Forum: IronRuby Contributing?

D91500918a07bfae5b38b8943fd42fbe?d=identicon&s=25 Mike Moore (Guest)
on 2010-10-23 02:19
(Received via mailing list)
With a brave new world ahead for IronRuby, what do you all think about
the
following ideas?

1) Move to GitHub - I won't likely contribute if I have to use CodePlex.
And
I'd like to contribute. Can we do like the rest of Mono and make GitHub
the
central repository and source of all truth? Please?

2) Detach IronPython - Looking through the code, it seems that
IronPython is
in the same repo. I'm sure this is a holdover of how Microsoft had
everything organized. And I'm sure it made sense then, but I don't think
it
makes sense now. I'd like to work on IronRuby, and I honestly don't care
about having to check IronPython. Now that the DLR API is fixed, and
there
is no longer one Microsoft group in charge of both projects, can we move
these into separate repos?

3) Detach from Visual Studio - I'm not on Windows, and I won't be using
Visual Studio. I'd prefer everything to build with rake tasks on Windows
and
Mono. I assume things are the way they are right now because of
Microsoft's
build process. And right now it does not seem possible for me to build
without Visual Studio. What about devs who are using MonoDevelop? Or any
of
the more powerful text editors? I'm convinced this is a major hurdle to
getting many of the other Rubyist language implementors involved.
Assuming I
am not alone on this, is it possible we can make this project reflect
the
realities of the majority of folks that would like to contribute?

4) Remove ancillary projects into their own repos - IronRuby.Rack should
live in its own repo, IMO. So should the Visual Studio tools. This is
similar to separating IronPython into its own repo, but it needs to go
further. Are there any reasons this should not be done?

What else do you guys think needs to change now that IronRuby is free
from
Microsoft's processes?
68f1bd9e9a7d9d10f0278c6535f700fe?d=identicon&s=25 Michael Letterle (mletterle)
on 2010-10-23 03:50
(Received via mailing list)
A couple thoughts below:

On Fri, Oct 22, 2010 at 8:10 PM, Mike Moore <blowmage@gmail.com> wrote:
> With a brave new world ahead for IronRuby, what do you all think about the
> following ideas?
> 1) Move to GitHub - I won't likely contribute if I have to use CodePlex. And
> I'd like to contribute. Can we do like the rest of Mono and make GitHub the
> central repository and source of all truth? Please?

http://github.com/ironruby/ironruby

My understanding is that Codeplex is used only for end user
distribution and bug tracking

> 3)Detachfrom Visual Studio - I'm not on Windows, and I won't be using
> Visual Studio. I'd prefer everything to build with rake tasks on Windows and
> Mono. I assume things are the way they are right now because of Microsoft's
> build process. And right now it does not seem possible for me to build
> without Visual Studio. What about devs who are using MonoDevelop? Or any of
> the more powerful text editors? I'm convinced this is a major hurdle to
> getting many of the other Rubyist language implementors involved. Assuming I
> am not alone on this, is it possible we can make this project reflect the
> realities of the majority of folks that would like to contribute?

You can certainly build without VS, this has always been the case.
All you need is msbuild or xbuild for the solution files, in the past
there have been Rake files as part of the repo from very early on..
but they seem to have disappeared.. anyone know what happened?

I'm taking a look now on building under mono on linux, it's something
me and Ivan were working on earlier in the year and this seems like a
good time to revisit.

--
Michael Letterle
IronRuby MVP
http://blog.prokrams.com
Cb51033949ffccd982ae32c9f890f25a?d=identicon&s=25 Tomas Matousek (Guest)
on 2010-10-23 03:55
(Received via mailing list)
1) IronRuby is on github. I has been there for a while. There is no
source code on CodePlex. The only CodePlex service we use is the bug
tracker.

2) It is not a holdover. It makes a lot of sense actually for at least
the following reasons:
a) Some IronRuby tests test interop between these languages. So there is
a direct dependency. When you debug issues in language interop you need
to have IronPython source code as well to step thru and make sense of
the interactions.
b) DLR has two parts - the "inner ring" that shipped in .NET Framework
4.0 and the "outer ring", which hasn't shipped. Although the outer ring
is pretty stable there are still many improvements that can/should be
done. Obviously when you change the DLR you should run tests for both
languages so that you don't break anything. Thus IronPython's test suite
in the repo is handy. Also, if you change public API in the outer ring
you might need to change both IronPython and IronRuby. All of this could
be done in stages across different repos and even source control
systems, but that's obviously much more complicated than having it just
work.

Is size of the repo really an issue? If not, what is?

3) It is absolutely possible to build without Visual Studio. On Mono you
can build using xbuild. Sometimes Mono's xbuild might have a bug but I
can assure you that Mono guys are more than happy to fix it asap if it
breaks IronRuby or IronPython builds. Doesn't MonoDevelop support
.csproj files? I have no experience with it, but I would assume it does.
The only part of the repo that is Visual Studio specific is (not
surprisingly) Visual Studio integration in the Tools directory. Building
a separate rake build system is just a redundancy that is difficult to
maintain. Trust me, we tried. The less duplication in the build
infrastructure the better.

4) We have an installer builder in the MSI directory that builds the
entire IronRuby.msi and IronPython.msi from command line. These
installers include Visual Studio integration so it needs to be there. Of
course we can come up with some sub-module structuring of the repo that
allows you to make a local enlistment with VS and without VS. That might
be the way to go. Though, again, why spend time on changing the
infrastructure? Shouldn't we focus on implementing features and fixing
bugs instead? Does the size of the repo bother you?

Be assured that me and Dino have spent a lot of time to make sure no
unnecessary process (Microsoft's or not) is needed when working on
IronRuby and IronPython.
There might be a few small things to clean up that were needed at
Microsoft and are not needed now. I'll be happy to make the changes. Two
of them that I can think of right now are removing "LCA_RESTRICTED" and
TFS bindings from the solutions. Other than that there is not much.

Any ideas how to make the development easier on Windows, Linux or MacOS
are indeed welcome. Just be aware there might be good reasons behind the
current repo setup and virtually none of them would be Microsoft process
requirements. I'll be happy to explain them should you have any
questions.

Thanks,
Tomas

From: ironruby-core-bounces@rubyforge.org
[mailto:ironruby-core-bounces@rubyforge.org] On Behalf Of Mike Moore
Sent: Friday, October 22, 2010 5:10 PM
To: ironruby-core@rubyforge.org
Subject: [Ironruby-core] Contributing?

With a brave new world ahead for IronRuby, what do you all think about
the following ideas?

1) Move to GitHub - I won't likely contribute if I have to use CodePlex.
And I'd like to contribute. Can we do like the rest of Mono and make
GitHub the central repository and source of all truth? Please?

2) Detach IronPython - Looking through the code, it seems that
IronPython is in the same repo. I'm sure this is a holdover of how
Microsoft had everything organized. And I'm sure it made sense then, but
I don't think it makes sense now. I'd like to work on IronRuby, and I
honestly don't care about having to check IronPython. Now that the DLR
API is fixed, and there is no longer one Microsoft group in charge of
both projects, can we move these into separate repos?

