< roasbeef>
so i've noticed that at times the DNS seeds (when using filtering) will return results for a particualr service bit, but once connected out, the peers advertise a distinct set (bit not actually there)
< roasbeef>
is this a result of the polling interval? or some filtering bug in the seed itself?
< roasbeef>
an example is asking for x49 (full node, segwit, bip 158), but then getting (full node, segwit, partial block history)
< TD-Linux>
sipa, your miniscript page doesn't explain the difference between the policy language and miniscript
< TD-Linux>
might be nice to add a little blurb about it before you hit the policy compiler box
< sipa>
yeah good point
< harding>
luke-jr: do you still have a git repository with your seeder index results over time? If so, maybe send roasbeef the URL so he can look up the history of the nodes with inconsistent service bits.
< luke-jr>
harding: yes, but GitLab limited its size so there's no way to get the last few months
< luke-jr>
it's 13 GB now
< bitcoin-git>
[bitcoin] fanquake opened pull request #16665: scripts: move update-translations.py to maintainer-tools repo (master...move_maintainer_scripts) https://github.com/bitcoin/bitcoin/pull/16665
< fanquake>
Had more spam PRs than legit PRs over the past few days
< tryphe>
fanquake, sure seems like it. i wonder if there's an easy way to make the git bot just cancel commits that container the default github PR message.
< tryphe>
s/container/contain*
< bitcoin-git>
[bitcoin] fanquake opened pull request #16667: build: remove mingw linker workaround from win gitian descriptor (master...remove_mingw_linker_workaround) https://github.com/bitcoin/bitcoin/pull/16667
< bitcoin-git>
[bitcoin] fanquake opened pull request #16669: [WIP] build: use new fork of osslsigncode for windows gitian signing (master...osslsigncode_2_0) https://github.com/bitcoin/bitcoin/pull/16669
< luke-jr>
instagibbs: if say someone had a PR that added a new TxMempoolInfo construction with the current fields, they'd end up with something very broken silently compiled?
< luke-jr>
or the opposite, someone based the new construction on this PR, and it got backported sloppily
< instagibbs>
ok is there something that would make you feel better? The current logic is broken.
< instagibbs>
wouldn't the increased number of fields make it impossible to compile if someone was using the old version?
< luke-jr>
at least in C, zeros are implicitly added on the end :/
< instagibbs>
CFeeRate field is also gone, so typing should also save the implementer?>
< luke-jr>
eh? it isn't gone
< luke-jr>
you just renamed it
< instagibbs>
(yuck re:c)
< luke-jr>
is it supposed to be a CAmount?
< instagibbs>
wat, let me re-read my code
< instagibbs>
maybe o_0
< luke-jr>
(in which case, I'm not sure how implicit conversions work here)
< luke-jr>
using an explicit constructor would probably solve at least backports
< instagibbs>
yeah that's a bug(well it "works" but is the wrong type)
< kallewoof>
GW-BASIC would hang my dad's 286 if I ever accidentally hit shift 4 instead of alt-gr 4 to get $. Sad times writing code and have to restart the machine all the time.
< kallewoof>
I can't believe anybody else knows about GW-BASIC tho. I honestly thought it was my own personal secret. Or a bad nightmare. Or both.
< sipa>
kallewoof: i learned to program in it
< kallewoof>
Me too! I guess it was more popular than I realized.
< luke-jr>
were there other options than GW-BASIC back then? :p
< luke-jr>
AFAIK QBASIC only came a few years later
< kallewoof>
luke-jr: Without internet and other ways to communicate, I'm afraid my knowledge on what was available was very limited (I was also single-digit years old). ;P
< luke-jr>
same
< luke-jr>
CompuServ and such came later
< luke-jr>
in the meantime, I had to write BASIC programs to simulate basically a BBS
< kallewoof>
I think I tried to write a bot that would talk back. I had great plans for my future AI overlords back then.
< luke-jr>
XD
< luke-jr>
kallewoof: better than Dr. Sbaitso?
< kallewoof>
luke-jr: Haha, most likely not
< luke-jr>
HELLO KALLE. MY NAME IS DOCTOR SBAITSO. I AM HERE TO HELP YOU. SAY WHATEVER IS ON YOUR MIND FREELY. OUR CONVERSATION WILL BE KEPT IN STRICT CONFIDENCE. MEMORY CONTENTS WILL BE WIPED OFF AFTER YOU LEAVE. SO, TELL ME ABOUT YOUR PROBLEMS.
< BlueMatt>
aj: lol nahhhhh anything less than 420x100 is a joke of a terminal :p
< sipa>
190x58 here
< andytoshi>
254x76 ... but my matt-style terminal (which slows my reading rate by well over 50%, the font is so small) is 318x98
< andytoshi>
but i still try to keep all my lines under 80.
< bitcoin-git>
[bitcoin] MarcoFalke opened pull request #16670: util: Add Join helper to join a list of strings (master...1908-utilStringJoin) https://github.com/bitcoin/bitcoin/pull/16670
< luke-jr>
pretty sure we already had one of those?