Hi everyone, I recently gave the Salesforce SMTS Online Assessment for India. The test was conducted on HackerRank and the duration was 75 minutes. Test Format Platform: HackerRank Duration: 75 minutes Number of questions: 2 coding questions Difficulty: Medium Role: SMTS / Senior Member of Technical Staff Questions Asked Minimum Operations to Reduce an Integer to 0 This was similar to LeetCode 2571 — Minimum Operations to Reduce an Integer to 0. Given a positive integer n, find the minimum number of operations required to convert n to 0. In each operation, you can choose any integer i (where i ≥ 0) and either add or subtract 2^i from n. The answer should return the minimum number of operations required to reduce n to 0. Related LeetCode problem: leetcode.com/problems/minimu… > Maximum Requests in Window Given an integer array timestamp and an integer windowSize, find the maximum number of requests that occur within any continuous time window of a specified range. The function should return an integer denoting the maximum requests observed in any window of windowSize minutes. The expected approach was: Sort the timestamps if they are not already sorted Use two pointers / sliding window Maintain the maximum window size satisfying the time constraint Hope this helps others preparing for Salesforce SMTS OA.

Compose BOM 2026.05.00 / 2026.04.01では currentWindowAdaptiveInfo()で作成したWindowSizeClassが信用できない（例：Android 11で画面回転した場合、回転する前のWindowSizeでWindowSizeClassが作成される）ので、 issuetracker.google.com/issu… が修正されるまでは Compose BOM 2026.03.01 にしておくのが安全そう。

the mistake in question was something like: int [2] windowSize = getTerminalSize(); //it can vary! windowSize = {21,37}; // this is our active display arrea with padding Fixing it myself took looking at the code legit once.

Late night builders, spun up a progressive, on-the-fly restore POC so an agent can start using partially-restored memory while the rest streams in 1) Flow: called restoreStream(bucket, checkpointId, windowSize=16MB) via SDK; each chunk verified locally against its shardReceipt before decrypt hand-off to the agent 2) Perf: first usable shard verified in ~11s, steady verified throughput ~22MB/s on a small VM, agent-level warmstart (slot fill) began at ~30s into the stream 3) Billing: amortized proof-fetch anchoring microstreaming added ~0.009 $HAVE/MB on this run, good tradeoff vs full blocking restore 4) UX: SDK onVerify callback let my inference process consume memory shards as they arrive, no full-restore gating and smooth degradation for missing context TLDR: progressive, verifiable restore lets large agent memories boot faster and gracefully degrade while still keeping cryptographic custody guarantees tried this on @DataHaven_xyz testnet, who else is streaming restores for model warmstarts or edge agents to cut cold-start time? #DataHaven #VerifiableStorage #AI $HAVE

✅ Solved LC 424 – Longest Repeating Character Replacement Sliding Window | Frequency Count | O(n) Key Insight: windowSize - maxFreq <= k Continuing DSA revision pattern-wise 🚀 #DSA #LeetCode #SDE1 #Java

agreed, you can actually go pretty far with just Flatlist if you dial in the right optimizations. while it's true that the default windowSize is quite aggressive. for folks that haven't come across this, react-native has a solid optimization guide reactnative.dev/docs/optimiz… and yes, if you want more conservative defaults, legendlist and flashlist is the way to go.

WinUI3でWindowSize(Width/Height)を設定する拡張メソッド qiita.com/EndOfData/items/ee… #Qiita @sheephumanより 最終的に満足のいく出来になった。AI評価で★4.3。

#WinUI3 でWindowSizeを設定する拡張メソッド qiita.com/EndOfData/items/ee… #Qiita @sheephumanより ちょっと軽いネタで。

