librelist archives

« back to archive

Ruby in 2012

Ruby in 2012

From:
David Eastman
Date:
2012-05-02 @ 00:48
Nice article about ruby today
http://rubylearning.com/blog/2012/04/23/ruby-in-2012/

At the end of the article, interesting lack of belief in ruby GUIs. I
wonder if shoes may be a little more major here than was first imagined.

Re: [shoes] Ruby in 2012

From:
Cecil Coupe
Date:
2012-05-02 @ 03:24
Shoes, The GUI/DSL certainly could fit in with mruby/mobiruby. Just
another fork off of Green Shoes if we knew enough about mobiruby API's. 
The bytecode compiler of mruby solves many of the packaging issues with
Shoes. 

From a Red shoes developers perspective there is a lot to like *if*
everything comes to pass and the VM API's are documented in English so
Shoes can mingle with them. 

If you believe Shoe's purpose is learning to program, then
mruby/mobiruby announcement won't change anything since few few tablets
and smart phones are tethered to keyboards and functioning text
editors. 

Mostly off topic, I really wish we had a Qt version of Green Shoes.  I
happen to like lost puppies and I installed Haiku (Be OS) on a
partition. There are things to like about BeOS (and many rough edges).
It's Posix enough; someone has implemented python and someone has
implemented Qt. That and mruby just feels like a black hole filled with
shiny things sucking me into it. 

--Cecil


On Wed, 2012-05-02 at 01:48 +0100, David Eastman wrote:
> Nice article about ruby
> today http://rubylearning.com/blog/2012/04/23/ruby-in-2012/ 
> 
> 
> At the end of the article, interesting lack of belief in ruby GUIs. I
> wonder if shoes may be a little more major here than was first
> imagined.

Re: [shoes] Ruby in 2012

From:
Peter Fitzgibbons
Date:
2012-05-02 @ 14:11
My kids tell me "Don't look at the shiny things!".  Usually in reference to
a new espresso maker though.

I feel the fire fanning for movement around the BrownShoes multi-framework
implementation, which would lend itself to the bendings of what you
describe.

Go Shoes!

Peter Fitzgibbons
(847) 859-9550
Email: peter.fitzgibbons@gmail.com
IM GTalk: peter.fitzgibbons
IM AOL: peter.fitzgibbons@gmail.com


On Tue, May 1, 2012 at 10:24 PM, Cecil Coupe <ccoupe@cableone.net> wrote:

> Shoes, The GUI/DSL certainly could fit in with mruby/mobiruby. Just
> another fork off of Green Shoes if we knew enough about mobiruby API's.
> The bytecode compiler of mruby solves many of the packaging issues with
> Shoes.
>
> >From a Red shoes developers perspective there is a lot to like *if*
> everything comes to pass and the VM API's are documented in English so
> Shoes can mingle with them.
>
> If you believe Shoe's purpose is learning to program, then
> mruby/mobiruby announcement won't change anything since few few tablets
> and smart phones are tethered to keyboards and functioning text
> editors.
>
> Mostly off topic, I really wish we had a Qt version of Green Shoes.  I
> happen to like lost puppies and I installed Haiku (Be OS) on a
> partition. There are things to like about BeOS (and many rough edges).
> It's Posix enough; someone has implemented python and someone has
> implemented Qt. That and mruby just feels like a black hole filled with
> shiny things sucking me into it.
>
> --Cecil
>
>
> On Wed, 2012-05-02 at 01:48 +0100, David Eastman wrote:
> > Nice article about ruby
> > today http://rubylearning.com/blog/2012/04/23/ruby-in-2012/
> >
> >
> > At the end of the article, interesting lack of belief in ruby GUIs. I
> > wonder if shoes may be a little more major here than was first
> > imagined.
>
>
>

Re: [shoes] Ruby in 2012

From:
Jenna Fox
Date:
2012-05-02 @ 06:36
Kids are already making games using Lua on iPads. They easily connect to 
usb keyboards with cheap adaptors, and connect to any bluetooth keyboards,
or apple's keyboard docks, but for experienced tablet users (kids 
especially) the touch keyboards are quite usable and fast, and are often 
augmented with dedicated programming keys for characters like '{' ';', 
'do', 'end', 'void', 'function() {'  

Go to any shop, you'll see the number one thing kids want today, is an 
iPad. Recent nelson studies 
(http://www.electronista.com/articles/11/11/17/nielsen.shows.kids.teens.want.ipad.for.christmas/)
suggest as much as 44% of kids wanted an iPad for christmas last year, and
demand has only grown since then. Among teenagers, iPads still took the 
lead ahead of every other sort of gadget. Why would kids want to make apps
for computers? They don't even want to use computers. A shoes for iOS 
would need to include it's own code editor functionality - this is true. 
It's a shame nobody's made anything like that (http://hackety.com/) in 
shoes before.  

—
Jenna


On Wednesday, 2 May 2012 at 1:24 PM, Cecil Coupe wrote:

> Shoes, The GUI/DSL certainly could fit in with mruby/mobiruby. Just
> another fork off of Green Shoes if we knew enough about mobiruby API's.  
> The bytecode compiler of mruby solves many of the packaging issues with
> Shoes.  
>  
> > From a Red shoes developers perspective there is a lot to like *if*
> everything comes to pass and the VM API's are documented in English so
> Shoes can mingle with them.  
>  
> If you believe Shoe's purpose is learning to program, then
> mruby/mobiruby announcement won't change anything since few few tablets
> and smart phones are tethered to keyboards and functioning text
> editors.  
>  
> Mostly off topic, I really wish we had a Qt version of Green Shoes. I
> happen to like lost puppies and I installed Haiku (Be OS) on a
> partition. There are things to like about BeOS (and many rough edges).
> It's Posix enough; someone has implemented python and someone has
> implemented Qt. That and mruby just feels like a black hole filled with
> shiny things sucking me into it.  
>  
> --Cecil
>  
>  
> On Wed, 2012-05-02 at 01:48 +0100, David Eastman wrote:
> > Nice article about ruby
> > today http://rubylearning.com/blog/2012/04/23/ruby-in-2012/  
> >  
> >  
> > At the end of the article, interesting lack of belief in ruby GUIs. I
> > wonder if shoes may be a little more major here than was first
> > imagined.
> >  
>  
>  
>  

Re: [shoes] Ruby in 2012

From:
Cecil Coupe
Date:
2012-05-02 @ 07:41
Jenna,
  You should speak out more often.

On Wed, 2012-05-02 at 16:36 +1000, Jenna Fox wrote:
> Kids are already making games using Lua on iPads. They easily connect
> to usb keyboards with cheap adaptors, and connect to any bluetooth
> keyboards, or apple's keyboard docks, but for experienced tablet users
> (kids especially) the touch keyboards are quite usable and fast, and
> are often augmented with dedicated programming keys for characters
> like '{' ';', 'do', 'end', 'void', 'function() {' 

Yikes, I am corrected and that is a good thing. 

> 
> Go to any shop, you'll see the number one thing kids want today, is an
> iPad. Recent nelson studies suggest as much as 44% of kids wanted an
> iPad for christmas last year, and demand has only grown since then.
> Among teenagers, iPads still took the lead ahead of every other sort
> of gadget. Why would kids want to make apps for computers? They don't
> even want to use computers. A shoes for iOS would need to include it's
> own code editor functionality - this is true. It's a shame nobody's
> made anything like that in shoes before.

That's my worry about mruby/mobiruby. Way too late to the game. By the
time they are ready for a Shoes fitting, the game was over and the
winners smirk at us. Lua vs python vs Ruby for the cool kids mind share
prize?  I wouldn't pick Shoes & Ruby to win.


Re: [shoes] Ruby in 2012

From:
Jenna Fox
Date:
2012-05-02 @ 08:07
Yeah, I wouldn't guess ruby to be number one either, but you know what? At
the end of the day, python is still python, lua is still lua, and I still 
like ruby most of all. iOS is a neat environment because it's very easy to
get an app and play around with it. There's not much friction in it. If 
you don't like the app, you delete it and it goes away, as well as all of 
it's files. No consequences! Maybe some python thing will be big and take 
off, but that doesn't mean there's no place for us. Ruby is a really nice 
language and it has some very playful interesting libraries and people! I 
think what is important is that we are a place for the sort of people who 
like ruby, people like us. I don't do much computer stuff these days, but 
I'd do a lot less if there was no ruby community. :)  

MacRuby, ARC, and Objective C's slow and steady rubification, make me 
think Ruby will have a place in Apple's walled gardens soon. MacRuby runs 
already on iOS except for garbage collection. ARC removes the need for 
garbage collection with Objective C via static analysis. Automatic 
Reference Counting is a language specific feature currently, but it shows 
Apple's developers are playing around with making languages like ruby work
well on iOS, while they continue to push MacRuby's performance to the 
limits.

Google is definitely python-town. HP loves javascript, microsoft has C#. 
Apple I think will be a place of ruby.

So I don't think we are late to the game. If anything, we're a bit too 
early! Regardless, while there are lua based game makers on the App Store,
they have not yet been massively popular (they're not very good). Apple 
hasn't brought out an Xcode for iPad yet, so as a development platform, 
it's all a bit too soon to make any conclusions like that. The race has 
yet to begin.

iOS is an OpenGL/OpenAL place, so keeping those technologies in mind could
make things easier down the track, if for instance we wanted to port a 
C-based shoes to MacRuby, and could just convert the code across while 
doing the same OpenGL calls and things like that.  

—
Jenna


On Wednesday, 2 May 2012 at 5:41 PM, Cecil Coupe wrote:

> Jenna,
> You should speak out more often.
>  
> On Wed, 2012-05-02 at 16:36 +1000, Jenna Fox wrote:
> > Kids are already making games using Lua on iPads. They easily connect
> > to usb keyboards with cheap adaptors, and connect to any bluetooth
> > keyboards, or apple's keyboard docks, but for experienced tablet users
> > (kids especially) the touch keyboards are quite usable and fast, and
> > are often augmented with dedicated programming keys for characters
> > like '{' ';', 'do', 'end', 'void', 'function() {'  
> >  
>  
>  
> Yikes, I am corrected and that is a good thing.  
>  
> >  
> > Go to any shop, you'll see the number one thing kids want today, is an
> > iPad. Recent nelson studies suggest as much as 44% of kids wanted an
> > iPad for christmas last year, and demand has only grown since then.
> > Among teenagers, iPads still took the lead ahead of every other sort
> > of gadget. Why would kids want to make apps for computers? They don't
> > even want to use computers. A shoes for iOS would need to include it's
> > own code editor functionality - this is true. It's a shame nobody's
> > made anything like that in shoes before.
> >  
>  
>  
> That's my worry about mruby/mobiruby. Way too late to the game. By the
> time they are ready for a Shoes fitting, the game was over and the
> winners smirk at us. Lua vs python vs Ruby for the cool kids mind share
> prize? I wouldn't pick Shoes & Ruby to win.
>  
>  

Re: [shoes] Ruby in 2012

From:
Steve Klabnik
Date:
2012-05-02 @ 05:55
> The bytecode compiler of mruby solves many of the packaging issues with
> Shoes.

Basically every Ruby, including MRI, has bytecode.


> Mostly off topic, I really wish we had a Qt version of Green Shoes.  I
> happen to like lost puppies and I installed Haiku (Be OS) on a
> partition. There are things to like about BeOS (and many rough edges).
> It's Posix enough; someone has implemented python and someone has
> implemented Qt. That and mruby just feels like a black hole filled with
> shiny things sucking me into it.

I had the teeny tiny bit done in blue, long long ago...

Re: [shoes] Ruby in 2012

From:
Cecil Coupe
Date:
2012-05-02 @ 06:33
On Wed, 2012-05-02 at 01:55 -0400, Steve Klabnik wrote:
> > The bytecode compiler of mruby solves many of the packaging issues with
> > Shoes.
> 
> Basically every Ruby, including MRI, has bytecode.

Which version of Ruby loads byte code from a file produced by a Ruby far
away in time and space ? mruby is (potentially) like the Java JVM. More
Python, more Java like. Kind of like JRuby w/o the extra Java to Ruby
translation performance hit and the huge jars. 

It's worth watching. I have my doubts it will succeed but I wouldn't
ignore it. 

Re: [shoes] Ruby in 2012

From:
Charles Oliver Nutter
Date:
2012-05-02 @ 17:15
What performance hit are you talking about? JRuby is probably the fastest
Ruby right now, bandit is precisely because we can translate to JVM
bytecode that we are able to run so fast.

And jar-wise, if you assume JRuby is already on device (same assumption
you'd make with mruby) you can ship a compressed archive of .rb code or
preconpiled bytecode that's tiny...a few kB for simple apps.

