* [bitcoindev] Does anyone still need testnet3?
@ 2025-03-14 8:52 Sjors Provoost
2025-03-14 9:12 ` 'Andreas Schildbach' via Bitcoin Development Mailing List
0 siblings, 1 reply; 3+ messages in thread
From: Sjors Provoost @ 2025-03-14 8:52 UTC (permalink / raw)
To: Bitcoin Development Mailing List
Dear list,
Testnet 4 was proposed last year in BIP94 [0] and is supported by Bitcoin Core since v28.
To make Bitcoin Core easier to maintain we would like to drop support for Testnet3 in a future release. This will happen no earlier than v30 this fall. The upcoming v29 release still supports it, and will throughout its maintenance period [1].
The network itself can't be deleted, so anyone willing to maintain node software themselves can keep using testnet3 until the end of time.
One of the motivations for introducing testnet4 was that testnet3 became increasing impractical to use due to block storms. So perhaps everyone already moved on to using testnet4 or a signet.
As an aside, it's possible to create a pure proof-of-work permissionless signet without any signatures, by setting the challenge to OP_TRUE. [2]
However, if anyone is still using testnet3 and needs more time to move away from it, please let us know. E.g. as a reply to the list or a comment on the Github issue [3]. This could include libraries, staging environments, testnet wallets, etc.
Although the v29 maintenance window will last for a few years, it's also important to know if dropping testnet3 would prevent you from upgrading to the next major node version. In other words, if migrating away from testnet3 would take you longer than all of 2025.
Although "I don't feel like it" isn't a great technical argument, it's still useful to know what to expect and perhaps others can help.
- Sjors
[0] https://github.com/bitcoin/bips/blob/master/bip-0094.mediawiki
[1] https://bitcoincore.org/en/lifecycle/#maintenance-period
[2] https://github.com/bitcoin/bips/blob/master/bip-0325.mediawiki#specification
[3] https://github.com/bitcoin/bitcoin/issues/31975
--
You received this message because you are subscribed to the Google Groups "Bitcoin Development Mailing List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to bitcoindev+unsubscribe@googlegroups•com.
To view this discussion visit https://groups.google.com/d/msgid/bitcoindev/9FAA7EEC-BD22-491E-B21B-732AEA15F556%40sprovoost.nl.
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [bitcoindev] Does anyone still need testnet3?
2025-03-14 8:52 [bitcoindev] Does anyone still need testnet3? Sjors Provoost
@ 2025-03-14 9:12 ` 'Andreas Schildbach' via Bitcoin Development Mailing List
2025-03-14 9:31 ` Sjors Provoost
0 siblings, 1 reply; 3+ messages in thread
From: 'Andreas Schildbach' via Bitcoin Development Mailing List @ 2025-03-14 9:12 UTC (permalink / raw)
To: Sjors Provoost, Bitcoin Development Mailing List
Hi Sjors,
thanks for asking.
Bitcoinj has just started to implement testnet4, but it's not finished
yet. And the next version 0.18 will probably not be released within 12
months, maybe longer.
By the way, testnet3 works great for us (for testing). I consider block
storms a feature, because it tests our code in rough conditions. We had
planned to support both testnets in the forseeable future.
https://github.com/bitcoinj/bitcoinj/
Then, there is btc-rpc-explorer, a self-hostable block explorer. It
doesn't yet support testnet4. I'm not affiliated to it, I'm just a user.
https://github.com/janoside/btc-rpc-explorer
Cheers,
Andreas
On 14/03/2025 09.52, Sjors Provoost wrote:
> Dear list,
>
> Testnet 4 was proposed last year in BIP94 [0] and is supported by Bitcoin Core since v28.
>
> To make Bitcoin Core easier to maintain we would like to drop support for Testnet3 in a future release. This will happen no earlier than v30 this fall. The upcoming v29 release still supports it, and will throughout its maintenance period [1].
>
> The network itself can't be deleted, so anyone willing to maintain node software themselves can keep using testnet3 until the end of time.
