kadmin-remote branch

["Zheng, Kai" <ka...@intel.com>]

Note kadmin-remote branch is created for the remote kadmin support work.

-----Original Message-----
From: drankye@apache.org [mailto:drankye@apache.org] 
Sent: Friday, January 08, 2016 10:57 PM
To: commits@directory.apache.org
Subject: [directory-kerby] Git Push Summary

Repository: directory-kerby
Updated Branches:
  refs/heads/kadmin-remote [created] bb9b05bc1

RE: kadmin-remote branch

["Zheng, Kai" <ka...@intel.com>]

>>At the end of the day, I would say : 'whatever works'.
Yeah, it's my end of the day. Thanks anyway for the discussion, because discussion is good than none, even not smoothly!

Kindly regards,
Kai

-----Original Message-----
From: Emmanuel Lécharny [mailto:elecharny@gmail.com] 
Sent: Monday, January 11, 2016 8:03 PM
To: kerby@directory.apache.org
Subject: Re: kadmin-remote branch

Le 11/01/16 12:12, Zheng, Kai a écrit :
>>> your initial message was "kadmin-remote branch is created for the remote kadmin support work.".
> My message here may be a little simple, but I based on the assumption and fact that we have discussed quite much about the feature. 

'Quite much' is an overstatement. Beside one mail you sent in january, 3th, there was no discussion about it.

Not that is is a big problem, though, I just wanted some clarification.
I'm not per se opposed to the crzation of a branch, I'm just asking for clarification. As I said, creating a branch for such a thing is a real problem for those using an IDE. But still, it may be a good thing to do.

Frankly, if it has to become a separate tool, I'd rather ask for a new repo to be created, as we have done in Fortress : that let you work easily on various projects without dealing with the pain of switching the branch when you have to work on some parts that are not in the branch or in the trunk.

Don't get me wrong : this is not a philosophical discussion, nor is it a blame whatsoever, it's mainly a way to address a potential issue beforhand. At the end of the day, I would say : 'whatever works'.

It's just that this morning I saw some Jenkcins failure on this branch, and I told myself : 'uh-oh, why a branch ?' and I tried to check if it was a good idea, a bad idea, or simply something that has to be accepted, with the small inconveniences that comes with.


Also don't underestimate the double language barrier : you are not a native english speaker, I'm nt either a naitve english speaker, so
*many* sentences and words we both are using my be misintepreted... This makes it difficult to communicate, but we have to gate use to it, it's not a problem that can be solved magically ! (When I was a kid, I thought that soon enough we will have soime 'mind reader' machines that will automatically translate in perfect native language whet someone else says in his/her own language... Seems like I'll pass away before I will see these kind of machines !!!)



Anyway, keep going with the good work, and let's go with a branch, if you thinks it's the way to go !

Re: kadmin-remote branch

[Emmanuel Lécharny <el...@gmail.com>]

Le 11/01/16 12:12, Zheng, Kai a écrit :
>>> your initial message was "kadmin-remote branch is created for the remote kadmin support work.".
> My message here may be a little simple, but I based on the assumption and fact that we have discussed quite much about the feature. 

'Quite much' is an overstatement. Beside one mail you sent in january,
3th, there was no discussion about it.

Not that is is a big problem, though, I just wanted some clarification.
I'm not per se opposed to the crzation of a branch, I'm just asking for
clarification. As I said, creating a branch for such a thing is a real
problem for those using an IDE. But still, it may be a good thing to do.

Frankly, if it has to become a separate tool, I'd rather ask for a new
repo to be created, as we have done in Fortress : that let you work
easily on various projects without dealing with the pain of switching
the branch when you have to work on some parts that are not in the
branch or in the trunk.

Don't get me wrong : this is not a philosophical discussion, nor is it a
blame whatsoever, it's mainly a way to address a potential issue
beforhand. At the end of the day, I would say : 'whatever works'.

