DDraceNetwork - developer

4af4f0f44cd921556932d01e4d0e57963e715eda90af5e8aa1a9de475d2edbc5 4af4f0f44441544104000000c1600400b40e0000080000003600000029000000 What happened here even

00:22

I io_write the top one, I get the bottom one when i io_read it

I was only writing the first 4 bytes

@heinrich5991 your blog could do with a <link> for the rss feed, then some feed readers can detect it automatically

07:32

If anyone hasn't seen it: https://heinrich5991.github.io/blog/blog/one-tick-unfreeze

The Anatomy of a One Tick Unfreeze

I saw the deepfly bind the other day:

def-

[ddnet/ddnet] Pull request opened: #1940 Update ddnet-libs version

Was forgotten in the libcurl linux removal I think?

bors[bot]

[ddnet/ddnet] Pull request closed: #1933 Fix two unintentional behavior changes of the stopper rewrite

bors[bot]

[ddnet/ddnet] Issue closed: #1931 Crashbug in CEntity::GetNearestAirPos()

08:26

bors[bot]

[ddnet/ddnet] Issue closed: #1797 Players should be considered idle when first entering server (or after map change), until their first movement

08:26

bors[bot]

[ddnet:master] 4 new commits

84f1542 Add a limit to GetNearestAirPos. Fix #1931 - Learath2 547f751 Mark players as afk initially. Fix #1797 - Learath2 5f0cc93 Update ddnet-libs version - def- aa79d81 Merge #1935 #1936 #1940 - bors[bot]

08:26

bors[bot]

[ddnet/ddnet] Pull request closed: #1940 Update ddnet-libs version

08:26

bors[bot]

[ddnet/ddnet] Pull request closed: #1936 Mark players as afk initially. Fix #1797

@deen so do we know why ddnet7 disappeared? and is it coming back?

it disappeared because foko deleted it

Wait there isn't any back-up/ bin where you can pull it back from the death?

13:06

I mean why has he authorisation to delete such a big file?

@HMH what was the issue with your queue implementation?

because I gave him the permissions, removed them now, will restore the repo later

do we have any proof he did it? because he told me he didnt.

@Learath2 as far as I remember it worked fine but it forces you to wrap everything into some pointertype like std::unique_ptr.

Hmm, why is it so complex?

Also, for my rpc stuff I implemented something simpler: https://github.com/H-M-H/ddnet/blob/rpc/src/base/tl/queue.h

H-M-H/ddnet

DDraceNetwork, a mod of Teeworlds. Contribute to H-M-H/ddnet development by creating an account on GitHub.

Wouldn't a lock and a queue be enough?

Probably because I wanted it to do too much at the time

One more question, what is the point of having a mutex and a cv?

@ChillerDragon yes github has an audit log

@HMH Oh actually I have one more question about your last queue, to push you need to acquire the lock, but if the consumer is stuck waiting on a condition variable, the queue will never be non empty, how does that work?

13:56

oh the wait unlocks doesn't it

The mutex is used for syncronization of the cv, e.g. cv.wait(lck) changes state of the cv and then releases the lock.

13:59

So, what do you want to do with a threadsafe queue ?

I was doing some work on sql (im not anymore since you are doing it)

13:59

Was just curious after that

14:02

TIL you can also do m_CV.wait(Lock, [this]{return m_Queue.empty()});

12pm

[ddnet/ddnet] Pull request opened: #1941 Fix IntersectCharacter

This makes hitting yourself when shooting through other tees work

Hitting yourself when shooting through other tees?

14:06

Wouldn't you just hit the other tee @jao?

with hit others disabled obviously

Wait, when hit others is disabled it should just go through the other tees, no?

14:07

What happens currently?

yes but if it then hits yourself it has no effect

14:08

with the fix it works like it wouldnt go through a tee

hm, I thought it should go through

it does