3) Detach from Visual Studio - I'm not on Windows, and I won't be using
Visual Studio. I'd prefer everything to build with rake tasks on Windows
and Mono. I assume things are the way they are right now because of
Microsoft's build process. And right now it does not seem possible for
me to build without Visual Studio. What about devs who are using
MonoDevelop? Or any of the more powerful text editors? I'm convinced
this is a major hurdle to getting many of the other Rubyist language
implementors involved. Assuming I am not alone on this, is it possible
we can make this project reflect the realities of the majority of folks
that would like to contribute?

4) Remove ancillary projects into their own repos - IronRuby.Rack should
live in its own repo, IMO. So should the Visual Studio tools. This is
similar to separating IronPython into its own repo, but it needs to go
further. Are there any reasons this should not be done?

What else do you guys think needs to change now that IronRuby is free
from Microsoft's processes?
71c1027d2483fe242b0affc5e59df647?d=identicon&s=25 Jonathan Allen (grauenwolf)
on 2010-10-23 03:56
(Received via mailing list)
I would caution you against moving too far away from Microsoft
Technology,
lest you alienate Windows developers.

1. If this project is going to be owned by Mono, then it would be
reasonable
to move to GitHub. But to drop CodePlex just because you feel like being
stubborn isn't.

3. One of the biggest barriers for Windows developers who want to
contribute
to Mono is the project/build system. When looking at it for the first
time
it is really hard to just figure out where to start. And they certainly
don't want to go back to editing source files with notepad. So while
being
able to build without Visual Studio is important, be careful not to
remove
the VS/MSBuild option.


Jonathan Allen
D91500918a07bfae5b38b8943fd42fbe?d=identicon&s=25 Mike Moore (Guest)
on 2010-10-23 06:03
(Received via mailing list)
On Fri, Oct 22, 2010 at 7:02 PM, Jonathan Allen
<grauenwolf@gmail.com>wrote:

> I would caution you against moving too far away from Microsoft Technology,
> lest you alienate Windows developers.
>

I don't think Windows developers could be anymore alienated from
IronRuby.


> 1. If this project is going to be owned by Mono, then it would be
> reasonable to move to GitHub. But to drop CodePlex just because you feel
> like being stubborn isn't.
>

Dropping CodePlex because it requires some Uber Visual Studio product to
access the code repo and make code contributions is reasonable, and not
some
emotional reaction. Gladly this doesn't seem to be the case anymore.
D91500918a07bfae5b38b8943fd42fbe?d=identicon&s=25 Mike Moore (Guest)
on 2010-10-23 06:12
(Received via mailing list)
On Fri, Oct 22, 2010 at 6:59 PM, Tomas Matousek <
Tomas.Matousek@microsoft.com> wrote:

>  2) It is not a holdover. It makes a lot of sense actually for at least
> the following reasons:
>
> a) Some IronRuby tests test interop between these languages. So there is a
> direct dependency. When you debug issues in language interop you need to
> have IronPython source code as well to step thru and make sense of the
> interactions.
>

I don't understand why the code needs to exist within the same git repo
for
this then. Can you not have a Visual Studio solution that includes
multiple
projects with their own repository? Can't you have your automated build
system pull both the IronRuby and IronPython projects to run the
integration
tests?


>  b) DLR has two parts  the inner ring that shipped in .NET Framework
> 4.0 and the outer ring, which hasnt shipped. Although the outer ring is
> pretty stable there are still many improvements that can/should be done.
> Obviously when you change the DLR you should run tests for both languages so
> that you dont break anything. Thus IronPythons test suite in the repo is
> handy. Also, if you change public API in the outer ring you might need to
> change both IronPython and IronRuby. All of this could be done in stages
> across different repos and even source control systems, but thats obviously
> much more complicated than having it just work.
>

I assumed the DLR was fixed. If that is not the case then shouldn't the
DLR
should be its own separate git repo as well?

Is size of the repo really an issue? If not, what is?
>

No, its not the size of the repo, rather the amount of tangental code.
As
someone who hasn't looked at the code for well over a year, I am struck
by
the amount of orthogonal concerns that are front and center. I find it
too
busy and off-putting as a developer looking to become familiar with the
code
or make a small contribution. I don't doubt there aren't reasons it is
the
way it is, just that it can be better if you want to really open the
floodgates of contributions.
68f1bd9e9a7d9d10f0278c6535f700fe?d=identicon&s=25 Michael Letterle (mletterle)
on 2010-10-23 06:16
(Received via mailing list)
>> 1. If this project is going to be owned by Mono, then it would be
>> reasonable to move to GitHub. But to drop CodePlex just because you feel
>> like being stubborn isn't.
>
> Dropping CodePlex because it requires some Uber Visual Studio product to
> access the code repo and make code contributions is reasonable, and not some
> emotional reaction. Gladly this doesn't seem to be the case anymore.
> _______________________________________________


Uh.. you know it doesn't, right? Codeplex supports SVN and Mercurial
and there are free TFS tools available in any case.

http://codeplex.codeplex.com/wikipage?title=Source...

--
Michael Letterle
IronRuby MVP
http://blog.prokrams.com
68f1bd9e9a7d9d10f0278c6535f700fe?d=identicon&s=25 Michael Letterle (mletterle)
on 2010-10-23 06:30
(Received via mailing list)
Along the lines of IronRuby on mono, a small change was required for
now since Mono doesn't implement one of the newer Monitor.TryEnter
overloads, see my branch here:
http://github.com/mletterle/ironruby/tree/linux

compiles with:

xbuild /property:TreatWarningsAsErrors=false
/property:DefineConstants="DEBUG MONO" Ruby.csproj

Working on getting IronRuby.Console compiled now... already ran into a
casing issue >.<

On Fri, Oct 22, 2010 at 9:02 PM, Jonathan Allen <grauenwolf@gmail.com>
wrote:
> don't want to go back to editing source files with notepad. So while being
>> 1) Move to GitHub - I won't likely contribute if I have to use CodePlex.
>> Visual Studio. I'd prefer everything to build with rake tasks on Windows and
>> further. Are there any reasons this should not be done?
> --
> Jonathan Allen
> 619-933-8527
>
> _______________________________________________
> Ironruby-core mailing list
> Ironruby-core@rubyforge.org
> http://rubyforge.org/mailman/listinfo/ironruby-core
>
>



--
Michael Letterle
IronRuby MVP
http://blog.prokrams.com
D91500918a07bfae5b38b8943fd42fbe?d=identicon&s=25 Mike Moore (Guest)
on 2010-10-23 06:31
(Received via mailing list)
On Fri, Oct 22, 2010 at 7:06 PM, Michael Letterle <
michael.letterle@gmail.com> wrote:

>
> http://github.com/ironruby/ironruby
>
> My understanding is that Codeplex is used only for end user
> distribution and bug tracking
>

Good to know. Last time I was paying attention this was not the case.
The
commits I've seen through twitter seemed to be syncing commits from
Microsoft's internal system out. I readily admin I am ignorant of the
truth
however. :)

You can certainly build without VS, this has always been the case.
> All you need is msbuild or xbuild for the solution files, in the past
> there have been Rake files as part of the repo from very early on..
> but they seem to have disappeared.. anyone know what happened?
>
> I'm taking a look now on building under mono on linux, it's something
> me and Ivan were working on earlier in the year and this seems like a
> good time to revisit.
>

Excellent. I'd prefer to have rake be the official way to test and
build,
since that is what every other Ruby project I work on does.
Ade8632553a9243ae05fc920f68644c1?d=identicon&s=25 Jim Deville (Guest)
on 2010-10-23 07:06
(Received via mailing list)
The Rake files were deleted because they were hard to maintain and
redundant.
D91500918a07bfae5b38b8943fd42fbe?d=identicon&s=25 Mike Moore (Guest)
on 2010-10-23 08:01
(Received via mailing list)
On Oct 22, 2010, at 10:00 PM, Jim Deville <jdeville@microsoft.com>
wrote:

> The Rake files were deleted because they were hard to maintain and redundant.