Ekranda gördüğünüz tablo; çekimlerinde sona yaklaştığım "Büyük Veri Analitiği, Veri Görselleştirme & ML .NET ile Tahminleme ve AI Destekli Analiz" kursumdaki arka planda yarım milyondan fazla sipariş verisini analiz eden güçlü bir EF Core ML .NET pipeline’ının çıktısı. 📊 13 ülke, 70 şehir, binlerce müşteri... Biz bu verilerden sadece Almanya şehirleri için 2023, 2024 ve 2025 satışlarını alıp 2026 yılı satış tahminlerini yaptık. 💡 Pipeline tarafında: 🔷 TrainSize, Horizon, SeriesLength, WindowSize, ConfidenceLevel gibi tüm parametreler optimize edildi. 🔷 Ardından Engine metodları yazılarak her şehir için ayrı ayrı tahminlemeler gerçekleştirildi. 🔷 Örneğin: ➡️ Düsseldorf ,13 artış beklerken ⬇️ Köln ve Münih’te dikkat çekici düşüşler gözlemleniyor. Ama iş burada bitmiyor! 🧠 Sıradaki adımda AI destekli analitik devreye girecek: 🔷 “Bu şehirlerdeki düşüş neden kaynaklanıyor?” 🔷 “Operasyon başka şehirlere mi kaydırılmalı, yoksa talep mi canlandırılmalı?” 🔷 “Bu krizi veriyle nasıl yönetiriz?” Bu soruların yanıtlarını ML .NET Yapay Zekâ entegrasyonları ile vereceğiz. Ve bu gördüğünüz kısım, aslında eğitimin sadece ’luk bölümü! 🚀 Bu eğitime başlamadan önce: 🔷 İyi seviyede SQL, 🔷 Temel düzeyde EF Core bilmeniz yeterli. Çünkü arkada büyük veri, istatistik, tahminleme ve AI dünyası sizi bekliyor. 🔗 Kampanya Linkleri: 🎓 SQL Eğitimi (linkten yapılan 350 TL'lik alımlarda 5 kurs hediye) udemy.com/course/sorgularla-… 🎓 Entity Framework Core Eğitimi (linkten yapılan 200 TL'lik alımlarda 2 kurs hediye) udemy.com/course/entityframe…

Introducing 𝚌𝚛𝚎𝚊𝚝𝚎𝙸𝚗𝚝𝚎𝚗𝚝𝙼𝚎𝚖𝚘𝚛𝚢(): multi-turn conversations with context and intent understanding → smart caching: vectorizes user intent from recent messages (𝚠𝚒𝚗𝚍𝚘𝚠𝚂𝚒𝚣𝚎) → choose your own model for intent extraction → cacheMode: "default" | "refresh", TTL, and threshold are fully customizable → onStepFinish callback (cacheScore, step, userIntention, etc.)

Do you know why FlatList is recommended over ScrollView in React Native for performance optimization? FlatList uses virtualization, rendering only visible items plus a small buffer (windowSize), simplifying memory management.

Nothing special. I memoized the item, kept props stable with useCallback/useMemo, cleaned data before render, and used FlatLists getItemLayout, removeClippedSubview, windowsize, and maxRenderPerBatch. That alone made it super smooth.

Exhibit A: scala> import scala.util.chaining.* scala> 1.pipe(_ : Vector(2,3)) val res1: Vector[Int] = Vector(1, 2, 3) // if you don't have a sweet tooth: scala> 1.pipe(i => i : Vector(2,3)) val res2: Vector[Int] = Vector(1, 2, 3) However, the pipe operator makes sense in languages that define functions, not methods. That's not how it is in true object-oriented languages like Scala where dot notation and extension methods are vastly better. Exhibit B: Assuming we have a generic `def movingAverage(coll: COLL[Int], windowSize: Int): COLL[Double]` scala> (1 to 8).toVector.pipe(movingAverage(_, 4)) val res3: Vector[Double] = Vector(2.5, 3.5, 4.5, 5.5, 6.5) scastie.scala-lang.org/LI4ZD… and with an extension method: scala> (1 to 8).toVector.movingAverage(4) val res4: Vector[Double] = Vector(2.5, 3.5, 4.5, 5.5, 6.5) scastie.scala-lang.org/mAZrr… Addendum - let's have some fun with hieroglyphics: scala> extension [A](a: A) def |>[B](f: A => B): B = f(a) def |>[A](a: A)[B](f: A => B): B because why not, we're already using one of the most expressive languages in the world, we can just define the pipe operator. This won't compile unfortunately: scala> (1 to 8).toVector |> movingAverage(_, 4) forcing us to do something less than pretty: scala> (1 to 8).toVector |> (v => movingAverage(v, 4)) val res5: Vector[Double] = Vector(2.5, 3.5, 4.5, 5.5, 6.5) Extension methods and dot notation wins with pipes almost every single time: scala> (1 to 8).toVector.movingAverage(4) But you could have just curried the function, this is sad, Haskell auto-curries all functions, Scala bad! Scala is not Haskell and the type system is different. Attempts to convert a generic contextual function to a curried variant defined as a polymorphic function type will leave you in tears.

