diff --git a/netdb.i2p2/fixedapp.py b/netdb.i2p2/fixedapp.py
new file mode 100644
index 00000000..5d11dcb2
--- /dev/null
+++ b/netdb.i2p2/fixedapp.py
@@ -0,0 +1,203 @@
+#
+# Required software:
+# sqlite3 3.6.23.1 or newer
+# http://www.sqlite.org/
+# unixodbc
+# http://www.unixodbc.org/
+# pyodbc-3.0.6 or newer
+# http://www.ch-werner.de/sqliteodbc/
+# werkzeug
+# http://werkzeug.pocoo.org/
+# python 2.6 or 2.7
+# http://python.org/
+#
+
+
+# Modify as needed, or use a symlink.
+netdbdir = 'netdb'
+database = 'Driver=SQLite;DATABASE=I2PnetDb'
+
+
+from werkzeug import BaseRequest, BaseResponse, ETagResponseMixin, escape, SharedDataMiddleware
+from werkzeug.exceptions import HTTPException
+import os
+from time import time, gmtime
+import calendar
+#import time
+from random import choice
+import pyodbc
+
+
+#
+# Not needed?
+#
+#import sha
+
+
+# db format
+#
+# table client ipaddress, time of access, file list index
+#
+# table html index, time of creation, html
+
+
+#
+# Init database if it does not exist.
+#
+
+def checkdatabase():
+ cnxn = pyodbc.connect(database)
+ cur = cnxn.cursor()
+
+ try:
+ cur.execute("select * from client")
+ except pyodbc.Error:
+ #cur.execute("create table peer (ip string, time string, index string)")
+ print ("Creating new table 'client'")
+ cur.execute("create table client (ip string, whn float)")
+ cnxn.commit()
+ try:
+ cur.execute("select * from entry")
+ except pyodbc.Error:
+ #cur.execute("create table peer (ip string, time string, index string)")
+ print ("Creating new table 'entry'")
+ cur.execute("create table entry (whn float, wht string)")
+ cnxn.commit()
+ cnxn.close()
+
+
+
+
+# try:
+# cur.execute("select * from html")
+# except pyodbc.Error:
+# cur.execute("create table client (index, time, html)")
+
+
+# cur.execute("insert into foo values(1, 'Me')")
+# cnxn.commit()
+# cur.execute("select * from foo")
+# cnxn.close()
+
+class Request(BaseRequest):
+ """Useful subclass of the default request that knows how to build urls."""
+
+ def __init__(self, environ):
+ BaseRequest.__init__(self, environ)
+
+
+class Response(BaseResponse, ETagResponseMixin):
+ """Subclass of base response that has a default mimetype of text/html."""
+ default_mimetype = 'text/html'
+
+def app(environ, start_response):
+ """The WSGI application that connects all together."""
+ req = Request(environ)
+ remote = req.remote_addr
+ now = time()
+ old = now - 3600
+ oldest = now - 7200
+ t = gmtime(now)
+ nowtag = calendar.timegm((t.tm_year, t.tm_mon, t.tm_mday, t.tm_hour, 0, 0, t.tm_wday, t.tm_yday, t.tm_isdst))
+
+ cnxn = pyodbc.connect(database)
+ cur = cnxn.cursor()
+ #
+ # Prune database from peers that are over 1 hour old
+ #
+ cur.execute("delete from client where whn < ?", old)
+ cnxn.commit()
+ #
+ # Prune database from entries that are over 2 hours old
+ #
+ cur.execute("delete from entry where whn < ?", oldest)
+ cnxn.commit()
+
+ #
+ # Get peer info
+ #
+ cur.execute("select * from client where ip = ?", remote)
+ info = cur.fetchall()
+ # page
+ path = req.path[1:]
+
+
+ if path == '':
+ page = u'NetDB'
+
+ if len(info) == 0:
+ # tag the ip as new
+ cur.execute("insert into client values (?, ?)", (remote, now))
+ # see if we have a list already, and use that
+ # generate links
+ entries = os.listdir(netdbdir)
+ if len(entries) > 150:
+ # select some randomly
+ new = []
+ for i in range(100):
+ while True:
+ sel = choice(entries)
+ if not sel.startswith('routerInfo-'):
+ continue
+ if sel not in new:
+ new.append(sel)
+ cur.execute("insert into entry values (?, ?)", (nowtag, sel))
+ break
+ entries = new
+ else:
+ # use old list based on date in database, i.e. sends the same as before.
+ junk, last = info[0]
+ t = gmtime(last)
+ oldtag = calendar.timegm((t.tm_year, t.tm_mon, t.tm_mday, t.tm_hour, 0, 0, t.tm_wday, t.tm_yday, t.tm_isdst))
+ cur.execute("select * from entry where whn = ?", oldtag)
+ stuff = cur.fetchall()
+ entries = []
+ for junk,sel in stuff:
+ entries.append(sel)
+
+ res = ''
+ for entry in entries:
+ if not entry.startswith('routerInfo-'):
+ continue
+ res += '%s' % (entry, entry)
+ resp = Response(page % res, mimetype='text/html')
+ resp.add_etag()
+
+ elif path == 'robots.txt':
+ dat = u"User-agent: *\nDisallow: /routerInfo-*.dat$\n"
+ resp = Response(dat, mimetype='text/plain')
+ resp.add_etag()
+
+ else:
+ if len(info) == 0:
+ # 404 This is to prevent cross seed site scrapes
+ resp = Response("Hi! Having fun?", status=403, mimetype='text/plain')
+ else:
+ #
+ # zap badness from path
+ #
+ chkpath = path.replace("'","").replace("/","")
+ # check to see if we have this entry in the database at all.
+ cur.execute("select * from entry where wht = ?", chkpath)
+ chk = cur.fetchall()
+ if len(chk) == 0:
+ resp = Response("Do you not have anything better to do?", status=400, mimetype='text/plain')
+ else:
+ try:
+ # load file
+ f = open(os.path.join(netdbdir, path), 'rb')
+ resp = Response(f.read(), mimetype='application/octet-stream')
+ f.close()
+ resp.add_etag()
+ except IOError:
+ # 404
+ resp = Response("Owch! Could not find file!", status=404, mimetype='text/plain')
+ cnxn.commit()
+ cnxn.close()
+ return resp(environ, start_response)
+
+checkdatabase()
+
+if __name__ == '__main__':
+ from werkzeug import run_simple
+ run_simple('localhost', 5007, app)