Hacker News new | past | comments | ask | show | jobs | submit login
Microsoft Loop brings back Google Wave? (techcrunch.com)
121 points by borisjabes 83 days ago | hide | past | favorite | 109 comments

> Microsoft is bringing back Google Wave, the doomed real-time messaging and collaboration platform Google launched in 2009 and prematurely shuttered in 2010.

IIRC wasn't it only "doomed" because it was such a resource hog people could barely use it? From what I remember the people who could get it working loved it.

> Loop components, “atomic units of productivity”

OLE! But for the web! Seriously anyone else remember when OLE was a thing and you could drop a print shop pro banner in Word, click it, and print shop pro toolbars would appear? (I think it was print shop pro, if not some similar bit of software, its been 25 years so my memory isn't perfect on this exact point!)

> Google Wave was clearly ahead of its time.

So were the million other attempts to do this over the years, but hopefully the experience will get incrementally better with each go at it our industry tries.

This is how progress happens, one step at a time. Learn from the past and make brand new mistakes that the next team gets to learn from.

It was also doomed because of Google's terrible invite system.

The few of my friends that got in early couldn't do anything because there were so few others, and everyone else just forgot about it months later after the hype died down.

No idea why they keep doing this.

It worked for Gmail, but Gmail was so far ahead of everything else at the time that people really wanted in.

Limited invites can create buzz, Facebook did it well by inviting entire universities at once, that got around the problem of "no one else you know is on it."

Inviting purely random people to a collaborative system? Eh, not such a good idea!

It was doomed because you couldn't mix gmail and waves. That meant any real conversations couldn't be had. Waves ended up being just horsing around with your colleagues.

"This is neat!"



"That was cool."


"K. Bye."

Yes, I do remember OLE. My colleague and I were looking forward to CORBA which was promising to do similar things in a standardized way cross platform AFAIR. Did not gain traction.

The nice thing about OLE was that it was a relatively clean standard. You could use ActiveX in the browser and had multiple ways of creating them, but all those ways were focused on the programming language, rather than on frameworks.

Nowadays you have Webpack and React and Node and all that stuff where you rarely get to be coding but most of the time you're spending at managing all those tools.

I was into Google Wave robots development at that time and I had trouble seeing how this could scale. There was a ton of data sent to developer's App Engine instance for every edit in a conversation.

It also was a privacy nightmare. IMO it was just a test from Google to see how reliably this would work, I doubt they had the intention to let it run as a product.

But I really loved it.

I still use it at work, mathcad embeded in word... :(

Eh, it was a cool idea. Way ahead of its time, like a ton of 90s MS technology.

Hell they tried remote RPC stuff and what, 30 years later protobufs comes along and everyone declare's Google's library to be LOL amazing.

TBF I've used both COM and protobufs (in the same team!!!) and and protobufs was 100x easier to work with.

But still, some of the those old ideas of MS were really good, just IMHO they were limited by the languages of the time. Programming in raw C or 90s C++ was hard enough as is, trying to add entire new technologies on top of that, eh, not nice.

And of course back then all software releases were big bang versions every 3 or 4 years and you to support all previous versions basically forever, so yeah, difficult limitations to create new technologies under.

Lots of good lessons to be learned though!

While Google Wave obviously wasn't able to catch on from a product perspective, its tech DNA is in every real-time collaborative work/productivity app released in the last decade – Google Docs, Quip, OneNote, Slack, Figma, Notion. Loop is just next in line, and from what I can see doesn't bring anything crazy new to the table.

Etherpad predates google wave by like a year. I’m sure there were others.

Back in the early 1990s we used a collaborative text editor on SparcStations running OSF/motif. I can’t remember the program name but it was pretty incredible for the time and useful for CS classes. I think it was limited to 4 concurrent editors.

There will be something massive one day, because fun and productive human collaboration, so far, is a massive source of friction and waste. A lot of offices spend tons of times synchronizing, organizing shared work.. even if it's all for a one page document. One day this will be as fun as launching a Tetris game.

The interface looks much more similar to Notion than Google Wave; I don't see the similarity to Wave.

Microsoft is trying to bring the OLE (or whatever it was called) to the web. The technology that allowed you to drop for example Visio picture inside Word document and edit it there.

