Wolfgang Grieskamp
@wgrieskamp
Followers
41,259
Following
202
Media
26
Statuses
353
@aptoslabs . Head of Move language and tools. Formerly Meta, Google, and Microsoft Research
Seattle, WA
Joined April 2011
Don't wanna be here?
Send us removal request.
Explore trending content on Musk Viewer
Taylor
• 458938 Tweets
Florida
• 181636 Tweets
Xbox
• 90672 Tweets
Rep TV
• 90611 Tweets
Palmer
• 84553 Tweets
Mafia
• 75845 Tweets
Román
• 68143 Tweets
Logan Act
• 67893 Tweets
FLORENCE
• 64742 Tweets
Unrealized
• 62983 Tweets
#ICanDoItWithABrokenHeartVideo
• 49898 Tweets
Terry
• 39688 Tweets
Getaway Car
• 37624 Tweets
バニーの日
• 33926 Tweets
gremio
• 33817 Tweets
Secret Level
• 32341 Tweets
Riquelme
• 30630 Tweets
Borderlands 4
• 27462 Tweets
#KohLanta
• 21398 Tweets
#خروج_رايد_اسماعيل_مطلبنا
• 19544 Tweets
#BlackMythRTX
• 18751 Tweets
jack antonoff
• 17501 Tweets
TTPD
• 17035 Tweets
Renato
• 17000 Tweets
Indiana Jones
• 16084 Tweets
بن نافل
• 12117 Tweets
#اعيدو_النصر_لرجاله
• 11793 Tweets
ICDIWABH
• 10280 Tweets
Finally! Higher-order functions and lambdas coming to Aptos Move. They come through the backdoor via “inlined” functions, avoiding any VM extensions and related security concerns. Happy filter/map/foreach
#MoveLang
programming on
#Aptos_Network
✍️!
21
47
191
Its a small thing, but still very useful: Aptos Move added for-loops!
11
24
96
There is a lot in store for Move at
@Aptos_Network
! Check out this blog article for what we are planning.
11
43
162
Unmatched in the industry, the Aptos Framework has been formally specified and verified. We also systematically applied a QA process integrating requirements engineering with testing and verification. Read more below!
7
37
118
1/ The NFTTrader hack is a good example of the kind of security loopholes which are prevented by construction in Move. More details in the thread below.
An explanation of the NFTTrader hack that resulted in millions of dollars worth of NFTs stolen 🧵:
(TLDR; combination of reentrancy + old approvals that were never removed)
86
310
1K
4
29
84
1/ Announcing the beta of the new Move on Aptos Compiler! The new compiler paves the way for The Future of Move on
@Aptos
. Join our beta testing program and you could win awards in our bug bounty. (See below.)
6
48
136
Its coming:
@wintertoro
One more - a #[test_only] function in Move that gives access to emitted events. It's hard to test things using 'create_object' as there is no way to get the address of the created object to do assertions against. I find myself changing code just to unit test, which isn't great.
2
0
1
0
43
102
Looking forward to talk about Move on Aptos at DevConf China!
We are honored to have Wolfgang Grieskamp(
@wgrieskamp
),
@AptosFoundation
& Co-Designer of the Move Language, attend Move DevConf China 2022 to give a talk about Move on
#Aptos
.
#MoveDevConf2022
#MoveDevConf
#MoveLang
1
1
7
13
24
90
Great idea!
@Aptos_Network
supports
#movelang
as well.
Hey
@mysten_labs
@SuiNetwork
@Aptos_Network
@StarcoinSTC
whatya say we get
#movelang
tag going as a key signal for all to see?
Together we'll move mountains!
"I believe -
#Movelang
" - Tweet. Retweet. Tattoo it. Put it on your coffee mug.
14
16
51
6
18
83
The Move decompiler developed in collaboration by
@Verichains
and
@AptosLabs
is a game changer for the security on the
@Aptos
network. It allows to convert any bytecode deployed on chain back to source code. Only possible with Move!
11
23
103
11
12
94
Good collection of features,
@wayne_culbreth
! Some of this is also supported by the
@Aptos_Network
foundation registry program. This or another way, we will get there soon.