It's just that this morning I saw some Jenkcins failure on this branch,
and I told myself : 'uh-oh, why a branch ?' and I tried to check if it
was a good idea, a bad idea, or simply something that has to be
accepted, with the small inconveniences that comes with.


Also don't underestimate the double language barrier : you are not a
native english speaker, I'm nt either a naitve english speaker, so
*many* sentences and words we both are using my be misintepreted... This
makes it difficult to communicate, but we have to gate use to it, it's
not a problem that can be solved magically ! (When I was a kid, I
thought that soon enough we will have soime 'mind reader' machines that
will automatically translate in perfect native language whet someone
else says in his/her own language... Seems like I'll pass away before I
will see these kind of machines !!!)



Anyway, keep going with the good work, and let's go with a branch, if
you thinks it's the way to go !

RE: kadmin-remote branch

["Zheng, Kai" <ka...@intel.com>]

>> your initial message was "kadmin-remote branch is created for the remote kadmin support work.".
My message here may be a little simple, but I based on the assumption and fact that we have discussed quite much about the feature. It's not a simple fix, but rather a significant effort that may take months.

>> why do we need a branch for something that deserves to be in trunk ? Or maybe it does not deserve to be in trunk, but I don't see why. Can you explain us ?
I thought I had explained before and just don't know how to explain it better. It's not a small fix or refactoring stuff, it may take months. It should be in trunk, because it evolves faster. It's a common practice to do such things in a feature branch. Nothing needless to say. The "deserve" word looks strange to me. The feature branch looks equally important in my point of view.

Regards,
Kai

-----Original Message-----
From: Emmanuel Lécharny [mailto:elecharny@gmail.com] 
Sent: Monday, January 11, 2016 6:22 PM
To: kerby@directory.apache.org
Subject: Re: kadmin-remote branch

Le 11/01/16 10:56, Zheng, Kai a écrit :
>>> Even if it's a brand new piece of code, I don't see why it should be seggregated to a branch, instead of being part of the trunk.
> I was surprised now you changed to think in this way, actually you said branch is cheap in git and we should use new branches for features. 

For features that need to be tested. But here, your initial message was "kadmin-remote branch is created for the remote kadmin support work.".
Does not look like a 'feature addition' to me.

Don't get me wrong, I have nothing about creating branches for experimentation, or for fixes done on existing released, but here, it's not clear we are in any of those two possibilities...



> If you remembered, in previous discussion before, I suggested we use trunk for development, and cutting separate branch for RC2 release. 
I don't think we need to cut a branch for a release. The maven build
*will* create them automatically. Those wanting to work on teh code while it's being released can create a branch and merge it back to the trunk at will.

What I don't get here is why do we need a branch for something that deserves to be in trunk ? Or maybe it does not deserve to be in trunk, but I don't see why. Can you explain us ?

thanks !

Re: kadmin-remote branch

[Emmanuel Lécharny <el...@gmail.com>]

Le 11/01/16 10:56, Zheng, Kai a écrit :
>>> Even if it's a brand new piece of code, I don't see why it should be seggregated to a branch, instead of being part of the trunk.
> I was surprised now you changed to think in this way, actually you said branch is cheap in git and we should use new branches for features. 

For features that need to be tested. But here, your initial message was
"kadmin-remote branch is created for the remote kadmin support work.".
Does not look like a 'feature addition' to me.

Don't get me wrong, I have nothing about creating branches for
experimentation, or for fixes done on existing released, but here, it's
not clear we are in any of those two possibilities...



> If you remembered, in previous discussion before, I suggested we use trunk for development, and cutting separate branch for RC2 release. 
I don't think we need to cut a branch for a release. The maven build
*will* create them automatically. Those wanting to work on teh code
while it's being released can create a branch and merge it back to the
trunk at will.

What I don't get here is why do we need a branch for something that
deserves to be in trunk ? Or maybe it does not deserve to be in trunk,
but I don't see why. Can you explain us ?

thanks !