Their vision is that developers are creating these "smart objects" which you can then put in various places. You can have them in Teams discussion, Loop, Outlook email client etc. The contents is synced between users in real-time. Instead of passing links to various applications where the work is performed, you bring the work to (for example) Teams.

Sounds like the start of another hive of security vulnerabilities.

Which looks like Quip, which looks like OneNote, which looks like a million other note-taking apps before it.

I think the more interesting aspect is the extreme coauthoring; it looks like the underlying tech is open source as well which is pretty interesting for a lot of applications: https://github.com/microsoft/FluidFramework

The similarity seems more than skin deep though; the "Loop components" data model as it's described in the linked article sounds similar to the underlying data model in Notion.

Notion doesn't have the multi-cursor live coauthoring feature though; that's new.

Do you know if the Fluid Framework is built on top of CRDTs?

Didn't do a deep dive, but it looks like they have distributed data structures (DDSes) on that layer, which appear different to CRDTs (merging on the client, not on the server).

I can imagine someone could build a multiplayer game on this thing (though I doubt that's what MS designed for).

From what I recall of a BUILD talk once on the deep dive fundamentals, Fluid is based on CRDTs under the hood (or possibly the CRDT predecessor OT like Wave was), but Fluid believes writing CRDTs is hard (it is), especially writing them that obey the math laws CRDTs are supposed to, and then getting that to perform well (including and especially things like catch-up/replay) is also hard, so yes the actual programming interface Fluid presents is high level "distributed data structures" though the wire format is closer to CRDTs with some cheats from assumed knowledge of the data structures (things like sending the latest contents of a list for catch-up rather than replaying the entire CRDT chain for that list).

It's a bit of a shallow perspective to say that Google shut down Wave. Sure, they shut down the branding and distinct feature that was called wave, but it was essentially just the experimentation environment for synchronous editing that was then deeply integrated into the rest of the GSuite. It didn't die, it just went through metamorphosis.

Do you have a source for this? I don't think it played out that way. Google Docs and Sheets both had synchronous editing before Wave launched. At some point Docs was rewritten to use the underlying mechanism from Sheets. I'm not aware of any cross-pollination from Wave. The Docs and Sheets teams were fairly closely connected; Wave was developed in an entirely separate part of the organization, I believe the team was mostly in Australia, and there was never much communication AFAIK.

(Source: I was co-author of the original Google Docs implementation aka Writely; had some visibility into later iterations. I left Google in 2010 but have some idea of what went on subsequently.)

Does no one remember Etherpad? It had collaborative editing and Google snatched them up.

A year later they released Google Wave and Google Docs also got collaborative editing at some point.

Even then, Etherpad was not the first collaborative editor. It was just the first (to my knowledge) online one but I believe desktop ones existed for years.

And of course, the principle used behind collaborative editing (operational transform) is from at least 1989.

CRDT is the newer game in town now and it may replace OT entirely.

As a third party observer, I take it they both Google Wave and the collaborative editing in GSuite derive from Etherpad, and Etherpad itself is built upon the academic papers that predate it. If that is the case, while I’m sure some code may have migrated over, Google Wave “died” but Etherpad lives on.

Wave launched internally at Google to much fanfare at TGIF, and the whole company was somewhat amazed at the time by its realtime collaborative editing. Wave was built as a secret project in Australia, so when it did launch, internally there was understandably some rife / confusion as to what the strategy was.

The tech for appjet was primarily Javascript server-side, and I believe it used GWT for Java->Javascript client-side. This was seen as very odd to many inside Google, as javascript on the server was a new concept back then, and a far cry from the "officially supported" languages for development at Google.

While the initial launch of Wave went ~okay, the product itself had massive scalability issues. The Appjet/Etherpad team was then aqui-hired and quickly relocated to Australia. While one would think they were acquired for the similarities between Wave and Etherpad, they were instead tasked with "fixing" the javascript server-side performance situation. This made sense to management as appjet had been pioneering the concept since before Node was a thing.

In the end this wasn't a great outcome for the acquired team, but is typical of the aqui-hiring Google does.

Doc and sheets later added comments and realtime collaborative editing, but it was all reimplemented and never directly lifted from Wave's implementation.

s/tech for appjet/tech for wave/

From what I've heard: Google docs had multi-user editing, but it wasn't real time. Periodically the document would upload, be synchronized with other copies, and then update on the clients. Wave's real time cursor, synchronization, and conflict resolution algorithms were integrated into docs.