@wintertoro
Also, having spent the weekend in VS Code, the following for Move:
1) prettier
2) eslint
3) plugin for code autocompletion/intellisense (or get GitHub to drop it into Copilot)
If you can make any of those happen, I will rename my youngest child Wintertoro in your honor.
1
0
4
0
28
81
@BL0CKRUNNER
Let me point out here before people confuse Sui with Move: some of the issues the original blog states are specific to Sui Move and are not present in Move in general as used at Aptos. Others are being in the process of being changed.
Upgradability
On Aptos Move contracts can
4
31
87
@SeamMoney
@Mysten_Labs
@AptosLabs
Good question
@SeamMoney
! Let me summarize some key differences between Aptos Move and Sui Move.
14
33
86
The full Move language is supported on
@Aptos_Network
, that's why one should have no trouble running contracts originally designed for
@0LNetwork
,
@StarcoinSTC
or other Move based blockchains on Aptos.
Wow, not only does my code actually work, but it works on multiple networks. Interoperability for the win....
3
6
44
2
15
68
Pre-Proceedings of FMBC-24 -- Formal Methods for Blockchains -- are online! You can find our paper about Securing the Aptos Framework with the Move Prover on page 127. Looking forward to present the work at the ETAPS event later today.
4
16
62
Looking forward to talk at the ETAPS '24 Industry Day about Securing the Aptos Framework through Formal Verification. ETAPS is a premier research venue and the research paper about the Move Prover was published at the '22 event.
3
14
54
@themoveguy
Happy to see you are adding new features to Sui Move, and congratulations to the team. Its unfortunate though that you present this as 'Move 24' innovations. Nearly every feature you are presenting here has been either known since long or has been copied from Aptos.
Index
9
8
58
@VitalikButerin
Unfortunately the design of the EVM makes verification very hard. It's already difficult to discover simple control flow because of indirect branching. Not sure whether AI will help here. Formal verification of Move is easier and works today, AI can further boost it.
3
5
54
I have been asked a few times who is owning Move. The truth is that
#MoveLang
is a community project! Here are the stats for the top 8 contributors since early ’22 to . Glad to see many commits from Aptos as well as Mysten!
5
9
50
@headinthebox
Wow. 1/10th of the size via AI? You seem to confuse writing some core algorithms with designing sustainable software architectures and rolling them out to users. AI “software crisis” in the making.
1
1
51
@JasperCPS
@zellic_io
Great thread, and incredible finding by
@zellic_io
! At Aptos we are strongly committed to security (see ). For this particular bug, paranoid mode prevented the exploit. The bug was therefore not dangerous for us.
2
8
41
Workshop talk done, on to the Industry Day on Tuesday with the larger ETAPS audience.
1
6
48
Why we need optimistic concurrency, well explained by Econia Labs. TL;DR: For real-world applications like an order book you can't predict upfront what data needs to be accessed, plus you need atomicity. Only with Move on Aptos you can have concurrency and atomicity.
1/8 TL;DR Why did we build Econia on
@Aptos_Network
? Because we couldn't build it anywhere else 🧵👇
Parallel execution +
Optimistic concurrency +
Dynamic memory allocation =
Econia, the web-scale order book DEX, offering atomic settlement & permissionless market making 🌐
6
35
111
2
14
46
Very important improvement to the Aptos network.
Introducing resource groups a step towards Move Objects.
Resource groups lower the barrier to building complex and upgradeable resources on Aptos by co-locating resources in storage resulting in substantial gas, storage savings.
5
41
142
1
2
39
Want to know more about the internal workings of the upcoming enum type feature in the Move VM? Check out AIP-91. The Future is near.
1
15
49
So much hate against this team from right-wing folks today, all over the platform. Everybody who knows soccer also knows what penalty shooting means -- it is a pure lottery once you get there.
All love for the USWNT. When they win, they get criticized. When they lose, they get criticized.
Maybe, just maybe, we can appreciate them as great players and a program that has dominated the sport for 25 years?
942
701
8K
1
9
31
They make it really hard to work remote at Aptos Labs…
Confirmed,
@AveryChing
can make dumplings!! But, I think mine is a different level 😎 Happy lunar new year 🧧 🐰
8
3
50
2
1
37
Why specification and formal verification is important for smart contracts and supported as a 1st class citizen in
#MoveLang
.