>
> One of the motivations for introducing testnet4 was that testnet3 became increasing impractical to use due to block storms. So perhaps everyone already moved on to using testnet4 or a signet.
>
> As an aside, it's possible to create a pure proof-of-work permissionless signet without any signatures, by setting the challenge to OP_TRUE. [2]
>
> However, if anyone is still using testnet3 and needs more time to move away from it, please let us know. E.g. as a reply to the list or a comment on the Github issue [3]. This could include libraries, staging environments, testnet wallets, etc.
>
> Although the v29 maintenance window will last for a few years, it's also important to know if dropping testnet3 would prevent you from upgrading to the next major node version. In other words, if migrating away from testnet3 would take you longer than all of 2025.
>
> Although "I don't feel like it" isn't a great technical argument, it's still useful to know what to expect and perhaps others can help.
>
> - Sjors
>
> [0] https://github.com/bitcoin/bips/blob/master/bip-0094.mediawiki
> [1] https://bitcoincore.org/en/lifecycle/#maintenance-period
> [2] https://github.com/bitcoin/bips/blob/master/bip-0325.mediawiki#specification
> [3] https://github.com/bitcoin/bitcoin/issues/31975
>
--
You received this message because you are subscribed to the Google Groups "Bitcoin Development Mailing List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to bitcoindev+unsubscribe@googlegroups•com.
To view this discussion visit https://groups.google.com/d/msgid/bitcoindev/7c28f8e9-d221-4633-8b71-53b4db07fa78%40schildbach.de.
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [bitcoindev] Does anyone still need testnet3?
2025-03-14 9:12 ` 'Andreas Schildbach' via Bitcoin Development Mailing List
@ 2025-03-14 9:31 ` Sjors Provoost
0 siblings, 0 replies; 3+ messages in thread
From: Sjors Provoost @ 2025-03-14 9:31 UTC (permalink / raw)
To: Bitcoin Development Mailing List; +Cc: Andreas Schildbach
Hi Andreas,
> By the way, testnet3 works great for us (for testing). I consider block storms a feature, because it tests our code in rough conditions. We had planned to support both testnets in the forseeable future.
>
> https://github.com/bitcoinj/bitcoinj/
The "good" news is that testnet4 has very frequent short reorgs. [0]
It appears that people are trying to CPU mine by exploiting the rule that difficulty drops to 1 after 20 minutes. For every real difficulty block that's mined, they instantly publish 5 or 6 blocks, right up against the rule that blocks can't be more than 2 hours in the future.
Another side-effect of that behavior is that on testnet4 MTP is typically in the future, rather than an hour in the past. Combined with the timewarp attack mitigation rule helped reveal some bugs. [1, 2, 3]
There is a potential counter measure [4] that involves re-orging out those low difficulty blocks, though it's just a hackathon proof-of-concept and not likely to be included in Bitcoin Core itself. If anyone were to run it, you'd get even more reorgs to test :-)
- Sjors
[0] https://fork.observer/?network=4
[1] https://github.com/benjamin-wilson/public-pool/commit/4282233d2f11ceecbd0d142e8292ccc9c37ea999
[2] https://github.com/stratum-mining/stratum/pull/1325
[3] https://github.com/bitcoin/bitcoin/pull/31600
[4] https://github.com/bitcoin/bitcoin/pull/31117
--
You received this message because you are subscribed to the Google Groups "Bitcoin Development Mailing List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to bitcoindev+unsubscribe@googlegroups•com.
To view this discussion visit https://groups.google.com/d/msgid/bitcoindev/C899E966-6444-41EC-B977-96CFFCBF936A%40sprovoost.nl.
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2025-03-14 13:01 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2025-03-14 8:52 [bitcoindev] Does anyone still need testnet3? Sjors Provoost
2025-03-14 9:12 ` 'Andreas Schildbach' via Bitcoin Development Mailing List
2025-03-14 9:31 ` Sjors Provoost
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox