Ko-Windowソースのwsrv.cのGetRequestに"WindowServerのVersionが違います"を発見しました。そこから先は分かりません。(そもそもですが、ウィンドウサーバのコンパイルが難しいのであきらめモードです)

Nope it's correct. The docs here communicate that any imports from the server package (e.g. `getRequest`) are denied by default on the client. I suppose, more accurately there are 3 ways denial can happen: file patterns, specifier patterns, and marker specifiers.

Server: LiteSpeed Linux 4.15 - 5.8 (97%), Linux 5.0 - 5.4 (97%) GetRequest, HTTPOptions: | HTTP/1.0 200 OK | connection: close | this domain is not configured on this server. They just blocked the domain. I wonder how they knew. It sounds like they may be complicit.

Sadly @predictdotfun doesn't use @UMAprotocol. They decide about the outcome. Analyzed the contracts in question about the "Lighter Governance Token case". Here is the outcome. Resolver Contract: bscscan.com/address/0x242E1B… Oracle Contract: bscscan.com/address/0x76F42e… ChatGPT Analysis: That output is great — it means you’ve found the correct on-chain question record for the Lighter market and it tells you exactly how it will be judged. Here’s what each part means: What you’re looking at You queried the adapter/resolver mapping: questions[questionID] → QuestionData So this is the canonical stored definition of the market question that the oracle will resolve. 1) requestTimestamp = 1765125698 This is when the oracle request for this question was created (Unix time, seconds). If you convert it, it’s around early December 2025 (UTC). (Exact minute/zone isn’t critical; the key is: this is the timestamp you must use to locate the oracle request.) 2) reward = 0, proposalBond = 0, liveness = 0 This is a big deal. It confirms what we discussed earlier: this system is UMA-interface compatible but not UMA’s real dispute game. No reward for proposers No bond required No liveness window configured here In a standard UMA Optimistic Oracle market, these values are non-zero and are what enable public proposing/disputing economics. Here they are pinned to zero (their design). 3) resolved = false The question is not resolved yet. No final YES/NO has been written back. So currently: the market is still awaiting final settlement, or it’s pending proposal/settlement at the oracle layer. 4) creator = 0x193b96CcFdDEcb25077285A1Cbec01195a253c80 That is the address that initialized/created the question in the adapter. This is not necessarily the “oracle proposer,” but it’s the on-chain creator for the question record. 5) ancillaryData (THIS IS THE MOST IMPORTANT PART) This hex decodes to the exact market rules. Here’s what it says (plain English, based on your bytes): ✅ Title “Will Lighter launch a governance token before 2026?” ✅ Critical rule: memecoin airdrop does not count It explicitly says: The market is about Lighter’s governance token launch Whether Lighter launches a memecoin in the timeframe has no bearing on resolution So if they airdrop a token that’s not governance, that’s strongly aligned with NO under these written rules. ✅ Resolution condition and deadline It resolves to YES only if: Lighter launches a governance token by December 31, 2025, 11:59 PM ET Otherwise NO. ✅ Extra exclusion It says: “locked tokens or non-swappable tokens will not be considered a launch.” So even if a token exists but isn’t freely transferable/tradable, it shouldn’t count. ✅ Payout mapping (important) It defines: p1: 0 = NO p2: 1 = YES p3: 0.5 = unknown / 50-50 What you can do next (to see what the oracle will/has proposed) To see if someone has already proposed YES/NO in the oracle, you now use: requester = adapter contract 0x242E1Ba24f6fC524bfb410062Ca5689A9622613d identifier = YES_OR_NO_QUERY (likely) timestamp = 1765125698 ancillaryData = exactly that bytes blob Then on the oracle contract (0x76F42e...2531) you call getRequest(...) (or similarly named read method). That will show: proposer address proposed price (0 or 1) expiration time / state resolved price (if settled) Bottom line meaning for your original scenario “They airdropped a coin without governance capability.” Based on the ancillaryData rules you pasted, that should not resolve YES, because: it’s explicitly about a governance token it explicitly says memecoin launch doesn’t affect resolution “launch” excludes locked/non-swappable tokens So unless the token actually provides governance rights (and is considered “launched” by these rules), the written resolution logic strongly points to NO. If you paste the oracle getRequest(...) output for timestamp 1765125698, I’ll tell you whether anyone already proposed YES/NO and whether it’s final yet.

