# Trying to use testnet to reproduce a bug

**URL:** <https://bisq.community/t/trying-to-use-testnet-to-reproduce-a-bug/454>\
**Category:** Support\
**Created:** [August 12, 2016, 7:50am UTC](https://bisq.community/t/trying-to-use-testnet-to-reproduce-a-bug/454 "2016-08-12T07:50:03Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![rik](https://bisq.community/letter_avatar/rik/32/5_5575768a8748004e209b776fc1b2916d.png) [@rik](https://bisq.community/u/rik)\
**Post date:** [August 12, 2016, 7:50am UTC](https://bisq.community/t/trying-to-use-testnet-to-reproduce-a-bug/454/1 "2016-08-12T07:50:03Z")

</div>

Hello all,

I’m currently in a trade that’s stuck becaue of a bug (will inform the arbitrator in about 10 hours).

In the meantime I want to reproduce this bug on testnet. Unfortunately it can’t find any peers (my two peers can’t find eachother). My question is: how can I get my two testnet BitSquares to talk to eachother (and find an arbitrator) so that I can try to reproduce the bug?

Update: I specified --seedNodes blabla.onion:9999, now it works, but since there are no arbitrators I still can’t create an offer.

Update2: will try to recompile BitSquare to become arbitrator on my private BitSquare testnet.

Update3: I found the needed documentation in doc/build.md to use a ‘fake’ arbitrator

Greetings,

Rik.

---

<div class="post-metadata">

**Author:** ![ManfredKarrer](https://bisq.community/user_avatar/bisq.community/manfredkarrer/32/26_2.png) [@ManfredKarrer](https://bisq.community/u/ManfredKarrer)\
**Post date:** [August 12, 2016, 9:47am UTC](https://bisq.community/t/trying-to-use-testnet-to-reproduce-a-bug/454/2 "2016-08-12T09:47:58Z")

</div>

Hi @Rik, great to hear that you are investigating and help out here!!!  
I did not maintain testnet anymore as nobody used it. But seems you have found all info how to set it up and get it running.  
If you can send me a detailed description of the bug, maybe I can help. You also can open a support ticked before the trade period is over by clicking cmd+o or ctrl+o at the trade screen.  
Thanks for your help, very welcome to get some support!

---

<div class="post-metadata">

**Author:** ![rik](https://bisq.community/letter_avatar/rik/32/5_5575768a8748004e209b776fc1b2916d.png) [@rik](https://bisq.community/u/rik)\
**Post date:** [August 12, 2016, 11:09am UTC](https://bisq.community/t/trying-to-use-testnet-to-reproduce-a-bug/454/3 "2016-08-12T11:09:35Z")

</div>

Thanks to Manfred for his respone. I’ve sent a description of the problem to him privately.

---

<div class="post-metadata">

**Author:** ![rik](https://bisq.community/letter_avatar/rik/32/5_5575768a8748004e209b776fc1b2916d.png) [@rik](https://bisq.community/u/rik)\
**Post date:** [August 12, 2016, 1:16pm UTC](https://bisq.community/t/trying-to-use-testnet-to-reproduce-a-bug/454/4 "2016-08-12T13:16:57Z")

</div>

I have reproduced the bug partially:

Alice: marketmaker, sells ETH for BTC  
Bob: taker, buys ETH for BTC

- Alice creates offer 0.01 BTC @ 50.00 ETH/BTC
- Bob takes offer, deposits BTC.
- Alice pays and clicks ‘Payment started’
- Alice creates new offer 0.01 BTC @ 50.00 ETH/BTC
- Bob acknowledges payment, moves funds to bitsquare wallet
- Alice moves funds to bitsquare wallet (Open Trades screen now empty)
- Bob takes new offer, deposits BTC, looks at ‘Open Trades’ screen and sees new deposit txid)
- Alice looks at ‘Open Trades’ screen and sees the new trade with the deposit txid from the previous trade. This is a bug and it also happened during the real trade.
- Deposit gets confirmed. All turns out well.

During the real trade:

- a timeout was received (probably the node gave up waiting on relay of the new deposit txid becaus it was somehow looking for the old one)
- confirmation of the new deposit tx was never acknowledged

Next: I will test what happens if i start the next trade after the previous deposit transaction has 6 confirmation. This may trigger the worse bug. I need to wait for that, since I am on TESTNET (and not REGTEST, what was recommended in the docs).

---

<div class="post-metadata">

**Author:** ![rik](https://bisq.community/letter_avatar/rik/32/5_5575768a8748004e209b776fc1b2916d.png) [@rik](https://bisq.community/u/rik)\
**Post date:** [August 12, 2016, 5:40pm UTC](https://bisq.community/t/trying-to-use-testnet-to-reproduce-a-bug/454/5 "2016-08-12T17:40:49Z")

</div>

Test ‘start trade after previous deposit transaction has 6 confirmations’. Same results as before. (done in REGTEST, it’s quite nice to be able to create REGTEST coins just by a simple keypress)

I will post if I have more results.
