diff --git a/papi/utils.py b/papi/utils.py
new file mode 100644
index 0000000..fbc646e
--- /dev/null
+++ b/papi/utils.py
@@ -0,0 +1,144 @@
+import requests
+import sys
+from time import sleep
+from copy import deepcopy
+
+import math
+import itertools
+import datetime
+import csv
+import json
+import logging as logger
+
+
+def prepare(record):
+
+    newdict = deepcopy(record)
+    newdict["channels"] = dict(
+        sorted(channel2tuple(channel) for channel in record["channels"])
+    )
+    return newdict
+
+
+def channel2tuple(channel):
+    return "%s#%s" % (channel["id"], channel["name"]), channel
+
+
+def get_status(channels, key):
+    d = channels["channels"].get(key, {"status": "absent"})["status"]
+    return d
+
+
+def compare(channels, previous, current):
+    changes, states = [], []
+    for key in channels:
+        pstatus = get_status(previous, key)
+        cstatus = get_status(current, key)
+        state = key, pstatus, cstatus
+        if pstatus != cstatus:
+            changes.append(state)
+        states.append(state)
+
+    res = {}
+    if not changes:
+        return {}
+    disparus = [state for state in states if state[2] == "absent"]
+    if disparus:
+        res["disparus"] = disparus
+    apparus = [state for state in states if state[1] == "absent"]
+    if apparus:
+        res["apparus"] = apparus
+
+    res["changements"] = changes
+    res["etats"] = states
+    return res
+
+
+def list_channels(p, c):
+    all_channels = sorted(set((*p["channels"], *c["channels"])))
+    return all_channels
+
+
+def main():
+    loopcount = itertools.count().__next__
+
+    login()
+    previous = load_or_fetch()
+
+    historique = []
+    while loopcount() < maxloop:
+        current = fetch().json()
+        forward(current)
+        with open("last.json", "w") as f:
+            json.dump(current, f)
+
+        with open(raw_filename(), "a") as f:
+            json.dump(current, f)
+
+        current, previous = prepare(current), prepare(previous)
+        all_channels = sorted(set((*previous["channels"], *current["channels"])))
+
+        savelog2csv(current)
+
+        diff = compare(all_channels, previous, current)
+
+        savediff(date=current["date"], diff=diff)
+
+        if diff:
+            print("**********")
+            print(diff["changements"])
+            print("!!!!!!!!!!")
+            historique.append(diff)
+            previous = current
+        sleep(20)
+    return historique
+
+
+def make_id_key(channel, keys=None, sep="#", tuple_=False):
+    """
+    This takes out the concatenation of keys, value to use is as a pair.
+
+    >>> sample = {'id': 6, 'name': 'foo'}
+    >>> make_id_key(sample)
+    {'6#foo': {'id': 6, 'name': 'foo'}}
+    """
+    if not keys:
+        keys = ["id", "name"]
+    kvalue = sep.join(str(channel[k]) for k in keys)
+    if tuple_:
+        return kvalue, channel
+    return {kvalue: channel}
+
+
+def raw_filename():
+    return "raw_" + str(datetime.date.today()).replace("-", "_") + ".json"
+
+
+def log_filename():
+    return "log_" + str(datetime.date.today()).replace("-", "_") + ".csv"
+
+
+def diff_filename():
+    return "diff_" + str(datetime.date.today()).replace("-", "_") + ".csv"
+
+
+def savelog2csv(alert, *, filename_f=log_filename):
+    keys = "date id name type status".split(" ")
+    with open(filename_f(), "a") as f:
+        writer = csv.DictWriter(f, keys)
+        if f.tell() == 0:
+            writer.writeheader()
+        for a in status2list(alert):
+            writer.writerow(a)
+
+
+def savediff(date, diff, *, filename=diff_filename):
+    keys = "date name before after".split(" ")
+    with open(filename(), "a") as f:
+        writer = csv.DictWriter(f, keys)
+        if f.tell() == 0:
+            writer.writeheader()
+        for d in diff:
+            data = {"date": date}
+            data.update(zip(("name", "before", "after"), d))
+            writer.writerow(data)