Your facts about JRuby seem a bit off.
On May 2, 2012 1:34 AM, "Cecil Coupe" <ccoupe@cableone.net> wrote:

> On Wed, 2012-05-02 at 01:55 -0400, Steve Klabnik wrote:
> > > The bytecode compiler of mruby solves many of the packaging issues with
> > > Shoes.
> >
> > Basically every Ruby, including MRI, has bytecode.
>
> Which version of Ruby loads byte code from a file produced by a Ruby far
> away in time and space ? mruby is (potentially) like the Java JVM. More
> Python, more Java like. Kind of like JRuby w/o the extra Java to Ruby
> translation performance hit and the huge jars.
>
> It's worth watching. I have my doubts it will succeed but I wouldn't
> ignore it.
>
>
>

Re: [shoes] Ruby in 2012

From:
Eric Watson
Date:
2012-05-03 @ 14:45
> And jar-wise, if you assume JRuby is already on device (same assumption 
you'd make with mruby) you can ship a compressed archive of .rb code or 
preconpiled bytecode that's tiny...a few kB for simple apps.

This would be great! If we didn't assume JRuby or Java is already on 
device (i.e. we bundle Java JRE, JRuby, and Shoes), what would you 
estimate for the download size?

Eric

Re: [shoes] Ruby in 2012

From:
Jenna Fox
Date:
2012-05-03 @ 14:51
Is it possible to run a jruby application in some sort of bundle, without 
installing Java in to the system? I wouldn't personally run a java app on 
my portable computer, as I sometimes spent time around security 
professionals who are sometimes inclined to pull a prank here and there, 
and java frightens me. Even if I disable java plugins in my web browser, 
other apps like email and feed readers still get it.


—
Jenna


On Friday, 4 May 2012 at 12:45 AM, Eric Watson wrote:

> > And jar-wise, if you assume JRuby is already on device (same 
assumption you'd make with mruby) you can ship a compressed archive of .rb
code or preconpiled bytecode that's tiny...a few kB for simple apps.
>  
>  
> This would be great! If we didn't assume JRuby or Java is already on 
device (i.e. we bundle Java JRE, JRuby, and Shoes), what would you 
estimate for the download size?
>  
> Eric  

Re: [shoes] Ruby in 2012

From:
Peter Fitzgibbons
Date:
2012-05-03 @ 16:13
Java apps (not the JRE itself) should be considered with the same control
as native apps.  If you'd be wary of a specific native app, you would be
equally wary of an equivalent Java app.  OTOH, If you trust a certain
native app, you should trust the equivalent java app (from same developer
:).

JRE is not evil.  Developers use the JRE to do evil.

I think we need to limit the scope of this discussion to desktop and
desktop/workstation-laptops.
IOS/Android is out of scope, and ChromeBook is out of scope (for now
anyway)... so "desktop only".

Thoughts?

Peter Fitzgibbons
(847) 859-9550
Email: peter.fitzgibbons@gmail.com
IM GTalk: peter.fitzgibbons
IM AOL: peter.fitzgibbons@gmail.com


On Thu, May 3, 2012 at 9:51 AM, Jenna Fox <a@creativepony.com> wrote:

>  Is it possible to run a jruby application in some sort of bundle, without
> installing Java in to the system? I wouldn't personally run a java app on
> my portable computer, as I sometimes spent time around security
> professionals who are sometimes inclined to pull a prank here and there,
> and java frightens me. Even if I disable java plugins in my web browser,
> other apps like email and feed readers still get it.
>
>
> —
> Jenna
>
> On Friday, 4 May 2012 at 12:45 AM, Eric Watson wrote:
>
> And jar-wise, if you assume JRuby is already on device (same assumption
> you'd make with mruby) you can ship a compressed archive of .rb code or
> preconpiled bytecode that's tiny...a few kB for simple apps.
>
>
> This would be great! If we didn't assume JRuby or Java is already on
> device (i.e. we bundle Java JRE, JRuby, and Shoes), what would you estimate
> for the download size?
>
> Eric
>
>
>

Re: [shoes] Ruby in 2012

From:
Jenna Fox
Date:
2012-05-03 @ 22:48
Sure, yes, but what I was asking is can java be bundled in to the shoes 
app so it doesn't infect the rest of the system? Java is a very complex 
piece of software with many moving parts, and like flash it introduces a 
large attack surface, and like flash it's pretty keen to expose that 
surface to the web, among other things. In other words, can the java go 
away entirely when the app isn't running, and can it be deleted from the 
computer entirely when the .app or .exe or whatever is deleted?  

—
Jenna


On Friday, 4 May 2012 at 2:13 AM, Peter Fitzgibbons wrote:

> Java apps (not the JRE itself) should be considered with the same 
control as native apps.  If you'd be wary of a specific native app, you 
would be equally wary of an equivalent Java app.  OTOH, If you trust a 
certain native app, you should trust the equivalent java app (from same 
developer :).
>  
> JRE is not evil.  Developers use the JRE to do evil.
>  
> I think we need to limit the scope of this discussion to desktop and 
desktop/workstation-laptops.
> IOS/Android is out of scope, and ChromeBook is out of scope (for now 
anyway)... so "desktop only".
>  
> Thoughts?
>  
> Peter Fitzgibbons
> (847) 859-9550
> Email: peter.fitzgibbons@gmail.com (mailto:peter.fitzgibbons@gmail.com)
> IM GTalk: peter.fitzgibbons
> IM AOL: peter.fitzgibbons@gmail.com (mailto:peter.fitzgibbons@gmail.com)
>  
>  
> On Thu, May 3, 2012 at 9:51 AM, Jenna Fox <a@creativepony.com 
(mailto:a@creativepony.com)> wrote:
> > Is it possible to run a jruby application in some sort of bundle, 
without installing Java in to the system? I wouldn't personally run a java
app on my portable computer, as I sometimes spent time around security 
professionals who are sometimes inclined to pull a prank here and there, 
and java frightens me. Even if I disable java plugins in my web browser, 
other apps like email and feed readers still get it.  
> >  
> >  
> > —
> > Jenna
> >  
> >  
> > On Friday, 4 May 2012 at 12:45 AM, Eric Watson wrote:
> >  
> > > > And jar-wise, if you assume JRuby is already on device (same 
assumption you'd make with mruby) you can ship a compressed archive of .rb
code or preconpiled bytecode that's tiny...a few kB for simple apps.
> > > >  
> > >  
> > >  
> > > This would be great! If we didn't assume JRuby or Java is already on
device (i.e. we bundle Java JRE, JRuby, and Shoes), what would you 
estimate for the download size?
> > >  
> > > Eric  
> >  
>  

Re: [shoes] Ruby in 2012

From:
Peter Fitzgibbons
Date:
2012-05-04 @ 10:52
I have another thought in response.

Have you considered the parallel of JRE and .NET ?  Both are architectural
libraries.  There are no running parts of either of these systems until a
dependent application (jar or exe) is "run".

As far as attack surfaces, since a running jar or exe are
full-security/fully-authenticated desktop executables, they have the same
attack surface as the user his/herself.  Sadly malware is in our culture
and it must be carefully avoided, though hopefully not to the limitation of
usability.

Kindest Regards,
Peter Fitzgibbons
(847) 859-9550
Email: peter.fitzgibbons@gmail.com
IM GTalk: peter.fitzgibbons
IM AOL: peter.fitzgibbons@gmail.com


On Thu, May 3, 2012 at 5:48 PM, Jenna Fox <a@creativepony.com> wrote:

>  Sure, yes, but what I was asking is can java be bundled in to the shoes
> app so it doesn't infect the rest of the system? Java is a very complex
> piece of software with many moving parts, and like flash it introduces a
> large attack surface, and like flash it's pretty keen to expose that
> surface to the web, among other things. In other words, can the java go
> away entirely when the app isn't running, and can it be deleted from the
> computer entirely when the .app or .exe or whatever is deleted?
>
> —
> Jenna
>
> On Friday, 4 May 2012 at 2:13 AM, Peter Fitzgibbons wrote:
>
> Java apps (not the JRE itself) should be considered with the same control
> as native apps.  If you'd be wary of a specific native app, you would be
> equally wary of an equivalent Java app.  OTOH, If you trust a certain
> native app, you should trust the equivalent java app (from same developer
> :).
>
> JRE is not evil.  Developers use the JRE to do evil.
>
> I think we need to limit the scope of this discussion to desktop and
> desktop/workstation-laptops.
> IOS/Android is out of scope, and ChromeBook is out of scope (for now
> anyway)... so "desktop only".
>
> Thoughts?
>
> Peter Fitzgibbons
> (847) 859-9550
> Email: peter.fitzgibbons@gmail.com
> IM GTalk: peter.fitzgibbons
> IM AOL: peter.fitzgibbons@gmail.com
>
>
> On Thu, May 3, 2012 at 9:51 AM, Jenna Fox <a@creativepony.com> wrote:
>
>  Is it possible to run a jruby application in some sort of bundle, without
> installing Java in to the system? I wouldn't personally run a java app on
> my portable computer, as I sometimes spent time around security
> professionals who are sometimes inclined to pull a prank here and there,
> and java frightens me. Even if I disable java plugins in my web browser,
> other apps like email and feed readers still get it.
>
>
> —
> Jenna
>
> On Friday, 4 May 2012 at 12:45 AM, Eric Watson wrote:
>
> And jar-wise, if you assume JRuby is already on device (same assumption
> you'd make with mruby) you can ship a compressed archive of .rb code or
> preconpiled bytecode that's tiny...a few kB for simple apps.
>
>
> This would be great! If we didn't assume JRuby or Java is already on
> device (i.e. we bundle Java JRE, JRuby, and Shoes), what would you estimate
> for the download size?
>
> Eric
>
>
>
>
>

Re: [shoes] Ruby in 2012

From:
Tobias Pfeiffer
Date:
2012-05-04 @ 09:51
I don't know if this has been mentioned here before but in general 
running programs in a runtime environment (like the JRE) is much safer 
than normal executables. It's pretty easy to write a C/C++ application 
with significant security flaws,  like a buffer overflow, that more or 
less easily grant admin/root privileges to an attacker. Runtime 
Environments usually have built in protection against that.

On 04.05.2012 14:52, Peter Fitzgibbons wrote:
> I have another thought in response.
>
> Have you considered the parallel of JRE and .NET ?  Both are 
> architectural libraries.  There are no running parts of either of 
> these systems until a dependent application (jar or exe) is "run".
>
> As far as attack surfaces, since a running jar or exe are 
> full-security/fully-authenticated desktop executables, they have the 
> same attack surface as the user his/herself.  Sadly malware is in our 
> culture and it must be carefully avoided, though hopefully not to the 
> limitation of usability.
>
> Kindest Regards,
> Peter Fitzgibbons
> (847) 859-9550
> Email: peter.fitzgibbons@gmail.com <mailto:peter.fitzgibbons@gmail.com>
> IM GTalk: peter.fitzgibbons
> IM AOL: peter.fitzgibbons@gmail.com <mailto:peter.fitzgibbons@gmail.com>
>
>
> On Thu, May 3, 2012 at 5:48 PM, Jenna Fox <a@creativepony.com 
> <mailto:a@creativepony.com>> wrote:
>
>     Sure, yes, but what I was asking is can java be bundled in to the
>     shoes app so it doesn't infect the rest of the system? Java is a
>     very complex piece of software with many moving parts, and like
>     flash it introduces a large attack surface, and like flash it's
>     pretty keen to expose that surface to the web, among other things.
>     In other words, can the java go away entirely when the app isn't
>     running, and can it be deleted from the computer entirely when the
>     .app or .exe or whatever is deleted?
>
>     —
>     Jenna
>
>     On Friday, 4 May 2012 at 2:13 AM, Peter Fitzgibbons wrote:
>
>>     Java apps (not the JRE itself) should be considered with the same
>>     control as native apps.  If you'd be wary of a specific native
>>     app, you would be equally wary of an equivalent Java app.  OTOH,
>>     If you trust a certain native app, you should trust the
>>     equivalent java app (from same developer :).
>>
>>     JRE is not evil.  Developers use the JRE to do evil.
>>
>>     I think we need to limit the scope of this discussion to desktop
>>     and desktop/workstation-laptops.
>>     IOS/Android is out of scope, and ChromeBook is out of scope (for
>>     now anyway)... so "desktop only".
>>
>>     Thoughts?
>>
>>     Peter Fitzgibbons
>>     (847) 859-9550 <tel:%28847%29%20859-9550>
>>     Email: peter.fitzgibbons@gmail.com
>>     <mailto:peter.fitzgibbons@gmail.com>
>>     IM GTalk: peter.fitzgibbons
>>     IM AOL: peter.fitzgibbons@gmail.com
>>     <mailto:peter.fitzgibbons@gmail.com>
>>
>>
>>     On Thu, May 3, 2012 at 9:51 AM, Jenna Fox <a@creativepony.com
>>     <mailto:a@creativepony.com>> wrote:
>>>     Is it possible to run a jruby application in some sort of
>>>     bundle, without installing Java in to the system? I wouldn't
>>>     personally run a java app on my portable computer, as I
>>>     sometimes spent time around security professionals who are
>>>     sometimes inclined to pull a prank here and there, and java
>>>     frightens me. Even if I disable java plugins in my web browser,
>>>     other apps like email and feed readers still get it.
>>>
>>>
>>>     —
>>>     Jenna
>>>
>>>     On Friday, 4 May 2012 at 12:45 AM, Eric Watson wrote:
>>>
>>>>>     And jar-wise, if you assume JRuby is already on device (same
>>>>>     assumption you'd make with mruby) you can ship a compressed
>>>>>     archive of .rb code or preconpiled bytecode that's tiny...a
>>>>>     few kB for simple apps.
>>>>
>>>>     This would be great! If we didn't assume JRuby or Java is
>>>>     already on device (i.e. we bundle Java JRE, JRuby, and Shoes),
>>>>     what would you estimate for the download size?
>>>>
>>>>     Eric
>>>
>>
>
>

Re: [shoes] Ruby in 2012

From:
Jenna Fox
Date:
2012-05-04 @ 22:39
JRE exposes it's own code (presumably written in C or C++) to the web via 
plugins, and registers itself to automatically run when various sorts of 
files are opened. That sort of thinking only works if you go all the way 
up the chain. That is why computer scientists are Kernels and Operating 
Systems which can be mathematically proven to be bug free and 
vulnerability free.  