Not all of them.

http://github.com/ironruby/ironruby/blob/master/Rakefile
D91500918a07bfae5b38b8943fd42fbe?d=identicon&s=25 Mike Moore (Guest)
on 2010-10-23 08:05
(Received via mailing list)
On Oct 22, 2010, at 10:13 PM, Michael Letterle
<michael.letterle@gmail.com> wrote:

> Uh.. you know it doesn't, right?

Obviously I don't. :)
71c1027d2483fe242b0affc5e59df647?d=identicon&s=25 Jonathan Allen (grauenwolf)
on 2010-10-23 08:35
(Received via mailing list)
> Dropping CodePlex because it requires some Uber Visual Studio product to
access the code repo and make code contributions is reasonable, and not
some
emotional reaction.

While academic at this point, I would like to reiterate that you don't
need
Visual Studio for CodePlex, Subversion works just fine. In fact,
Subversion
is preferred over VS because it allows anonymous access the source
control
repository while VS requires at least developer status on the project.

Jonathan
99b1f0c67bec23747d007e27d000487b?d=identicon&s=25 Ryan Riley (Guest)
on 2010-10-23 09:30
(Received via mailing list)
On Oct 22, 2010, at 8:27 PM, Mike Moore <blowmage@gmail.com> wrote:
>
> Dropping CodePlex because it requires some Uber Visual Studio product to access
the code repo and make code contributions is reasonable, and not some emotional
reaction. Gladly this doesn't seem to be the case anymore.

Or, has never been the case. I've used vim and mono on some pieces
before, so the new arrangement doesn't change anything except allow us
to contribute directly to parts of the codebase we couldn't before. :)

Ryan
Cb51033949ffccd982ae32c9f890f25a?d=identicon&s=25 Tomas Matousek (Guest)
on 2010-10-23 09:54
(Received via mailing list)
What do you need the "DEBUG MONO" constant for? The default "Debug"
configuration doesn't work?

What version of Mono do you use? I think 2.8 should be the target for
IronRUby.

Tomas
Cb51033949ffccd982ae32c9f890f25a?d=identicon&s=25 Tomas Matousek (Guest)
on 2010-10-23 09:55
(Received via mailing list)
I don't understand how three distinct github repos that I need to map
into some directories on my disk whose relative location to each other
is hardcoded in some scripts in each are better than a single repo that
has a well-defined structure.

The amount of code you need to know depends on the part you contribute
to. If you work on IronRuby libraries you don't need to even look at
IronPython. It is indeed better to look there because similar
functionality might already be implemented there. For example, if you
worked on FFI you might want to check out IronPython's CTypes and
perhaps reuse some code. Why would you need to understand the entire
code base if you needed to do a local change? Just don't look where you
don't need to :). Building IronRuby is also simple - you just run xbuild
Ruby.sln from Solution directory. Subsequent builds are incremental, so
you don't even need to build everything if you don't change core
components. I don't see how this could be any simpler.

Tomas

From: ironruby-core-bounces@rubyforge.org
[mailto:ironruby-core-bounces@rubyforge.org] On Behalf Of Mike Moore
Sent: Friday, October 22, 2010 8:27 PM
To: ironruby-core@rubyforge.org
Subject: Re: [Ironruby-core] Contributing?

On Fri, Oct 22, 2010 at 6:59 PM, Tomas Matousek
<Tomas.Matousek@microsoft.com<mailto:Tomas.Matousek@microsoft.com>>
wrote:
2) It is not a holdover. It makes a lot of sense actually for at least
the following reasons:
a) Some IronRuby tests test interop between these languages. So there is
a direct dependency. When you debug issues in language interop you need
to have IronPython source code as well to step thru and make sense of
the interactions.

I don't understand why the code needs to exist within the same git repo
for this then. Can you not have a Visual Studio solution that includes
multiple projects with their own repository? Can't you have your
automated build system pull both the IronRuby and IronPython projects to
run the integration tests?

b) DLR has two parts - the "inner ring" that shipped in .NET Framework
4.0 and the "outer ring", which hasn't shipped. Although the outer ring
is pretty stable there are still many improvements that can/should be
done. Obviously when you change the DLR you should run tests for both
languages so that you don't break anything. Thus IronPython's test suite
in the repo is handy. Also, if you change public API in the outer ring
you might need to change both IronPython and IronRuby. All of this could
be done in stages across different repos and even source control
systems, but that's obviously much more complicated than having it just
work.

I assumed the DLR was fixed. If that is not the case then shouldn't the
DLR should be its own separate git repo as well?

Is size of the repo really an issue? If not, what is?

No, its not the size of the repo, rather the amount of tangental code.
As someone who hasn't looked at the code for well over a year, I am
struck by the amount of orthogonal concerns that are front and center. I
find it too busy and off-putting as a developer looking to become
familiar with the code or make a small contribution. I don't doubt there
aren't reasons it is the way it is, just that it can be better if you
want to really open the floodgates of contributions.
99b1f0c67bec23747d007e27d000487b?d=identicon&s=25 Ryan Riley (Guest)
on 2010-10-23 10:23
(Received via mailing list)
Jim, Jimmy, Tomas, and Miguel:

Shri used to offer pair programming sessions to people interested in
helping out with the project. I know at least a few participated, and I
found it really helpful. Is there any chance of starting this up again?
I don't mean for you to necessarily be the ones to do it, but you might
know some food candidates. This might help onboard people like Mike who
are interested but unsure where to begin.

Ryan

Sent from my iPhone
68f1bd9e9a7d9d10f0278c6535f700fe?d=identicon&s=25 Michael Letterle (mletterle)
on 2010-10-23 13:40
(Received via mailing list)
Well, it's just the DEBUG constant and a MONO constant.

I'm building with mono trunk, I added the MONO to the ifdef to use the
Monitor.TryEnter(object) overload that is also used for CLR2.  I
couldn't just use  the CLR2 constant because that broke everything
else.  I'm going to look into adding the Monitor.TryEnter(object,
bool) overloads to mono proper so this is just temporary.



