@Andy Perhaps, but I hate having to wait 30 seconds for the CI to build just so I can pull my commit. Making it faster while still retaining the tests would be nice.
I don't see 30 seconds of build time on travis. I see the 5 seconds you mention for the cache, but everything else is time amounts of time. Are you sure the 30 seconds isn't related to the delay between GitHub and Travis realizing there is a new commit to build against?
10ms probably isn't worth half an hour of time, but what I learn from it is more valuable than what I would learn from starting video games or drugs or whatever else :P
The following extra packages will be installed:
libc-bin libc-dev-bin libc6-dev
Suggested packages:
glibc-doc
The following packages will be upgraded:
libc-bin libc-dev-bin libc6 libc6-dev
4 upgraded, 0 newly installed, 0 to remove and 226 not upgraded.
Need to get 8,840 kB of archives.
After this operation, 14.3 kB disk space will be freed.
Get:1 us.archive.ubuntu.com/ubuntu precise-updates/main libc6-dev amd64 2.15-0ubuntu10.15 [2,943 kB]
Get:2 us.archive.ubuntu.com/ubuntu precise-updates/main libc-dev-bin amd64 2.15-0ubuntu10.15 [84.7 kB]
I'll admit I don't know a lot about Travis's infrastructure, but do they ever roll those into the images so it's already there and this is just temporary? Or is this a permanent thing?
The clone and submodule initialize add another chunk of time too.
@Undo Push access would be useful for me, I am usually online at the moment when no other contributors are online, so that is useful for quickly blacklisting a website
File "/media/sda2/SmokeDetector/ChatExchange/chatexchange/client.py", line 125, in get_me
assert self._br.user_id is not None
File "/media/sda2/SmokeDetector/ChatExchange/chatexchange/_utils.py", line 78, in __get__
method()
File "/media/sda2/SmokeDetector/ChatExchange/chatexchange/browser.py", line 235, in _update_chat_fkey_and_user
self._load_user(favorite_soup)
File "/media/sda2/SmokeDetector/ChatExchange/chatexchange/browser.py", line 218, in _load_user
user_id, user_name = self.user_id_and_name_from_link(user_link_soup)
Recovered from ValueError: No JSON object could be decoded
UnicodeEncodeError: 'ascii' codec can't encode character u'\u0ca0' in position 0: ordinal not in range(128)
2016-05-27 19:46:01.852236 UTC
File "/media/sda2/SmokeDetector/excepthook.py", line 46, in run_with_except_hook
run_old(*args, **kw)
File "/usr/lib/python2.7/threading.py", line 505, in run
self.__target(*self.__args, **self.__kwargs)
File "/media/sda2/SmokeDetector/ChatExchange/chatexchange/browser.py", line 696, in _runner
self.on_activity(json.loads(a))
File "/media/sda2/SmokeDetector/ChatExchange/chatexchange/rooms.py", line 81, in on_activity
It tried to parse formats such as sd 2- into formats formats where every chat message is on a separate line, for the ease of checking the individual commands later, but gave an error when encountering a unicode symbol on those lines. Thats why it crashed
> Url in title, blacklisted website in body, blacklisted website in title ---------- Title - Position 13-31: homeworkhelponline Body - Position 13-31: homeworkhelponline Title - Position 1-35: https://www.homeworkhelponline.net
> Url in title, blacklisted website in body, blacklisted website in title ---------- Title - Position 8-20: vbuycars.com Body - Position 8-20: vbuycars.com Title - Position 1-20: http://vbuycars.com
@sandwich We could add some code that the manually reported reason is always the first (and make it optionally bold), so people see its a manual report, and after that, the normal reasons should go
So it could work like: a post is reported, run the usual spam-checking process and add "manually reported" to the reasons, then display in the chat normally.
@sandwich ProTip: I've been on flights (and a few airports) where you don't actually have to pay - they just block DNS. So if you know the IP address of a server you can use as a proxy...
@Undo That's because there pretty much is no multithreading. Currently, since we're using threading, multithreading equals "if one thread is doing I/O another can do computations"
We should probably switch over to multiprocessing sometime but I didn't want to do that unless I had a decent amount of time to fix stuff. :P
