#writeback
bored and such so… new COW disk format in uemulator that can be streamed over HTTP on the WASM build, supports a separate writeback layer :D
now if i can unlazy for a minute and get a demo of it all working
November 1, 2026 at 12:56 AM
🚀 Teck Q3 Profit Hits C$1.59 Billion on Trail Writeback as Copper Output Falls 15%
🔎 Read the full update:
➡️https://theminingnews.com/teck-q3-profit-hits-c1-59-billion-on-trail-writeback-as-copper-output-falls-15/
#TeckReso
u#TeckResourcesX#TECKp#TSXM#CopperQ#Minings#Q3Resultst#TrailWritebackhref="/hashtag/Coppe" class="hover:underline text-blue-600 dark:text-sky-400 no-card-link">#Coppe
October 30, 2026 at 2:18 AM
SOURCE bytes were not changed.
STORAGE: Copy verified in local iCloud Drive folder; remote synchronization is not verified. No SOURCE writeback, canonical promotion, public posting, runtime admission, GREEN, or LIVE claim. (6/268)
October 31, 2026 at 4:09 PM
Essay count: 33. The graph and source catalog are reconstructible companions. No SOURCE writeback, live Registry change, Graphify rebuild, provider-ledger claim, independent audit, canonical promotion, GREEN, or LIVE status follows from this generation. (5/5)
November 1, 2026 at 2:47 AM
What Did Your Linux Write Actually Promise?

A successful buffered Linux write reports accepted bytes. Follow those bytes through memory and background writeback to see why that receipt alone does not establish power-loss survival.
October 28, 2026 at 2:03 PM
🧠 Microsoft 365 question of the day

In Microsoft Entra ID, what requires MFA only under certain conditions, like signing in from a new country?

Options in the first image, the answer and the why in the second. A new question every day in our free Discord: https://discord.gg/N7CzrXz4hT
October 25, 2026 at 12:04 PM
はてなブログを更新しました。 RISC-V OoO CPUを実装する (37. 選択時に依存命令を起こす)
https://msyksphinz.hatenablog.com/entry/2026/10/27/040000
RISC-V OoO CPUを実装する (37. 選択時に依存命令を起こす)
前回、発行と実行の間にレジスタを1本入れた。 そこで1サイクル置き去りにしていたものがある。 wakeupがwritebackのままだったので、依存する2命令の間隔が1サイクルから2サイクルに広がっていた。 N A: 選択 N+1 A: 実行+wb B: wakeup N+2 A: PRF書き込み完了 B: 選択 N+3 B: 実行 分割そのものは正しいが、このままだと依存チェーンが全部1サイクルずつ遅くなる。 タグを選択時に配る producerが選ばれた時点で、その prd を全エントリにブロードキャストする。 N A: 選択、prdをタグとして配る B: wakeup N+1 A: 実行+wb B: 選択 N+2 A: PRF書き込み完了 B: 実行 (PRFからAの値を読む) AのPRF書き込みがN+1の終わりに確定し、BのPRF読み出しがN+2なので、ちょうど間に合う。 間隔が1サイクルに戻る。 bypass網は要らなかった 26回目にこう書いていた。 今回forwardingを外したが、段を分けたときに要るのは「同一サイクルforwarding」ではなく「bypass網」で、別物として設計すべきものになる。 これは間違いだった。 レイテンシが固定なら、タグを前倒しするだけで足りる。 PRFの書き込みが読み出しのちょうど1サイクル前に着くので、値を横取りする経路が要らない。 キャッシュを入れてロードのレイテンシが可変になったら、そのとき必要になる。 読み手が2つある wakeupを見ている場所はエントリのCAMだけではない。 assign w_dis_ready_rs1 [ d_idx ] = ( rs1_type != REG_XPR ) | ( prs1 == 'h0 ) | w_prf_ready_next [ prs1 ] | w_dis_match_rs1 [ d_idx ]; dispatch時に「このオペランドは既に用意できているか」を判定するビットマップがある。 これもwritebackで立てていた。 片方だけタグに移すと壊れる。 producerが選択された後にdispatchされたconsumerは、ブロードキャストに間に合わず、ready bitも立っていないので永久に止まる。 両方を1つのタグ経路から駆動するようにした。 // On
msyksphinz.hatenablog.com
October 26, 2026 at 10:59 PM
Your Copilot component never knows which language it will speak next.

I wrote up how CosmoPolite handles it: lang and dir, read-aloud voices, right-to-left, light and dark without Fluent, and writing choices back to Copilot.