On Sat, Oct 23, 2010 at 3:19 AM, Tomas Matousek
<Tomas.Matousek@microsoft.com> wrote:
> Subject: Re: [Ironruby-core] Contributing?
>
>> the first time it is really hard to just figure out where to start.
>>> about the following ideas?
>>> 3)Detachfrom Visual Studio - I'm not on Windows, and I won't be
>>> This is similar to separating IronPython into its own repo, but it
>>
>
>
> _______________________________________________
> Ironruby-core mailing list
> Ironruby-core@rubyforge.org
> http://rubyforge.org/mailman/listinfo/ironruby-core
>



--
Michael Letterle
IronRuby MVP
http://blog.prokrams.com
D91500918a07bfae5b38b8943fd42fbe?d=identicon&s=25 Mike Moore (Guest)
on 2010-10-23 15:05
(Received via mailing list)
On Oct 23, 2010, at 1:30 AM, Tomas Matousek
<Tomas.Matousek@microsoft.com> wrote:

> I don’t understand how three distinct github repos that I need to map into some
directories on my disk whose relative location to each other is hardcoded in 
some
scripts in each are better than a single repo that has a well-defined structure.
>
You are speaking like someone responsible for both languages and the
DLR. I'm making a suggestion as someone really only interested in
IronRuby. The repo isn't called "DynamicLanguages", it's called
"IronRuby", which is at best confusing. If only git had some way to
define a link to another repository as some sort of sub module...

As a Rubyist I'd like to run a rake task to build to each defined target
and run the RubySpecs. It wouldn't replace xbuild, just automate it. I
don't understand the pushback to this idea. Why not make a dedicated
repo for IronRuby free of the ancillary projects and geared to someone
like me? And likewise make the IronPython repo friendly to our Pythonic
friends?
> The amount of code you need to know depends on the part you contribute to. If
you work on IronRuby libraries you don’t need to even look at IronPython. It is
indeed better to look there because similar functionality might already be
implemented there. For example, if you worked on FFI you might want to check out
IronPython’s CTypes and perhaps reuse some code. Why would you need to 
understand
the entire code base if you needed to do a local change? Just don’t look where 
you
don’t need to J. Building IronRuby is also simple – you just run xbuild Ruby.sln
from Solution directory. Subsequent builds are incremental, so you don’t even 
need
to build everything if you don’t change core components. I don’t see how this
could be any simpler.
>
Unfortunately I do see how this could be simpler.
F10efb534989afcadb26ffbc858fc3ab?d=identicon&s=25 Douglas Husemann (Guest)
on 2010-10-23 17:19
(Received via mailing list)
Maybe it is time to come up with the why's??

This is how I see the Iron languages.

The languages are built on top of the DLR for Interoperability between
the languages and
to Ensure that the DLR offers the basic building blocks for Dynamic
languages and to the base CLR.
(be that mono or the CLR now that it is oss)

the Iron languages are built so that a Rubiest will feel comfortable in
the .net platform or a pythonist for that matter.

But more importantly that the libraries built can be used clr/dlr wide.
IOW a Ruby library can be used (in theory) by
a rubiest for sure, but also a pythonist, a c sharper, a Vber, and for
that matter a Fsharper.

Now to get working on my programming skills so I can contribute in the
future.  took a 17 year diversion into equipment
 repair.

But to me the single most important part of the project is the
interoperability it offers.

D


From: blowmage@gmail.com
Date: Sat, 23 Oct 2010 06:44:16 -0600
To: ironruby-core@rubyforge.org
Subject: Re: [Ironruby-core] Contributing?



On Oct 23, 2010, at 1:30 AM, Tomas Matousek
<Tomas.Matousek@microsoft.com> wrote:



I dont understand how three distinct github repos that I need to map
into some directories on my disk whose relative location to each other
is hardcoded in some scripts in each are better than a single repo that
has a well-defined structure.
You are speaking like someone responsible for both languages and the
DLR. I'm making a suggestion as someone really only interested in
IronRuby. The repo isn't called "DynamicLanguages", it's called
"IronRuby", which is at best confusing. If only git had some way to
define a link to another repository as some sort of sub module...


As a Rubyist I'd like to run a rake task to build to each defined target
and run the RubySpecs. It wouldn't replace xbuild, just automate it. I
don't understand the pushback to this idea. Why not make a dedicated
repo for IronRuby free of the ancillary projects and geared to someone
like me? And likewise make the IronPython repo friendly to our Pythonic
friends?

The amount of code you need to know depends on the part you contribute
to. If you work on IronRuby libraries you dont need to even look at
IronPython. It is indeed better to look there because similar
functionality might already be implemented there. For example, if you
worked on FFI you might want to check out IronPythons CTypes and perhaps
reuse some code. Why would you need to understand the entire code base
if you needed to do a local change? Just dont look where you dont need
to J. Building IronRuby is also simple  you just run xbuild Ruby.sln
from Solution directory. Subsequent builds are incremental, so you dont
even need to build everything if you dont change core components. I dont
see how this could be any simpler.
Unfortunately I do see how this could be simpler.

_______________________________________________ Ironruby-core mailing
list Ironruby-core@rubyforge.org
http://rubyforge.org/mailman/listinfo/ironruby-core
68f1bd9e9a7d9d10f0278c6535f700fe?d=identicon&s=25 Michael Letterle (mletterle)
on 2010-10-23 18:30
(Received via mailing list)
mletterle@michael2:~/src/ironruby/Languages/Ruby/bin/Debug$ mono
ir64.exe
IronRuby 1.1.1.0 on 2.8.1 (master/e8a3aab Fri Oct 22 22:15:50 EDT 2010)
Copyright (c) Microsoft Corporation. All rights reserved.

>>> print RUBY_PLATFORM
i386-linux>>>
>>> [1,2,3,4].reverse.each{|x| print "Hello! #{x}"}
=> nil
Hello! 4Hello! 3Hello! 2Hello! 1=> [4, 3, 2, 1]>>>

Mono didn't like how some stuff was done in Enumerable.cs, but a
slight refactoring (eliminate gotos, move an anonymous method out of a
method argument) did the trick.  Working on running the tests now.

Changes are here: http://github.com/mletterle/ironruby/tree/linux

On Sat, Oct 23, 2010 at 7:02 AM, Michael Letterle
<michael.letterle@gmail.com> wrote:
> On Sat, Oct 23, 2010 at 3:19 AM, Tomas Matousek
>> To: ironruby-core@rubyforge.org
>> Working on getting IronRuby.Console compiled now... already ran into a casing
issue >.<
>>> contribute to Mono is the project/build system. When looking at it for
>>>> With a brave new world ahead for IronRuby, what do you all think
>>>> charge of both projects, can we move these into separate repos?
>>>> should live in its own repo, IMO. So should the Visual Studio tools.
>>>
>>>
>> http://rubyforge.org/mailman/listinfo/ironruby-core
> Michael Letterle
> IronRuby MVP
> http://blog.prokrams.com
>



--
Michael Letterle
IronRuby MVP
http://blog.prokrams.com
Cb51033949ffccd982ae32c9f890f25a?d=identicon&s=25 Tomas Matousek (Guest)
on 2010-10-23 19:00
(Received via mailing list)
I see. "DEBUG" seems to be redundant since it's set by the default Debug
configuration in the project files.
Strange, I think I tried running IronRuby on Mono 2.8 and most of it
worked. If TryEnter implementation was missing something should have
failed. But this was compiled using our compiler not Mono's. So maybe
the problem is just at build time?

We should work with JB to get any issues fixed.