—
Jenna


On Friday, 4 May 2012 at 7:51 PM, Tobias Pfeiffer wrote:

> I don't know if this has been mentioned here before but in general 
running programs in a runtime environment (like the JRE) is much safer 
than normal executables. It's pretty easy to write a C/C++ application 
with significant security flaws,  like a buffer overflow, that more or 
less easily grant admin/root privileges to an attacker. Runtime 
Environments usually have built in protection against that.
>  
> On 04.05.2012 14:52, Peter Fitzgibbons wrote:  
> > I have another thought in response.  
> >  
> > Have you considered the parallel of JRE and .NET ?  Both are 
architectural libraries.  There are no running parts of either of these 
systems until a dependent application (jar or exe) is "run".  
> >  
> > As far as attack surfaces, since a running jar or exe are 
full-security/fully-authenticated desktop executables, they have the same 
attack surface as the user his/herself.  Sadly malware is in our culture 
and it must be carefully avoided, though hopefully not to the limitation 
of usability.  
> >  
> > Kindest Regards,
> > Peter Fitzgibbons
> > (847) 859-9550
> > Email: peter.fitzgibbons@gmail.com (mailto:peter.fitzgibbons@gmail.com)
> > IM GTalk: peter.fitzgibbons
> > IM AOL: peter.fitzgibbons@gmail.com (mailto:peter.fitzgibbons@gmail.com)
> >  
> >  
> > On Thu, May 3, 2012 at 5:48 PM, Jenna Fox <a@creativepony.com 
(mailto:a@creativepony.com)> wrote:
> > > Sure, yes, but what I was asking is can java be bundled in to the 
shoes app so it doesn't infect the rest of the system? Java is a very 
complex piece of software with many moving parts, and like flash it 
introduces a large attack surface, and like flash it's pretty keen to 
expose that surface to the web, among other things. In other words, can 
the java go away entirely when the app isn't running, and can it be 
deleted from the computer entirely when the .app or .exe or whatever is 
deleted?  
> > >  
> > > —  
> > > Jenna
> > >  
> > >  
> > > On Friday, 4 May 2012 at 2:13 AM, Peter Fitzgibbons wrote:
> > >  
> > > > Java apps (not the JRE itself) should be considered with the same 
control as native apps.  If you'd be wary of a specific native app, you 
would be equally wary of an equivalent Java app.  OTOH, If you trust a 
certain native app, you should trust the equivalent java app (from same 
developer :).  
> > > >  
> > > > JRE is not evil.  Developers use the JRE to do evil.  
> > > >  
> > > > I think we need to limit the scope of this discussion to desktop 
and desktop/workstation-laptops.  
> > > > IOS/Android is out of scope, and ChromeBook is out of scope (for 
now anyway)... so "desktop only".
> > > >  
> > > > Thoughts?  
> > > >  
> > > > Peter Fitzgibbons
> > > > (847) 859-9550 (tel:%28847%29%20859-9550)
> > > > Email: peter.fitzgibbons@gmail.com (mailto:peter.fitzgibbons@gmail.com)
> > > > IM GTalk: peter.fitzgibbons
> > > > IM AOL: peter.fitzgibbons@gmail.com (mailto:peter.fitzgibbons@gmail.com)
> > > >  
> > > >  
> > > > On Thu, May 3, 2012 at 9:51 AM, Jenna Fox <a@creativepony.com 
(mailto:a@creativepony.com)> wrote:
> > > > > Is it possible to run a jruby application in some sort of 
bundle, without installing Java in to the system? I wouldn't personally 
run a java app on my portable computer, as I sometimes spent time around 
security professionals who are sometimes inclined to pull a prank here and
there, and java frightens me. Even if I disable java plugins in my web 
browser, other apps like email and feed readers still get it.  
> > > > >  
> > > > >  
> > > > > —  
> > > > > Jenna
> > > > >  
> > > > >  
> > > > > On Friday, 4 May 2012 at 12:45 AM, Eric Watson wrote:
> > > > >  
> > > > > > > And jar-wise, if you assume JRuby is already on device (same
assumption you'd make with mruby) you can ship a compressed archive of .rb
code or preconpiled bytecode that's tiny...a few kB for simple apps.
> > > > > > >  
> > > > > > >  
> > > > > >  
> > > > > >  
> > > > > > This would be great! If we didn't assume JRuby or Java is 
already on device (i.e. we bundle Java JRE, JRuby, and Shoes), what would 
you estimate for the download size?  
> > > > > >  
> > > > > > Eric  
> > > > >  
> > > >  
> > >  
> >  
>  

Re: [shoes] Ruby in 2012

From:
Steve Klabnik
Date:
2012-05-05 @ 01:16
> That is why computer scientists are Kernels and Operating Systems
> which can be mathematically proven to be bug free and vulnerability free.

uhhhhhhhh

Re: [shoes] Ruby in 2012

From:
Cecil Coupe
Date:
2012-05-05 @ 03:07
On Fri, 2012-05-04 at 18:16 -0700, Steve Klabnik wrote:
> > That is why computer scientists are Kernels and Operating Systems
> > which can be mathematically proven to be bug free and vulnerability free.
> 
> uhhhhhhhh

Yup. Just yesterday, Ubuntu wanted me install security upgrades to the
kernel. Even though the kernel was perfect per the definition above, I
decided to install the patches to the perfect OS.

Re: [shoes] Ruby in 2012

From:
Dave Wilkinson
Date:
2012-05-05 @ 04:28
> > That is why computer scientists [create? study? research?] Kernels and
Operating Systems
> > which can be mathematically proven to be bug free and vulnerability
free.
>
> uhhhhhhhh

Anything computational can be proven to be bug free and vulnerability free
by using mathematical models. If you are worried about upper/lower layers
being incorrect, you are generally out of luck. Formally verified kernels
do not verify the hardware. They only verify the computation the kernel
performs. Anything else is out of scope. You may do the same, and people do
especially for crypto libraries, in the application layer. It has nothing
to do with virtual machines or hypervisors (sandboxing.) Those are not
inherently safer, but the argument can be made that they *could* reduce the
trusted computing base. However, in practical settings they trust the same
things a normal application does: the kernel, OS, and hardware. They merely
access those things through an abstraction. So, you protect against...
invalid memory accesses that are not caught by the underlying kernel or
hardware. Sandboxes are generally good for graceful handling of application
errors and debugging. [see seL4 and pretty much anything else NICTA does]

And. Well. You know, us 'computer scientist' 'operating system' peoples
care about applications, too. That's what it is all about, in the end. :) <3

I'm not sure what the original topic was. The JRE was discussed. You can
just web install that. I don't see why you cannot. You can also distribute
it with a JRE package and then install that package as part of the setup
process. As long as the user sees the same EULA, Oracle cannot complain. In
fact, they encourage it in their README as long as the JRE is not beta or
pre-release:

http://www.oracle.com/technetwork/java/javase/jrereadme-182762.html

At the root, this is probably the greatest argument for different fittings
of shoes since some runtime environments are just more distributable on
certain platforms. It is great that we can have a native JVM version of
shoes for devices that work really well with a native JVM application. Just
offer up a better tool to determine which is best to download along with a
shoes API that is _the same_ and well specified for each color. That is:
one documentation, one interface, one specification, many implementations.
Deviation means redundant effort.

Re: [shoes] Ruby in 2012

From:
Cecil Coupe
Date:
2012-05-05 @ 05:25
On Sat, 2012-05-05 at 00:28 -0400, Dave Wilkinson wrote:
> You can also distribute it with a JRE package and then install that
> package as part of the setup process. As long as the user sees the
> same EULA, Oracle cannot complain. In fact, they encourage it in their
> README as long as the JRE is not beta or pre-release:

It's the system wide install of JRE with su/root/admin that offends some
folks. I have the open-jdk in my /usr/bin/java
> 
> http://www.oracle.com/technetwork/java/javase/jrereadme-182762.html

Thanks, Dave. It's well worth reading the T&C's. It's also a list of
tasks to be done if Shoes was to include Oracles JRE. 
> 
> At the root, this is probably the greatest argument for different
> fittings of shoes since some runtime environments are just more
> distributable on certain platforms. It is great that we can have a
> native JVM version of shoes for devices that work really well with a
> native JVM application. Just offer up a better tool to determine which
> is best to download along with a shoes API that is _the same_ and well
> specified for each color. That is: one documentation, one interface,
> one specification, many implementations. Deviation means redundant
> effort. 

I agree on the concept. Sadly, we are no where close to that ideal and
the definition of Shoes varies wildly.

Re: [shoes] Ruby in 2012

From:
Cecil Coupe
Date:
2012-05-04 @ 02:44
Java apps are distributed in jars (.jar) which is closely related
to .tar and .zip.  The user tells the pre-installed JRE to run one of
the files/classes in the jar. You have to trust the .jar and the folks
who wrote it.  The JRE and Java design does have some additional
security that aren't found in Ruby or Python (ask anyone who has written
a custom class loader in java).

Red Shoes is a different matter. It doesn't install Ruby in the
*system* , it's a private Ruby installed in User directories (for OSX
and Linux). if you run that private ruby as root it can do anything a
system installed ruby can do. I've never tried, but I'll bet a Red Shoes
app can wreck a Windows system if it wants to. 

Theoretically one could package Shoes, jruby and Java/JRE and run it in
user space but it's a monster task for each platform and a huge PITA for
the end user to wait for the huge installation file to unpack. Not to
mention the licensing issues of distributing a private copy of Java. Ask
Oracle and Google lawyers if they can think of anything that might go
wrong. 

Purple Shoes assumes the user has already installed Java JRE in their
system. Just as Green Shoes assumes Ruby MRI has been installed in their
system. Red Shoes scripts can only use pure ruby gems. Shoes scripts in
Purple/Green/Brown can use the system gem libraries. Thats a good option
for complex Desktop Shoes apps. 

As for mobile apps and Ruby it's already possible if you play by Apple
and Google rules.  I'd had forgotten about http://ruboto.org/ for
Android.  For iOS:

http://arstechnica.com/business/news/2012/05/exclusive-building-ruby-ios-applications-with-rubymotion.ars

Neither one is Shoes easy or casual developer friendly but Ruboto isn't
that different from the Shoes DSL. A smart Android developer could put a
Shoes face on if google allows. We'll just have to wait to see if
mruby/mobiruby gets any traction. As long as mobile devices are
constrained to their walled gardens, cross platform GUI is unlikely. 

