X 🫐 Profile Banner
X 🫐 Profile
X 🫐

@xdev_10

Followers
10,682
Following
2,401
Media
0
Statuses
95
Explore trending content on Musk Viewer
@xdev_10
X 🫐
2 years
Addressing some of the issues brought up about @GMX_IO 1. Timelock contracts 2. Price feeds
22
120
631
@xdev_10
X 🫐
1 year
Received a few questions about this, providing some context: - There are caps on long / short open interest - If you execute a trade on an order book and the total fee + price impact is 0.1% that's similar to paying a fee of 0.1% and having zero price impact
@degentradingLSD
degentrading
1 year
1/n - Why @GMX_IO will fail
186
271
1K
11
57
269
@xdev_10
X 🫐
1 year
Been almost two years since then Some may not know but Dovey was one of the earliest supporters of GMX back when it was called XVIX in Dec 2020 Even though the project had a very short track record and founded by an anon team
@DoveyWan
Dovey "Rug the fiat" Wan (hiring)
1 year
GMX has hit a total of $100B volume across Arbitrum and Avalanche, a casual 1000x up in less than 2 years Wen $100B party? @xdev_10 😈
Tweet media one
4
5
63
21
29
212
@xdev_10
X 🫐
2 years
We are currently working closely with Chainlink so that eventually the custom price feeds and keeper to prevent front-running will no longer be needed, once their solution is ready we will be switching over
5
16
133
@xdev_10
X 🫐
1 year
Really like the work that @guardianaudits has been doing, they have audited multiple iterations of the GMX Synthetics codebase, and their format gives better confidence that the code was thoroughly evaluated
@0xOwenThurm
Owen | Guardian
1 year
Audits are not accessible to the entrepreneurial class of crypto and its a huge problem for the entire industry. Here’s how we’re going to solve it. 🧵
3
16
56
5
14
87
@xdev_10
X 🫐
1 year
@CL207 @smolgrrr @adamscochran it is limited in total at the moment, could consider time window quite a lot of improvements coming in the next update, multi-collateral, positive funding fees would be good to talk more, will dm
3
6
70
@xdev_10
X 🫐
1 year
- If there are any suggestions on how the system can be improved or made safer, we are definitely open to them, I can be contacted on Telegram, my handle is @xdev_10
4
2
65
@xdev_10
X 🫐
1 year
Thankful to have been able to work with Dovey and the many GMX contributors and supporters all this while
2
1
60
@xdev_10
X 🫐
8 months
agree with @BFreshHB 's thoughts on this for trading incentives in particular, it may be worth considering limiting total incentives to less than the total fees paid
@BFreshHB
𝗵𝘂𝗻𝘁𝗲𝗿
8 months
aaaaahhhh im writiiiiiiiiiinggggggg check out this piece i wrote about how community on Arbitrum has been built over the past 2 years and my take on the impact that incentives may have on it going forward
28
40
176
10
8
61
@xdev_10
X 🫐
1 year
Thinking back, the project was very rough and unpolished at the start, so I felt surprised by the people willing to support it There's still a lot more to polish but I'm glad for the iterations and improvements made so far
3
0
60
@xdev_10
X 🫐
2 years
We would like to use Chainlink instead for pricing, Chainlink has made a lot of improvements to the speed of price updates
2
2
49
@xdev_10
X 🫐
1 year
- These changes will help reduce risks, but it will not eliminate it
2
2
42
@xdev_10
X 🫐
2 years
1. Timelock contracts GMX contracts have a Timelock contract to manage configuration, this was raised as a medium risk issue in the Quantstamp audit
1
1
39
@xdev_10
X 🫐
1 year
- Adding positive funding fees to incentivize balancing of long / short open interest and reduce net exposure of LPs
1
1
41
@xdev_10
X 🫐
2 years
Timelocks can be monitored by anyone, we know that community members monitor them as we have received questions for clarification about signalled transactions Reporting of a malicious timelock transaction is also covered in the GMX Immunefi bounty:
1
0
39
@xdev_10
X 🫐
1 year
She helped the project through the early stages, constantly looking out for it and helping to connect partners
1
0
37
@xdev_10
X 🫐
8 months
not doing so may overly attract volume that only has farming tokens as its goal increases in metrics help to increase awareness of a network, but the cons of crowding out genuine users and obscuring the effectiveness of incentives outweigh the pros imo
0
1
36
@xdev_10
X 🫐
2 years
All systems have their trade-offs and having to deal with front-running is one of the bigger issues for oracle based protocols, the post from Synthetix describes this in detail:
1
3
34
@xdev_10
X 🫐
2 years
The current iteration uses a two step transaction to prevent front-running: 1. User sends a transaction to request a position to be increased or decreased 2. A keeper fetches the median prices of exchanges, then executes the user's request
1
0
34
@xdev_10
X 🫐
2 years
The improved contracts automate most processes, isolate risks for LPs, provide better protection against price manipulation, allowing active adjustments to no longer be needed for things like weights and caps
1
0
33
@xdev_10
X 🫐
1 year
- We agree that as we hit these limitations it becomes more difficult to scale, the pool can be a large size but max allowed open interest is limited