Tomas
99b1f0c67bec23747d007e27d000487b?d=identicon&s=25 Ryan Riley (Guest)
on 2010-10-23 19:01
(Received via mailing list)
On Sat, Oct 23, 2010 at 5:44 AM, Mike Moore <blowmage@gmail.com> wrote:

> On Oct 23, 2010, at 1:30 AM, Tomas Matousek <<Tomas.Matousek@microsoft.com>
> Tomas.Matousek@microsoft.com> wrote:
>
> I don’t understand how three distinct github repos that I need to map into
> some directories on my disk whose relative location to each other is
> hardcoded in some scripts in each are better than a single repo that has a
> well-defined structure.
>
> You are speaking like someone responsible for both languages and the DLR.
>

And you are speaking like someone who has tried hard several times to
contribute to IronRuby and failed because of a bloated project
structure.
I'm sure there are several people who would be willing to help you
figure
out what's wrong. Where's the repo with your contribution?


> I'm making a suggestion as someone really only interested in IronRuby. The
> repo isn't called "DynamicLanguages", it's called "IronRuby", which is at
> best confusing. If only git had some way to define a link to another
> repository as some sort of sub module...
>

Ah yes, and if only github had something like forking ...

>
> As a Rubyist I'd like to run a rake task to build to each defined target
> and run the RubySpecs. It wouldn't replace xbuild, just automate it. I don't
> understand the pushback to this idea.
>

If you want to create and maintain these, I'm sure no one would
complain. I
don't understand the push back to the idea that the three core
contributors
were a little tired of building IronRuby and maintaining two build
approaches. I also don't understand a Rubyist's failure to see an
opportunity to contribute rake tasks to a project.


> Why not make a dedicated repo for IronRuby free of the ancillary projects
> and geared to someone like me? And likewise make the IronPython repo
> friendly to our Pythonic friends?
>

IronPython already has a separate repo at
http://ironpython.codeplex.com/.

> be any simpler.
>
> Unfortunately I do see how this could be simpler.
>

Great! Fork the project and show everyone!

Sorry for the tone. If you were intending to come across differently, it
would help offer help and not critique the great work Tomas and the rest
have done. Try working on the project. Ask for help. You'll receive it
readily. You may even find some people to help you restructure the
project.
If it works better, I'm sure you'll get support for moving the project
to
that structure. Just don't demand "your way right away" when you haven't
contributed and don't understand the history. It's not your project yet.
Get
involved and make it yours.

Ryan Riley

Email: ryan.riley@panesofglass.org
LinkedIn: http://www.linkedin.com/in/ryanriley
Twitter: @panesofglass
Blog: http://wizardsofsmart.net/
Website: http://panesofglass.org/
740c85b27e1a80c5f908d81dca5d4fbf?d=identicon&s=25 Jb Evain (Guest)
on 2010-10-23 19:27
(Received via mailing list)
Hey,

On Sat, Oct 23, 2010 at 6:01 PM, Michael Letterle
<michael.letterle@gmail.com> wrote:
> Mono didn't like how some stuff was done in Enumerable.cs, but a
> slight refactoring (eliminate gotos, move an anonymous method out of a
> method argument) did the trick. Working on running the tests now.

When stumbling about cases like that, it would be really nice to
extract repros and file bugs for Mono.

Thanks!
68f1bd9e9a7d9d10f0278c6535f700fe?d=identicon&s=25 Michael Letterle (mletterle)
on 2010-10-23 20:17
(Received via mailing list)
Working on it actually... ;)  Maybe, possibly, with a patch as well...
(perhaps).

On Sat, Oct 23, 2010 at 12:34 PM, Jb Evain <jb@nurv.fr> wrote:
>
> Thanks!
>
> --
> Jb Evain <jb@nurv.fr>
> _______________________________________________
> Ironruby-core mailing list
> Ironruby-core@rubyforge.org
> http://rubyforge.org/mailman/listinfo/ironruby-core
>



--
Michael Letterle
IronRuby MVP
http://blog.prokrams.com
68f1bd9e9a7d9d10f0278c6535f700fe?d=identicon&s=25 Michael Letterle (mletterle)
on 2010-10-23 22:33
(Received via mailing list)
I have to respecify DEBUG since I'm redefining the DefineConstants
property.

Running would have worked as long as you didn't try to call that method
:)

On Sat, Oct 23, 2010 at 12:29 PM, Tomas Matousek
<Tomas.Matousek@microsoft.com> wrote:
> To: ironruby-core@rubyforge.org
> On Sat, Oct 23, 2010 at 3:19 AM, Tomas Matousek <Tomas.Matousek@microsoft.com>
wrote:
>> Sent: Friday, October 22, 2010 8:02 PM
>>
>>>
>>> On Fri, Oct 22, 2010 at 5:10 PM, Mike Moore <blowmage@gmail.com> wrote:
>>>> and I honestly don't care about having to check IronPython. Now that
>>>> this, is it possible we can make this project reflect the realities of the
majority of folks that would like to contribute?
>>>>
>>> http://rubyforge.org/mailman/listinfo/ironruby-core
>> Ironruby-core mailing list
>
> Ironruby-core mailing list
> Ironruby-core@rubyforge.org
> http://rubyforge.org/mailman/listinfo/ironruby-core
>



--
Michael Letterle
IronRuby MVP
http://blog.prokrams.com
D91500918a07bfae5b38b8943fd42fbe?d=identicon&s=25 Mike Moore (Guest)
on 2010-10-24 01:31
(Received via mailing list)
I don't want a confrontation, I was just trying to voice some concerns I
have. If everyone is in agreement that the way things have been done is
the
best way they could have possibly been done, and that nothing should be
changed now that the project owners and the rules for making
contributions
are fundamentally different, then by all means continue.

On Sat, Oct 23, 2010 at 10:32 AM, Ryan Riley
<ryan.riley@panesofglass.org>wrote:

>> You are speaking like someone responsible for both languages and the DLR.
>>
>
> And you are speaking like someone who has tried hard several times to
> contribute to IronRuby and failed because of a bloated project structure.
> I'm sure there are several people who would be willing to help you figure
> out what's wrong. Where's the repo with your contribution?
>

I don't understand. Is this some sort of challenge?


> I'm making a suggestion as someone really only interested in IronRuby. The
>> repo isn't called "DynamicLanguages", it's called "IronRuby", which is at
>> best confusing. If only git had some way to define a link to another
>> repository as some sort of sub module...
>>
>
> Ah yes, and if only github had something like forking ...
>

I don't think changing the structure in forked repos would do anyone any
good, as it would make sharing contributions between repos difficult.


>
I think you are confused to where I am puzzled about resistance. It is
not
about having rake tasks. I agree that they are easy enough to add and
maintain, and that whining about them would be quite ridiculous. That's
not
my point, however. My point is that there would be more contributions if
it
were not a single monolithic repository. I also think most of the
historic
benefits of having a monolithic repo can be mitigated with submodules
and an
automated build and integration server.

Feel free to disagree.