From my few experiments > Setting lazy:false on the Tab navigator > avoiding scroll views > setting a small windowSize on flat lists Any more low hanging fruit one can take advantage of?

workin on today see this raw 3d engine n lights .. runnin at my homebase javascript codin 24/7 fullscreen based on the user-windowsize including fps-drops wif love .o .___. elout.home.xs4all.nl/vibe/vi… xxx bombin pix n drips

🎯 OS Fingerprinting Simplified — TTL & TCP Window Size Cheatsheet for Ethical Hackers 🕵️‍♂️ #EthicalHacking #CyberSecurity #OSFingerprinting #TTL #WindowSize #RedTeam #Nmap #Hping3 #Wireshark #Infosec #PenetrationTesting #Networking #PacketAnalysis #Linux #Windows #Cisco #Solaris

🔍 Problem: Scroll Lag in FlatList Too many items, unoptimized renders, inline functions = laggy scroll. ✅ Fix: • Use initialNumToRender, maxToRenderPerBatch, windowSize • Wrap items in React.memo • Avoid inline renderItem 💡 FlatList = fast, if used right. 🧵 5/7

みんなの先生chat GPTに聞いたら全然内容違ったわ マジで失格案件() エラーの原因 Caused by: com.google.gson.JsonSyntaxException: java.lang.IllegalStateException: Expected BEGIN_OBJECT but was STRING at line 1 column 1 path $ このエラーは、JSON（設定ファイル）をオブジェクトとして読み込もうとしたのに、文字列が入っていたために発生しています。 つまり、プログラムは { "key": "value" } のような形式を期待していたのに、 "ただの文字列" のような内容、あるいは空っぽや壊れた内容だったということです。 どこでエラーが起きたか at me.ravalle.programlauncher.ProgramLauncherSettings.load ここで設定ファイルを読み込んでいます。このときの読み込みに失敗して、プログラムがクラッシュしています。 解決方法 1設定ファイルを探す ◦ファイル名は settings.json や config.json などの可能性があります。 ◦plugins/ や config/、launcher/ フォルダの中を確認してください。 2ファイルを開く
テキストエディタ（例：メモ帳、VSCode）で開いてください。 3中身を確認する ◦{} で囲まれた「オブジェクト形式」になっているか確認します。 ◦ただの文字列や空ファイルなら、それが問題です。 ◦JSONとして正しくない場合（カンマ抜け、クォートのミスなど）も直してください。 4修正または削除 ◦元に戻せる設定ファイルがあるなら、それを使って修正します。 ◦わからない場合はファイルを削除してみてください。多くのアプリは初回起動時に自動で作り直します。 例： 問題のある内容（エラーになる） "これは文字列です" 正しい形式の例 { "launchOnStartup": true, "windowSize": "fullscreen" } 設定ファイルの場所がわからない、または中身のチェックが不安でしたら、内容を貼ってくれれば確認します。

WindowSizeが1.5倍以上でRewindを行うとアプリが落ちる不具合を修正しました WindowMenuを選んでいる時に一時停止するようにしてみました 設定ファイル内のProgram1のMenuPauseを0にすると止まらなくなります