1
2
33
@xdev_10
X 🫐
2 years
High impact functions require two transactions to execute: 1. Signal transaction 2. 24 hour delay 3. Finalize transaction
1
0
31
@xdev_10
X 🫐
2 years
To reduce the volume from front-running bots we introduced a requirement that price had to move by more than 1.5% or 3 hours had to pass before a position could be closed in profit
3
0
31
@xdev_10
X 🫐
2 years
Based on our analysis we have not observed any front-running volume with this change
1
0
30
@xdev_10
X 🫐
2 years
This is to allow the platform to adapt quickly if needed, ideally we would want to minimize the need for active adjustments and we are working towards that, we've spent a lot of time improving the contracts towards this goal
1
0
29
@xdev_10
X 🫐
2 years
Within the next few months the price feed will be iterated upon again, the majority of the work for this has already been completed and is undergoing audit as previously mentioned
2
0
29
@xdev_10
X 🫐
2 years
Contracts have been updated to allow for multiple oracle signers similar to Chainlink:
1
1
29
@xdev_10
X 🫐
1 year
- Isolated markets so that LPs are only exposed to the market they provide liquidity for
3
2
31
@xdev_10
X 🫐
2 years
While LPs were still in profit overall due to fees from regular users, this front-running issue would lead to a gradual draining of LP funds, this would be an issue especially in times of high volatility
1
0
30
@xdev_10
X 🫐
2 years
Chainlink price feeds are currently being used as a fallback and safety check, in case the keeper price deviates by a configured percentage
1
0
29
@xdev_10
X 🫐
1 year
- We're making a few changes to improve things:
3
2
30
@xdev_10
X 🫐
1 year
- Adding positive / negative price impact that is based on the skew of long / short open interest
1
1
30
@xdev_10
X 🫐
2 years
It was difficult to use the platform because of the spread at times, but more importantly, based on our analysis we found a significant amount of volume was from front-running bots, using the two price feed sources was insufficient to guard against this
1
0
27
@xdev_10
X 🫐
2 years
If the keeper stops functioning for some reason, the second step can be executed by the user themselves at the Chainlink price, three minutes after the request has been created More details:
2
0
27
@xdev_10
X 🫐
1 year
- The limitation of zero price impact is that allowed capacity at any point cannot stray too far from liquidity in the order books of reference exchanges
1
2
29
@xdev_10
X 🫐
2 years
We are analysing the differences to be sure that we can safely switch over in its current state and not have LP funds be at risk of draining from front-running
1
0
27
@xdev_10
X 🫐
2 years
The Timelock contracts allow some configuration to be changed without a two step process, for example, setting of open interest caps and token weights
1
0
26
@xdev_10
X 🫐
2 years
or if the cumulative change in price deviates more than the configured percentage, there will be a spread from the keeper price to the Chainlink price
1
0
27
@xdev_10
X 🫐
2 years
It reduced front-running volume by a bit, but was not fully effective because bots would still open their positions then hedge it on the platform to get around the requirement, the only difference is that they would have to pay more fees for that
1
0
26
@xdev_10
X 🫐
2 years
2. Price feeds The price feed source used in GMX has been iterated upon a few times The earlier version of GMX was Gambit (now deprecated) and was deployed on BNB Chain, the Chainlink price and current price on PancakeSwap were used, creating a spread between the two prices
3
1
27
@xdev_10
X 🫐
2 years
The signal transactions can be initiated by the admin of the Timelock If the admin signals a malicious transaction, during the 24 hour delay, the multi-sig of the Timelock can change the admin, which will prevent the malicious transaction from being finalized
1
0
26
@xdev_10
X 🫐
2 years
The current multi-sig holders:
1
1
23
@xdev_10
X 🫐
1 year
They are looking to expand their team, definitely worth considering if their vision of improving the accessibility of quality audits for small teams resonates with you
@0xOwenThurm
Owen | Guardian
1 year
If you have any smart contract security expertise or are interested in becoming an auditor join so we can tackle this mission head-on. If this issue resonates with you, retweet 🔁 this thread so we can reach more auditors and align the community.
4
8
18
0
2
15
@xdev_10
X 🫐
2 years
@SamuelAlpha77 @ChainLinkGod @n0madcapital @DefiDrew @tanoeth @GMX_IO @0xKepler @Riley_gmi @nullpackets for volatile tokens, would recommend the front running resistant feeds, those do not use Chainlink at the moment, but we are discussing how we can work together with them more
1
0
4
@xdev_10
X 🫐
2 years
@SamuelAlpha77 @ChainLinkGod @n0madcapital @DefiDrew @tanoeth @GMX_IO @0xKepler @Riley_gmi @nullpackets the oracle contracts allow two types of prices to be used, one would be the off chain oracle system, and the other would be Chainlink price feeds, Chainlink price feeds can be used for stablecoin prices
1
0
4