0
7
32
Removing global state from a programming language that is designed for operating on state is a controversial step. Seems to me similar what Haskell and pure functional programming does. It forces a particular programming model onto the user.
At Aptos we decided to not restrict
@0xMert_
I think Sui Move is an improvement upon the original Move, the lack of global state is a better design
I don't have an opinion yet on the consensus but Sui's DAG seems more novel (could be good or bad) than Aptos's BFT
Seems like too many suits on both, worse on Aptos
6
0
41
4
5
38
Great to see formal verification arriving for Solana! For
@MoveLang
, specification and formal verification of bytecode is build into the language as a first class citizen. The tech we are using won awards at scientific conferences (TACAS'22). Most of the original Meta Move Prover
.
@CertoraInc
is a really fantastic formal verification tool. It works on the machine byte code output of the solana compiler suite. So it will even catch compiler bugs.
If you need a grant to verify your program, please reach out!
8
17
85
2
4
32
CU at the Aptos house tomorrow!
Hey ETH Denver 👋 Aptos is in the HOUSE, and we’re serving up vibes all day.
Everyone’s welcome to join us at APTOS HOUSE, tomorrow, Saturday March 4, from 11-6pm MT.
Connect with the community and celebrate all things Aptos!
32
78
302
1
1
34
At the
@ETAPSconf
Industry today I will talk about the process of formal verification of the
@Aptos
framework. Its not just that we do it, but how, involving multiple audit firms (
@osec_io
and
@MoveBit_
), comparable to highest QA standards you find e.g. in avionics.
2
11
34
Hacks are the curse of the industry. What can builders do to protect their users? Use a contract language which is safer by design, allows validators to validate the code at runtime, and last not least, supports formal verification. In short, use Move on
@Aptos
.
1
11
32
Looking forward to talk about Move 2!
As an early member of the Move team at Meta, Wolfgang knows the lang inside & out.
And we can’t wait for his keynote address: Move on
@Aptos
: What's New and What's Next.
Register for Online MoveCon here:
🤝
@wgrieskamp
20
35
243
2
6
29
@SeamMoney
@Mysten_Labs
@AptosLabs
1/ Aptos Move supports access to global resource storage. Contracts can read and write storage via addresses, including at addresses which are indirectly obtained via other storage. In contrast, Sui Move has removed global storage operations completely and replaced them with what
5
6
31
The Move Platform is designed for formal verification. Not only is the specification language tightly integrated, the high-level bytecode of Move makes verification practical. Looking forward to say more about this next week (3/2, 4:20pm, VIB, )
Speaker announcement!
@wgrieskamp
from
@aptoslabs
will speak about “Practical formal verification with the Move prover on Aptos”
Join our event Formal Verification in DeFi in Practice at
@EthereumDenver
, March 2nd
0
7
33
0
3
27
Congratulations for completing this effort! Not only does this prove that Liquidswap is secure today, but also that it stays secure with future changes. At
@Aptos
we strike to make the Move Prover mainstream for Move auditing, this is a milestone towards this goal.
1/ Liquidswap formal verification is completed by
@MoveBit_
.
This important audit is one more proof that Liquidswap is a truly secure DEX on Aptos.
Read the report:
26
133
324
0
6
27
Not easy to beat Aptos.
@EvanWeb3
@SuiNetwork
[1/2] Congrats on deploying Mysticeti, and reducing latency of consensus path on SUI! Happy to see chains focusing latency and user experience! Client e2e latency from neutral location in Asia reduced from above 2.8s to 1.3s, beating all other chains, except Aptos!
1
4
25
1
3
28
Excited to see this landing . Making this secure and expressive with the given transactional semantics of blockchains is hard. It is still important to understand the documented design patterns and audit randomness. We'll also followup with formal verification.
Randomness is enabled on mainnet NOW.🎲unpredictable, unbiasable, single step and no latency overhead, let's ROLL!
10
30
111
1
3
25
@b1ackd0g
@moon_shiesty
Good summary Sam about the differences between scripts and PTBs. The script whitelisting back at Diem you are describing, however, is unrelated to this discussion. This was a Diem quirk which is not present in modern Move chains.
