Monday, January 4, 2021

Project Monarch

Zac Bowden:

Project Monarch is the end-goal for Microsoft’s “One Outlook” vision, which aims to build a single Outlook client that works across PC, Mac, and the Web. Right now, Microsoft has a number of different Outlook clients for desktop, including Outlook Web, Outlook (Win32) for Windows, Outlook for Mac, and Mail & Calendar on Windows 10.

Microsoft wants to replace the existing desktop clients with one app built with web technologies. The project will deliver Outlook as a single product, with the same user experience and codebase whether that be on Windows or Mac.

Via Steve Troughton-Smith:

Sad to see Microsoft consciously eroding its native platforms, drawn to the idea of a local ecosystem full of web apps like a moth to a flame.

Previously:

Update (2021-01-05): John Gruber:

I have no idea how big the “footprint” is for the current Mac Outlook app, but if it’s based on Electron how could the footprint not be enormous? “Hello World” in Electron is huge.

Version 16.44 of Mac Outlook is 1.98 GB. The Electron version would only be a fat binary for the non-JavaScript portions, and it would likely remove lots of features and code, so it certainly could be slimmer despite using Electron.

See also: MacRumors.

Update (2021-01-06): Colin Cornaby:

The Electron runtime is about 1.2 gigs alone combining the Intel and ARM versions. It’s so large that Electron apps aren’t distributed as universal so far.

Hard to see how it gets slimmer on Mac.

Microsoft’s plan on Windows is to build these things against the version of Chromium now built into Windows (via Edge.) So things might get slimmer. On Windows.

Nick Heer:

Some reports have interpreted this as though Microsoft will discard the Mac app redesign it previewed in September. I am not sure that is the case. The new version of Outlook for Mac looks an awful lot like an Electron app already.

Like most web apps in a native wrapper, this sounds like a stopgap way of easing cross-platform development at the cost of usability, quality, speed, and platform integration. To be fair, I am not sure that anyone would pitch today’s desktop Outlook apps as shining examples of quality or speed, but I spend a lot of time from Monday through Friday in the Outlook web app and it is poor.

13 Comments RSS · Twitter

If MS can scale this down to replace their mobile apps, it’d be an immense achievement to go from five (iOS, Android, Win, Mac, web) versions of Outlook down to one.

@Chris. An immense achievement of the "Yes, but why?" type. They're throwing out decades of bugfixed code and sacrificing all the benefits of having native apps (speed, using standard OS conventions and integrations, etc). In exchange for... what, exactly? This is Microsoft. Issues of not having the resources to support multiple versions of the app do not apply. I really don't see the advantage here, except to be able to check off certain trendy buzz words on someone's list or to claim (falsely) lower development costs that amount to pocket change for the company.

@Glaurung
I can’t speak for the Windows side, but Apple’s pace of OS development demands a lot of resources just to keep your apps running on the latest versions. We’ve transitioned from PPC to Intel, to ARM. From 32 bit to 64 bit. From “lickable” GUIs to flat design. From the Wild West to the Sandbox. From standard res to Retina. Once they’re off that treadmill maybe they’ll have time for more and better features in Outlook and not worry about having to painstakingly port them to every OS they support.

Old Unix Geek

Why?

Costs: one codebase to rule them all. (potentially iOS, Android, Linux, MacOS, Windows and the Web).

Access to programmers: There are also a lot of UX programmers who only know Web Technologies these days.

Ease of use for users: one UI to learn across all platforms.

... and an email client / calendar doesn't need to be very CPU intensive, except perhaps for search.

@Chris All the thing you mentioned happened over a 20 years life span. And they are not difficult. Especially for Outlook which requires very little if any ISA specific code.

>Sad to see Microsoft consciously eroding its
>native platforms

Nobody cares about native platforms. People care that when the IT department tells them on Yammer how to do something in Outlook, it requires five different explanations, depending on what exact combination of operating systems and Outlook versions people use, and then still only works for half the people.

An immense achievement of the “Yes, but why?” type. They’re throwing out decades of bugfixed code and sacrificing all the benefits of having native apps (speed, using standard OS conventions and integrations, etc). In exchange for… what, exactly?

I’m guessing two big driving factors here are:

the Windows version is full of Win32 legacy code. It’s newer than Word and Excel (Outlook only appeared in the mid-90s, although I’m not sure how much code from its predecessors Schedule+ and Mail it inherited), but still very old. Making fancy UI in Win32 is exhausting.
in contrast, Teams has been evolving far more quickly. There’s a lot about it I’m not happy with, and it does have that reek of “this is a web app” (for example, try scrolling through message history… the local caching is high-latency and buggy, with images frequently disappearing), but I can’t argue that things that used to be hard to do, like video conferencing and screen sharing, work just fine, across multiple platforms.

I imagine there’s a bit of an internal competition where the Outlook team got jealous of the Teams team (durr), perhaps, or more explicitly got told by a higher-up to pick a modern UI framework and move forward. It’s just a bummer that, once again, Microsoft’s own frameworks fall by the wayside. They could’ve made the Windows app UWP. They could’ve waited for MAUI to ship and make the Windows, macOS, and mobile apps MAUI.

(Perhaps MAUI being delayed-ish was another factor in this decision.)

Nobody cares about native platforms. People care that when the IT department tells them on Yammer how to do something in Outlook, it requires five different explanations, depending on what exact combination of operating systems and Outlook versions people use, and then still only works for half the people.

Thirty-seven years into the Mac, then, we’ve come full circle: the user no longer matters as long as the IT department can cut costs.

>the user no longer matters as long as the IT department can cut costs.

Users do matter. It matters that features only ship to half or a third of Outlook's users at a time, because all versions of Outlook are built on different technologies, have different release cycles, and different feature backlogs, which leads to confusion and unhappiness.

Users do matter. It matters that features only ship to half or a third of Outlook’s users at a time, because all versions of Outlook are built on different technologies, have different release cycles, and different feature backlogs, which leads to confusion and unhappiness.

That’s a false dilemma, though. “We either want to half-ass all platforms or do one shared platform well” is a choice Microsoft made that they didn’t have to make.

>That’s a false dilemma

No, it's really not. Managing complex software like Outlook, which has a ton of external dependencies, is already difficult. Managing it across five different platforms (Mac, Windows, Android, iOS, web, some of which have different sub-version, i.e. there are at least two different web versions, one built on modern technologies and a fallback version for people on older systems), using five different technologies, and thus multiple different development teams, with their own PMs and POs, and then trying to keep release cycles in sync, is essentially impossible. I can't think of a single company that does this.

This is just the reality of software development.

The idea that Microsoft is "half-assing" this because they can't keep these platforms in sync is just not a realistic assessment of the difficulty of this problem.

"Users do matter. It matters that features only ship to half or a third of Outlook's users at a time, because all versions of Outlook are built on different technologies, have different release cycles, and different feature backlogs, which leads to confusion and unhappiness."

...and as everything gets moved to web tech, we'll have confusion and unhappiness because "features" will get rolled out with little or no warning to entire enterprises all at once, regardless of platforms, so help desks will get slammed with "button was there yesterday, but not today" questions every few days because some "design" expert wants to play with pixels. ...and if you think a new "feature" hurts productivity, then you'll likely be told that you're a technophobe who "hates change." Harrumph.

Old Unix Geek

A 2 Gb email client? That's hilarious. I really did join the wrong timeline.

These are funny times. The full-fledged win64 version of Office 2007 (Word, Excel, Powerpoint, Access, Outlook, Published) which I still use because IT JUST WORKS requires just over one-third of the space.

Leave a Comment