[mod_python] PSP bug report

Conrad Steenberg conrad at hep.caltech.edu
Thu Dec 4 10:04:12 EST 2003


Hi

Hmmm, you're right, I should have read the documentation again before
posting that. Treating %> as an unindent would also eliminate the
looping html construct.

Cheers

Conrad

On Thu, 2003-12-04 at 09:50, Dan W. wrote:
> That would eliminate conditional html.  However, it might be interesting if 
> PSP had a special directive for conditional html such as two percent signs.
> 
> <%%
> for x in xrange(10):
> %>
> 
> -Dan
> 
> At 09:35 AM 12/4/2003 -0800, you wrote:
> >Hi Grisha
> >
> >This may not be a bug, but it could still be worth the effort to make an
> >%> be considered a newline as well - purely from a support perspective
> >and the amount of annoyance it might save you and countless new users of
> >PSP :-)
> >
> >Cheers
> >
> >Conrad
> >
> >On Wed, 2003-12-03 at 22:00, Gregory (Grisha) Trubetskoy wrote:
> > > I think the best practice is to use a comment rather than a blank line:
> > >
> > > <%
> > > for x in range(0):
> > >       pass
> > > # end for
> > > %>
> > >
> > > Grisha
> > >
> > > On Wed, 3 Dec 2003, Dan W. wrote:
> > >
> > > >
> > > > I realized my error too late.  I needed a blank line after the loop.  PSP
> > > > was considering the rest of the page to be part of the loop and since the
> > > > loop did not execute neither did the html.
> > > >
> > > > -Dan
> > > >
> > > >
> > > >
> > > > At 09:14 PM 12/3/2003 -0600, you wrote:
> > > >
> > > > >It turns out that my problem wasn't related to the worker MPM after
> > > > >all.  After minimizing the code down I've found that PSP fails on zero
> > > > >element for loops.  The following PSP page should reproduce the problem.
> > > > >
> > > > ><%
> > > > >for x in range(0):
> > > > >     pass
> > > > >%>
> > > > >Hello world
> > > > >
> > > > >Does the PSP maintainer read this list?
> > > > >
> > > > >-Dan
> > > > >
> > > > >
> > > > >
> > > > >At 05:54 PM 12/3/2003 -0600, you wrote:
> > > > >
> > > > > >I'm running into a weird problem with 3.1.2b and Apache 2.0.47 
> > compiled
> > > > > >under the worker MPM.  I have two Apache processes, each with 25
> > > > > >threads.  I have a very simple PSP page which imports a simple .py 
> > module
> > > > > >and prints out a value.  The problem is that the page is only 
> > served part
> > > > > >of the time.  When it is served, it is correct (no tracebacks), 
> > but if I
> > > > > >do a browser reload a couple times it will eventually serve a blank
> > > > > >page.  I can immediately do a browser reload a few more times and 
> > get a
> > > > > >few more blank pages.  Then if I wait a few seconds and do a browser
> > > > > >reload again then it will serve up the proper page.
> > > > > >
> > > > > >I'm not quite sure how to verify it, but it appears as though one 
> > process
> > > > > >is having a problem with mod_python while the other process is 
> > functioning
> > > > > >properly.  Any suggestions on how I could verify this or other ideas?
> > > > > >
> > > > > >I wasn't able to find anything in the documentation on MPM 
> > compatibility.
> > > > > >
> > > > > >Thanks!
> > > > > >
> > > > > >-Dan
> > > > > >
> > > > > >_______________________________________________
> > > > > >Mod_python mailing list
> > > > > >Mod_python at modpython.org
> > > > > >http://mailman.modpython.org/mailman/listinfo/mod_python
> > > > >
> > > > >_______________________________________________
> > > > >Mod_python mailing list
> > > > >Mod_python at modpython.org
> > > > >http://mailman.modpython.org/mailman/listinfo/mod_python
> > > >
> > > > _______________________________________________
> > > > Mod_python mailing list
> > > > Mod_python at modpython.org
> > > > http://mailman.modpython.org/mailman/listinfo/mod_python
> > > >
> > > _______________________________________________
> > > Mod_python mailing list
> > > Mod_python at modpython.org
> > > http://mailman.modpython.org/mailman/listinfo/mod_python
> >--
> >Conrad Steenberg <conrad at hep.caltech.edu>
> 
> _______________________________________________
> Mod_python mailing list
> Mod_python at modpython.org
> http://mailman.modpython.org/mailman/listinfo/mod_python
-- 
Conrad Steenberg <conrad at hep.caltech.edu>



More information about the Mod_python mailing list