< jnewbery>
It's a bit pedantic - probably no-one is running pre-0.8 anymore. I have a slight preference to include the upgrade advice, but won't be too upset if other people vote to remove it.
< sipa>
oh, it was actually removed!
< sipa>
ok
< sipa>
jnewbery: i've also changed all v0.15 notation into 'version' or 'Bitcoin Core'
< jnewbery>
I saw. Looks good to me.
< Emcy_>
'On Windows, do not forget to uninstall all earlier versions of the Bitcoin client first.' Is this really *necessary*. No one ever does this
< Emcy_>
over the top installs happen
< luke-jr>
Emcy_: I have no idea, but we said that in older versions
< luke-jr>
Emcy_: if you can confirm it isn't needed, maybe it can be removed
< Emcy_>
it seems to be one of those fluff recommendations that windows software makes, like 'reboot your machine after install' as if anyone really does that.
< Emcy_>
If its just copying new files over the top i dont see why its necessary. Ill try it when 0.15 comes out
< theymos>
I've never given the advice to uninstall first. In fact, that could conceivably cause problems because uninstalling sometimes creates a one-time run-at-boot task which removes some files that the uninstaller couldn't immediately remove.
< theymos>
(Speaking generally here -- I've never looked at the Bitcoin Core installer.)
< Emcy_>
thats true
< Emcy_>
tahts what it says in those draft rls notes though, twice in fact and the second time makes it sound like not-boilerplate
< theymos>
The uninstall note must've been added recently. Maybe there's some reason for it, or maybe it was added in error.
< gmaxwell>
I've always told people to overinstall.
< gmaxwell>
I would be kinda worried about someone managing to delete their chain or wallet if I instructed them to remove!
< gmaxwell>
might be useful to investigate the history to see if there was a reason or if it was a copyediting error.
< gmaxwell>
generally I'm super scared about telling users to remove or delete anything; lest the next thing I see from them was "okay, I used the system recovery DVD"
< Emcy_>
lol
< luke-jr>
Emcy_: when 0.15 comes out, it will be too late to change..
< luke-jr>
theymos: I copied the uninstall bit from 0.8 or 0.9 IIRC
< luke-jr>
where are we at with a 0.14.3 with BIP148? I'm going to miss next meeting most likely again :/
< BitMEX_Wally>
There is a chainsplit in testnet and my 14.1 node is refusing to follow the longest chain after I ran invalidateblock on the short chain
< BitMEX_Wally>
getchaintips shows the longest chain as 'headers-only'
< gmaxwell>
it's really common for the testnet chain with the most work to have fewer blocks FWIW.
< BitMEX_Wally>
So how do I get my node to download those blocks? It is currently stuck on 1155805 / 00000000000005757375cdb666a319cf6758d54af7a19da42a96a0551f342d6b and not updating
< BitMEX_Wally>
This chain has the most work as far as I can tell
< gmaxwell>
000000000000066cd029ad78432f1c8d30e24459e407cc8f834a4b251bd83640 is in my best chain here...
< gmaxwell>
BitMEX_Wally: so now you need to look in your debug.log to see why 000000000000066cd029ad78432f1c8d30e24459e407cc8f834a4b251bd83640 was rejected as invalid.
< BitMEX_Wally>
1155846 / 00000000c164f6ac32ee23d104a81d7308d23b7a54745b4d613442c8a6c6ee1e has "chainwork": "00000000000000000000000000000000000000000000002872d1f10daeee43fa",
< gmaxwell>
look at debug.log and search for 000000000000066cd029ad78432f1c8d30e24459e407cc8f834a4b251bd83640
< BitMEX_Wally>
I marked 000000000000066cd029ad78432f1c8d30e24459e407cc8f834a4b251bd83640 as invalid because the other chain had more work
<@sipa>
you can undo invalidateblock using reconsiderblock
< BitMEX_Wally>
Yeah, but then I would be on chain 1155831 again, which is not most work?
< gmaxwell>
my nodes have never seen 00000000c164f6ac32ee23d104a81d7308d23b7a54745b4d613442c8a6c6ee1e
< bitcoin-git>
[bitcoin] theuni opened pull request #10971: build: fix missing warnings and sse42 in depends builds (master...fix-config-override) https://github.com/bitcoin/bitcoin/pull/10971
< bitcoin-git>
[bitcoin] practicalswift opened pull request #10972: Check return value of addr.GetKeyID(keyid) on custom change address change (master...GetKeyID-assertion) https://github.com/bitcoin/bitcoin/pull/10972
< earlz>
Has anyone successfully cross-compiled Bitcoin Core for a raspberry pi? If so, is that process documented somewhere?
< earlz>
I saw one tutorial on how to compile it on-device, but I'd rather spend less than 1 day compiling
< sipa>
earlz: more, there are release binaries for ARM