--Cecil




On Fri, 2012-05-04 at 08:48 +1000, Jenna Fox wrote:
> Sure, yes, but what I was asking is can java be bundled in to the
> shoes app so it doesn't infect the rest of the system? Java is a very
> complex piece of software with many moving parts, and like flash it
> introduces a large attack surface, and like flash it's pretty keen to
> expose that surface to the web, among other things. In other words,
> can the java go away entirely when the app isn't running, and can it
> be deleted from the computer entirely when the .app or .exe or
> whatever is deleted? 
> 
> 
> —
> Jenna
> 
> 
> 
> On Friday, 4 May 2012 at 2:13 AM, Peter Fitzgibbons wrote:
> 
> > Java apps (not the JRE itself) should be considered with the same
> > control as native apps.  If you'd be wary of a specific native app,
> > you would be equally wary of an equivalent Java app.  OTOH, If you
> > trust a certain native app, you should trust the equivalent java app
> > (from same developer :).
> > 
> > 
> > JRE is not evil.  Developers use the JRE to do evil.
> > 
> > 
> > I think we need to limit the scope of this discussion to desktop and
> > desktop/workstation-laptops.
> > IOS/Android is out of scope, and ChromeBook is out of scope (for now
> > anyway)... so "desktop only".
> > 
> > 
> > Thoughts?
> > 
> > Peter Fitzgibbons
> > (847) 859-9550
> > Email: peter.fitzgibbons@gmail.com
> > IM GTalk: peter.fitzgibbons
> > IM AOL: peter.fitzgibbons@gmail.com
> > 
> > 
> > On Thu, May 3, 2012 at 9:51 AM, Jenna Fox <a@creativepony.com>
> > wrote:
> > > Is it possible to run a jruby application in some sort of bundle,
> > > without installing Java in to the system? I wouldn't personally
> > > run a java app on my portable computer, as I sometimes spent time
> > > around security professionals who are sometimes inclined to pull a
> > > prank here and there, and java frightens me. Even if I disable
> > > java plugins in my web browser, other apps like email and feed
> > > readers still get it.
> > > 
> > > 
> > > 
> > > 
> > > —
> > > Jenna
> > > 
> > > 
> > > 
> > > On Friday, 4 May 2012 at 12:45 AM, Eric Watson wrote:
> > > 
> > > > > And jar-wise, if you assume JRuby is already on device (same
> > > > > assumption you'd make with mruby) you can ship a compressed
> > > > > archive of .rb code or preconpiled bytecode that's tiny...a
> > > > > few kB for simple apps.
> > > > 
> > > > 
> > > > This would be great! If we didn't assume JRuby or Java is
> > > > already on device (i.e. we bundle Java JRE, JRuby, and Shoes),
> > > > what would you estimate for the download size?
> > > > 
> > > > 
> > > > Eric
> > > 
> > > 
> > 
> > 
> 
> 

Re: [shoes] Ruby in 2012

From:
Jenna Fox
Date:
2012-05-04 @ 07:08
I don't get you guys. I was just asking if you would. I've no interest in 
a philosophical debate or arguing the finer points of software bundling 
legalities. The answer is clearly no, it would have not taken nearly so 
many words to just say that.  


—
Jenna


On Friday, 4 May 2012 at 12:44 PM, Cecil Coupe wrote:

> Java apps are distributed in jars (.jar) which is closely related
> to .tar and .zip. The user tells the pre-installed JRE to run one of
> the files/classes in the jar. You have to trust the .jar and the folks
> who wrote it. The JRE and Java design does have some additional
> security that aren't found in Ruby or Python (ask anyone who has written
> a custom class loader in java).
>  
> Red Shoes is a different matter. It doesn't install Ruby in the
> *system* , it's a private Ruby installed in User directories (for OSX
> and Linux). if you run that private ruby as root it can do anything a
> system installed ruby can do. I've never tried, but I'll bet a Red Shoes
> app can wreck a Windows system if it wants to.  
>  
> Theoretically one could package Shoes, jruby and Java/JRE and run it in
> user space but it's a monster task for each platform and a huge PITA for
> the end user to wait for the huge installation file to unpack. Not to
> mention the licensing issues of distributing a private copy of Java. Ask
> Oracle and Google lawyers if they can think of anything that might go
> wrong.  
>  
> Purple Shoes assumes the user has already installed Java JRE in their
> system. Just as Green Shoes assumes Ruby MRI has been installed in their
> system. Red Shoes scripts can only use pure ruby gems. Shoes scripts in
> Purple/Green/Brown can use the system gem libraries. Thats a good option
> for complex Desktop Shoes apps.  
>  
> As for mobile apps and Ruby it's already possible if you play by Apple
> and Google rules. I'd had forgotten about http://ruboto.org/ for
> Android. For iOS:
> 
http://arstechnica.com/business/news/2012/05/exclusive-building-ruby-ios-applications-with-rubymotion.ars
>  
> Neither one is Shoes easy or casual developer friendly but Ruboto isn't
> that different from the Shoes DSL. A smart Android developer could put a
> Shoes face on if google allows. We'll just have to wait to see if
> mruby/mobiruby gets any traction. As long as mobile devices are
> constrained to their walled gardens, cross platform GUI is unlikely.  
>  
> --Cecil
>  
>  
>  
>  
> On Fri, 2012-05-04 at 08:48 +1000, Jenna Fox wrote:
> > Sure, yes, but what I was asking is can java be bundled in to the
> > shoes app so it doesn't infect the rest of the system? Java is a very
> > complex piece of software with many moving parts, and like flash it
> > introduces a large attack surface, and like flash it's pretty keen to
> > expose that surface to the web, among other things. In other words,
> > can the java go away entirely when the app isn't running, and can it
> > be deleted from the computer entirely when the .app or .exe or
> > whatever is deleted?  
> >  
> >  
> > —
> > Jenna
> >  
> >  
> >  
> > On Friday, 4 May 2012 at 2:13 AM, Peter Fitzgibbons wrote:
> >  
> > > Java apps (not the JRE itself) should be considered with the same
> > > control as native apps. If you'd be wary of a specific native app,
> > > you would be equally wary of an equivalent Java app. OTOH, If you
> > > trust a certain native app, you should trust the equivalent java app
> > > (from same developer :).
> > >  
> > >  
> > > JRE is not evil. Developers use the JRE to do evil.
> > >  
> > >  
> > > I think we need to limit the scope of this discussion to desktop and
> > > desktop/workstation-laptops.
> > > IOS/Android is out of scope, and ChromeBook is out of scope (for now
> > > anyway)... so "desktop only".
> > >  
> > >  
> > > Thoughts?
> > >  
> > > Peter Fitzgibbons
> > > (847) 859-9550
> > > Email: peter.fitzgibbons@gmail.com (mailto:peter.fitzgibbons@gmail.com)
> > > IM GTalk: peter.fitzgibbons
> > > IM AOL: peter.fitzgibbons@gmail.com (mailto:peter.fitzgibbons@gmail.com)
> > >  
> > >  
> > > On Thu, May 3, 2012 at 9:51 AM, Jenna Fox <a@creativepony.com 
(mailto:a@creativepony.com)>
> > > wrote:
> > > > Is it possible to run a jruby application in some sort of bundle,
> > > > without installing Java in to the system? I wouldn't personally
> > > > run a java app on my portable computer, as I sometimes spent time
> > > > around security professionals who are sometimes inclined to pull a
> > > > prank here and there, and java frightens me. Even if I disable
> > > > java plugins in my web browser, other apps like email and feed
> > > > readers still get it.
> > > >  
> > > >  
> > > >  
> > > >  
> > > > —
> > > > Jenna
> > > >  
> > > >  
> > > >  
> > > > On Friday, 4 May 2012 at 12:45 AM, Eric Watson wrote:
> > > >  
> > > > > > And jar-wise, if you assume JRuby is already on device (same
> > > > > > assumption you'd make with mruby) you can ship a compressed
> > > > > > archive of .rb code or preconpiled bytecode that's tiny...a
> > > > > > few kB for simple apps.
> > > > > >  
> > > > >  
> > > > >  
> > > > >  
> > > > > This would be great! If we didn't assume JRuby or Java is
> > > > > already on device (i.e. we bundle Java JRE, JRuby, and Shoes),
> > > > > what would you estimate for the download size?
> > > > >  
> > > > >  
> > > > > Eric  

Re: [shoes] Ruby in 2012

From:
Cecil Coupe
Date:
2012-05-04 @ 07:48
But If I had just said "NO NOT EVER" you would think I speak for every
one. I DONT'T. Then you would ask why not.  Why should you  believe what
one prickly prick (I am) says?  And then I or someone else would write
the same thing I wrote about why not. I think its called education or
teaching or something. Or your suggestion would have no response at all
from the mailing list. Not answering questions and suggestions is a bad
way to build a community. My Apologies for explaining things too much.

No harm, No Foul.
--Cecil
On Fri, 2012-05-04 at 17:08 +1000, Jenna Fox wrote:
> I don't get you guys. I was just asking if you would. I've no interest
> in a philosophical debate or arguing the finer points of software
> bundling legalities. The answer is clearly no, it would have not taken
> nearly so many words to just say that. 
> 
> 
> 
> 
> —
> Jenna
> 
> 
> 
> On Friday, 4 May 2012 at 12:44 PM, Cecil Coupe wrote:
> 
> > Java apps are distributed in jars (.jar) which is closely related
> > to .tar and .zip. The user tells the pre-installed JRE to run one of
> > the files/classes in the jar. You have to trust the .jar and the
> > folks
> > who wrote it. The JRE and Java design does have some additional
> > security that aren't found in Ruby or Python (ask anyone who has
> > written
> > a custom class loader in java).
> > 
> > 
> > Red Shoes is a different matter. It doesn't install Ruby in the
> > *system* , it's a private Ruby installed in User directories (for
> > OSX
> > and Linux). if you run that private ruby as root it can do anything
> > a
> > system installed ruby can do. I've never tried, but I'll bet a Red
> > Shoes
> > app can wreck a Windows system if it wants to. 
> > 
> > 
> > Theoretically one could package Shoes, jruby and Java/JRE and run it
> > in
> > user space but it's a monster task for each platform and a huge PITA
> > for
> > the end user to wait for the huge installation file to unpack. Not
> > to
> > mention the licensing issues of distributing a private copy of Java.
> > Ask
> > Oracle and Google lawyers if they can think of anything that might
> > go
> > wrong. 
> > 
> > 
> > Purple Shoes assumes the user has already installed Java JRE in
> > their
> > system. Just as Green Shoes assumes Ruby MRI has been installed in
> > their
> > system. Red Shoes scripts can only use pure ruby gems. Shoes scripts
> > in
> > Purple/Green/Brown can use the system gem libraries. Thats a good
> > option
> > for complex Desktop Shoes apps. 
> > 
> > 
> > As for mobile apps and Ruby it's already possible if you play by
> > Apple
> > and Google rules. I'd had forgotten about http://ruboto.org/ for
> > Android. For iOS:
> > 
http://arstechnica.com/business/news/2012/05/exclusive-building-ruby-ios-applications-with-rubymotion.ars
> > 
> > 
> > Neither one is Shoes easy or casual developer friendly but Ruboto
> > isn't
> > that different from the Shoes DSL. A smart Android developer could
> > put a
> > Shoes face on if google allows. We'll just have to wait to see if
> > mruby/mobiruby gets any traction. As long as mobile devices are
> > constrained to their walled gardens, cross platform GUI is
> > unlikely. 
> > 
> > 
> > --Cecil
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > On Fri, 2012-05-04 at 08:48 +1000, Jenna Fox wrote:
> > > Sure, yes, but what I was asking is can java be bundled in to the
> > > shoes app so it doesn't infect the rest of the system? Java is a
> > > very
> > > complex piece of software with many moving parts, and like flash
> > > it
> > > introduces a large attack surface, and like flash it's pretty keen
> > > to
> > > expose that surface to the web, among other things. In other
> > > words,
> > > can the java go away entirely when the app isn't running, and can
> > > it
> > > be deleted from the computer entirely when the .app or .exe or
> > > whatever is deleted? 
> > > 
> > > 
> > > 
> > > 
> > > —
> > > Jenna
> > > 
> > > 
> > > 
> > > 
> > > 
> > > 
> > > On Friday, 4 May 2012 at 2:13 AM, Peter Fitzgibbons wrote:
> > > 
> > > 
> > > > Java apps (not the JRE itself) should be considered with the
> > > > same
> > > > control as native apps. If you'd be wary of a specific native
> > > > app,
> > > > you would be equally wary of an equivalent Java app. OTOH, If
> > > > you
> > > > trust a certain native app, you should trust the equivalent java
> > > > app
> > > > (from same developer :).
> > > > 
> > > > 
> > > > 
> > > > 
> > > > JRE is not evil. Developers use the JRE to do evil.
> > > > 
> > > > 
> > > > 
> > > > 
> > > > I think we need to limit the scope of this discussion to desktop
> > > > and
> > > > desktop/workstation-laptops.
> > > > IOS/Android is out of scope, and ChromeBook is out of scope (for
> > > > now
> > > > anyway)... so "desktop only".
> > > > 
> > > > 
> > > > 
> > > > 
> > > > Thoughts?
> > > > 
> > > > 
> > > > Peter Fitzgibbons
> > > > (847) 859-9550
> > > > Email: peter.fitzgibbons@gmail.com
> > > > IM GTalk: peter.fitzgibbons
> > > > IM AOL: peter.fitzgibbons@gmail.com
> > > > 
> > > > 
> > > > 
> > > > 
> > > > On Thu, May 3, 2012 at 9:51 AM, Jenna Fox <a@creativepony.com>
> > > > wrote:
> > > > > Is it possible to run a jruby application in some sort of
> > > > > bundle,
> > > > > without installing Java in to the system? I wouldn't
> > > > > personally
> > > > > run a java app on my portable computer, as I sometimes spent
> > > > > time
> > > > > around security professionals who are sometimes inclined to
> > > > > pull a
> > > > > prank here and there, and java frightens me. Even if I disable
> > > > > java plugins in my web browser, other apps like email and feed
> > > > > readers still get it.
> > > > > 
> > > > > 
> > > > > 
> > > > > 
> > > > > 
> > > > > 
> > > > > 
> > > > > 
> > > > > —
> > > > > Jenna
> > > > > 
> > > > > 
> > > > > 
> > > > > 
> > > > > 
> > > > > 
> > > > > On Friday, 4 May 2012 at 12:45 AM, Eric Watson wrote:
> > > > > 
> > > > > 
> > > > > > > And jar-wise, if you assume JRuby is already on device
> > > > > > > (same
> > > > > > > assumption you'd make with mruby) you can ship a
> > > > > > > compressed
> > > > > > > archive of .rb code or preconpiled bytecode that's
> > > > > > > tiny...a
> > > > > > > few kB for simple apps.
> > > > > > 
> > > > > > 
> > > > > > 
> > > > > > 
> > > > > > This would be great! If we didn't assume JRuby or Java is
> > > > > > already on device (i.e. we bundle Java JRE, JRuby, and
> > > > > > Shoes),
> > > > > > what would you estimate for the download size?
> > > > > > 
> > > > > > 
> > > > > > 
> > > > > > 
> > > > > > Eric
> 
> 

Re: [shoes] Ruby in 2012

From:
Peter Fitzgibbons
Date:
2012-05-04 @ 10:47
Whoa Whoa.