Here are some examples from my side 1. src/start.ts is only mentioned in the global middleware page. This makes you confused what the file is and if it should be included all the time due to its name 2. In the server functions page, there's no links to getRequest, getRequestHeader, etc. Even though it says you can get more info here 3. Code splitting is documented in TS Router, but not in TS start in terms of how to pass config in vite.config.ts. I had to assume there was a "router" property on "tanstackStart()" 4. Env variables runtime validation section I believe is incorrect. I tried that code and it lead to zod errors when importing any env, due to the other env also being there. I might have done something else wrong though and I fixed it with T3 env. Maybe worth mentioning T3 env? 5. I haven't seen in the docs to not use a server fn withing another one. I am still not 100% sure this is not allowed, but it doesn't seem to work for me 6. This is for TS router but still applies. In the Tanstack Query integration page, at the bottom it says redirects are auto-handled unless otherwise configured. I haven't been able to get this to work reliably and there's no examples or anything else to go on Overall, the docs lack depth. They are just direct to the point, e.g. "This is X, you can use it for Y". It lacks a lot of the time "Here's how" and "If you use Z, then also add this" and a link to more info This makes anyone using it have a bunch of question marks, and it makes it very difficult to have confidence that you're using TS Start the right way

I think there’s a “getRequest()” function you can import from react-start/server

🚀 @tan_stack Start is now in RC! If you're upgrading from devinxi@v1.131.50 to RC v1.132.6, here are the key breaking changes: - ServerRoute / createServerFileRoute → Route / createFileRoute - clientOnly → createClientOnlyFn - serverOnly → createServerOnlyFn - getWebRequest → getRequest - In router.tsx: export function createRouter() → getRouter() - Env setup: Add config() from "dotenv" to load env vars on the server For better SSR handling, use ClientOnly and createClientOnlyFn (see docs – Execution Model) ✅ Upgrade now available in the TSStarter repo for local development (Prod is still on v1.131.50) 📝 Blog post may come later!

Either way, an implementation detail I never plan on exposing. Just the abstractions like `getRequest` or `setHeader` a la Vinxi/h3

You're right that it's not required and could even get a big unhinged, but it is also very tedious *only* using function arguments to do all of your DI on async stuff. Most of what I'm referring to is something like a `getCtx` or `getRequest` that hides the underlying context.

Using async store is great DX and to me the "right type" of magic - I kinda wish it was `getRequest()` rather than `headers()` tho

If getting the request object is integral of being "magical", then I would do my best to make that magic explicit. // a.js const request = getRequest() b(request) // b.js function b(request) {} As a framework author, clarity would be my primary concern here. But if magic is optional, I would expose it as an argument at the top-most surface and let the user do whatever they want with it.

GETRequestバージョンか、、アドリブの掛け合いほんと毎回違うんだなぁーーー カウントに毎回しぼう。 あーーー好き💕💕

🙋‍♀️Anyone noticed that if you observe the event "controller_action_predispatch" in #Magento 2.4.4 $observer->getEvent()->getRequest()->getFullActionName(); 👉 fullActionName always gives "___" (three underscores) for CMS & category pages Bug or Feature? 🤔

🙋‍♀️Anyone noticed that if you observe the event "controller_action_predispatch" in #Magento 2.4.4 $observer->getEvent()->getRequest()->getFullActionName(); 👉 fullActionName always gives "___" (three underscores) for CMS & category pages Bug or Feature? 🤔

『セイリング』 またgetRequest第三段があれば、この曲入れてほしいなぁ、、 #k_tra

🙋‍♀️Anyone noticed that if you observe the event "controller_action_predispatch" in #Magento 2.4.4 $observer->getEvent()->getRequest()->getFullActionName(); 👉 fullActionName always gives "___" (three underscores) for CMS & category pages Bug or Feature? 🤔

Horde Webmail #opensource email client remains vulnerable to #RCE flaw allow server takeover. Details: ow.ly/GTON50JoeVo #Horde #vulnerability #CSRF #GetRequest #emailsecurity #webmail #appsec #Threatfeeds #SecureBlink

-> arrow <- $get->This$getRequest() $get = function -> = arrow This = Key word $getRequest() = building function

It's injected into the context later on. So inside a Vue component, I can just do "this.$getRequest" to send a proxied request. Pretty simple.