#9154: Huge improvements in templates rendering
---------------------------------------------+------------------------------
Reporter: msaelices | Owner: msaelices
Status: new | Milestone:
Component: Template system | Version: SVN
Resolution: | Keywords:
Stage: Design decision needed | Has_patch: 1
Needs_docs: 0 | Needs_tests: 0
Needs_better_patch: 0 |
---------------------------------------------+------------------------------
Changes (by mtredinnick):
* stage: Unreviewed => Design decision needed
Comment:
It simply '''must''' be possible for this new behaviour ot be disabled (in
fact, it should be disabled by default). You're proposing quite a large
change in behaviour here: whenever a developer changes a template to fix a
typo or something, the web server must be restarted (in effect; whatever
is needed to cause the Django process to restart). That is quite a change
in operational practices for a lot of places where designers can currently
update templates and have them rolled out without a restart being
required.
I'm also not amazingly happy with effectively adding a new cache here that
tries to hold things in memory. Why not using Django's existing caching
framework and, if you're going to do that, why not just supply a caching
wrapperaround all existing loaders? By storing stuff in memory you are
increasing Django's in-memory footprint (there could be hundreds or
thousands of templates involved) and there's no way to do resource
management -- as opposed to using a proper cache which manages the amount
of memory it will use. The amount of memory used by the urlresolvers is
tiny compared to templates and we cache only as last resource because the
overhead of the parsing was becoming unacceptably slow when it was done on
every request. Template parsing is proportionally much faster (since you
only have to parse the the template you requested, not potentially every
template in the system as is the case for URL reversing).
This whole ticket feels like something designed for a small-scale
situation that simply won't scale well. I'm still -1 on the proposed
implementation.
--
Ticket URL: <http://code.djangoproject.com/ticket/9154#comment:16>
Django <http://code.djangoproject.com/>
The Web framework for perfectionists with deadlines.
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"Django updates" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at
http://groups.google.com/group/django-updates?hl=en
-~----------~----~----~----~------~----~------~--~---