Why not make a dedicated repo for IronRuby free of the ancillary
projects
>> and geared to someone like me? And likewise make the IronPython repo
>> friendly to our Pythonic friends?
>>
>
> IronPython already has a separate repo at http://ironpython.codeplex.com/
> .
>

I dunno, it looks really, really similar to the IronRuby repo on GitHub
to
me. Is this synched with the GitHub repo? Is this where all the "Project
Merlin" changes are coming from?
99b1f0c67bec23747d007e27d000487b?d=identicon&s=25 Ryan Riley (Guest)
on 2010-10-24 02:17
(Received via mailing list)
On Sat, Oct 23, 2010 at 4:09 PM, Mike Moore <blowmage@gmail.com> wrote:

>>> You are speaking like someone responsible for both languages and the DLR.
>>>
>>
>> And you are speaking like someone who has tried hard several times to
>> contribute to IronRuby and failed because of a bloated project structure.
>> I'm sure there are several people who would be willing to help you figure
>> out what's wrong. Where's the repo with your contribution?
>>
>
> I don't understand. Is this some sort of challenge?
>

No, that's just how I read your message. You seem to indicate that the
current repo structure causes difficulty when you want to contribute. If
you
need help finding your way around, I'm happy to help, as I'm sure are
many
others. (Of course, you may want the help of others as I'm a bit rusty
myself.) :)


> I don't think changing the structure in forked repos would do anyone any
> good, as it would make sharing contributions between repos difficult.
>

True, but if you were able to create a repo structure you think would
allow
more freedom to contribute, then I think everyone would move to that
structure. Changing the structure is going to take work, and I think
most
people who have been contributing are fine with the current structure
since
they've been used to it for some time. Actually, several layers of the
repo
have already been removed, so it's better than it was six months ago.

I think you are confused to where I am puzzled about resistance. It is
not
> about having rake tasks. I agree that they are easy enough to add and
> maintain, and that whining about them would be quite ridiculous. That's not
> my point, however. My point is that there would be more contributions if it
> were not a single monolithic repository. I also think most of the historic
> benefits of having a monolithic repo can be mitigated with submodules and an
> automated build and integration server.
>
> Feel free to disagree.
>

I don't really disagree with you at all. I use that same strategy myself
on
my own projects. I just don't want to volunteer to restructure the
entire
project, especially if people are actively working on integration
scenarios.
Yes, the project was originally in both TFS and git, so it had to play
nice
with both. Other options are now open. Hence why I suggested forking it
and
showing a better structure.

I dunno, it looks really, really similar to the IronRuby repo on GitHub
to
> me. Is this synched with the GitHub repo? Is this where all the "Project
> Merlin" changes are coming from?
>

I believe the sync is one-way, IronRuby includes IronPython. I'm sure
the
similarity in structure comes from the internal TFS repos that were used
and
the fact that the teams were both working on the DLR together. Someone
please correct me if this is wrong.

Cheers,
Ryan
68f1bd9e9a7d9d10f0278c6535f700fe?d=identicon&s=25 Michael Letterle (mletterle)
on 2010-10-24 05:22
(Received via mailing list)
FWIW having separate IronRuby, IronPython, and Common repos that are
sub moduled(is that a word?) would make sense, that way changes that
are done in Common by both people working on Ruby and Python are
easily shared.. the current configuration feels.. fragile.

Perhaps the BEST thing to do is to layout what a more reasonable
structure could look like in concrete terms rather then just
complaining about the current setup.

The repo in general needs some love though, build instructions should
be in the README, and the current layout could use some explanations,
some files that seem to be needed are incorrect or in places different
than project files expect (specifically thinking of App.config here)

I think the fundamental problem is the current repo was never really
put together with the idea of someone coming in blind to it.. the only
way to change that is to.. well start changing it. :)

On the subject of the rake tasks, when the rake files were originally
there xbuild was nowhere near the state it is now, and it WAS a pain
to maintain, especially across platforms, we probably should have some
linux scripts to call xbuild though.

The most important thing, in my opinion, is that these discussions are
now occurring.

On Sat, Oct 23, 2010 at 7:09 PM, Mike Moore <blowmage@gmail.com> wrote:
>>>
>> And you are speaking like someone who has tried hard several times to
>>> repository as some sort of sub module...
>>
> were not a single monolithic repository. I also think most of the historic
>
> I dunno, it looks really, really similar to the IronRuby repo on GitHub to
> me. Is this synched with the GitHub repo? Is this where all the "Project
> Merlin" changes are coming from?
> _______________________________________________
> Ironruby-core mailing list
> Ironruby-core@rubyforge.org
> http://rubyforge.org/mailman/listinfo/ironruby-core
>
>



--
Michael Letterle
IronRuby MVP
http://blog.prokrams.com
48bf247037f46e594599b6a2c9532b66?d=identicon&s=25 Alistair Bush (Guest)
on 2010-10-24 13:48
(Received via mailing list)
> I have to respecify DEBUG since I'm redefining the DefineConstants
> property.
>
> Running would have worked as long as you didn't try to call that method :)
>

I would love to see your work.   Ive been meaning to package ironruby
for
gentoo linux for ages,  but so far have not got around to it.  The main
reason
for this is that it doesn't build "naked".   Gentoo has a policy of
staying as
close to upstream as possible combined with me being lazy and until a
couple
of days ago was unsure of whether I could contribute patches upstream.
none of
those made me particularly  motivated to create patches :)

If you have made changes I would

1)  like to get them merged into ironruby.  so #if MONO's would be nice
instead of a straight fix. (assuming they are acceptable)
2)  like to get the issue fixed in mono (if its that doc one for which
there is
already a bug,  but also that cast from dynamic to object one)  so
please can
we file bugs upstream too.

On a more general note.   if you are interested in ruby on gentoo I
suggest
you have a look at a fellow dev's blog [1].   Its an interesting read
and you
won't be disappointed :)  Really please read it,  you might learn some
of the
challenges gentooers currently face.

>From a linux packaging perspective.  Ultimately I would hope that we can
acheive nice, easy (for everyone), reproducable builds.  Im perfectly
happy
with msbuild,  but could we add bat/sh files that call msbuild/xbuild to
build
specific parts of the project.  Could we document all the options
available to
configure the build (including defines if necessary, etc).

-Alistair.

[1] http://blog.flameeyes.eu/tag/rubyng
[2] https://bugs.gentoo.org/show_bug.cgi?id=94374

ps.  Does ironruby have a irc channel,  other place where dev's hang
out?
9f7229220b13ec7922f6d826ef16ee2d?d=identicon&s=25 Andrius Bentkus (Guest)
on 2010-10-25 11:11
(Received via mailing list)
The project directory structure is a mess. Dozens of executable files in
the
versioning and what not.
9f7229220b13ec7922f6d826ef16ee2d?d=identicon&s=25 Andrius Bentkus (Guest)
on 2010-10-25 13:55
(Received via mailing list)
On Sun, Oct 24, 2010 at 1:57 AM, Michael Letterle <
michael.letterle@gmail.com> wrote:

> FWIW having separate IronRuby, IronPython, and Common repos that are
> sub moduled(is that a word?) would make sense, that way changes that
> are done in Common by both people working on Ruby and Python are
> easily shared.. the current configuration feels.. fragile.
>

There is a major problem: different vcs tools. I guess the IronPython
project will stay with TFS/SVN while IronRuby will use git(hub). Having
common submodule repos managed by different VCS would be a world of
pain.
There is no way of dividing the project into submodules if IronPython
doesn't move to github/git. Maybe some git-svn magic would help and
mirror
versions on github of the svn repositories would be needed.