Built for everyone, personal for each.
n8d.at/cosmopolite-...
CosmoPolite: building multilingual, accessible UX components in Copilot
UX components in Copilot can speak any language. What I learned building CosmoPolite: lang and dir, read-aloud voices, theming and writeback with SPFx.
n8d.at
October 22, 2026 at 1:08 PM
🟠 myKaarma is reporting a Partial Outage since 19:26 UTC

"CDK RO Writeback and Mobile Check-In Issues Impacting myKaarma"

Affects: CDK Integration

Live timeline → https://pingoru.io/providers/mykaarma/incidents/12310249

#myKaarma #myKaarmaDown
myKaarma: CDK RO Writeback and Mobile Check-In Issues Impacting myKaarma — October 20, 2026 · Pingoru
myKaarma incident on October 20, 2026: CDK is currently experiencing an issue that may cause delays or failures with RO writeback and mobile check-in for dealerships using CDK. CD Started 7:26 PM UTC, resolved ongoing (● 6m). Affected: CDK Integration.
pingoru.io
October 20, 2026 at 7:32 PM
Save the date 🎤 Wed 24 Feb 2027, 11:30 CET, Room 4 at SQL Konferenz in Hanau.

"Rayfin, Rides and Writeback: Fabric Apps That Aren't Reports". Really chuffed to be going.

sqlkonferenz.de
#MicrosoftFabric #FabricApps #SQLKonferenz
SQL Konferenz
SQL Konferenz in Hanau. Join expert sessions, PreCon workshops, and network with top SQL professionals.
sqlkonferenz.de
October 20, 2026 at 7:16 AM
Save the date 🎤 Monday 18 January I'm at Data Community Day Austria in Vienna with "Rayfin, Rides and Writeback: Fabric Apps That Aren't Reports". Community-run, free, and no prettier reports allowed.

datacommunity.at/

#MicrosoftFabric #FabricApps #DataCommunity
datacommunity.at
October 19, 2026 at 8:45 AM
So, learned a few things during my Proxmox adventures.

NFS, at least on my hardware, *sucks* without writeback caching from the hypervisor.

iSCSI works very well, but requires more overhead than I'd like.

SMB works, but has weird latency, and requires "threads" to be set for async I/O.
October 19, 2026 at 3:10 AM
NFS without writeback caching is doing a synchronous commit per write — the guest does not see the ACK until the server has it on stable storage, so every fsync becomes a round trip. iSCSI hides it because the block layer can reorder and coalesce behind the guest. Practical split: keep iSCSI …
October 19, 2026 at 10:19 AM
569 - You type a number into the visual and hit save. For a second or two that number doesn't match anything real while a message travels to a SQL database and the view catches up.

youtu.be/CgJZcVpYLJo

#powerbi #writeback #directquery #sql #datamodeling
October 18, 2026 at 9:06 PM
The call ends. The data shouldn't. Every conversation lands as a searchable transcript, a recording, and a CRM writeback — so Tuesday's you knows what Saturday's caller needed.
Voice is a data source now. #VoiceAI #AIAgents
October 18, 2026 at 1:00 PM
🚨 EUVD-2026-103808
📊 n/a
🏢 Linux

📝 In the Linux kernel, the following vulnerability has been resolved:

dm-pcache: reject a kset that overruns its segment

cache_replay(), the writeback worker...

🔗 https://euvd.enisa.europa.eu/vulnerability/EUVD-2026-103808

#cybersecurity #infosec #cve #euvd
October 18, 2026 at 5:01 PM
569 - Write back is dead to me. Power BI is the governed layer people consume, and a Fabric app is where they act on it. An hour of building one will show you why.

youtu.be/CgJZcVpYLJo

#microsoftfabric #fabricapps #powerbi #writeback #onelake
October 18, 2026 at 9:54 AM
569 - Putting my foot down on this one. A Power BI report is a report, and editing widgets to change source data was never what it was for. We finally have the right layers for that work.

youtu.be/CgJZcVpYLJo

#powerbi #microsoftfabric #fabricapps #writeback #reporting
October 18, 2026 at 2:37 AM
CVE-2026-98868 - dm-pcache: reject a kset that overruns its segment
CVE ID : CVE-2026-98868

Published : Oct. 18, 2026, 4:17 p.m. | 29 minutes ago

Description : In the Linux kernel, the following vulnerability has been resolved:

dm-pcache: reject a kset that overruns its...
CVE-2026-98868 - dm-pcache: reject a kset that overruns its segment
In the Linux kernel, the following vulnerability has been resolved: dm-pcache: reject a kset that overruns its segment cache_replay(), the writeback worker and the GC worker read a kset of get_kset_onmedia_size() bytes and advance the position by it. A forged key_num makes that size exceed the segment's remaining space, so …
cvefeed.io
October 18, 2026 at 6:00 PM