Switch to waitress wsgi server by default in rhodecode.

- waitress is fast, handles chunked encoding correctly (needed for git)
- waitress works on linux and windows
- it's now default because it can handle git properly out of the box.

development.ini

@@ -30,15 +30,16 b' pdebug = false'
 
 [server:main]
 ##nr of threads to spawn
-threadpool_workers = 5
+#threadpool_workers = 5
 
 ##max request before thread respawn
-threadpool_max_requests = 10
+#threadpool_max_requests = 10
 
 ##option to use threads of process
-use_threadpool = true
+#use_threadpool = true
 
-use = egg:Paste#http
+#use = egg:Paste#http
+use = egg:waitress#main
 host = 0.0.0.0
 port = 5000
 
@@ -47,7 +48,7 b' use = egg:rhodecode'
 full_stack = true
 static_files = true
 # Optional Languages
-# en, fr, pt_BR, zh_CN, zh_TW
+# en, fr, ja, pt_BR, zh_CN, zh_TW
 lang = en
 cache_dir = %(here)s/data
 index_dir = %(here)s/data/index

production.ini

@@ -30,15 +30,16 b' pdebug = false'
 
 [server:main]
 ##nr of threads to spawn
-threadpool_workers = 5
+#threadpool_workers = 5
 
 ##max request before thread respawn
-threadpool_max_requests = 10
+#threadpool_max_requests = 10
 
 ##option to use threads of process
-use_threadpool = true
+#use_threadpool = true
 
-use = egg:Paste#http
+#use = egg:Paste#http
+use = egg:waitress#main
 host = 127.0.0.1
 port = 8001
 
@@ -47,7 +48,7 b' use = egg:rhodecode'
 full_stack = true
 static_files = true
 # Optional Languages
-# en, fr, pt_BR, zh_CN, zh_TW
+# en, fr, ja, pt_BR, zh_CN, zh_TW
 lang = en
 cache_dir = %(here)s/data
 index_dir = %(here)s/data/index

rhodecode/config/deployment.ini_tmpl

@@ -30,15 +30,16 b' pdebug = false'
 
 [server:main]
 ##nr of threads to spawn
-threadpool_workers = 5
+#threadpool_workers = 5
 
 ##max request before thread respawn
-threadpool_max_requests = 10
+#threadpool_max_requests = 10
 
 ##option to use threads of process
-use_threadpool = true
+#use_threadpool = true
 
-use = egg:Paste#http
+#use = egg:Paste#http
+use = egg:waitress#main
 host = 127.0.0.1
 port = 5000
 

rhodecode/model/pull_request.py

@@ -25,19 +25,17 b''
 
 import logging
 import binascii
+
 from pylons.i18n.translation import _
 
+from rhodecode.model.meta import Session
 from rhodecode.lib import helpers as h
 from rhodecode.model import BaseModel
-from rhodecode.model.db import PullRequest, PullRequestReviewers, Notification,\
-    ChangesetStatus
+from rhodecode.model.db import PullRequest, PullRequestReviewers, Notification
 from rhodecode.model.notification import NotificationModel
 from rhodecode.lib.utils2 import safe_unicode
 
 from rhodecode.lib.vcs.utils.hgcompat import discovery
-from rhodecode.model.changeset_status import ChangesetStatusModel
-from rhodecode.model.comment import ChangesetCommentsModel
-from rhodecode.model.meta import Session
 
 log = logging.getLogger(__name__)
 

setup.py

@@ -33,6 +33,7 b" callback = lambda V: ('.'.join(map(str, "
 is_windows = __platform__ in _get_meta_var('PLATFORM_WIN', _metadata)
 
 requirements = [
+    "waitress==0.8.1",
     "Pylons==1.0.0",
     "Beaker==1.6.3",
     "WebHelpers==1.3",
@@ -49,7 +50,7 b' requirements = ['
     "markdown==2.1.1",
     "docutils==0.8.1",
     "simplejson==2.5.2",
-    "mock"
+    "mock",
 ]
 
 if sys.version_info < (2, 6):

test.ini

@@ -30,15 +30,16 b' pdebug = false'
 
 [server:main]
 ##nr of threads to spawn
-threadpool_workers = 5
+#threadpool_workers = 5
 
 ##max request before thread respawn
-threadpool_max_requests = 2
+#threadpool_max_requests = 2
 
 ##option to use threads of process
-use_threadpool = true
+#use_threadpool = true
 
-use = egg:Paste#http
+#use = egg:Paste#http
+use = egg:waitress#main
 host = 127.0.0.1
 port = 5000