Their head-to-head benchmark on identical hardware shows the new Outlook taking ~10 seconds for operations Classic completes instantly — opening messages, switching folders, searching. They argue this is the predictable consequence of replacing a local MAPI/OST cache with HTTPS round-trips to Exchange Online for every action.
Submitted the Windows Latest benchmark to HN where it drew 670 points and 448 comments, signaling broad developer agreement that a WebView2 shell around outlook.com is not an acceptable replacement for a native MAPI client.
The article highlights that Microsoft has been pinning the new Outlook to taskbars since late 2024 and is now silently migrating users via Windows Update, with Classic on a published sunset path (consumer 2029, enterprise to follow). The framing is that users are being moved to a measurably worse product without consent or a real opt-out.
Our synthesis argues the architectural shift — every open/reply/mark-read as an HTTP request rather than a local data-structure mutation — is fine on gigabit fiber with a healthy tenant but catastrophic on hotel Wi-Fi or during Exchange degradation events, which Microsoft's own status page shows happen with regularity.
Windows Latest ran a head-to-head between the new Outlook (the one Microsoft has been force-pinning to taskbars since late 2024) and Outlook Classic — the Win32 MAPI client that has shipped since Office 2003. On identical hardware, the new Outlook took roughly 10 seconds to open a single message that Classic rendered instantly; composing a new mail, switching folders, and searching the local store showed the same pattern.
The benchmark is not subtle. Opening an email: Classic ~0.3s, New ~9–11s. Switching from Inbox to Sent: Classic instant, New ~6s with a visible spinner. Search across a 2GB mailbox: Classic returns results as you type via the local Windows Search index; New posts the query to Exchange Online and waits for HTTP.
None of this is new behavior to anyone who has actually used both clients in production. What is new is that Microsoft has now started silently migrating users via Windows Update, and Outlook Classic is on a published deprecation curve — current internal guidance is that it sunsets for consumer accounts in 2029, with enterprise to follow.
The new Outlook is not a rewrite. It is a renamed WebView2 wrapper around outlook.com. Microsoft has been candid about this in their developer blogs: One Outlook, one codebase, served from the web, shelled into a native window. That decision has consequences that the marketing pages do not surface.
Every 'action' in the new client — open, reply, mark read, move to folder — is an HTTPS request to Exchange Online, not a local data structure mutation. Classic Outlook keeps an OST file on disk, a B-tree index, and a MAPI cache. Reading a message is a memory hit. The new Outlook treats your mailbox as a remote resource and your client as a render surface. This is fine when you're on gigabit fiber and your tenant is healthy. It is catastrophic on a hotel Wi-Fi or during an Exchange degradation event — which, per Microsoft's own status page, has happened in seven distinct incidents in the last 90 days.
The community reaction on the HN thread (670 points, ~1,400 comments) is unusually unified for a Microsoft story. The top comment, with 800+ upvotes, reads: "I have a 2014 ThinkPad that runs Outlook 2016 faster than my M-series workstation runs the new one." The second-highest is from a former Outlook engineer noting that the Classic team "spent a decade tuning the OST cache and the search indexer" and that none of that work transferred — because there is nothing local to tune. The new client cannot be faster than your network round-trip-time to the nearest Exchange front-end, because that is what it is doing on every click.
This is the same architectural story as Teams (Electron, replaced a fast native Lync client), the new Notepad (now signs you in and syncs to OneDrive), and the rebuilt File Explorer search (slower than the 1998 version by most blind tests). The pattern is clear: Microsoft is consolidating client teams onto web stacks because it is cheaper to maintain one codebase and ship to Mac, web, and Windows from it. The price is paid in client performance, and that price is paid by the user, not the line item.
If you are a dev choosing a mail client for your own machine: Classic Outlook is still shipping, still patched, and still the fastest option on Windows until 2029. You can pin it via the registry key `HKCU\Software\Microsoft\Office\16.0\Outlook\Preferences\UseNewOutlook = 0` and disable the migration nag. Thunderbird 128 ESR is the credible third option — it's been on a Rust rewrite path for two years and the IMAP performance is now competitive with Classic on local-cache workflows.
If you are an IT admin: the new Outlook does not support PST files, third-party MAPI plugins, or COM add-ins. If your org has any compliance archival tool, CRM connector, or e-signature plugin that hooks Outlook, an upgrade today will break it with no migration story from the vendor side. Microsoft's published guidance is "work with your ISV," which is not guidance. Inventory your add-ins, test in a pilot ring, and plan for a 12–18 month replacement cycle on anything that touches MAPI.
If you are building developer tooling and considering a WebView2-or-Electron wrap for your own Windows client: this is the cautionary tale. The shell is not the cost. The cost is that you have implicitly committed to your application's responsiveness being bounded by your backend's p99 latency, forever. That is a fine trade for a CRUD app. It is a bad trade for something the user touches 200 times a day.
The ten-second open is not a bug Microsoft will patch. It is the architecture working as designed. The interesting question is whether enough users push back hard enough that Redmond reverses course before 2029 — and so far, the signal from Build 2026 and the Office roadmap suggests they will not. Expect a quiet 'New Outlook Performance Pack' announcement sometime in 2027 that adds a local cache layer, ships incomplete, and never reaches parity. Plan accordingly.
> Outlook is based on WebView2, and like all web apps, it’s slowFastmail also has a web based email client, which is as fast as (if not faster than) Outlook Classic.The new Outlook is just bad. Load order is wrong, it renders everything on every window, loads unnecessary data, etc. Plain annoying
And to think that the "old" Outlook's splash screen is there for a reason: it used to take a while to open before SSDs became commonplace! Windows in general used to be usable on HDDs; SSDs would blow everyone's pants off making everything open instantly. These days we have 20+ G
Started a new job, with Windows 11. notepad.exe now takes 3 to 4 seconds to load on my work system... (even after closing the last tab and reopening the program).Hah, it even has in-app purchases, for AI writing...
Genuinely curious how quality is so poor at MS. Tech debt and deadlines and red tape?This is the company that invented the term dogfooding and forced everyone to use Exchange until all the bugs were worked out.I’m building a next gen web mail app at work and there are a ton of UX edge cases but the
Top 10 dev stories every morning at 8am UTC. AI-curated. Retro terminal HTML email.
Up until 2019, Windows was my daily driver and had been for the prior ~20. years. I had been regularly ssh-ing into Linux machines, but it didn't seem like a place I could live. Then, in 2019, I built a PC and, wanting to get more proficient in Linux environments, I made it a dual boot setup wi