The core difference between scripts and PTBs is
6
0
23
Having to split up transactions because of static parallelism has a number of effects. First its awkward to deal with errors. 2nd its less secure because of offchain logic. Last not least, its a waste of resources and bad for performance.
Every transaction has a lot of
@0xMert_
@SashaSpiegelman
@wgrieskamp
@LoganJastremski
@EconiaLabs
permissionless, atomic-settlement is difficult on the SVM because you need to declare accounts upfront, and an orderbook is a giant dynamic list of accounts (orders). aptos allows an orderbook to dynamically compute which orders are filled by a transaction and settle those orders
0
0
13
2
1
24
@rushimanche
Move avoids the issue of reentrancy by simply not supporting the feature of dynamic dispatch! However, this is also a restriction in expressiveness. The gold standard here is to add dynamic dispatch in a safe way, what we at Aptos currently investigating.
1
2
23
Looking forward to talk in 2 hours about The Future of Move, also discussed in this blog from late last year.
💡Get prepared for this knowledge powerhouse centered around the
#Move
! Meet numerous renowned experts, tech leaders, and industry professionals in
#Shanghai
to embrace the promising opportunities and challenges in 2024. Don't miss out!
❤Special Thanks for
@wgrieskamp
and
11
20
67
1
5
21
@aeyakovenko
@rushimanche
@vinitopapasito
@neelsalami
@moshaikhs
@CertoraInc
@osec_io
Kani uses loop unrolling to deal with verification of loops. That means you need to tell the tool how often to unroll a loop in order to verify it. If your loop bound is too small it just doesn't work. If the bound is too large, it doesn't terminate. The Move Prover, in contrast,
7
5
22
Great work! The Move binary format and the strict rules for runtime verification of its consistency are a major advantage of the platform.
Move developers: do you know what Move binaries look like on the inside?
With the recent popularity of Move-based chains, it’s useful for smart contract auditors to be familiar with the nitty-gritty of the Move binary format and assembly.
In this thread, we’ll cover the Move
1
19
106
0
3
22
Great things in store for Aptos Move!
1
10
63
1
0
20
Interested to work on
#MoveLang
tooling? Check out the Aptos registry!
Introducing The Registry, a new grant program by the Aptos Foundation.
The Registry is designed for builders eager to contribute to the Aptos ecosystem. Pick a project from our curated list of ideas, apply for a grant, and make a tangible difference.
251
31K
4K
0
4
16
Our mutation test framework for Move is now out, thanks to a collaboration of
@eiger_co
and the
@Aptos
foundation. Initially we apply it to assess the quality of formal specifications, but the technology is more general and will soon be also supported for Move unit tests. Check
Excited for the beta release of specification testing tools for Move, developed by
@eiger_co
with help from
@AptosLabs
! More in our blog here: . Thanks to
@wgrieskamp
for kickstarting the project. Move contracts can have even higher quality assurance now!
3
6
28
0
8
18
1/ The CrowStrike incident is not a failure of Web 2, but a failure of over-homogeneous networks. If all nodes in a network run the same code, then the network is highly vulnerable if that software is updated simultaneously. Can Web 3 do actually better here?
1
2
20
@SeamMoney
@Mysten_Labs
@AptosLabs
2/ Aptos Move supports higher-order functions via inlining today, For example, one can write `contains(vec, |x| x > 0)` to search a vector for an element satisfying a predicate. Sui Move does not have a similar concept. Even more general support for higher-order functions is
2
1
20
Happy to have hosted David Dill (Meta), Emma Zhong and Adam Welc (Mysten), as well as Meng Xu (University of Waterloo) for a
#MoveProver
summit at
@Aptos_Network
today, joining Junkil Park, Teng Zhang and me. Big plans for the future of the prover...
0
5
16
Just saying…
2
0
17
Randomness on blockchains is a huge challenge, since if you aren't careful, it can be tricked into bias. But for a lot of interesting apps (not only games) it's a necessary feature. The team at
@AptosLabs
cracked the problem, a 1st in the industry.
