[Mono-list] I give up / Mac OS X PPC support

Erik Dasque edasque@ximian.com
Sun, 11 Apr 2004 12:33:10 -0400


--Apple-Mail-1-314775400
Content-Transfer-Encoding: 7bit
Content-Type: text/plain;
	charset=US-ASCII;
	format=flowed

Hi all,

Developing and delivering software like Mono is not an easy task and a 
complex engineering project. We're committed to delivering Mono 1.0 in 
June and supporting the Linux x86, MacOS X platforms, have no doubt 
about that. As you know we also intend to provide support for Solaris 
and Linux PPC. We're also committed to delivering GTK# 1.0 in this 
release and SWF by the end of the year.

We are privileged to observe the evolution of Mono on a daily basis. In 
my previous job, building from the source was something you didn't want 
to do unless you had to, builds were often broken and unreliable. A 
good build, one QA could work on, was a rare commodity. Most of us 
build Mono from CVS everyday with its share of frustration and 
disillusions but all in all, it often works. My main machine is a 
Powerbook running MacOS X but I also run Linux PPC, Linux x86 and 
Windows XP ; I have to because like Miguel, our Ximian engineers and 
all the good people contributing to Mono, I want to make sure we 
deliver on our promises. And because I mainly use MacOS X, I harass 
Miguel everyday about bugs I find. Mono on PPC has been frustrating in 
the last few weeks but I run 0.31 with some success. Recently, Miguel 
asked Martin to help out on PPC ; with his involvement in addition to 
Paolo's and Miguel's, I am confident we'll be successful.

Erik Dasque
Product Manager for Mono, Ximian.
Office: (+1) 617 613 2009
Cell:    (+1) 617 953 9104
edasque@ximian.com

On Apr 9, 2004, at 1:07 AM, Steve Mentzer wrote:

>
> Sadly, I must agree. The mono team have made excellent progress on 
> bringing PPC into the JIT era.... but.....
>
> On a whim, I decided to install fedora core 1 and build mono. It built 
> and installed without incident. Truly remarkable. I must say that I 
> was impressed. XSP kept crashing, but that is a different story...
>
> On x86 hardware, I prefer windows xp/2003. Sorry, I love *nix, but 
> linux doesn't do much for me on x86, especially when x86/windows 
> offers 100% compatibility and killer dev tools. Frankly, there is no 
> real reason to host asp.net apps under apache, when my XP box does 
> better after locking it down. No religion here folks, just reality.
>
> OS/X and PPC is a different story. This is a situation that is DYING 
> for an x86 crossover platform. Don't talk to me about Java.... if I 
> wanted java, I wouldn't be using mono or c#.
>
> C++/wxWindows/Qt/GTK/etc... yawn. Once again, I want c#.
>
> Mono on the PPC is painful. There is no *documented*, stable or 
> official GUI toolkit support. Hell, even the core runtime is about as 
> stable as IIS 4. :)
>
> Beyond that, the mono releases are hit and miss. You have about a 75% 
> chance of the build failing or getting the dreaded 'bus error'.
>
> I appreciate the new features and the roadmap, but if all mono has to 
> offer is x86/linux stability, then mono has only succeeded in doing is 
> providing a 'free' implementation of the .net runtime that runs on the 
> same commodity hardware that a fully supported and commercial version 
> runs on.
>
> Mono cannot win if it sticks to the 'linux vs. windows' card. Linux 
> hacks hate MS and C# and .NET. Corporate windows users will be 
> reluctant to adopt linux/mono when they can get a fully supported, 
> commercial version from MS that will run on the same hardware. Believe 
> it or not, most windows shops adopting .NET don't care about OS 
> licensing costs or security. They are interested in RAD. So mono is a 
> little like selling ice cubes to the eskimos.
>
> Now, if mono ran reliably on HPUX, sparc, linux, os/x, windows, *bsd 
> and some other *nix variants, then you would have succeeded in beating 
> MS at their own game. The only argument against .net now is that it 
> isn't cross platform. If mono fills that niche, then the sky is the 
> limit.
>
> No offense to the mono team, but you should either drop support for 
> PPC entirely or actually concentrate on getting it as stable as 
> x86/linux.
>
> Sorry for the long rant guys... its been a long day... :)
>
>
> Original Message -----------------------
>> my  problem with gtk# personally is that mac os x is not supported.
>> although you can find people who hacked it together, if  you co it
>> from cvs, and install it, the samples don't work. i don't have a linux
>> box, only a mac, and i would love to play with mono on it, but it's
>> been two months now, and i don't feel any closer.
>>
>> there are also problems with xsp, and sometimes mono doesn't compile
>> either (i mean cvs version).
>> regards,
>
> Grudgingly, I have to agree with above statement(s).  Mono support for
> Mac OS X/ PPC is fragile at best.  I even tried to put together a page
> on the basic steps (
> http://homepage.mac.com/griffincaprio/mono/RunningMonoOnMacOSX.html ),
> but I have all but given up trying to maintain that page.  This is
> because mono not reliable from day to day, release to release, on the
> PPC / Mac OS X platform.
>
> I am not saying that running off the CVS tree is the most stable way to
> work, but this goes for the releases also.  0.29 worked on the PPC,
> 0.30 didn't  without some major hacking.  0.31 doesn't work reliably,
> either.
>
> Some of the steps I have taken to compile mono include:
> - configure switches
> - external, 3rd party source downloads
> - editing actual code/headers to accommodate the PPC platform.
>
> What's worse, is that when I encounter errors, i receive almost no help
> on the mailing lists.  Some of these are not doubt strange errors, and
> some are very common, but my posts go unanswered either way.
>
> I haven't even gotten to gtk#...
>
> Segmentation faults, bus errors, frozen compiles, etc....  The list
> goes on and on.  I hope to come back to mono in a few months, when/if
> PPC support is better.  But for now, I feel that it's a lost cause.
>
> - Griffin
>
> _______________________________________________
> Mono-list maillist  -  Mono-list@lists.ximian.com
> http://lists.ximian.com/mailman/listinfo/mono-list
>
> _______________________________________________
> Mono-list maillist  -  Mono-list@lists.ximian.com
> http://lists.ximian.com/mailman/listinfo/mono-list

--Apple-Mail-1-314775400
Content-Transfer-Encoding: 7bit
Content-Type: text/enriched;
	charset=US-ASCII

Hi all,


Developing and delivering software like Mono is not an easy task and a
complex engineering project. We're committed to delivering Mono 1.0 in
June and supporting the Linux x86, MacOS X platforms, have no doubt
about that. As you know we also intend to provide support for Solaris
and Linux PPC. We're also committed to delivering GTK# 1.0 in this
release and SWF by the end of the year.


We are privileged to observe the evolution of Mono on a daily basis.
In my previous job, building from the source was something you didn't
want to do unless you had to, builds were often broken and unreliable.
A good build, one QA could work on, was a rare commodity. Most of us
build Mono from CVS everyday with its share of frustration and
disillusions but all in all, it often works. My main machine is a
Powerbook running MacOS X but I also run Linux PPC, Linux x86 and
Windows XP ; I have to because like Miguel, our Ximian engineers and
all the good people contributing to Mono, I want to make sure we
deliver on our promises. And because I mainly use MacOS X, I harass
Miguel everyday about bugs I find. Mono on PPC has been frustrating in
the last few weeks but I run 0.31 with some success. Recently, Miguel
asked Martin to help out on PPC ; with his involvement in addition to
Paolo's and Miguel's, I am confident we'll be successful.


<fontfamily><param>Helvetica</param><x-tad-smaller>Erik Dasque

Product Manager for Mono, Ximian.

Office: (+1) 617 613 2009

Cell:    (+1) 617 953 9104

edasque@ximian.com</x-tad-smaller></fontfamily>


On Apr 9, 2004, at 1:07 AM, Steve Mentzer wrote:


<excerpt>

Sadly, I must agree. The mono team have made excellent progress on
bringing PPC into the JIT era.... but.....


