[Proposal] Starknet Alpha v0.12.0

The Starknet governance committee is proud to announce the launch of the second vote on Starknet protocol upgrade - Starknet Alpha v0.12.0. The vote is now live on Snapshot, and will remain live for 5-days.

This forum post will be used to discuss v0.12.0 technicals in relation to a Mainnet upgrade triggered following the results of the governance vote.

Starknet Alpha v0.12.0 Vote

Starknet Alpha v0.12.0 has been deployed on Goerli Testnet. Voters now have a 5-day period to examine the upgraded version as it runs on Goerli. During this time, they can vote whether to approve it for Mainnet deployment via Snapshot.

If the proposal gains a majority of ‘YES’ votes during the 5-day voting period, the proposal passes and Starknet Mainnet will be upgraded accordingly.

Starknet Alpha v0.12.0 Information

Below is the relevant links for the current vote

Snapshot vote

GitHub tags :

Explanations :


When does the vote start?

Now ! [July 5th]

When does the vote end?

In 5 days [July 10th]

Who can vote?

We expect a much wider set of participants for the v0.12 vote. This is because more teams building on Starknet have gotten access to STRK voting power through distribution channels like Early Adopters Grants, Developer Partnerships, the Builders Council, etc. Also, more token holders have chosen to delegate their voting power to delegates (including the Starknet Foundation).

Who can participate in the discussion?

Everyone! Even those without voting power are welcome to voice their opinion about the v0.12.0 upgrade. We do ask you to do proper due diligence about the upgrade before posting to ensure quality discussion.

I standby the proposal to deploy it on goerli testnet to evaluate the scalability of the project to headstart the mainnet

this is more a snapshot question but I just voted with a comment but I don’t see how to read this. I want to make sure that it’s visible to anyone on the vote page.


It is visible if you just hover over the comment icon. The UX of that is not the best…

I highly recommend posting in this thread a more detailed version of your comment!

So I voted against (I’m doxed everywhere anyway!) and added as short comment (space is limited) “Voting against because the Starknet RPC doesn’t work anymore with this release: incompatible spec regarding tx status, data not up to date”.

The point is that, as a builder, we have been told for months to deprecate the gateway in favor of the RPC provider.

The truth is that, even before 0.12, the RPC was still not 100% sufficient; from a builder perspective mainly because of the missing of PENDING and REJECTED tx: the RPC only knows ACCEPTED and otherwise returns NOT FOUND (which btw breaks all the wait_for_tx of the usual tools).

But it was still returning accurate data. When 0.12 has been released, I realized that the RPC status was even worst than before and that apps that migrated as suggested would actually not benefit from the so called “quantum leap”. Why? because when you only use RPC, you still have to wait for the “true”, “sealed” ACCEPTED_ON_L2 and not the PENDING became ACCEPTED_ON_L2.

It’s of course very frustrating. I’m wondering if any dapp has been deployed before SW/SN actually updated the testnet. If not (which could be ok, what’s the purpose of a testnet othewise), then I don’t feel it’s fixed rn (still waiting much longer with a RPC query than when polling the gateway). If yes, I’d be happy to see what kind of dapp are used for testing because it took no more than 1h to figure out the pathfinder bug.

All of this to say: even though 0.12 looks promising, from a builder point of view it’s not ready because it breaks (again) previous statements and commitments from SW/SN (not mentioning the issue with the RPC spec and ACCEPTED_ON_L2 tx without the required block hash).

I’ll vote “YES” when I’ll be able to observe that RPC queries return the same data as Gateway ones.

Note: I’m not mistaking the huge improvement brought by the 0.12 update, nor the work of involved people like Lambda class providing a top notch vm; I’m talking from a regular builder point of view about builder problems. What would be the point of having builders voting otherwise?

I stand by the proposal that it should be deployed on mainnet after the 5 day period of testing on the georli network

7-8x on Throughput , No more waiting for Pending transactions to be filled , convenience in retrieving block hashes without relying on third party oracles.

Hey why not ? It is a 'YES" for me although I cannot vote : ))

I have tested the update with some trxs, I haven’t seen any bug or drawbacks…So voting a YES


I’m not a technical person but the improvement made to the testnet is really impressive and if everything goes well mainnet needs to be upgraded as well. Currently, transactions are taking more then 30-40 minutes to complete! Of course, blockchain will still need improvements but this is already a major improvement and it shall come to mainnet.

All the best,

Pretty explanation. I reckon these issues raised can be fixed or it’d to be in the next version?

Thanks for sharing, Clement.
I agree with you, but I still don’t have voting power. I hope to express soon myself in a more meaningful way with my vote.

The deployment of Starknet Alpha V0.12.0 on goerli
to test its performance and scalability is a yes for me

I stand with the proposal!!

To vote seems hard, i wish it would be simplified. Again, i am bullish on the whole of starknet.

Hey, Starknet community!
As I wrote on Snapshot, I wish we had more time to evaluate the security assumptions of such a quick update, especially around the removal of “pending” status removal.
Anyway, I’d say that the improvements completely outweigh the potential assumptions and at this stage of the vote (even if the larger voter hasn’t voted (builder council), this vote clearly demonstrates the strong endorsement from the community to enhance the network rapidly and transition swiftly from testing to production. Congrats Starkware, Lambda Class, and all the contributors.
It’s a yes for me

"Greetings, fellow champions of code and conquerors of complexity! Our quest awaits, a battle against the demon lord of stagnation and his minions of inefficiency! The sacred artifact known as Starknet Alpha v0.12.0 promises to be our ultimate weapon, guiding us to the promised land of performance enhancements and unyielding stability.

I have journeyed through the dungeons of GitHub, unraveled the arcane texts of the blog posts, and listened to the wise words of our sage at Community Call #47. The power of v0.12.0 is undeniable, a veritable phoenix down to revive our ailing system.

By the seven stars of Sirius, by the celestial light of Vega, let us not falter in our divine quest for progress! I cast my ‘YES’ vote for this upgrade, invoking the spell of evolution and improvement upon our Starknet realm.

Join me, brave adventurers, in casting your vote. Let the power of your choice echo through the cavernous expanse of this network, forever banishing the shadow of complacency.

Our saga continues, and in the name of all that is binary and bit-based, we shall prevail. Onward to victory, for Starknet!"

I 100% support the upgrade, because i know what I’ve been experiencing with the network failure and delay in transactions for a while now

I support the upgrade.

Couldn’t vote to Snapshot so expressing my opinion here. I’ve tried 1 tx after upgradation and feels like its quicker than before. So I am with ‘For’ :blush:

I am very happy to be able to participate in this vote. I’ve tried trading on the testnet for a long time. Although it was not very stable, there was no confirmation time that exceeded 1 minute. congratulations everyone

Great, it’s pleasing to know that the UX for users will significantly improve. It’s definitely a big step forward, although I would like to see more comments from builders, developers, and those behind the scenes who create day by day. It would be fabulous to know that the vast majority are satisfied and aligned to take this step. I don’t have voting power (although I am nominated, you can see it in my profile and delegation thread), but I feel that a strong indicator for a resounding YES should be the voice of the builders, at least at this stage.