I think that another major problem the IronRuby project has are the 4
sites
with IronRuby content. There are like 4 sites now on github, rubyforge,
ironruby.net and the codeplex with different content on ironruby. This
is
madness, IronPython as only 2 sites, ironpython.net and codeplex  and
that
makes sence. When I looked into the project I was just confused, because
I
couldn't find any information and the little bits of Information were
scattered and outdated. And this is a real dilemma, because you just
can't
move away from any of these sites: you have to stay at codeplex at
because
it is an Iron project, you have to stay at rubyforge because it is ruby
afterall, you can't move from github, because all the ruby kids use git,
so
there is only ironruby.net left, but you can't get rid of that either,
it's
after all the ironruby domain.
The purposes of the sites need to be trimmed down: use codeplex and
rubyforge only for binary distribution, github for versioning, issue
tracking and wiki and ironruby.net as a presentation site just like
ironpython.net is, cut the documentation out of it and stuff it in the
github wiki, redudancy is hard to version.
I don't think that keeping the issue tracking system on codeplex really
helps in any way: people who are interested only in IronRuby have to
register now on codeplex and github, people who are interested in bot
iron
projects will have to register on both anyway.

I guess splitting up this this project is a nasty nasty dilemma, because
the
project tries to unite different communities which have different tool
preferences.
D91500918a07bfae5b38b8943fd42fbe?d=identicon&s=25 Mike Moore (Guest)
on 2010-10-25 20:05
(Received via mailing list)
On Mon, Oct 25, 2010 at 4:22 AM, Andrius Bentkus <
Andrius.Bentkus@rwth-aachen.de> wrote:

> There is a major problem: different vcs tools. I guess the IronPython
> project will stay with TFS/SVN while IronRuby will use git(hub). Having
> common submodule repos managed by different VCS would be a world of pain.
> There is no way of dividing the project into submodules if IronPython
> doesn't move to github/git. Maybe some git-svn magic would help and mirror
> versions on github of the svn repositories would be needed.
>

I don't think this would be too difficult to work around. There is
already
some process that replicates changes from the IronPython's CodePlex repo
to
IronRuby's GitHub repo. If the current monolithic project structure were
broken up into submoldules, you could have just IronPython's CodePlex
being
replicated to an IronPython git repo.

I think that another major problem the IronRuby project has are the 4
sites
> it's after all the ironruby domain.
> The purposes of the sites need to be trimmed down: use codeplex and
> rubyforge only for binary distribution, github for versioning, issue
> tracking and wiki and ironruby.net as a presentation site just like
> ironpython.net is, cut the documentation out of it and stuff it in the
> github wiki, redudancy is hard to version.
> I don't think that keeping the issue tracking system on codeplex really
> helps in any way: people who are interested only in IronRuby have to
> register now on codeplex and github, people who are interested in bot iron
> projects will have to register on both anyway.
>

I don't see the number of content sites as a major problem. Rubyforge is
being phased out in favor of better tools, so I don't think its a long
term
solution. (Even gem hosting has moved to rubygems.org instead of
gems.rubyforge.org.) I don't think the ironruby.net site is holding the
project back at all, but I agree it could be better. I think a better
solution would be to replace it with a jekyll site running on GitHub.
Just
point the ironruby.net domain to GitHub and you're done. The reason I
like
the jekyll approach is because it makes it much easier for folks to
create
and improve web content. Its just a pull request away from being
published.
I think that system works really well, and its free.

It doesn't really matter where downloads are hosted, as long as
ironruby.netlinks to them. But they could just as easily be hosted on
GitHub.

The only thing that I'm aware is being used at CodePlex is the
ticketing.
Again, I think there are better ticketing solutions out there, most of
them
free for open source projects. I don't really see a need for CodePlex
myself, but I understand the desire to stay on it for some things. Its
probable that you could also push the jekyll content to CodePlex (and
Rubyforge) if that was desired.


> I guess splitting up this this project is a nasty nasty dilemma, because
> the project tries to unite different communities which have different tool
> preferences.
>

All the more reason to have separate repositories, IMO.
Cb51033949ffccd982ae32c9f890f25a?d=identicon&s=25 Tomas Matousek (Guest)
on 2010-10-25 20:41
(Received via mailing list)
Can you be more specific? What's wrong with the structure (other than
LCA_RESTRICTED directories, which I agree were there only to satisfy our
lawyers but can be now merged into other directories)?

Yes, there are executable files checked in. These are tools that are
needed for running scripts, tests and various code generators.

Tomas

From: ironruby-core-bounces@rubyforge.org
[mailto:ironruby-core-bounces@rubyforge.org] On Behalf Of Andrius
Bentkus
Sent: Monday, October 25, 2010 12:51 AM
To: ironruby-core@rubyforge.org
Subject: Re: [Ironruby-core] Contributing?

The project directory structure is a mess. Dozens of executable files in
the versioning and what not.
22743df4d03210598447e912bd5dcbd4?d=identicon&s=25 Christopher Bennage (Guest)
on 2010-10-25 20:42
(Received via mailing list)
As someone who attempted to dive into IronRuby a couple of months ago, I
found it difficult to discern the "IronRuby Story" from the various
websites.
By that I mean that was not immediately obvious to me what information
was
current and what was outdated. I'd like to see an executive summary of
the
project that is updated at least once a month and perhaps a
more conspicuous hub for FeRb activity. (For example, see the one page
site
we have for Caliburn -> http://www.caliburnproject.com/)

I'm .NET dev, but a ruby beginner. However, we're also a git shop,
OSS enthusiasts, and so on. That's just to say that tooling and culture
were
not a roadblock.

I'd also be interested in contributing to the effort of web presence.

Christopher
68f1bd9e9a7d9d10f0278c6535f700fe?d=identicon&s=25 Michael Letterle (mletterle)
on 2010-10-25 20:43
(Received via mailing list)
My only comment on the sites:

1) github = source control,
2) rubyforge = mailing list,
3) codeplex = issue tracker/binary distro,
4) ironruby.net = documentaiton

1 and 2 are pretty set, I see no reason to change from them.

3 is probably fine as well, though having source control and issue
tracking in one location may be desirable.

4) Using gh-pages is an interesting idea, is ironruby.net being hosted
at Microsoft's expense or one of the team members?  That would
probably have some bearing on that.  Regardless, I'd rather see a nice
README first. :)



On Mon, Oct 25, 2010 at 1:27 PM, Mike Moore <blowmage@gmail.com> wrote:
>
> I don't think this would be too difficult to work around. There is already
> some process that replicates changes from the IronPython's CodePlex repo to
> IronRuby's GitHub repo. If the current monolithic project structure were
> broken up into submoldules, you could have just IronPython's CodePlex being
> replicated to an IronPython git repo.
>>

This I agree with.  The current repo structure is counter-intuitive
IMHO.

--
Michael Letterle
IronRuby MVP
http://blog.prokrams.com
D91500918a07bfae5b38b8943fd42fbe?d=identicon&s=25 Mike Moore (Guest)
on 2010-10-25 21:37
(Received via mailing list)
On Mon, Oct 25, 2010 at 12:26 PM, Michael Letterle <
michael.letterle@gmail.com> wrote:

> Regardless, I'd rather see a nice README first. :)
>

+1
Cb51033949ffccd982ae32c9f890f25a?d=identicon&s=25 Tomas Matousek (Guest)
on 2010-10-25 21:47
(Received via mailing list)
Yes, I agree our web sites need some work. Jimmy, what is that status of
the new design you started working on (the one IronPython.net has
already)?

Tomas

From: ironruby-core-bounces@rubyforge.org
[mailto:ironruby-core-bounces@rubyforge.org] On Behalf Of Christopher
Bennage
Sent: Monday, October 25, 2010 11:20 AM
To: ironruby-core@rubyforge.org
Subject: Re: [Ironruby-core] Contributing?

As someone who attempted to dive into IronRuby a couple of months ago, I
found it difficult to discern the "IronRuby Story" from the various
websites.
By that I mean that was not immediately obvious to me what information
was current and what was outdated. I'd like to see an executive summary
of the project that is updated at least once a month and perhaps a more
conspicuous hub for FeRb activity. (For example, see the one page site
we have for Caliburn -> http://www.caliburnproject.com/)

I'm .NET dev, but a ruby beginner. However, we're also a git shop, OSS
enthusiasts, and so on. That's just to say that tooling and culture were
not a roadblock.

I'd also be interested in contributing to the effort of web presence.

Christopher
Cb51033949ffccd982ae32c9f890f25a?d=identicon&s=25 Tomas Matousek (Guest)
on 2010-10-25 21:56
(Received via mailing list)
Let's assume IronPython moves to github.

There would be two options:

1)      We could just rename the current IronRuby repo to
"DynamicLanguages" repo and IronPython can use it as it is (more or
less).

2)      It might be possible to split the repo to 3 parts - IronRuby
specific, IronPython specific, and DLR, make a submodule for each and
combine those submodules into "DynamicLanguages" repo. So what's exactly
the effective difference among the repo built this way and 1)? AFAICT
it's only that super-module doesn't track the head of the sub-module
automatically. You need to manually update it to the latest version. How
does that help us? If we have a (single) CI server that makes sure that
both IronPython's and IronRuby's heads are passing all tests, what is
the advantage of not using the latest source code of each other?

Or am I missing something (maybe I misunderstand what git submodule can
do)?

PS: All this is orthogonal to minor refactoring of the current directory
structure that is a no-brainer and I already mentioned them (like
removing LCA_RESTRICTED).

Tomas

From: ironruby-core-bounces@rubyforge.org
[mailto:ironruby-core-bounces@rubyforge.org] On Behalf Of Mike Moore
Sent: Monday, October 25, 2010 10:28 AM
To: ironruby-core@rubyforge.org
Subject: Re: [Ironruby-core] Contributing?

On Mon, Oct 25, 2010 at 4:22 AM, Andrius Bentkus
<Andrius.Bentkus@rwth-aachen.de<mailto:Andrius.Bentkus@rwth-aachen.de>>
wrote:

On Sun, Oct 24, 2010 at 1:57 AM, Michael Letterle
<michael.letterle@gmail.com<mailto:michael.letterle@gmail.com>> wrote:
FWIW having separate IronRuby, IronPython, and Common repos that are
sub moduled(is that a word?) would make sense, that way changes that
are done in Common by both people working on Ruby and Python are
easily shared.. the current configuration feels.. fragile.

There is a major problem: different vcs tools. I guess the IronPython
project will stay with TFS/SVN while IronRuby will use git(hub). Having
common submodule repos managed by different VCS would be a world of
pain. There is no way of dividing the project into submodules if
IronPython doesn't move to github/git. Maybe some git-svn magic would
help and mirror versions on github of the svn repositories would be
needed.

I don't think this would be too difficult to work around. There is
already some process that replicates changes from the IronPython's
CodePlex repo to IronRuby's GitHub repo. If the current monolithic
project structure were broken up into submoldules, you could have just
IronPython's CodePlex being replicated to an IronPython git repo.

I think that another major problem the IronRuby project has are the 4
sites with IronRuby content. There are like 4 sites now on github,
rubyforge, ironruby.net<http://ironruby.net> and the codeplex with
different content on ironruby. This is madness, IronPython as only 2
sites, ironpython.net<http://ironpython.net> and codeplex  and that
makes sence. When I looked into the project I was just confused, because
I couldn't find any information and the little bits of Information were
scattered and outdated. And this is a real dilemma, because you just
can't move away from any of these sites: you have to stay at codeplex at
because it is an Iron project, you have to stay at rubyforge because it
is ruby afterall, you can't move from github, because all the ruby kids
use git, so there is only ironruby.net<http://ironruby.net> left, but
you can't get rid of that either, it's after all the ironruby domain.
The purposes of the sites need to be trimmed down: use codeplex and
rubyforge only for binary distribution, github for versioning, issue
tracking and wiki and ironruby.net<http://ironruby.net> as a
presentation site just like ironpython.net<http://ironpython.net> is,
cut the documentation out of it and stuff it in the github wiki,
redudancy is hard to version.
I don't think that keeping the issue tracking system on codeplex really
helps in any way: people who are interested only in IronRuby have to
register now on codeplex and github, people who are interested in bot
iron projects will have to register on both anyway.

I don't see the number of content sites as a major problem. Rubyforge is
being phased out in favor of better tools, so I don't think its a long
term solution. (Even gem hosting has moved to
rubygems.org<http://rubygems.org> instead of
gems.rubyforge.org<http://gems.rubyforge.org>.) I don't think the
ironruby.net<http://ironruby.net> site is holding the project back at
all, but I agree it could be better. I think a better solution would be
to replace it with a jekyll site running on GitHub. Just point the
ironruby.net<http://ironruby.net> domain to GitHub and you're done. The
reason I like the jekyll approach is because it makes it much easier for
folks to create and improve web content. Its just a pull request away
from being published. I think that system works really well, and its
free.

It doesn't really matter where downloads are hosted, as long as
ironruby.net<http://ironruby.net> links to them. But they could just as
easily be hosted on GitHub.

The only thing that I'm aware is being used at CodePlex is the
ticketing. Again, I think there are better ticketing solutions out
there, most of them free for open source projects. I don't really see a
need for CodePlex myself, but I understand the desire to stay on it for
some things. Its probable that you could also push the jekyll content to
CodePlex (and Rubyforge) if that was desired.

I guess splitting up this this project is a nasty nasty dilemma, because
the project tries to unite different communities which have different
tool preferences.

All the more reason to have separate repositories, IMO.
868cfafe762b3a534710098b0caa98e5?d=identicon&s=25 Jb Evain (Guest)
on 2010-10-25 22:52
(Received via mailing list)
Hey,

On Mon, Oct 25, 2010 at 8:19 PM, Tomas Matousek
<Tomas.Matousek@microsoft.com> wrote:
> 2)      It might be possible to split the repo to 3 parts – IronRuby
> specific, IronPython specific, and DLR, make a submodule for each and
> combine those submodules into “DynamicLanguages” repo. So what’s exactly the
> effective difference among the repo built this way and 1)?

The difference is that it then becomes easier for people that are only
interested in either IronPython or IronRuby to track commits. There
will be a timeline for each modules. And you get to follow the ones
you're interested in. And if you're only interested in one, your
timeline isn't «polluted» with comments or commits from the others.

Jb
Cb51033949ffccd982ae32c9f890f25a?d=identicon&s=25 Tomas Matousek (Guest)
on 2010-10-25 23:19
(Received via mailing list)
Makes sense. That seems like a goodness and I guess we should evolve the
repo that way. Let's keep everything working and get there one step at a
time. It doesn't feel like something that would be critical to do right
now though.

Tomas
This topic is locked and can not be replied to.