[Complete] Testing Feathercoin - 0.8.5.0 \* Released
-
[quote name=“wrapper0feather” post=“35521” timestamp=“1385075276”]
Hi Ghostlander, Just downloaded and tried to compile the new Pull. Build failed./home/gitwrap/Feathercoin-0.8.5/src/main.cpp:2331: warning: format ‘%u’ expects argument of type ‘unsigned int’, but argument 2 has type ‘int64 {aka long long int}’ [-Wformat=]
I’ll look again later…
[/quote]Well, that’s a trivial fix. Replace %u with %ld. What version of G++ do you use (32-bit/64-bit)?
-
Re: Ghostlander/Feathercoin0.8.5.0 Dev Fork.
It’s the same set-up as successful 0.8.5.0 build using Qt Creator 2.7.1 Based on Qt 5.0.2 (64 bit)
x64 13.10 KUbuntu.I’ve tried installing libstdc++5, still failing to build.
G++ 4.8.1
-
Don’t use Qt Creator. Use [i]qmake -makefile feathercoin-qt.pro[/i] and [i]make -f Makefile.Release[/i] or just [i]make[/i] if there is a single Makefile. It’s good to build the daemon first to be sure your tools have no issues with the base code.
BTW, the previous message was a warning, not an error. If the build fails, there must be something else with your configuration like unresolved dependencies.
-
Hi Ghostlander,
It just finished building, I’ll test it now. It is just a warning.
Looks like I stopped it too early last time, it took quite a lot longer to compile.
I’m using QT-Creator - so I can see the code, and build up to do some debugging.
I’d be interested to do the warning fix : how do I change the parameter definition?
CBlockIndex* pcheckpoint = Checkpoints::GetLastCheckpoint(mapBlockIndex);
if (pcheckpoint && pblock->hashPrevBlock != hashBestChain)
{if((pblock->GetBlockTime() - pcheckpoint->nTime) < 0) {
if(pfrom) pfrom->Misbehaving(100);
return error(“ProcessBlock() : block has a time stamp of %u before the last checkpoint of %u”, pblock->GetBlockTime(), pcheckpoint->nTime);
} -
[quote name=“ghostlander” post=“35543” timestamp=“1385084604”]
Don’t use Qt Creator. Use [i]qmake -makefile feathercoin-qt.pro[/i] and [i]make -f Makefile.Release[/i] or just [i]make[/i] if there is a single Makefile. It’s good to build the daemon first to be sure your tools have no issues with the base code.
[/quote]I can understand, just about, what you mean. I’ll try that out and add it to my “set up a test / compiled system in a Virtualbox thread”. I’m trying to widen the potential test base, so we’re probably better to let some testing stew on the forum before reporting it as an Github issue.
I knew it was a warning, but the compilation taking longer made my think it had failed. Sorry about that. I always like warnings fixed tho’. ++I was expecting there to be more errors than we’re getting so far as well…
-
[quote name=“Wellenreiter” post=“35455” timestamp=“1385062662”]
Started new ft-client on .0.6.4.4 wallet-
sync started and hung at block 34400 and all blocks seen where Seen as ORPHANED
-
tried stop/restart of client without success.
-
deleted local blockchain and started client again -> sync hung around block 30000 (didn’t check exact block number)
-
here I left the client run for ~ 2 hours without a any progress
-
also tried this 3 times ->always the sycn hang after 34400 Blocks
[/quote]
Wellenreiter, is it possible for you to run the above test again now that Ghostlander has put a fix in for this?
https://github.com/FeatherCoin/Feathercoin-0.8.5/commit/5c12e666cdad943a9e36cd2df8917fd33b7b6af0
-
-
Just checking the debug.log for “error”. I’m assuming this is because 0.8.5 is not backwardly compatible with version “60001”?
I was reading Litecoin will be version “70002” i.e. there will be a hard fork everyone must update to, to prevent old buggy / insecure versions?[quote]
receive version message: version 60001, blocks=88491, us=92.5.5.45:37481, them=90.154.132.74:9336, peer=90.154.132.74:9336
ERROR: CAlert::CheckSignature() : verify signature failed
Misbehaving: 90.154.132.74:9336 (0 -> 10)
[/quote] -
Just cloned and compiled the latest Litecoin 0.8.5.1 for comparison, version =“70001”. I found this error in the Litecoin Debug, I’m assuming it is just rejecting an orphan block.
[quote]
ProcessBlock: ORPHAN BLOCK, prev=ee6f8e23c4da61dbc19f0bf093a9066eb39020b46cd229e723e2a3fce5734d25
ERROR: CheckInputs() : tried to spend coinbase at depth 57
ERROR: CTxMemPool::accept() : ConnectInputs failed d9d8397f21a9d5183b4c332562a7ae55fdba096a7eb18b4fc0064b7d1ab61ec3
Misbehaving: 110.149.193.241:9333 (0 -> 0)
[/quote]The Feathercoin 0.8.5.0 version seems to be working well, all references / ports / wallet works. The default build file could be set to output to Feathercoin-qt.
Otherwise changes to 0.8.5 seems to be well tested by Bitcoin and Litecoin.
Are all the Feathercoin changes back in?
-
[quote name=“wrapper0feather” post=“35828” timestamp=“1385162234”]
Just checking the debug.log for “error”. I’m assuming this is because 0.8.5 is not backwardly compatible with version “60001”?
I was reading Litecoin will be version “70002” i.e. there will be a hard fork everyone must update to, to prevent old buggy / insecure versions?
