« first day (32 days earlier)      last day (42 days later) » 

1:16 AM
sorry, I have been very busy with finals soon
I'll try to get something on GH this weekend
 
@Downgoat Our finals come after winter break
 
 
2 hours later…
3:29 AM
@Pavel ours used to be like this but then admin thought that gave us too much time to study
 
 
2 hours later…
5:31 AM
How do you think PPCG v2 will deal with those things?
1. FGITW - faster answers get more +1s.
2. The highest upvoted answer will more likely to get +1.
3. Some answers are invalid with or without OP's knowledge. It definitely takes time to validate answers. Automatic testing are also mentioned previously, but it doesn't work well.
4. Sandbox questions are hardly noticed.
5. It's annoying that some comments are left unnoticed by OP, while they are important; and the answer still get +1 (in case of invalid answer) / the question still get answers (in case of unclear question).
 
1. That's a problem with any site with live voting. There are [ways to reduce the impact](https://redditblog.com/2009/10/15/reddits-new-comment-sorting-system/), but it will never go away.
2. That's sort of a problem, and like FGITW, it's not really solveable. I say "sort of", because if an answer is good and it continues to get +1s, that's not a bad thing.
3. I'm not sure. That's something that has been debated in here, and will continue to be debated.
4. Sandbox will be its own review queue, rather than a massive post on meta.
 
 
2 hours later…
7:48 AM
So we could have a review queue for challenges (the Sandbox) and a review queue for answers?
 
Yep
Answers wouldn't automatically go in the queue - just answers from new users (like the SE queue), and answers that have been flagged as possibly invalid (as a replacement for the VLQ queue)
The goal is to provide a way for experienced users to review possibly-invalid answers and remove them, other than the current method of 1. happen to see it, 2. comment, 3a. downvote into oblivion until it can be delete-voted, or 3b. raise a custom mod flag and make the mods have to decide whether or not it is valid
I want to empower the community by giving them more tools to be proactive about moderation in cases where having a mod step in isn't necessary
 
8:52 AM
Man, I am really glad that I decided to give each answer a unique id, rather than allowing each post to have its own set of answer ids (and thus requiring both the post id and the answer id to select a specific answer). It's making everything so much simpler.
 
9:43 AM
Are we repeating what SE did with having posts and answers in the same table?
 
Nope
Posts and answers are their own tables
Which is 4357984% smarter than combining them together, especially given that posts and answers are much more different here than on SE
@Downgoat PPCGv2 is currently untestable on Windows due to the n dependency: gist.github.com/Mego/77070d1eb83028dd4d343b4ba5fd9e01
Is there a way around that?
 
 
2 hours later…
12:19 PM
Messy tech graph I made for Sherlock. @Downgoat did I miss anything?
 
 
1 hour later…
1:44 PM
Doing unit tests which require a Flask session is only moderately hellish
 
2:27 PM
@Mego n shouldnt be a dependency not sure why that’s there
 
Nov 22 at 1:07, by Downgoat
@Mego sudo npm i -g n
 
Yeah n is just a utility to help install nodejs itself
 
3:09 PM
Note to self: use unicodetext instead of LONGTEXT
 
 
2 hours later…
4:57 PM
Does challenge posting work with noscript?
 
Ugh, login doesn't work for me
 
 
4 hours later…
9:22 PM
I just configured PostgreSQL to use with a personal project of mine. It was wonderful and amazing and everything worked on the first try. 10/10 would use over MySQL.
 

« first day (32 days earlier)      last day (42 days later) »