Where did this conversation derail?

Jenna, I hear your security question around Java.  Java/JRE is almost NEVER
packaged "inside" the app/jar.  There are installers that "include" a
download-on-the-fly install of JRE if it doesn't exist, and some installers
even have a "non-JRE" and "includes-JRE" downloads that contain the JRE
installer.  In all these situations, the fact is that the JRE is being
installed _apart_from_ the app/jar.  The JRE is a high-security application
environment, and I encourage you to look toward Java docs and whitepapers
on how security is achieved by apps/jars.

You may note that since any standard "desktop" java app is used by
downloading a Jar and running it.  This makes the app a first-class member
of the system, no sandbox, and therefore has full control of the system as
it is running.  Standard app-level security and awareness applies just like
any other compiled app that is downloaded.  Hopefully this is no surprise
to you.

I'm wondering if you're thinking of JavaWebStart, which is a browser based
system to launch apps specifically crafted to run within a browser
sesstion.  JWS is basically another version of installer, though in this
case, the "installer" is the entry point to running the program every time.
 This is equivalent to .NET "OneClick" installer (might be outdated now).

I personally hope you're ok with the basic direction of Shoes-on-Java.
 We'd like to keep you around.  You sond like you're very thoughtful and
thorough and I encourage you to step in for development if you have time!

Kindest Regards,
Peter Fitzgibbons
(847) 859-9550
Email: peter.fitzgibbons@gmail.com
IM GTalk: peter.fitzgibbons
IM AOL: peter.fitzgibbons@gmail.com


On Fri, May 4, 2012 at 2:48 AM, Cecil Coupe <ccoupe@cableone.net> wrote:

> But If I had just said "NO NOT EVER" you would think I speak for every
> one. I DONT'T. Then you would ask why not.  Why should you  believe what
> one prickly prick (I am) says?  And then I or someone else would write
> the same thing I wrote about why not. I think its called education or
> teaching or something. Or your suggestion would have no response at all
> from the mailing list. Not answering questions and suggestions is a bad
> way to build a community. My Apologies for explaining things too much.
>
> No harm, No Foul.
> --Cecil
> On Fri, 2012-05-04 at 17:08 +1000, Jenna Fox wrote:
> > I don't get you guys. I was just asking if you would. I've no interest
> > in a philosophical debate or arguing the finer points of software
> > bundling legalities. The answer is clearly no, it would have not taken
> > nearly so many words to just say that.
> >
> >
> >
> >
> > —
> > Jenna
> >
> >
> >
> > On Friday, 4 May 2012 at 12:44 PM, Cecil Coupe wrote:
> >
> > > Java apps are distributed in jars (.jar) which is closely related
> > > to .tar and .zip. The user tells the pre-installed JRE to run one of
> > > the files/classes in the jar. You have to trust the .jar and the
> > > folks
> > > who wrote it. The JRE and Java design does have some additional
> > > security that aren't found in Ruby or Python (ask anyone who has
> > > written
> > > a custom class loader in java).
> > >
> > >
> > > Red Shoes is a different matter. It doesn't install Ruby in the
> > > *system* , it's a private Ruby installed in User directories (for
> > > OSX
> > > and Linux). if you run that private ruby as root it can do anything
> > > a
> > > system installed ruby can do. I've never tried, but I'll bet a Red
> > > Shoes
> > > app can wreck a Windows system if it wants to.
> > >
> > >
> > > Theoretically one could package Shoes, jruby and Java/JRE and run it
> > > in
> > > user space but it's a monster task for each platform and a huge PITA
> > > for
> > > the end user to wait for the huge installation file to unpack. Not
> > > to
> > > mention the licensing issues of distributing a private copy of Java.
> > > Ask
> > > Oracle and Google lawyers if they can think of anything that might
> > > go
> > > wrong.
> > >
> > >
> > > Purple Shoes assumes the user has already installed Java JRE in
> > > their
> > > system. Just as Green Shoes assumes Ruby MRI has been installed in
> > > their
> > > system. Red Shoes scripts can only use pure ruby gems. Shoes scripts
> > > in
> > > Purple/Green/Brown can use the system gem libraries. Thats a good
> > > option
> > > for complex Desktop Shoes apps.
> > >
> > >
> > > As for mobile apps and Ruby it's already possible if you play by
> > > Apple
> > > and Google rules. I'd had forgotten about http://ruboto.org/ for
> > > Android. For iOS:
> > >
> 
http://arstechnica.com/business/news/2012/05/exclusive-building-ruby-ios-applications-with-rubymotion.ars
> > >
> > >
> > > Neither one is Shoes easy or casual developer friendly but Ruboto
> > > isn't
> > > that different from the Shoes DSL. A smart Android developer could
> > > put a
> > > Shoes face on if google allows. We'll just have to wait to see if
> > > mruby/mobiruby gets any traction. As long as mobile devices are
> > > constrained to their walled gardens, cross platform GUI is
> > > unlikely.
> > >
> > >
> > > --Cecil
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > > On Fri, 2012-05-04 at 08:48 +1000, Jenna Fox wrote:
> > > > Sure, yes, but what I was asking is can java be bundled in to the
> > > > shoes app so it doesn't infect the rest of the system? Java is a
> > > > very
> > > > complex piece of software with many moving parts, and like flash
> > > > it
> > > > introduces a large attack surface, and like flash it's pretty keen
> > > > to
> > > > expose that surface to the web, among other things. In other
> > > > words,
> > > > can the java go away entirely when the app isn't running, and can
> > > > it
> > > > be deleted from the computer entirely when the .app or .exe or
> > > > whatever is deleted?
> > > >
> > > >
> > > >
> > > >
> > > > —
> > > > Jenna
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > > On Friday, 4 May 2012 at 2:13 AM, Peter Fitzgibbons wrote:
> > > >
> > > >
> > > > > Java apps (not the JRE itself) should be considered with the
> > > > > same
> > > > > control as native apps. If you'd be wary of a specific native
> > > > > app,
> > > > > you would be equally wary of an equivalent Java app. OTOH, If
> > > > > you
> > > > > trust a certain native app, you should trust the equivalent java
> > > > > app
> > > > > (from same developer :).
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > JRE is not evil. Developers use the JRE to do evil.
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > I think we need to limit the scope of this discussion to desktop
> > > > > and
> > > > > desktop/workstation-laptops.
> > > > > IOS/Android is out of scope, and ChromeBook is out of scope (for
> > > > > now
> > > > > anyway)... so "desktop only".
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > Thoughts?
> > > > >
> > > > >
> > > > > Peter Fitzgibbons
> > > > > (847) 859-9550
> > > > > Email: peter.fitzgibbons@gmail.com
> > > > > IM GTalk: peter.fitzgibbons
> > > > > IM AOL: peter.fitzgibbons@gmail.com
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > On Thu, May 3, 2012 at 9:51 AM, Jenna Fox <a@creativepony.com>
> > > > > wrote:
> > > > > > Is it possible to run a jruby application in some sort of
> > > > > > bundle,
> > > > > > without installing Java in to the system? I wouldn't
> > > > > > personally
> > > > > > run a java app on my portable computer, as I sometimes spent
> > > > > > time
> > > > > > around security professionals who are sometimes inclined to
> > > > > > pull a
> > > > > > prank here and there, and java frightens me. Even if I disable
> > > > > > java plugins in my web browser, other apps like email and feed
> > > > > > readers still get it.
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > —
> > > > > > Jenna
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > On Friday, 4 May 2012 at 12:45 AM, Eric Watson wrote:
> > > > > >
> > > > > >
> > > > > > > > And jar-wise, if you assume JRuby is already on device
> > > > > > > > (same
> > > > > > > > assumption you'd make with mruby) you can ship a
> > > > > > > > compressed
> > > > > > > > archive of .rb code or preconpiled bytecode that's
> > > > > > > > tiny...a
> > > > > > > > few kB for simple apps.
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > This would be great! If we didn't assume JRuby or Java is
> > > > > > > already on device (i.e. we bundle Java JRE, JRuby, and
> > > > > > > Shoes),
> > > > > > > what would you estimate for the download size?
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > Eric
> >
> >
>
>
>

Re: [shoes] Ruby in 2012

From:
Jenna Fox
Date:
2012-05-04 @ 02:58
So, the answer to my actual question is "probably possible, but it's very 
difficult so we probably won't do that".  

Thanks.


—
Jenna


On Friday, 4 May 2012 at 12:44 PM, Cecil Coupe wrote:

> Java apps are distributed in jars (.jar) which is closely related
> to .tar and .zip. The user tells the pre-installed JRE to run one of
> the files/classes in the jar. You have to trust the .jar and the folks
> who wrote it. The JRE and Java design does have some additional
> security that aren't found in Ruby or Python (ask anyone who has written
> a custom class loader in java).
>  
> Red Shoes is a different matter. It doesn't install Ruby in the
> *system* , it's a private Ruby installed in User directories (for OSX
> and Linux). if you run that private ruby as root it can do anything a
> system installed ruby can do. I've never tried, but I'll bet a Red Shoes
> app can wreck a Windows system if it wants to.  
>  
> Theoretically one could package Shoes, jruby and Java/JRE and run it in
> user space but it's a monster task for each platform and a huge PITA for
> the end user to wait for the huge installation file to unpack. Not to
> mention the licensing issues of distributing a private copy of Java. Ask
> Oracle and Google lawyers if they can think of anything that might go
> wrong.  
>  
> Purple Shoes assumes the user has already installed Java JRE in their
> system. Just as Green Shoes assumes Ruby MRI has been installed in their
> system. Red Shoes scripts can only use pure ruby gems. Shoes scripts in
> Purple/Green/Brown can use the system gem libraries. Thats a good option
> for complex Desktop Shoes apps.  
>  
> As for mobile apps and Ruby it's already possible if you play by Apple
> and Google rules. I'd had forgotten about http://ruboto.org/ for
> Android. For iOS:
> 
http://arstechnica.com/business/news/2012/05/exclusive-building-ruby-ios-applications-with-rubymotion.ars
>  
> Neither one is Shoes easy or casual developer friendly but Ruboto isn't
> that different from the Shoes DSL. A smart Android developer could put a
> Shoes face on if google allows. We'll just have to wait to see if
> mruby/mobiruby gets any traction. As long as mobile devices are
> constrained to their walled gardens, cross platform GUI is unlikely.  
>  
> --Cecil
>  
>  
>  
>  
> On Fri, 2012-05-04 at 08:48 +1000, Jenna Fox wrote:
> > Sure, yes, but what I was asking is can java be bundled in to the
> > shoes app so it doesn't infect the rest of the system? Java is a very
> > complex piece of software with many moving parts, and like flash it
> > introduces a large attack surface, and like flash it's pretty keen to
> > expose that surface to the web, among other things. In other words,
> > can the java go away entirely when the app isn't running, and can it
> > be deleted from the computer entirely when the .app or .exe or
> > whatever is deleted?  
> >  
> >  
> > —
> > Jenna
> >  
> >  
> >  
> > On Friday, 4 May 2012 at 2:13 AM, Peter Fitzgibbons wrote:
> >  
> > > Java apps (not the JRE itself) should be considered with the same
> > > control as native apps. If you'd be wary of a specific native app,
> > > you would be equally wary of an equivalent Java app. OTOH, If you
> > > trust a certain native app, you should trust the equivalent java app
> > > (from same developer :).
> > >  
> > >  
> > > JRE is not evil. Developers use the JRE to do evil.
> > >  
> > >  
> > > I think we need to limit the scope of this discussion to desktop and
> > > desktop/workstation-laptops.
> > > IOS/Android is out of scope, and ChromeBook is out of scope (for now
> > > anyway)... so "desktop only".
> > >  
> > >  
> > > Thoughts?
> > >  
> > > Peter Fitzgibbons
> > > (847) 859-9550
> > > Email: peter.fitzgibbons@gmail.com (mailto:peter.fitzgibbons@gmail.com)
> > > IM GTalk: peter.fitzgibbons
> > > IM AOL: peter.fitzgibbons@gmail.com (mailto:peter.fitzgibbons@gmail.com)
> > >  
> > >  
> > > On Thu, May 3, 2012 at 9:51 AM, Jenna Fox <a@creativepony.com 
(mailto:a@creativepony.com)>
> > > wrote:
> > > > Is it possible to run a jruby application in some sort of bundle,
> > > > without installing Java in to the system? I wouldn't personally
> > > > run a java app on my portable computer, as I sometimes spent time
> > > > around security professionals who are sometimes inclined to pull a
> > > > prank here and there, and java frightens me. Even if I disable
> > > > java plugins in my web browser, other apps like email and feed
> > > > readers still get it.
> > > >  
> > > >  
> > > >  
> > > >  
> > > > —
> > > > Jenna
> > > >  
> > > >  
> > > >  
> > > > On Friday, 4 May 2012 at 12:45 AM, Eric Watson wrote:
> > > >  
> > > > > > And jar-wise, if you assume JRuby is already on device (same
> > > > > > assumption you'd make with mruby) you can ship a compressed
> > > > > > archive of .rb code or preconpiled bytecode that's tiny...a
> > > > > > few kB for simple apps.
> > > > > >  
> > > > >  
> > > > >  
> > > > >  
> > > > > This would be great! If we didn't assume JRuby or Java is
> > > > > already on device (i.e. we bundle Java JRE, JRuby, and Shoes),
> > > > > what would you estimate for the download size?
> > > > >  
> > > > >  
> > > > > Eric  