On a whim, I decided to install fedora core 1 and build mono. It built
and installed without incident. Truly remarkable. I must say that I
was impressed. XSP kept crashing, but that is a different story...


On x86 hardware, I prefer windows xp/2003. Sorry, I love *nix, but
linux doesn't do much for me on x86, especially when x86/windows
offers 100% compatibility and killer dev tools. Frankly, there is no
real reason to host asp.net apps under apache, when my XP box does
better after locking it down. No religion here folks, just reality.


OS/X and PPC is a different story. This is a situation that is DYING
for an x86 crossover platform. Don't talk to me about Java.... if I
wanted java, I wouldn't be using mono or c#.


C++/wxWindows/Qt/GTK/etc... yawn. Once again, I want c#.


Mono on the PPC is painful. There is no *documented*, stable or
official GUI toolkit support. Hell, even the core runtime is about as
stable as IIS 4. :)


Beyond that, the mono releases are hit and miss. You have about a 75%
chance of the build failing or getting the dreaded 'bus error'.


I appreciate the new features and the roadmap, but if all mono has to
offer is x86/linux stability, then mono has only succeeded in doing is
providing a 'free' implementation of the .net runtime that runs on the
same commodity hardware that a fully supported and commercial version
runs on.


Mono cannot win if it sticks to the 'linux vs. windows' card. Linux
hacks hate MS and C# and .NET. Corporate windows users will be
reluctant to adopt linux/mono when they can get a fully supported,
commercial version from MS that will run on the same hardware. Believe
it or not, most windows shops adopting .NET don't care about OS
licensing costs or security. They are interested in RAD. So mono is a
little like selling ice cubes to the eskimos.


Now, if mono ran reliably on HPUX, sparc, linux, os/x, windows, *bsd
and some other *nix variants, then you would have succeeded in beating
MS at their own game. The only argument against .net now is that it
isn't cross platform. If mono fills that niche, then the sky is the
limit.


No offense to the mono team, but you should either drop support for
PPC entirely or actually concentrate on getting it as stable as
x86/linux.


Sorry for the long rant guys... its been a long day... :)



Original Message -----------------------

<excerpt>my  problem with gtk# personally is that mac os x is not
supported.

although you can find people who hacked it together, if  you co it

from cvs, and install it, the samples don't work. i don't have a linux

box, only a mac, and i would love to play with mono on it, but it's

been two months now, and i don't feel any closer.


there are also problems with xsp, and sometimes mono doesn't compile

either (i mean cvs version).

regards,

</excerpt>

Grudgingly, I have to agree with above statement(s).  Mono support for

Mac OS X/ PPC is fragile at best.  I even tried to put together a page

on the basic steps (

http://homepage.mac.com/griffincaprio/mono/RunningMonoOnMacOSX.html ),

but I have all but given up trying to maintain that page.  This is

because mono not reliable from day to day, release to release, on the

PPC / Mac OS X platform.


I am not saying that running off the CVS tree is the most stable way to

work, but this goes for the releases also.  0.29 worked on the PPC,

0.30 didn't  without some major hacking.  0.31 doesn't work reliably,

either.


Some of the steps I have taken to compile mono include:

- configure switches

- external, 3rd party source downloads

- editing actual code/headers to accommodate the PPC platform.


What's worse, is that when I encounter errors, i receive almost no help

on the mailing lists.  Some of these are not doubt strange errors, and

some are very common, but my posts go unanswered either way.


I haven't even gotten to gtk#...


Segmentation faults, bus errors, frozen compiles, etc....  The list

goes on and on.  I hope to come back to mono in a few months, when/if

PPC support is better.  But for now, I feel that it's a lost cause.


- Griffin


_______________________________________________

Mono-list maillist  -  Mono-list@lists.ximian.com

http://lists.ximian.com/mailman/listinfo/mono-list


_______________________________________________

Mono-list maillist  -  Mono-list@lists.ximian.com

http://lists.ximian.com/mailman/listinfo/mono-list

</excerpt>
--Apple-Mail-1-314775400--