[Pharo-project] Cog builds on Hudson
I modified the Cog config made by Marcus to sync with my changes in scripts. https://pharo-ic.lille.inria.fr/hudson/view/Cog/ i also added a new target for building - Stack VM (except that it not works now, because Hudson can't pick up the built artifacts despite VM built successfully -- lets wait for Marcus to fix that , since i'm not expert in Hudson :) ). Adding another two is also a piece of cake - just copy existing configuration and change destination directory and configuration class used for generating sources and build config. The script i made for Hudson can work even for macs, it takes 3 arguments: - git repo url - working subdirectory name where to place everything - cmake configuration class name (which build to perform) that's all you need to build VM from scratch using Hudson -- Best regards, Igor Stasenko AKA sig.
On 15 February 2011 18:13, Igor Stasenko <siguctua@gmail.com> wrote:
I modified the Cog config made by Marcus to sync with my changes in scripts.
https://pharo-ic.lille.inria.fr/hudson/view/Cog/
i also added a new target for building - Stack VM (except that it not works now, because Hudson can't pick up the built artifacts despite VM built successfully -- lets wait for Marcus to fix that , since i'm not expert in Hudson :) ).
Adding another two is also a piece of cake - just copy existing configuration and change destination directory and configuration class used for generating sources and build config.
oops.. i was to hurry typing. By "another two" i meant - Cog with debug info and - StackVM with debug info
The script i made for Hudson can work even for macs, it takes 3 arguments:  - git repo url  - working subdirectory name where to place everything  - cmake configuration class name (which build to perform)
that's all you need to build VM from scratch using Hudson
-- Best regards, Igor Stasenko AKA sig.
-- Best regards, Igor Stasenko AKA sig.
I love these mails. Nice music to my hear......agility, reproducibility, control of complexity.... Imagine building a new vm each time a package changes. Is it not a dream? Imagine soon we will be able to know when aperformance degrades because of a change and all that based on git hub and a nice hudson server. So people will be able to modify their vms and plugin and manage their life and publish improvements. Simply gorgeous. I think that what is important besides making VM compiling more mainstream is that it gives a good signal to the outside world. Yes investors/inventors/people on the edge yes we have processes, yes we are concerned by quality, yes we are concerned by reproducibility,... Stef
I modified the Cog config made by Marcus to sync with my changes in scripts.
https://pharo-ic.lille.inria.fr/hudson/view/Cog/
i also added a new target for building - Stack VM (except that it not works now, because Hudson can't pick up the built artifacts despite VM built successfully -- lets wait for Marcus to fix that , since i'm not expert in Hudson :) ).
Adding another two is also a piece of cake - just copy existing configuration and change destination directory and configuration class used for generating sources and build config.
The script i made for Hudson can work even for macs, it takes 3 arguments: - git repo url - working subdirectory name where to place everything - cmake configuration class name (which build to perform)
that's all you need to build VM from scratch using Hudson
-- Best regards, Igor Stasenko AKA sig.
participants (2)
-
Igor Stasenko -
Stéphane Ducasse