Re: [shoes] Ruby in 2012

From:
Cecil Coupe
Date:
2012-05-04 @ 03:42
Tell you what, Jenna. You hire the lawyers and negotiate with Oracle and
when they tell us we can ship a 100MB copy of JRE without their
copyright for each Shoes script any kid writes, then we'll consider it.
Shouldn't cost you more than $10K in legal fees. $20K tops. You'd
probably want could buy an insurance policy for the developers if your
lawyer is wrong. Still interested?

I said "technically possible" and pointed out what hurdles exist (and
those are only some of the known issues). We can't get a Red Shoes
maintenance release out (3.2) and you want this struggling community to
break the law and fight with Oracle, Google, and/or Apple later?  I
don't about you, but I'd rather program in Lua or SDK/ADK than break the
law. But hey, If you pay for it, someone will consider it.



On Fri, 2012-05-04 at 12:58 +1000, Jenna Fox wrote:
> So, the answer to my actual question is "probably possible, but it's
> very difficult so we probably won't do that". 
> 
> 
> Thanks.
> 
> 
> 
> 
> —
> Jenna
> 
> 
> 
> On Friday, 4 May 2012 at 12:44 PM, Cecil Coupe wrote:
> 
> > Java apps are distributed in jars (.jar) which is closely related
> > to .tar and .zip. The user tells the pre-installed JRE to run one of
> > the files/classes in the jar. You have to trust the .jar and the
> > folks
> > who wrote it. The JRE and Java design does have some additional
> > security that aren't found in Ruby or Python (ask anyone who has
> > written
> > a custom class loader in java).
> > 
> > 
> > Red Shoes is a different matter. It doesn't install Ruby in the
> > *system* , it's a private Ruby installed in User directories (for
> > OSX
> > and Linux). if you run that private ruby as root it can do anything
> > a
> > system installed ruby can do. I've never tried, but I'll bet a Red
> > Shoes
> > app can wreck a Windows system if it wants to. 
> > 
> > 
> > Theoretically one could package Shoes, jruby and Java/JRE and run it
> > in
> > user space but it's a monster task for each platform and a huge PITA
> > for
> > the end user to wait for the huge installation file to unpack. Not
> > to
> > mention the licensing issues of distributing a private copy of Java.
> > Ask
> > Oracle and Google lawyers if they can think of anything that might
> > go
> > wrong. 
> > 
> > 
> > Purple Shoes assumes the user has already installed Java JRE in
> > their
> > system. Just as Green Shoes assumes Ruby MRI has been installed in
> > their
> > system. Red Shoes scripts can only use pure ruby gems. Shoes scripts
> > in
> > Purple/Green/Brown can use the system gem libraries. Thats a good
> > option
> > for complex Desktop Shoes apps. 
> > 
> > 
> > As for mobile apps and Ruby it's already possible if you play by
> > Apple
> > and Google rules. I'd had forgotten about http://ruboto.org/ for
> > Android. For iOS:
> > 
http://arstechnica.com/business/news/2012/05/exclusive-building-ruby-ios-applications-with-rubymotion.ars
> > 
> > 
> > Neither one is Shoes easy or casual developer friendly but Ruboto
> > isn't
> > that different from the Shoes DSL. A smart Android developer could
> > put a
> > Shoes face on if google allows. We'll just have to wait to see if
> > mruby/mobiruby gets any traction. As long as mobile devices are
> > constrained to their walled gardens, cross platform GUI is
> > unlikely. 
> > 
> > 
> > --Cecil
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > 
> > On Fri, 2012-05-04 at 08:48 +1000, Jenna Fox wrote:
> > > Sure, yes, but what I was asking is can java be bundled in to the
> > > shoes app so it doesn't infect the rest of the system? Java is a
> > > very
> > > complex piece of software with many moving parts, and like flash
> > > it
> > > introduces a large attack surface, and like flash it's pretty keen
> > > to
> > > expose that surface to the web, among other things. In other
> > > words,
> > > can the java go away entirely when the app isn't running, and can
> > > it
> > > be deleted from the computer entirely when the .app or .exe or
> > > whatever is deleted? 
> > > 
> > > 
> > > 
> > > 
> > > —
> > > Jenna
> > > 
> > > 
> > > 
> > > 
> > > 
> > > 
> > > On Friday, 4 May 2012 at 2:13 AM, Peter Fitzgibbons wrote:
> > > 
> > > 
> > > > Java apps (not the JRE itself) should be considered with the
> > > > same
> > > > control as native apps. If you'd be wary of a specific native
> > > > app,
> > > > you would be equally wary of an equivalent Java app. OTOH, If
> > > > you
> > > > trust a certain native app, you should trust the equivalent java
> > > > app
> > > > (from same developer :).
> > > > 
> > > > 
> > > > 
> > > > 
> > > > JRE is not evil. Developers use the JRE to do evil.
> > > > 
> > > > 
> > > > 
> > > > 
> > > > I think we need to limit the scope of this discussion to desktop
> > > > and
> > > > desktop/workstation-laptops.
> > > > IOS/Android is out of scope, and ChromeBook is out of scope (for
> > > > now
> > > > anyway)... so "desktop only".
> > > > 
> > > > 
> > > > 
> > > > 
> > > > Thoughts?
> > > > 
> > > > 
> > > > Peter Fitzgibbons
> > > > (847) 859-9550
> > > > Email: peter.fitzgibbons@gmail.com
> > > > IM GTalk: peter.fitzgibbons
> > > > IM AOL: peter.fitzgibbons@gmail.com
> > > > 
> > > > 
> > > > 
> > > > 
> > > > On Thu, May 3, 2012 at 9:51 AM, Jenna Fox <a@creativepony.com>
> > > > wrote:
> > > > > Is it possible to run a jruby application in some sort of
> > > > > bundle,
> > > > > without installing Java in to the system? I wouldn't
> > > > > personally
> > > > > run a java app on my portable computer, as I sometimes spent
> > > > > time
> > > > > around security professionals who are sometimes inclined to
> > > > > pull a
> > > > > prank here and there, and java frightens me. Even if I disable
> > > > > java plugins in my web browser, other apps like email and feed
> > > > > readers still get it.
> > > > > 
> > > > > 
> > > > > 
> > > > > 
> > > > > 
> > > > > 
> > > > > 
> > > > > 
> > > > > —
> > > > > Jenna
> > > > > 
> > > > > 
> > > > > 
> > > > > 
> > > > > 
> > > > > 
> > > > > On Friday, 4 May 2012 at 12:45 AM, Eric Watson wrote:
> > > > > 
> > > > > 
> > > > > > > And jar-wise, if you assume JRuby is already on device
> > > > > > > (same
> > > > > > > assumption you'd make with mruby) you can ship a
> > > > > > > compressed
> > > > > > > archive of .rb code or preconpiled bytecode that's
> > > > > > > tiny...a
> > > > > > > few kB for simple apps.
> > > > > > 
> > > > > > 
> > > > > > 
> > > > > > 
> > > > > > This would be great! If we didn't assume JRuby or Java is
> > > > > > already on device (i.e. we bundle Java JRE, JRuby, and
> > > > > > Shoes),
> > > > > > what would you estimate for the download size?
> > > > > > 
> > > > > > 
> > > > > > 
> > > > > > 
> > > > > > Eric
> 
> 

Re: [shoes] Ruby in 2012

From:
Eric Watson
Date:
2012-05-04 @ 18:24

On May 3, 2012, at 10:42 PM, Cecil Coupe <ccoupe@cableone.net> wrote:

> Tell you what, Jenna. You hire the lawyers and negotiate with Oracle and
> when they tell us we can ship a 100MB copy of JRE without their
> copyright for each Shoes script any kid writes, then we'll consider it.

JRuby offers a windows download that comes with a JRE. I'm not sure 
whether the JRE gets installed system wide or whether it's "private" to 
the JRuby install. 

http://www.jruby.org/download

Maybe it's possible...

Re: [shoes] Ruby in 2012

From:
Khristian
Date:
2012-05-04 @ 18:57
The JRE installation is copied into JRuby's folder, and the installer
probably configures your PATH using jruby\bin and jruby\jre\bin (I
didn't find any difference in the .bat files for JRuby, so I assumed
it worked like this). This way, you would have a private JRE install
which is accessible for all processes, at least for those running
under your user (because of the PATH configuration done). This might
be less secure than a proper JRE install because it is not protected
by UAC's admin access restrictions.
I just downloaded it for testing, but can't test the PATH setup (don't
want to ruin the configuration on this machine right now).


On Fri, May 4, 2012 at 3:24 PM, Eric Watson <wasnotrice@gmail.com> wrote:
>
>
> On May 3, 2012, at 10:42 PM, Cecil Coupe <ccoupe@cableone.net> wrote:
>
>> Tell you what, Jenna. You hire the lawyers and negotiate with Oracle and
>> when they tell us we can ship a 100MB copy of JRE without their
>> copyright for each Shoes script any kid writes, then we'll consider it.
>
> JRuby offers a windows download that comes with a JRE. I'm not sure 
whether the JRE gets installed system wide or whether it's "private" to 
the JRuby install.
>
> http://www.jruby.org/download
>
> Maybe it's possible...



-- 
Khristian Alexander Schönrock
http://derkosak.blogspot.com - Meu blógue!

Re: [shoes] Ruby in 2012

From:
Charles Oliver Nutter
Date:
2012-05-05 @ 05:40
You are so full of misinformation you should probably stop now.

