stopstalk / stopstalk-deployment Goto Github PK

Things to consider -

• It will have to be hardcoded, since this was never the use-case and the current schema won't support it
• Most probably Github API does not support retrieving ALL the commits till date for a specific user

Python "iterators" are a type of "lazy-evaluation". They 'feed' data one step at time; traditional python loops create the entrire set of data in memory before looping.

The traditional use of select is:

for row in db(db.table.id>0).select():
    rtn = row

but for large numbers of rows, using an iterator-based alternative has dramatically lower memory use:

for row in db(db.table.id>0).iterselect():
    rtn = row

Testing shows this is around 10% faster as well, even on machines with large RAM.

Same dot as shown in Leaderboard page for -

• Any Submissions page
• Filter page
• Notification page

This should be added only if the query matches(example: enter some text in user name that is not valid we will get all the submissions of logged-in user)
Error -> here

Maybe all we need to do is strip the ending comma

Trending problems have different number of total submissions than the actual

We need HTML emails for the following purposes -

• On registration before complete website launch
• Welcoming the user when the website is up
• On receiving a friend request
• Acceptance of a friend request
• Rejectance of a friend request
• Unfriending a friend

We can show the custom friends in the search and an option to add that custom user to yours 😄

If this then redirect back to /user/profile/same_user -

If this then redirect back to /default/search -

Follow up on - https://github.com/stopstalk/stopstalk/issues/18
This idea works fine for deployment since the submissions are retrieved only after a day!
Idea -

• Create a database view for submissions of a user
• Drop this view and create new view only when
  • Refreshing entire database
  • A friend is added
  • A friend is unfriended

Need to look into how web2py supports database views.

• What page should it be?
• Navigation bar?
• Where the results shall be shown?

Saves a lot of extra space! ;)

http://www.codeforces.com/problemset/problem/650/B --> problemset/problem/650/B

How did I miss it! ;)

More than 0 submission in atleast 2 profile sites

authentic flag can take the following values -

• True - authentic user
• False - not updated / not authentic

Also for problem page!

Questions -

• Does topcoder have api?
• Does topcoder actually has public profile pages?
• Say above are true, how do we retrieve only the topcoder submissions given we have last_retrieved already updated for current users?

Waiting for CodeChef API :(

Since the submission table is growing considerably - We need to come up with an efficient script to update the trending problems. Also trending problems for friends has to be dealt separately to optimize page load time of trending page.

Instead of storing the handles and sites in the auth_user table have a table having site handles and site names and then make id of this table as the foreign key for auth_user table.

• Advantages -
  • Retrieve(update) submissions of 1 particular handle only once.
  • Lot of submissions that are duplicate can be saved
• Disadvantages
  • Migration is dangerous
  • It will take a bit of time to change the existing code(Surely more than a bit ;) )

Also make a problem table(extend current problem_tags table) and just include problem_id in a submission.