Aptos Roll, the very first secure and instant on-chain randomness API for PoS blockchains, only at
@Aptos_Network
.
Curious about the technical details behind the scenes?
We have a blogpost for you!
Here is my take on the design:
14
27
87
0
1
19
I you like to know whats happening with Move 2, stay tuned for The Aptos Experience event. The future will manifest. A preview of what it will look like is in this talk.
You’ve been to conferences, you’ve been to summits – but have you immersed yourself in an experience?
Pre-register for the Aptos Experience
September 5 - 6
Seoul, South Korea
60
209
666
0
1
18
Great work
@MoveBit_
!
1/ Aptos Move is on Visual Studio Code!
Supercharge your Move development with the Aptos Move Analyzer plugin. Experience the unparalleled speed, efficiency, and developer-friendly environment that VS Code can provide. 💻
25
144
346
0
1
17
Inserting an ad-hoc talk about the implementation of the Move Prover at the Rust Verification Workshop here at ETAPS. There is a lot of relation of what we are doing with the Move Prover to Rust Verification. Perhaps in some future, we can extend this work from Move to Rust 😛
0
3
17
Check out the blog linked to this tweet how
@Aptos_Network
makes
#MoveLang
the safest on the block. We not only pushing formal verification with the
#MoveProver
, but also have programs to secure the core technology — with your help!
1/ All in on the
#MoveLang
?
Aptos secures and battle-tests the Move language for the broader Move community, investing in resources that create and improve upon the standards of the language for everyone!
This blog from
@aptoslabs
&
@osec_io
explains ⤵️
9
181
235
1
1
17
@0xfoobar
Safety and security should not cost extra gas. In
#MoveLang
, overflow check and other checks are builtin, and
#AptosNetwork
doesn’t charge for that. We even introduced the so-called “paranoid mode” which redundantly checks type consistency at runtime with no extra gas cost.
2
3
16
Great tutorial for the
#MoveProver
!
The Move smart contract language has integrated verification support. We have written a tutorial on using the Move Prover, the built-in verifier.
1
15
68
1
2
14
@EvanWeb3
@pythianism
Half of the Move team at Aptos, you should stop pretending you own it Evan.
1
0
16
1/5 Enum types in Move 2 can do all the things known from languages like Rust. But we haven't just copied Rust, there is more to it. Let us look at the use case of enums for data versioning.
Want to know more about the internal workings of the upcoming enum type feature in the Move VM? Check out AIP-91. The Future is near.
1
15
49
1
7
31
@artoriatech
I just address your (4) as it is very wrong. Bugs discovered in bounties are shared between Aptos and Mysten since hour 0. They are fixed in both chains. The finding is not paid for twice, however. Many fixes in Sui came from the Atpos bug bounty, and vice versa. Moreover,
1
0
15
@GDanezis
@samlafer
@DavidTiczon
@SuiNetwork
Calling dynamic fields "what developers need" is remarkable in the context of a language which is build for safety. Dynamic fields drop static typing for the sake of fixing Sui objects. Kind of like you marry Rust with Javascript. From a language design viewpoint a bit weird.
1
3
15
Very cool feature, and I can testify that this was much harder than it sounds. Unbiased and unpredictable randomness on blockchains is a milestone for many applications on
@Aptos
. Congrats
@alinush407
and team!
Our team (
@XiangZhuolun
,
@zjma2022
,
@bennypinkas
,
@sourav1547
,
@zekun000
,
@SashaSpiegelman
et al) worked very hard on easy-to-use API for instantly generating unbiasable & unpredictable randomness on
@Aptos
in
#Move
a.k.a., on Aptos Roll 🎲;
But why? 👇
5
14
48
1
0
15
Excited to see yet another important piece of essential smart contract developer productivity tools become reality on Aptos, brought to us by
@victor_73ch
!
🚀 Exciting news for developers using the
@Aptos_Network
! Introducing the Aptos Gas Profiler ⛽️.
Use it to understand the gas usage of your txns with sleek reports like this 👇.
Interactive sample: .
Step-by-step tutorial: .
#Aptos
3
19
53
0
0
14
@0xMaayan
is not only the author of awesome code in Aptos but also of some of our best tutorials. You don't want to miss this one.