The JRE is maybe 20MB on the high end, and it is redistributable without
any special licensing. JRuby has distributed an all in one installer for
Windows for years.
On May 3, 2012 10:42 PM, "Cecil Coupe" <ccoupe@cableone.net> wrote:

> Tell you what, Jenna. You hire the lawyers and negotiate with Oracle and
> when they tell us we can ship a 100MB copy of JRE without their
> copyright for each Shoes script any kid writes, then we'll consider it.
> Shouldn't cost you more than $10K in legal fees. $20K tops. You'd
> probably want could buy an insurance policy for the developers if your
> lawyer is wrong. Still interested?
>
> I said "technically possible" and pointed out what hurdles exist (and
> those are only some of the known issues). We can't get a Red Shoes
> maintenance release out (3.2) and you want this struggling community to
> break the law and fight with Oracle, Google, and/or Apple later?  I
> don't about you, but I'd rather program in Lua or SDK/ADK than break the
> law. But hey, If you pay for it, someone will consider it.
>
>
>
> On Fri, 2012-05-04 at 12:58 +1000, Jenna Fox wrote:
> > So, the answer to my actual question is "probably possible, but it's
> > very difficult so we probably won't do that".
> >
> >
> > Thanks.
> >
> >
> >
> >
> > —
> > Jenna
> >
> >
> >
> > On Friday, 4 May 2012 at 12:44 PM, Cecil Coupe wrote:
> >
> > > Java apps are distributed in jars (.jar) which is closely related
> > > to .tar and .zip. The user tells the pre-installed JRE to run one of
> > > the files/classes in the jar. You have to trust the .jar and the
> > > folks
> > > who wrote it. The JRE and Java design does have some additional
> > > security that aren't found in Ruby or Python (ask anyone who has
> > > written
> > > a custom class loader in java).
> > >
> > >
> > > Red Shoes is a different matter. It doesn't install Ruby in the
> > > *system* , it's a private Ruby installed in User directories (for
> > > OSX
> > > and Linux). if you run that private ruby as root it can do anything
> > > a
> > > system installed ruby can do. I've never tried, but I'll bet a Red
> > > Shoes
> > > app can wreck a Windows system if it wants to.
> > >
> > >
> > > Theoretically one could package Shoes, jruby and Java/JRE and run it
> > > in
> > > user space but it's a monster task for each platform and a huge PITA
> > > for
> > > the end user to wait for the huge installation file to unpack. Not
> > > to
> > > mention the licensing issues of distributing a private copy of Java.
> > > Ask
> > > Oracle and Google lawyers if they can think of anything that might
> > > go
> > > wrong.
> > >
> > >
> > > Purple Shoes assumes the user has already installed Java JRE in
> > > their
> > > system. Just as Green Shoes assumes Ruby MRI has been installed in
> > > their
> > > system. Red Shoes scripts can only use pure ruby gems. Shoes scripts
> > > in
> > > Purple/Green/Brown can use the system gem libraries. Thats a good
> > > option
> > > for complex Desktop Shoes apps.
> > >
> > >
> > > As for mobile apps and Ruby it's already possible if you play by
> > > Apple
> > > and Google rules. I'd had forgotten about http://ruboto.org/ for
> > > Android. For iOS:
> > >
> 
http://arstechnica.com/business/news/2012/05/exclusive-building-ruby-ios-applications-with-rubymotion.ars
> > >
> > >
> > > Neither one is Shoes easy or casual developer friendly but Ruboto
> > > isn't
> > > that different from the Shoes DSL. A smart Android developer could
> > > put a
> > > Shoes face on if google allows. We'll just have to wait to see if
> > > mruby/mobiruby gets any traction. As long as mobile devices are
> > > constrained to their walled gardens, cross platform GUI is
> > > unlikely.
> > >
> > >
> > > --Cecil
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > > On Fri, 2012-05-04 at 08:48 +1000, Jenna Fox wrote:
> > > > Sure, yes, but what I was asking is can java be bundled in to the
> > > > shoes app so it doesn't infect the rest of the system? Java is a
> > > > very
> > > > complex piece of software with many moving parts, and like flash
> > > > it
> > > > introduces a large attack surface, and like flash it's pretty keen
> > > > to
> > > > expose that surface to the web, among other things. In other
> > > > words,
> > > > can the java go away entirely when the app isn't running, and can
> > > > it
> > > > be deleted from the computer entirely when the .app or .exe or
> > > > whatever is deleted?
> > > >
> > > >
> > > >
> > > >
> > > > —
> > > > Jenna
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > > On Friday, 4 May 2012 at 2:13 AM, Peter Fitzgibbons wrote:
> > > >
> > > >
> > > > > Java apps (not the JRE itself) should be considered with the
> > > > > same
> > > > > control as native apps. If you'd be wary of a specific native
> > > > > app,
> > > > > you would be equally wary of an equivalent Java app. OTOH, If
> > > > > you
> > > > > trust a certain native app, you should trust the equivalent java
> > > > > app
> > > > > (from same developer :).
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > JRE is not evil. Developers use the JRE to do evil.
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > I think we need to limit the scope of this discussion to desktop
> > > > > and
> > > > > desktop/workstation-laptops.
> > > > > IOS/Android is out of scope, and ChromeBook is out of scope (for
> > > > > now
> > > > > anyway)... so "desktop only".
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > Thoughts?
> > > > >
> > > > >
> > > > > Peter Fitzgibbons
> > > > > (847) 859-9550
> > > > > Email: peter.fitzgibbons@gmail.com
> > > > > IM GTalk: peter.fitzgibbons
> > > > > IM AOL: peter.fitzgibbons@gmail.com
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > On Thu, May 3, 2012 at 9:51 AM, Jenna Fox <a@creativepony.com>
> > > > > wrote:
> > > > > > Is it possible to run a jruby application in some sort of
> > > > > > bundle,
> > > > > > without installing Java in to the system? I wouldn't
> > > > > > personally
> > > > > > run a java app on my portable computer, as I sometimes spent
> > > > > > time
> > > > > > around security professionals who are sometimes inclined to
> > > > > > pull a
> > > > > > prank here and there, and java frightens me. Even if I disable
> > > > > > java plugins in my web browser, other apps like email and feed
> > > > > > readers still get it.
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > —
> > > > > > Jenna
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > On Friday, 4 May 2012 at 12:45 AM, Eric Watson wrote:
> > > > > >
> > > > > >
> > > > > > > > And jar-wise, if you assume JRuby is already on device
> > > > > > > > (same
> > > > > > > > assumption you'd make with mruby) you can ship a
> > > > > > > > compressed
> > > > > > > > archive of .rb code or preconpiled bytecode that's
> > > > > > > > tiny...a
> > > > > > > > few kB for simple apps.
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > This would be great! If we didn't assume JRuby or Java is
> > > > > > > already on device (i.e. we bundle Java JRE, JRuby, and
> > > > > > > Shoes),
> > > > > > > what would you estimate for the download size?
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > Eric
> >
> >
>
>
>

Re: [shoes] Ruby in 2012

From:
Cecil Coupe
Date:
2012-05-05 @ 06:57
On Sat, 2012-05-05 at 05:40 +0000, Charles Oliver Nutter wrote:
> You are so full of misinformation you should probably stop now.
> 
> The JRE is maybe 20MB on the high end, and it is redistributable
> without any special licensing. JRuby has distributed an all in one
> installer for Windows for years.

You didn't read the later messages were I was politely corrected, but
thanks for your broadside attack. Now that you're here to save the Shoes
project, I'll leave it in your good hands.


Re: [shoes] Ruby in 2012

From:
Charles Oliver Nutter
Date:
2012-05-05 @ 10:14
I don't take grossly overstated FUD well. The broadside attack was a bit
knee-jerk, I admit, but maybe you need to temper your enthusiasm for
slamming Java and the JVM a bit. Or at least get your facts straight before
you begin :-)

FWIW, what I know of the plans for OpenJDK tell me that JRuby is going to
become more and more viable for fast, small, portable GUI apps over the
next year as both straight-line and startup perf are improved by JVM
updates. If anything holds a promising future for Ruby in 2012 and beyond,
it's JRuby.

We'll keep improving JRuby and you folks can continue using it :-)
On May 5, 2012 1:58 AM, "Cecil Coupe" <ccoupe@cableone.net> wrote:

> On Sat, 2012-05-05 at 05:40 +0000, Charles Oliver Nutter wrote:
> > You are so full of misinformation you should probably stop now.
> >
> > The JRE is maybe 20MB on the high end, and it is redistributable
> > without any special licensing. JRuby has distributed an all in one
> > installer for Windows for years.
>
> You didn't read the later messages were I was politely corrected, but
> thanks for your broadside attack. Now that you're here to save the Shoes
> project, I'll leave it in your good hands.
>
>
>
>

Re: [shoes] Ruby in 2012

From:
Cecil Coupe
Date:
2012-05-07 @ 06:13
After Dave's suggested timeout. 

On Sat, 2012-05-05 at 10:14 +0000, Charles Oliver Nutter wrote:
> I don't take grossly overstated FUD well. The broadside attack was a
> bit knee-jerk, I admit, but maybe you need to temper your enthusiasm
> for slamming Java and the JVM a bit. Or at least get your facts
> straight before you begin :-) 

I don't slam Java. I started with Java in Apple Java 1.0.2 so I go way
back. When the generics syntax arrived I saw another C++ arriving. For
me, it was something I no longer cared to be involved with since python
and ruby could do what *I* wanted faster and easier. As I said before,
JRuby is an amazing bit of damn hard work. The JVM is amazing. I
embedded Rhino and Slisp in that Java/AWT 1.0.2 app and it ran on Win 98
laptop with 80MB of memory as well as the G4 Mac. Amazing. 

> 
> FWIW, what I know of the plans for OpenJDK tell me that JRuby is going
> to become more and more viable for fast, small, portable GUI apps over
> the next year as both straight-line and startup perf are improved by
> JVM updates. If anything holds a promising future for Ruby in 2012 and
> beyond, it's JRuby.
> 

You are more tuned into future JRE/JDK developments than I. I hope you
are correct about the future. We got cross wired because I was mostly
talking to Jenna about her visionary mobile Shoes dreams and one of us
dragged Java/JRE into that discussion and it caught your attention.

> We'll keep improving JRuby and you folks can continue using it :-) 
That's all anybody can do! When JRuby arrives on mobile devices in a
form Shoes can build upon, then the Shoes folks can figure out how to
use it on mobile device.

Regards,
--Cecil


Re: [shoes] Ruby in 2012

From:
David Eastman
Date:
2012-05-08 @ 00:28
This has been an interesting thread (excuse the British understatement) but
a few things are being said that are probably not in keeping with the list.

- Like most people here, I love shoes, and I recognise that shoes strength
and weakness is ruby.
- I am keen to see how ruby can be strengthened beyond the MRI
- The JVM and its related technologies may possibly be key to shoes
- Java as a language is a bit of a turd.

That's where I sit. But many people on this list are probably here to share
experiences about ruby and shoes, and are going to be highly alienated by
culture wars that originate elsewhere.

Please tread carefully.


On Mon, May 7, 2012 at 7:13 AM, Cecil Coupe <ccoupe@cableone.net> wrote:

> After Dave's suggested timeout.
>
> On Sat, 2012-05-05 at 10:14 +0000, Charles Oliver Nutter wrote:
> > I don't take grossly overstated FUD well. The broadside attack was a
> > bit knee-jerk, I admit, but maybe you need to temper your enthusiasm
> > for slamming Java and the JVM a bit. Or at least get your facts
> > straight before you begin :-)
>
> I don't slam Java. I started with Java in Apple Java 1.0.2 so I go way
> back. When the generics syntax arrived I saw another C++ arriving. For
> me, it was something I no longer cared to be involved with since python
> and ruby could do what *I* wanted faster and easier. As I said before,
> JRuby is an amazing bit of damn hard work. The JVM is amazing. I
> embedded Rhino and Slisp in that Java/AWT 1.0.2 app and it ran on Win 98
> laptop with 80MB of memory as well as the G4 Mac. Amazing.
>
> >
> > FWIW, what I know of the plans for OpenJDK tell me that JRuby is going
> > to become more and more viable for fast, small, portable GUI apps over
> > the next year as both straight-line and startup perf are improved by
> > JVM updates. If anything holds a promising future for Ruby in 2012 and
> > beyond, it's JRuby.
> >
>
> You are more tuned into future JRE/JDK developments than I. I hope you
> are correct about the future. We got cross wired because I was mostly
> talking to Jenna about her visionary mobile Shoes dreams and one of us
> dragged Java/JRE into that discussion and it caught your attention.
>
> > We'll keep improving JRuby and you folks can continue using it :-)
> That's all anybody can do! When JRuby arrives on mobile devices in a
> form Shoes can build upon, then the Shoes folks can figure out how to
> use it on mobile device.
>
> Regards,
> --Cecil
>
>
>
>

Re: [shoes] Ruby in 2012

From:
Peter Fitzgibbons
Date:
2012-05-08 @ 01:26
I'm mostly on board with your assessment.

I think that most programmers who focus on a non-Java (or C#) language,
especially dynamics, tend towards name-calling in expressing their
disappointment in Java.
I always take the opportunity to point out that JRuby, along with Jython,
Clojure, and Scala

(and<http://en.wikipedia.org/wiki/Java_Virtual_Machine#Support_for_dynamic_languages>
others <http://en.wikipedia.org/wiki/List_of_JVM_languages> I'm missing)
find benefit in the performance, multi-platform use, and wide, wide net of
libraries available for JVM.  Thanks are due to Java and the cloud  of
developers.

Thanks for your thoughts!

