I compiled the app locally using the web2py admin, edited the app.yaml to upload *.pyc, then I uploaded the app to see if the compiled/*.pyc worked and got the error. I assumed that Google removed get_magic for some reason, but I did not pursue it.
It would be interesting to skip get_magic and see if it loads. Robin On Oct 18, 6:52 pm, mdipierro <[EMAIL PROTECTED]> wrote: > Robin, > > If you could send me the exact error I will ask Guido. This must have > a very easy fix. > > Massimo > > On Oct 18, 6:47 pm, Robin B <[EMAIL PROTECTED]> wrote: > > > When I uploaded compiled *.pyc and tried to run it, an error was > > raised, something about imp.get_magic not existing. It might be > > possible to avoid imp.get_magic and get it loading, but I have not > > tried that. I did try putting the code into a dict, and that sped > > things up. > > > Robin > > > On Oct 18, 5:55 pm, mdipierro <[EMAIL PROTECTED]> wrote: > > > > Robin, > > > > is it sufficient to byte-code compile the app and make sure the pyc > > > files (including gluon/*.pyc) are uploaded too or do we need to modify > > > web2py? > > > > Massimo > > > > On Oct 18, 5:28 pm, yarko <[EMAIL PROTECTED]> wrote: > > > > > Sorry - I don't know about *.pyc files; I assume they're not platform > > > > dependent, so I'm assuming that uploading *.pyc files with your app > > > > doesn't work (?). > > > > > How is the Django support on GAE handling this issue? > > > > > On Oct 18, 1:51 pm, Robin B <[EMAIL PROTECTED]> wrote: > > > > > > In compileapp.py, after reading, parsing, and compiling a model, view, > > > > > or controller, if a .pyc cannot be written to disk, instead store the > > > > > compiled code, by file name/function, in a global dict so that next > > > > > request you can simply load the precompiled code directly from RAM. > > > > > > Since web2py does not have an environment (does not distinguish > > > > > between development and production etc), the only way to update the > > > > > cached code is to check the mtime of each file on every request which > > > > > is wasteful in production where the code does not change, but not > > > > > nearly as wasteful as repeatedly reading, parsing and compiling. > > > > > > Robin > > > > > > On Oct 18, 10:27 am, mdipierro <[EMAIL PROTECTED]> wrote: > > > > > > > On Oct 18, 9:51 am, Robin B <[EMAIL PROTECTED]> wrote: > > > > > > > > Web2py 'works' on appengine, but it is reading, parsing and > > > > > > > compiling > > > > > > > the models, controllers, and views on *every* request causing all > > > > > > > the > > > > > > > wasted CPU cycles. Normally, web2py caches code as .pyc files, > > > > > > > but > > > > > > > you cannot write the filesystem on appengine so nothing gets > > > > > > > cached by > > > > > > > default. It is trivial to cache the compiled code in a dict and > > > > > > > reuse > > > > > > > it on the next request. > > > > > > > Could you explain more? --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "web2py Web Framework" group. To post to this group, send email to web2py@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/web2py?hl=en -~----------~----~----~----~------~----~------~--~---