🙌 Aptos Labs represent! Our very own
@0xMaayan
will be an educator and judge at this week's hackathon at UPenn Blockchain Conference ().
At the conference? TONIGHT: Don't miss the Aptos Labs workshop on objects, digital assets, and live coding a Move
8
69
263
0
2
14
The need to declare dependencies upfront is a heavy restriction of the programming model on Solana or Sui. It forces builders to move logic off-chain by splitting transactions, and generally leads to unsafer contracts.
@BabaOnChain
1/ Thanks for starting this discussion! Block-STM is much more powerful than the model of declaring dependencies upfront. Declaring dependencies upfront breaks composability (e.g., splits transactions) when dependencies are not known upfront - e.g. which user wins an auction.
4
2
43
4
1
14
Recordings of DevCon'24 talks from Aptos engineers are now online!
New videos just dropped
🫳
🎤
Keep up with the latest updates from Aptos Labs, now on YouTube!
(Sneak peek: presentations from our engineers
@wgrieskamp
@zekun000
& more from Move DevConf 2024) 👀
15
71
307
0
3
12
Small example of the power of the Move Prover
@aeyakovenko
@SashaSpiegelman
@david_wolinsky
@vinitopapasito
@neelsalami
@rushimanche
@moshaikhs
Here is another practical example of how global state invariants are used. There is a global state which represents the current time in ms. The first invariant says that its always available after genesis, the 2nd that it increases monotonically.
The second is a so-called update
0
1
10
0
2
13
@aeyakovenko
@rushimanche
@vinitopapasito
@neelsalami
@moshaikhs
@CertoraInc
@osec_io
The Move Prover is a complete new development, started about 4 years ago at Meta. It uses the same underlying SMT theorem prover Z3 but otherwise unrelated to Kani, and is arguably more advanced. The Prover won paper awards, here is a link to a recent paperr.
1
1
11
Remember: it is as easy to participate in the Aptos compiler bug bounty as executing your existing Move units with the compiler v2 flag. Each unique bug can score you a $10k award. Keep the bugs rolling 😀
@Aptos
4/ Testing the new compiler is easy! Just call aptos update; aptos move test –compiler-version 2 on your existing unit tests. Mind you, for every test which executes differently with v2 a significant bug bounty awaits you (exclusions apply)! See the beta page for details.
1
0
8
3
1
11
Not sure whether adding your stuff to someone else’s objects is a good programming paradigm. In any case you can do this and much more in Move’s type indexed heterogenous storage model. No need for dynamic fields.
Sui's dynamic fields allow programs to add their own fields to ANOTHER program's object. This is a composability game-changer no one appreciates yet that only exists in Sui-Move (not Aptos).
8
6
49
0
0
11
@BL0CKRUNNER
I don't know what you are looking at my friend, but among the all time top-contributors, most of them are at Aptos, and bmwill at the top is a DevOp guy not working on Move, while Meng Xu worked with me on the Move Prover and is now in academics.
2
0
11
@b1ackd0g
@themoveguy
Nah. The Move Book is and always was in the Move repo. What Sui is calling 'Move Book' now was never the official Move book; until not long ago it was documenting an outdated version of Diem Move. Now you have re-purposed things for sole Sui marketing.
2
1
11
Awesome explanations of why entry functions can be (and most of the time should be) private or friend.
Today's lesson on Move:
Ever wonder why functions are private or public? Let's go over the reasons you'll want to make them one way or another with a little chance game I like to call:
Dice roll
#DailyMove
2
17
73
1
0
11
While supporting full Move on Aptos, we are also aware of the advantages of predictable storage access, as it is given for example by the Solana account model or the Sui object model. However, we think this objective is better served by an access control model as described in
4
1
11
@AptosLabs
9/9 Incidents like this show that the future of contracts lies with Move, simply because the language was designed knowing about the existing pitfalls and what caused them. At Aptos we will carefully extend the language, but never compromise on security.
