s) loading

2009-03-04 Thread Sean Hennessee




Yay!! +1

Updating ini files is often a pain. This is an excellent step in the
right direction.

In this scenario, my suggestion/preference would be to have the
bin/config directory contain all the SVN default values, (any number of
files), with NO bin/opensim.ini file present by default. Then I could
create a bin/opensim.ini file as a very small set of overriding values
specific to my grids. The inimaster really wouldn't be used since it
would always get overwritten.

Below are my suggestions instead of using this parsing order.

Looking at the three steps, I'm not sure I understand the logic of the
order chosen. The 'inimaster' sounds like it should be the "master" and
not ever be overwritten, i.e. it should be parsed last. To keep things
at least more familiar to what they are now, I would prefer that the
bin/opensim.ini file be the same one it is now, the default one that
comes with the SVN, and parsed first. The bin/config directory would
then be the second set of ini's parsed to allow my multiple ini files
to override the standard default. Then, if one specifically identifies
an inimaster, that should override everything else, especially since
you have to specify it on the command line. My line of thinking is that
if you are specifically saying "use this inimaster" on the command line
that would mean you really want to use it only, as in a testing
scenario without having to worry about default/other overrides. IIRC
other linux/server type programs, like sendmail, behave similar to this.

It might also be nice to have an "include" type of statement in the ini
file/url allowing you to reference files/urls without having them in a
special location as well as specifying the order you want them read
instead of having to make them alphabetical for order.

Anyway, my OS$2. And thanks for working on this!

Peace,
Sean

MW wrote:

  

  
Last week, I added the ability for opensim to search a
folder for ini files and load (and merge together) all of those files. 

By default it will look for the folder 'bin\config' and search that for
.ini files. The folder it searches can be changed by using the command
line argument -inidirectory=path.

So now our ini loading is a three step process:

Step 1: If the -inimaster=filepath or url command line argument
was used, load that file/url.

Step 2: Then check if the config directory (default 'bin\config' or one
set by -inidirectory=path) exists and search that folder for
ini files and load them. They are merged on top of the inimaster file.
So any duplicate entries will be overwrote.

Step 3: Load the inifile, which by default is 'bin\opensim.ini' but can
be changed by using the -inifile=filepath or url command line
argument. This again is merged on top of all previous settings so will
overwrite the older (from previous steps) settings.

Now the question is how we should set things up in SVN as the default.
It would be nice to split up the opensim.ini, into multiple ini files
in the config directory. And then just have a small opensim.ini that
contains the values likely to change ofter, like network settings etc. 

But because the config directory overwrites the master file, this could
cause problems for people who have already set things up for using a
master file/url. Of course the simple solution for them would be just
to delete the config directory.

But anyway what does everyone think? 

  

  
  
  

___
Opensim-dev mailing list
Opensim-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/opensim-dev
  


-- 

Sean Hennessee
Central Computing Support
Network  Academic Computing Services
UC Irvine


... . .- -. /   . -. -. . ... ... . .


___
Opensim-dev mailing list
Opensim-dev@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/opensim-dev


Re: [Opensim-dev] Ini file(s) loading

2009-03-04 Thread Mic Bowman
fwiw... we use a shared inimaster as the base for the configuration
of simulators running on multiple servers. then each simulator has a
very small opensim.ini, typically just to configure the network ports
and the storage parameters. that is, the local opensim.ini overwrites
the shared values in the inimaster. bulk changes are easy (just change
the inimaster) and customizations are very small.

--mic


On Wed, Mar 4, 2009 at 4:14 PM, MW michaelwr...@yahoo.co.uk wrote:
 Well I only inserted the loading from the config folder, inbetween the two
 existing steps. It has always been that the 'masterini' file was loaded
 first (if one was given as a argument) and then opensim.ini (or whatever was
 set with -inifile) was loaded.

 The original idea behind the master being that it could be a master file for
 say a whole grid, while each one had some local changes in the
 bin\opensim.ini.

 But I agree that it can be confusing, especially with the chosen names. So
 maybe we do need to rethink that.

 --- On Wed, 4/3/09, Sean Hennessee s...@uci.edu wrote:

 From: Sean Hennessee s...@uci.edu
 Subject: Re: [Opensim-dev] Ini file(s) loading
 To: michaelwr...@yahoo.co.uk, opensim-dev@lists.berlios.de
 Date: Wednesday, 4 March, 2009, 10:37 PM

 Yay!! +1

 Updating ini files is often a pain. This is an excellent step in the right
 direction.

 In this scenario, my suggestion/preference would be to have the bin/config
 directory contain all the SVN default values, (any number of files), with NO
 bin/opensim.ini file present by default. Then I could create a
 bin/opensim.ini file as a very small set of overriding values specific to my
 grids. The inimaster really wouldn't be used since it would always get
 overwritten.

 Below are my suggestions instead of using this parsing order.

 Looking at the three steps, I'm not sure I understand the logic of the order
 chosen. The 'inimaster' sounds like it should be the master and not ever
 be overwritten, i.e. it should be parsed last. To keep things at least more
 familiar to what they are now, I would prefer that the bin/opensim.ini file
 be the same one it is now, the default one that comes with the SVN, and
 parsed first. The bin/config directory would then be the second set of ini's
 parsed to allow my multiple ini files to override the standard default.
 Then, if one specifically identifies an inimaster, that should override
 everything else, especially since you have to specify it on the command
 line. My line of thinking is that if you are specifically saying use this
 inimaster on the command line that would mean you really want to use it
 only, as in a testing scenario without having to worry about default/other
 overrides. IIRC other linux/server type programs, like sendmail, behave
 similar to this.

 It might also be nice to have an include type of statement in the ini
 file/url allowing you to reference files/urls without having them in a
 special location as well as specifying the order you want them read instead
 of having to make them alphabetical for order.

 Anyway, my OS$2. And thanks for working on this!

 Peace,
 Sean

 MW wrote:

 Last week, I added the ability for opensim to search a folder for ini files
 and load (and merge together) all of those files.

 By default it will look for the folder 'bin\config' and search that for .ini
 files. The folder it searches can be changed by using the command line
 argument -inidirectory=path.

 So now our ini loading is a three step process:

 Step 1: If the -inimaster=filepath or url command line argument was used,
 load that file/url.

 Step 2: Then check if the config directory (default 'bin\config' or one set
 by -inidirectory=path)  exists and search that folder for ini files and
 load them. They are merged on top of the inimaster file. So any duplicate
 entries will be overwrote.

 Step 3: Load the inifile, which by default is 'bin\opensim.ini' but can be
 changed by using the -inifile=filepath or url command line argument. This
 again is merged on top of all previous settings so will overwrite the older
 (from previous steps) settings.

 Now the question is how we should set things up in SVN as the default. It
 would be nice to split up the opensim.ini, into multiple ini files in the
 config directory. And then just have a small opensim.ini that contains the
 values likely to change ofter, like network settings etc.

 But because the config directory overwrites the master file, this could
 cause problems for people who have already set things up for using a master
 file/url. Of course the simple solution for them would be just to delete the
 config directory.

 But anyway what does everyone think?

 
 ___
 Opensim-dev mailing list
 Opensim-dev@lists.berlios.de
 https://lists.berlios.de/mailman/listinfo/opensim-dev


 --

 Sean Hennessee
 Central Computing Support
 Network  Academic Computing Services
 UC Irvine