cfbolz changed the topic of #pypy to: #pypy PyPy, the flexible snake https://pypy.org | IRC logs: https://quodlibet.duckdns.org/irc/pypy/latest.log.html#irc-end and https://libera.irclog.whitequark.org/pypy | so many corner cases, so little time
Ninpo has quit [Ping timeout: 250 seconds]
Ninpo has joined #pypy
shodan45 has joined #pypy
otisolsen70 has joined #pypy
dmalcolm_ has joined #pypy
dmalcolm has quit [Ping timeout: 250 seconds]
slav0nic has joined #pypy
Ninpo has joined #pypy
Ninpo has quit [Changing host]
mattip has quit [Ping timeout: 240 seconds]
mattip has joined #pypy
Atque has quit [Quit: ...]
Atque has joined #pypy
<fijal> marmoute: how do I create an account?
<marmoute> fijal: you don't have to create an account, you use the sso of your existing foss.heptapod.net account
<fijal> when I click log in I get that error
<fijal> can we try to debug an actual issue and not the meta-issue?
<fijal> I can't essentially push to heptapod, which is a bit bad
Atque has quit [Quit: ...]
<mattip> cfbolz: could you tweet "We recently got a new aarch64 buildbot worker. Thanks to Huawei for sponsoring the VM and supporting PyPy"
<amazigh> I built an immutable binary search tree, see preliminary benchmarks at https://github.com/amirouche/lbst/pull/3
<amazigh> I am wondering how I can put it for good use, any idea?
<amazigh> It really shines with more than 1k key-value pairs.
otisolsen70 has quit [Ping timeout: 240 seconds]
<marmoute> amazigh: is it good on cpython too ?
<amazigh> The benchmark does the following: pick N random integers, stuff them into a dictionary, copy and sort as necessary, query every integers one by one without re-using a cursor, do the same with log-balanced search tree except there is no need for full copies all the time;
<amazigh> With CPython 3.9, starting with 1,000 key-value pairs it is as fast as copying.
<amazigh> With PyPy 3.7 that I downloaded a couple of weeks back, it is faster starting at 100 key-value pairs
<amazigh> I was thinking about building somekind of redis-clone or something like that.
<amazigh> get inspiration from https://github.com/yinqiwen/ardb/
<amazigh> marmoute: ^
<amazigh> I know nothing that requires than much items in dictionary, except databases.
<marmoute> okay, I suspect it the number of key necessary to start to pay off is too high for the usecae I have
<marmoute> we are using immutable dictionnary in Mercurial when agregating copy/rename information along the graph.
<marmoute> We do that for the Rust version.
<marmoute> using immutable objct in Rust
<marmoute> We do dict copying in Python
<amazigh> ok
<amazigh> It will remain on github with a sign saying: 'only useful for its hack-value'
Guest96 has joined #pypy
<cfbolz> marmoute: I still think it's sad that immutable dicts didn't make it into python
<cfbolz> (was discussed on python-dev a few times)
<cfbolz> there is a perfectly good implementation hidden in the contxtvars modules
Guest96 has quit [Quit: Textual IRC Client: www.textualapp.com]
<Corbin> Yeah, I'm still sore about frozendict too. At least we can say `frozenlist = tuple` and we have frozenset.
<marmoute> something that annoyes me is that making a dict/set frozen cannot be done in place
krono has quit [Read error: Connection reset by peer]
krono has joined #pypy
slav0nic has quit [Ping timeout: 252 seconds]