Hi, We are handling such problem as well. The solution is not that good, but still usable.
We programmers and tds are using Mercurial to maintain just one copy of scripts/plug-ins/py packages, on a centralized server. Those artists are loading those things remotely on maya/standalone startup. The idea is to make sure everybody on the same page. We have a maya plugin to config Python sys.path to locate necessary py packages, to source mel scripts, to load plug-ins, create menus. so on. On each workstation, just all MAYA_PLUG_IN_PATH to locate that config plugin, auto-load it will do the magic each time maya started. For standalone applications, we use the same idea. Most apps are just simple PyQt, do not start it directly, start a config tool first, setup sys.path, then run the main app. We managed to create a standardized py script to do that, to launch a specif app, just run python configStandalon -a appname Put it into a bat/sh file, so those artists can use an app by double-click.\ zhang On Wed, Jun 6, 2012 at 12:19 AM, Judah Baron <[email protected]> wrote: > I won't go into a lot of detail right now, but here are some of our > highlights: > > Version controlled distributions so users can sync to update, roll back to > escape problematic releases, etc. We use Perforce, but anything is better > than nothing here. Syncing also means everyone's system looks similar. > We void custom installations and configuration at all costs because it > makes maintaining numerous machines more difficult and very time consuming. > For instance, our tools do not rely on the version of Python that is > installed on the workstation. Instead, we maintain a copy of Python and all > of our dependencies in our version control system, and our tools call on > these dependencies explicitly. This allows us to have multiple versions of > Maya, running potentially multiple versions of Python, at the same time, > with [ideally] the same code. > > Ideal Setup For New Employee: > > 1. Fresh machine configured by IT with standard studio image as > starting point. > 2. Sync to latest release of tools. > 3. Run. > > This ideal setup is actually not just our ideal, but how we really > operate. If ever we have an issue with a newly configured machine not being > able to run any component, it servers as an indicator that our system needs > to be modified to handle such a situation. We still run into problems every > once in a while, but it generally has to do with the disk image not being > complete. > > Good luck, > -Judah > > > On Tue, May 29, 2012 at 5:43 AM, Alexander Kalychny <[email protected]>wrote: > >> Hi there >> We cg-animation studio with >100 people and now we deep in major >> pipeline cleaning for all departments, and wondering how guys at >> major studios distribute, update ( maybe auto-update?) tools for tds >> and artist in maya. >> Maybe using pip / easy_install with your own server serving eggs? or >> smth else? >> Im interested in tools at all scales and platforms >> (windows,linux,osx): from pyqt small dialog tools for maya, to big >> standalone pyqt package with a lot of C++ custom modules >> interdependencies.. >> How you guys approach these kind of problems? >> >> Thanks >> >> -- >> view archives: http://groups.google.com/group/python_inside_maya >> change your subscription settings: >> http://groups.google.com/group/python_inside_maya/subscribe >> > > -- > view archives: http://groups.google.com/group/python_inside_maya > change your subscription settings: > http://groups.google.com/group/python_inside_maya/subscribe > -- view archives: http://groups.google.com/group/python_inside_maya change your subscription settings: http://groups.google.com/group/python_inside_maya/subscribe