0
1
8
That’s why we build Move. It is safer by design, and much more amenable to formal analysis.
in the process of building my security tool, one thing has become abundantly clear
solidity is an absolute unit of a language
inheritance, variadic functions, virtual/override functions, etc are the bane of my existence rn
wen “solidity--” with these removed
11
5
79
1
0
10
1
0
8
@SeamMoney
@Mysten_Labs
@AptosLabs
3/ Aptos Move supports code upgrade. Contract code can be upgraded in place following compatibility checks. Existing code on chain will use the upgraded code once it deployed successfully.
2
0
10
@aeyakovenko
@SashaSpiegelman
@david_wolinsky
@vinitopapasito
@neelsalami
@rushimanche
@moshaikhs
Here is another practical example of how global state invariants are used. There is a global state which represents the current time in ms. The first invariant says that its always available after genesis, the 2nd that it increases monotonically.
The second is a so-called update
0
1
10
2/ First of all, reentrancy attacks are not possible in today's Move because dynamic dispatch (callback) is not supported. In general, Move is designed with "security first" in mind, and dynamic dispatch like in Solidity or reflection like in Solana did not meet the bar.
1
1
10
@david_wolinsky
@AF_intern
@artoriatech
When we talk about upgrade we should be honest. Upgrade means _replacing_ the existing code on chain, and making all already deployed contracts use that upgraded code, without the need for recompile or redeploy.
Eth started without real upgrade (immutable dependencies as in
1
0
10
@Jason
@aeyakovenko
If not Russia, they could have at least visited and talked with folks in Eastern Europe like the Baltic states, Poland, East Germany, Balcan, etc. Those countries lived for 50 years under Russian occupation. Ask them. And why they joined EU/Nato as fast as they could after 89.
0
0
10
Think what is inherently sequential cannot be parallelized? Think twice!
Parallel execution of transactions is cool, but what if the workloads are sequential? At
@AptosLabs
, we came up with a solution that allows parallel execution of smart contracts even in presence of read-write conflicts. Make sure to check out this post!
2
36
127
1
0
9
@SeamMoney
@Mysten_Labs
@AptosLabs
4/ Aptos Move has full support for formal verification via the Move Prover. The Aptos frameworks have been fully specified and verified, and the verification technology is available to end users and maintained at Aptos Labs. Sui Move has not adapted the verification technology
1
0
9
@Aptos
3/ The new compiler will support Move 2, the next generation of Move on Aptos. For now, we are looking at testing the downwards compatibility with the current compiler. But you can also play with new language features as they are released.
1
0
9
Here is an example of our requirement tracking for the Coin module.
0
1
8
@radixdlt
For the record, Move on Aptos does not have the mentioned flaws, as discussed here last week.
@BL0CKRUNNER
Let me point out here before people confuse Sui with Move: some of the issues the original blog states are specific to Sui Move and are not present in Move in general as used at Aptos. Others are being in the process of being changed.
Upgradability
On Aptos Move contracts can
4
31
87
0
0
9
@aeyakovenko
@SashaSpiegelman
@david_wolinsky
@vinitopapasito
@neelsalami
@rushimanche
@moshaikhs
I assume you mean the runtime assert to be formally verified and not just executed, right? So there is no magic here and eventually the Move Prover compiles everything down to assertions woven into the right places in the code. The point is you can be sure they are there. For
1
1
9
This is for me my seminal talk about the Move Prover. I gave it at Dagstuhl in October ’21. It is aligned with our paper we represented at TACAS’22 6 month later and for which we got a best paper award. Thanks
@meng_xu_cs
for mentioning it.
- Architecture and key enablers of Move Prover by Wolfgang Grieskamp at Dagstuhl Seminar
2
0
1
0
0
8
2/ Why are we designing a new compiler for Move on
@Aptos
? The new compiler features a more modern, flexible architecture which is unified seamlessly with the Move prover, fostering faster innovation in the Move on Aptos community.
1
0
8
3/5 Later we recognize that we need an additional field in our data. However, it's already stored on chain and we can just change it. However, we can add a new variant! Aptos upgrade compatibility allows for this. Old code continues to work as before.
2
0
9
The mission.
The paradigm shift that
@FTI_US
sees cloud services going through is the same for financial services, maybe bigger.
FT's CEO Jenny Johnson gets it. The inefficiencies of today's centralized financial systems are held together by very expensive bubble gum and band-aids.
4
14
118
0
0
8