[/quote]You are correct, 0.8.5 is only compatible with 0.6.4.4 (60002). I have now updated the PROTOCOL_VERSION to 60003. This allows us to message it separately with CAlert.
[quote author=wrapper0feather link=topic=4574.msg35854#msg35854 date=1385169989]
Just cloned and compiled the latest Litecoin 0.8.5.1 for comparison, version =“70001”. I found this error in the Litecoin Debug, I’m assuming it is just rejecting an orphan block.[quote]
ProcessBlock: ORPHAN BLOCK, prev=ee6f8e23c4da61dbc19f0bf093a9066eb39020b46cd229e723e2a3fce5734d25
ERROR: CheckInputs() : tried to spend coinbase at depth 57
ERROR: CTxMemPool::accept() : ConnectInputs failed d9d8397f21a9d5183b4c332562a7ae55fdba096a7eb18b4fc0064b7d1ab61ec3
Misbehaving: 110.149.193.241:9333 (0 -> 0)
[/quote]The Feathercoin 0.8.5.0 version seems to be working well, all references / ports / wallet works. The default build file could be set to output to Feathercoin-qt.
Otherwise changes to 0.8.5 seems to be well tested by Bitcoin and Litecoin.
Are all the Feathercoin changes back in?
[/quote]That is an orphan.
What file name is it outputting for you, I’ve got feathercoin-qt.exe lowercase?
The Feathercoin changes are all in. I will test CAlert on a local testnet to make sure that works properly.
I’m looking at changing the network magic numbers so that the client no longer sees other chains. I’ve had a bit of a Eureka moment on that one that trumps previous suggestions. Rather than change on block number change on epoch time like the fix for BIP0050. I will have to test this for sanity.
-
Yo!
Version 0.8.5 seems to be stable and i have not seen uncommon or critical errors so far in combination with p2pool. Are more tests required / desired?
Regards,
MrSheep -
I agree with MrSheep, we just need feathercoin specific tests (for this version), there’s too many changes to test individually, and 0.8 is already live on Litecoin and Bitcoin.
Unless there is anything been added to 0.8.5.0, that feathercoin should think about swtching off?
Just been trying libqrecode-dev install to see if QR codes appear. No luck. Did you say they are only available on all windows clients?
1.issue : No QR codes (0.6.4.4. & 0.8.5.0)
2. Just testing sending, what is the current Transaction fee policy?
send 0.01 (FTC)
[quote]
The total exceeds your balance when the 0.04 FTC transaction fee is included. [/quote] -
[quote name=“wrapper0feather” post=“35896” timestamp=“1385207113”]
Just been trying libqrecode-dev install to see if QR codes appear. No luck. Did you say they are only available on all windows clients?
1.issue : No QR codes (0.6.4.4. & 0.8.5.0)
[/quote]
I tried the same and also no luck :(
even no menu item appears offering the qr code.
Will check the qt part of the source next
-
nTime is declared in CBlock class as unsigned int and that’s a correct definition. Various functions in the code such as GetBlockTime() return it as long int (int64). This is not necessary absolutely, so either we a) fix these or b) set my printf to long int or c) type cast my printf to unsigned int. I’m inclined to fix them.
-
[quote name=“ghostlander” post=“35906” timestamp=“1385211340”]
nTime is declared in CBlock class as unsigned int and that’s a correct definition. Various functions in the code such as GetBlockTime() return it as long int (int64). This is not necessary absolutely, so either we a) fix these or b) set my printf to long int or c) type cast my printf to unsigned int. I’m inclined to fix them.
[/quote]I agree but,
Will we have to keep fixing GetBlockTime with every Fork / release?
If so should we ask for a fix, or offer patch to fix Bitcoin -> Litecoin -> us? For instance, I’ve posted bugs/issues to Litecoin before, I could raise it as an issue, with a patch to fix it. Which should make it easier to get included.
-
Feathercoin client 0.8.5 running successfully. Compiled on Windows 7.0 SP1 32bit.
Sync: Correct
Send: Correct
Received: Correct
New address: works
New sending address: works
Backup Wallet: works
User interface language : works
Chinese display: CorrectI think it can be released . ;)
-
[quote name=“lizhi” post=“36081” timestamp=“1385262603”]
Feathercoin client 0.8.5 running successfully. Compiled on Windows 7.0 SP1 32bit.[/quote]
Did you compile that to Windows? Did you manage to switch QR codes on?
-
QR-code enabled and working :)
Pull request is on Github.
I had some difficulties to get the fork and pull done.
was my first time to fork something on github and I hope, that I didn’t damage anything.Regarding functionality:
I created a QR code for payment on my PC and scanned the code with my tablet.
Worked absolutely fine. The Wallet on the tablet recognized the receiving address as well as the amount of FTC to be sent. -
Just noticed when testing the right click in “receive coins”, for QR codes.
When you mouse over (Linux Compile) it says “double click to edit address or label”. But double click doesn’t do anything. Same in 0.6.4.4. client. Same in Litecoin 0.8.5.1
-
Morning everyone. Considering that this release generally seems to be stable I have made the repo public so people can test without the need for explicit permissions. This was only to safeguard someone from putting this into a production environment.
-
Just reporting the results of my first software hack. As Wellenreiters patch didn’t work for my 13.10 Linux compile. I tried setting the if statement to what looks more logical to me.
That didn’t work either, tho’. QRs must be windows only.
addressbookpage.cpp
#ifndef USE_QRCODE
ui->showQRCode->setVisible(true);
#else
ui->showQRCode->setVisible(false);
#endif