RE: kadmin-remote branch

["Zheng, Kai" <ka...@intel.com>]

>> Even if it's a brand new piece of code, I don't see why it should be seggregated to a branch, instead of being part of the trunk.
I was surprised now you changed to think in this way, actually you said branch is cheap in git and we should use new branches for features. If you remembered, in previous discussion before, I suggested we use trunk for development, and cutting separate branch for RC2 release. 

>> Being in trunk does not mean you have to build it with the other modules (enough to comment it in the main pom)...
It would look rather messy in this way. It maybe works but still depends on the fact that there're only new codes that reside in the separate new module, no new codes and no modification codes into existing modules.

>> you have to reimport the full project everytime you switch
Yeah, I felt some inconvenience. I used a work around, how about two working copies.

Regards,
Kai

-----Original Message-----
From: Emmanuel Lécharny [mailto:elecharny@gmail.com] 
Sent: Monday, January 11, 2016 5:35 PM
To: kerby@directory.apache.org
Subject: Re: kadmin-remote branch

Le 11/01/16 10:18, Zheng, Kai a écrit :
> I don't quite understand your question. Apparently the codes are in some module(s), but isn't relevant to the question that whether it should stay in trunk or a branch. Kadmin-remote is a big new feature that involves significant new codes and currently it's still in its very early phase, therefore I don't think we should do it in trunk.
I probably wasn't clear.

Even if it's a brand new piece of code, I don't see why it should be seggregated to a branch, instead of being part of the trunk. Being in trunk does not mean you have to build it with the other modules (enough to comment it in the main pom), but having it in a branch froces the developers to switch to the branches to build it. Having it in a branch makes it painful because even if it's easy to switch from trunk to branch and back, it's not the same story when it comes to teh IDE integration (basically, you have to reimport the full project everytime you switch).

Re: kadmin-remote branch

[Emmanuel Lécharny <el...@gmail.com>]

Le 11/01/16 10:18, Zheng, Kai a écrit :
> I don't quite understand your question. Apparently the codes are in some module(s), but isn't relevant to the question that whether it should stay in trunk or a branch. Kadmin-remote is a big new feature that involves significant new codes and currently it's still in its very early phase, therefore I don't think we should do it in trunk.
I probably wasn't clear.

Even if it's a brand new piece of code, I don't see why it should be
seggregated to a branch, instead of being part of the trunk. Being in
trunk does not mean you have to build it with the other modules (enough
to comment it in the main pom), but having it in a branch froces the
developers to switch to the branches to build it. Having it in a branch
makes it painful because even if it's easy to switch from trunk to
branch and back, it's not the same story when it comes to teh IDE
integration (basically, you have to reimport the full project everytime
you switch).

RE: kadmin-remote branch

["Zheng, Kai" <ka...@intel.com>]

I don't quite understand your question. Apparently the codes are in some module(s), but isn't relevant to the question that whether it should stay in trunk or a branch. Kadmin-remote is a big new feature that involves significant new codes and currently it's still in its very early phase, therefore I don't think we should do it in trunk.

Regards,
Kai

-----Original Message-----
From: Emmanuel Lécharny [mailto:elecharny@gmail.com] 
Sent: Monday, January 11, 2016 4:51 PM
To: kerby@directory.apache.org
Subject: Re: kadmin-remote branch

Le 08/01/16 15:58, Zheng, Kai a écrit :
> Note kadmin-remote branch is created for the remote kadmin support work.

Why is that ? Can't it be seen as a sub-module of the Kerby project, instead of being moved to a branch ?

Re: kadmin-remote branch

[Emmanuel Lécharny <el...@gmail.com>]

Le 08/01/16 15:58, Zheng, Kai a écrit :
> Note kadmin-remote branch is created for the remote kadmin support work.

Why is that ? Can't it be seen as a sub-module of the Kerby project,
instead of being moved to a branch ?