Peter Fitzgibbons
(847) 859-9550
Email: peter.fitzgibbons@gmail.com
IM GTalk: peter.fitzgibbons
IM AOL: peter.fitzgibbons@gmail.com


On Mon, May 7, 2012 at 7:28 PM, David Eastman <deastman1@gmail.com> wrote:

> This has been an interesting thread (excuse the British understatement)
> but a few things are being said that are probably not in keeping with the
> list.
>
> - Like most people here, I love shoes, and I recognise that shoes strength
> and weakness is ruby.
> - I am keen to see how ruby can be strengthened beyond the MRI
> - The JVM and its related technologies may possibly be key to shoes
> - Java as a language is a bit of a turd.
>
> That's where I sit. But many people on this list are probably here to
> share experiences about ruby and shoes, and are going to be highly
> alienated by culture wars that originate elsewhere.
>
> Please tread carefully.
>
>
> On Mon, May 7, 2012 at 7:13 AM, Cecil Coupe <ccoupe@cableone.net> wrote:
>
>> After Dave's suggested timeout.
>>
>> On Sat, 2012-05-05 at 10:14 +0000, Charles Oliver Nutter wrote:
>> > I don't take grossly overstated FUD well. The broadside attack was a
>> > bit knee-jerk, I admit, but maybe you need to temper your enthusiasm
>> > for slamming Java and the JVM a bit. Or at least get your facts
>> > straight before you begin :-)
>>
>> I don't slam Java. I started with Java in Apple Java 1.0.2 so I go way
>> back. When the generics syntax arrived I saw another C++ arriving. For
>> me, it was something I no longer cared to be involved with since python
>> and ruby could do what *I* wanted faster and easier. As I said before,
>> JRuby is an amazing bit of damn hard work. The JVM is amazing. I
>> embedded Rhino and Slisp in that Java/AWT 1.0.2 app and it ran on Win 98
>> laptop with 80MB of memory as well as the G4 Mac. Amazing.
>>
>> >
>> > FWIW, what I know of the plans for OpenJDK tell me that JRuby is going
>> > to become more and more viable for fast, small, portable GUI apps over
>> > the next year as both straight-line and startup perf are improved by
>> > JVM updates. If anything holds a promising future for Ruby in 2012 and
>> > beyond, it's JRuby.
>> >
>>
>> You are more tuned into future JRE/JDK developments than I. I hope you
>> are correct about the future. We got cross wired because I was mostly
>> talking to Jenna about her visionary mobile Shoes dreams and one of us
>> dragged Java/JRE into that discussion and it caught your attention.
>>
>> > We'll keep improving JRuby and you folks can continue using it :-)
>> That's all anybody can do! When JRuby arrives on mobile devices in a
>> form Shoes can build upon, then the Shoes folks can figure out how to
>> use it on mobile device.
>>
>> Regards,
>> --Cecil
>>
>>
>>
>>
>

Re: [shoes] Ruby in 2012

From:
Cecil Coupe
Date:
2012-05-08 @ 03:23
On Mon, 2012-05-07 at 20:26 -0500, Peter Fitzgibbons wrote:

> 
> I always take the opportunity to point out that JRuby, along with
> Jython, Clojure, and Scala (and others I'm missing) find benefit in
> the performance, multi-platform use, and wide, wide net of libraries
> available for JVM.  Thanks are due to Java and the cloud  of
> developers.

The others:  http://en.wikipedia.org/wiki/List_of_JVM_languages




Re: [shoes] Ruby in 2012

From:
Dave Wilkinson
Date:
2012-05-05 @ 13:34
You both overreacted. So, let's all calm down in the meantime. :)

Re: [shoes] Ruby in 2012

From:
Cecil Coupe
Date:
2012-05-03 @ 00:52
Charles, 

  Many Shoes scripts don't run long enough or loop enough to get the
full hotspot love. If the GUI is sluggish or slow to start then it seems
slow. For Shoes, its slow to start.  No big deal to some, an annoyance
to others who constantly edit the script, crank up shoes to test it,
exit and repeat. 

  I've followed jruby's project since you announced the project. Awesome
job just getting to work. 



On Wed, 2012-05-02 at 17:15 +0000, Charles Oliver Nutter wrote:
> What performance hit are you talking about? JRuby is probably the
> fastest Ruby right now, bandit is precisely because we can translate
> to JVM bytecode that we are able to run so fast.
> 
> And jar-wise, if you assume JRuby is already on device (same
> assumption you'd make with mruby) you can ship a compressed archive
> of .rb code or preconpiled bytecode that's tiny...a few kB for simple
> apps.
> 
> Your facts about JRuby seem a bit off.
> 
> On May 2, 2012 1:34 AM, "Cecil Coupe" <ccoupe@cableone.net> wrote:
>         On Wed, 2012-05-02 at 01:55 -0400, Steve Klabnik wrote:
>         > > The bytecode compiler of mruby solves many of the
>         packaging issues with
>         > > Shoes.
>         >
>         > Basically every Ruby, including MRI, has bytecode.
>         
>         Which version of Ruby loads byte code from a file produced by
>         a Ruby far
>         away in time and space ? mruby is (potentially) like the Java
>         JVM. More
>         Python, more Java like. Kind of like JRuby w/o the extra Java
>         to Ruby
>         translation performance hit and the huge jars.
>         
>         It's worth watching. I have my doubts it will succeed but I
>         wouldn't
>         ignore it.
>         
>         

Re: [shoes] Ruby in 2012

From:
Peter Fitzgibbons
Date:
2012-05-03 @ 14:30
So you're talking about startup performance.

I may be a very patient man, I've noted vast improvement in jruby startup.
 Development flow on brown_shoes has been positive in the past months.
 Jruby 1.6.7 is an important version.. .many startup and profiling
improvements.

Cecil, have you worked with brown-shoes in the past months?

Peter Fitzgibbons
(847) 859-9550
Email: peter.fitzgibbons@gmail.com
IM GTalk: peter.fitzgibbons
IM AOL: peter.fitzgibbons@gmail.com


On Wed, May 2, 2012 at 7:52 PM, Cecil Coupe <ccoupe@cableone.net> wrote:

> Charles,
>
>  Many Shoes scripts don't run long enough or loop enough to get the
> full hotspot love. If the GUI is sluggish or slow to start then it seems
> slow. For Shoes, its slow to start.  No big deal to some, an annoyance
> to others who constantly edit the script, crank up shoes to test it,
> exit and repeat.
>
>  I've followed jruby's project since you announced the project. Awesome
> job just getting to work.
>
>
>
> On Wed, 2012-05-02 at 17:15 +0000, Charles Oliver Nutter wrote:
> > What performance hit are you talking about? JRuby is probably the
> > fastest Ruby right now, bandit is precisely because we can translate
> > to JVM bytecode that we are able to run so fast.
> >
> > And jar-wise, if you assume JRuby is already on device (same
> > assumption you'd make with mruby) you can ship a compressed archive
> > of .rb code or preconpiled bytecode that's tiny...a few kB for simple
> > apps.
> >
> > Your facts about JRuby seem a bit off.
> >
> > On May 2, 2012 1:34 AM, "Cecil Coupe" <ccoupe@cableone.net> wrote:
> >         On Wed, 2012-05-02 at 01:55 -0400, Steve Klabnik wrote:
> >         > > The bytecode compiler of mruby solves many of the
> >         packaging issues with
> >         > > Shoes.
> >         >
> >         > Basically every Ruby, including MRI, has bytecode.
> >
> >         Which version of Ruby loads byte code from a file produced by
> >         a Ruby far
> >         away in time and space ? mruby is (potentially) like the Java
> >         JVM. More
> >         Python, more Java like. Kind of like JRuby w/o the extra Java
> >         to Ruby
> >         translation performance hit and the huge jars.
> >
> >         It's worth watching. I have my doubts it will succeed but I
> >         wouldn't
> >         ignore it.
> >
> >
>
>
>

Re: [shoes] Ruby in 2012

From:
Steve Klabnik
Date:
2012-05-02 @ 15:28
> Which version of Ruby loads byte code from a file produced by a Ruby far
> away in time and space ?

Rubinius produces .rbc files. JRuby has 'jrubyc', which gives you a
.class that you can run as long as jruby.jar is on your $CLASSPATH.

In MRI, you can dump it, but the bytecode verifier isn't done, so they
don't expose the loader, which is a Ruby 2.0 target.

Re: [shoes] Ruby in 2012

From:
Date:
2012-05-02 @ 09:02
could not decode message

Re: [shoes] Ruby in 2012

From:
Jenna Fox
Date:
2012-05-02 @ 09:29
I have a raspberry pi on prder! I think they're way better than computers 
because you can make stuff out of them! Skype portals and robots and 
puzzles!  

—
Jenna Fox


On Wednesday, 2 May 2012 at 7:02 PM, deastman1@gmail.com wrote:

> In the UK, a lot of fuss has been made about the Raspberry Pi - a cheap 
computer board reminiscent of the old ZX spectrum days.
>  
> There will be a lot of dads buying these for kids, only to find the kids
want an iPad.
>  
>  
>  
> -- Sent from my HP Pre3
> On 2 May 2012 07:39, Jenna Fox <a@creativepony.com> wrote:  
>  
> Kids are already making games using Lua on iPads. They easily connect to
usb keyboards with cheap adaptors, and connect to any bluetooth keyboards,
or apple's keyboard docks, but for experienced tablet users (kids 
especially) the touch keyboards are quite usable and fast, and are often 
augmented with dedicated programming keys for characters like '{' ';', 
'do', 'end', 'void', 'function() {'  
>  
> Go to any shop, you'll see the number one thing kids want today, is an 
iPad. Recent nelson studies 
(http://www.electronista.com/articles/11/11/17/nielsen.shows.kids.teens.want.ipad.for.christmas/)
suggest as much as 44% of kids wanted an iPad for christmas last year, and
demand has only grown since then. Among teenagers, iPads still took the 
lead ahead of every other sort of gadget. Why would kids want to make apps
for computers? They don't even want to use computers. A shoes for iOS 
would need to include it's own code editor functionality - this is true. 
It's a shame nobody's made anything like that (http://hackety.com/) in 
shoes before.  
>  
> —
> Jenna
>  
>  
> On Wednesday, 2 May 2012 at 1:24 PM, Cecil Coupe wrote:
>  
> > Shoes, The GUI/DSL certainly could fit in with mruby/mobiruby. Just
> > another fork off of Green Shoes if we knew enough about mobiruby API's.  
> > The bytecode compiler of mruby solves many of the packaging issues with
> > Shoes.  
> >  
> > > From a Red shoes developers perspective there is a lot to like *if*
> > everything comes to pass and the VM API's are documented in English so
> > Shoes can mingle with them.  
> >  
> > If you believe Shoe's purpose is learning to program, then
> > mruby/mobiruby announcement won't change anything since few few tablets
> > and smart phones are tethered to keyboards and functioning text
> > editors.  
> >  
> > Mostly off topic, I really wish we had a Qt version of Green Shoes. I
> > happen to like lost puppies and I installed Haiku (Be OS) on a
> > partition. There are things to like about BeOS (and many rough edges).
> > It's Posix enough; someone has implemented python and someone has
> > implemented Qt. That and mruby just feels like a black hole filled with
> > shiny things sucking me into it.  
> >  
> > --Cecil
> >  
> >  
> > On Wed, 2012-05-02 at 01:48 +0100, David Eastman wrote:
> > > Nice article about ruby
> > > today http://rubylearning.com/blog/2012/04/23/ruby-in-2012/  
> > >  
> > >  
> > > At the end of the article, interesting lack of belief in ruby GUIs. I
> > > wonder if shoes may be a little more major here than was first
> > > imagined.
> > >  
> >  
> >  
> >  
> >  
>  
>  

Re: [shoes] Ruby in 2012

From:
Jesse Cooke
Date:
2012-05-02 @ 05:57
On Tue, May 1, 2012 at 8:24 PM, Cecil Coupe <ccoupe@cableone.net> wrote:

> Shoes, The GUI/DSL certainly could fit in with mruby/mobiruby. Just
> another fork off of Green Shoes if we knew enough about mobiruby API's.
> The bytecode compiler of mruby solves many of the packaging issues with
> Shoes.
>
> >From a Red shoes developers perspective there is a lot to like *if*
> everything comes to pass and the VM API's are documented in English so
> Shoes can mingle with them.
>
> If you believe Shoe's purpose is learning to program, then
> mruby/mobiruby announcement won't change anything since few few tablets
> and smart phones are tethered to keyboards and functioning text
> editors.
>
> Mostly off topic, I really wish we had a Qt version of Green Shoes.  I
> happen to like lost puppies and I installed Haiku (Be OS) on a
> partition. There are things to like about BeOS (and many rough edges).
> It's Posix enough; someone has implemented python and someone has
> implemented Qt. That and mruby just feels like a black hole filled with
> shiny things sucking me into it.
>
Someone got rubinius running on Haiku. Just chatted with Brian Ford about
Haiku this evening.

>
> --Cecil
>
>
> On Wed, 2012-05-02 at 01:48 +0100, David Eastman wrote:
> > Nice article about ruby
> > today http://rubylearning.com/blog/2012/04/23/ruby-in-2012/
> >
> >
> > At the end of the article, interesting lack of belief in ruby GUIs. I
> > wonder if shoes may be a little more major here than was first
> > imagined.
>
>
>

Re: [shoes] Ruby in 2012

From:
Cecil Coupe
Date:
2012-05-02 @ 06:55
On Tue, 2012-05-01 at 22:57 -0700, Jesse Cooke wrote:

> Someone got rubinius running on Haiku. Just chatted with Brian Ford
> about Haiku this evening.  
>         

Damn thee, you shiny objects! Be gone you demons of progress! Tempt me
naught with Qt.