[Pharo-project] Problem with Pharo 1.3 in Windows
Stef wrote
Now my main questions are
Hey ... dont take me wrong. I want the build server to build the VMs too, I want a repeatable process and a reliable VM. I appreciate that we change this!!! We all know that if VM development is dependent on a single maintainer person then nobody would invest into Pharo, Squeak or Cuis since the truck factor* is one. I support all this. Remember: I was the one who wrote the first ConfigurationOfVMMaker so that it is easy to get all the bits and pieces. I want to make it easier for people to dive into VM building. The windows build slave is a huge step forward ... we can not only build the VM and installer ... we can also run the tests on Windows too to see if there are platform specific issues. Cool and hardly needed! But someone (person or team) has to wear the Win32 VM "hat" to check contributions or help fixing a bug. So may question is what is the direction we will take here? It looks like the standard SqueakWin32 VM is unmaintained since the last release is from July 2010. We still have no Alien, we still have no callbacks. Yes, there is Cog whose code base is newer but I think Eliots focus is on speed rather than functionality. Will we fork from standard VM, will we participate in Eliots work or will we fork an own using git and maintain it by picking pieces from the other VMs? Also: what can we do to reach the goal of feature complete VM's with aligned version numbers. I think this means to participate on lists like "vm-dev"... But maybe I'm just too impatient... Bye T. *the truck factor = The number of people on your team who have to be hit with a truck before the project is in serious trouble" -- Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de
On May 19, 2011, at 1:25 PM, Torsten Bergmann wrote:
Stef wrote
Now my main questions are
Hey ... dont take me wrong.
I did not :) I just took your mail as a pretext to make us all think :)
I want the build server to build the VMs too, I want a repeatable process and a reliable VM. I appreciate that we change this!!!
We all know that if VM development is dependent on a single maintainer person then nobody would invest into Pharo, Squeak or Cuis since the truck factor* is one.
I support all this. Remember: I was the one who wrote the first ConfigurationOfVMMaker so that it is easy to get all the bits and pieces. I want to make it easier for people to dive into VM building.
I know :)
The windows build slave is a huge step forward ... we can not only build the VM and installer ... we can also run the tests on Windows too to see if there are platform specific issues. Cool and hardly needed!
But someone (person or team) has to wear the Win32 VM "hat" to check contributions or help fixing a bug. So may question is what is the direction we will take here?
I have the impression that igor will have to have a look at it :) We did not tell him yet :) but there is a high chance that igor should do something :)
It looks like the standard SqueakWin32 VM is unmaintained since the last release is from July 2010. We still have no Alien, we still have no callbacks. Yes, there is Cog whose code base is newer but I think Eliots focus is on speed rather than functionality.
Will we fork from standard VM, will we participate in Eliots work or will we fork an own using git and maintain it by picking pieces from the other VMs?
this is not the answer that we can give alone :)
Also: what can we do to reach the goal of feature complete VM's with aligned version numbers. I think this means to participate on lists like "vm-dev"...
stef
On 19 May 2011 13:25, Torsten Bergmann <astares@gmx.de> wrote:
Stef wrote
Now my main questions are
Hey ... dont take me wrong. I want the build server to build the VMs too, I want a repeatable process and a reliable VM. I appreciate that we change this!!!
We all know that if VM development is dependent on a single maintainer person then nobody would invest into Pharo, Squeak or Cuis since the truck factor* is one.
I support all this. Remember: I was the one who wrote the first ConfigurationOfVMMaker so that it is easy to get all the bits and pieces. I want to make it easier for people to dive into VM building.
The windows build slave is a huge step forward ... we can not only build the VM and installer ... we can also run the tests on Windows too to see if there are platform specific issues. Cool and hardly needed!
But someone (person or team) has to wear the Win32 VM "hat" to check contributions or help fixing a bug. So may question is what is the direction we will take here?
I could help a lot with Win32, because i know it well. And i want to help. The only problem is time. Because if you have too many hats, it is hard to decide how long you should wear one, before switching to another, and hard to decide which one should be worn next. :)
It looks like the standard SqueakWin32 VM is unmaintained since the last release is from July 2010. We still have no Alien, we still have no callbacks. Yes, there is Cog whose code base is newer but I think Eliots focus is on speed rather than functionality.
Will we fork from standard VM, will we participate in Eliots work or will we fork an own using git and maintain it by picking pieces from the other VMs?
Branding VM is not a fork. It just putting a label on already manufactured product. And i think the right answer is to do it together. Eliot cannot do everything, neither do i, or anyone else. But if we merge our efforts, then we can do everything.
Also: what can we do to reach the goal of feature complete VM's with aligned version numbers. I think this means to participate on lists like "vm-dev"...
yes.
But maybe I'm just too impatient...
For the impatient: register on gitorious and clone blessed repository and then put your modifications, which you described in another thread into your clone. A little, simple step. But then i will not be forced to search mailing list in order to harvest these bits.
Bye T.
*the truck factor = The number of people on your team who  have to be hit with a truck before the project is in serious  trouble"
-- Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de
-- Best regards, Igor Stasenko AKA sig.
participants (3)
-
Igor Stasenko -
Stéphane Ducasse -
Torsten Bergmann