You're correct regarding the original Docs implementation. I had thought that the rewrite (launched around 2010 or possibly 2011) was primarily based on the architecture of Sheets. But it's possible there was some Wave heritage I'm unaware of.

Additionally, I believe the commenting and suggestion features were added to Docs after Wave was turned down; my recollection is that that bit is one of the main things that changed about Docs after Wave.

Google Wave was sold to us as a better replacement for e-mail. That was its main feature! It definitely died.

Sure they took some of the pieces and reused them in other products, but Wave definitely died.

10 years ago, I collaborated on Word docs by making changes and emailing it back. We’d look at revision histories to keep track.

Now the only email is the one that gives the link to the online document.

And I was the person who had to merge two peoples work if they both made changes on the same day with the 'master' copy.

Sometimes I wonder how we got out of that era alive.

But collaborative editing predates Google Wave. The innovation that Google Wave presented was using a unstructured shared document for all communication. That idea is pretty dead right now, and given that that is Wave’s main innovation, I’d consider Wave dead.

GSuite might have some code from Wave, but collaborative editing came from existing implementations.

It is well integrated in their office products, but Robots and Gadgets went away. Those were core to Google Wave, where you could search for games which you'd drop into a conversation and could play them right there inline with your chat partner(s).

I would agree here. Much of Wave was put into Google Docs, Drive, etc., as well as various other Google office collaboration products. Yes, Wave itself died, but it didn't fail.

Google does this with a lot of their stuff, it seems, too. Google Glass, for example, spread into so much of Android and further: Ok Google for the Assistant came from Ok Glass; the timeline made its way into Google Now and eventually into Assistant again; heck, you can see the beginnings of Material Design in the Glass UI. Google makes a ton of experimental products only to axe them swiftly and put their remains into new products.

Has anyone tried to develop anything with MSFT's Fluid Framework [1], which is used to build Loop? I have not got a chance to check details. But it seems quite nice. Any general remarks are welcome.

Particularly, I'm interested in understanding customizability of Fluid components, integration with React, and any concerns about Azure service lock-in.

[1] "The Fluid Framework is a library for building distributed, real-time collaborative web applications using JavaScript or TypeScript." https://github.com/microsoft/FluidFramework

I haven't used it, only read through documentation, but IMO Fluid's problem is not so much lock-in as an embrace of old-school columnar storage and handle-based object manipulation. An experienced Windows developer or game dev might feel entirely at home with the tradeoffs/footguns implied by https://fluidframework.com/docs/build/dds/#picking-the-right... ... but show that to a junior React developer and they're likely to be fundamentally confused, or worse assume that the only code example shown is a valid code example. (People writing documentation: please do not make one of the most prominent code examples in your Getting Started an example of what not to do!). And on the handle front, https://fluidframework.com/docs/build/data-modeling/#using-h... is similarly counterintuitive, to say the least.

Comparatively, I'm much more excited about Automerge https://github.com/automerge/automerge, which promises much friendlier developer ergonomics as simple as:

    doc1 = Automerge.change(doc1, 'Mark card as done', doc => {
      doc.cards[0].done = true
Contributor Martin Kleppman (of Designing Data-Intensive Applications fame) has great overview slides here: https://martin.kleppmann.com/2021/06/04/craft-conf.html . If anything, Automerge suffers from a "there's multiple ways to have a server backend, including P2P and centralized, and no one right way" anti-lockin problem, which is refreshing and also frustrating for people who just want to try something out. This is a solvable problem though!

Thanks for introducing Automerge, and pointing out the pros/cons! Will dig deeper.

I just took a look at the FAQ:


I think you have to consider your use case. If you want to build pure online collaboration tool it is fine. But if you want to build something that supports offline collaboration it seems not to be ready yet:

> Clients do have to be connected to the Fluid service. Fluid can tolerate brief network outages and continue operating but eventually the promise of being able to merge local changes weakens. We are investigating ways to improve this using other merging techniques designed to reason over large deltas but no final solution is in place today.

However, what I do like about their approach that they support multiple modern front-end frameworks from React to Vue to Web Components.

Usually projects like this are meant to be used together with Office 365 or for extending Microsoft ecosystem.

Straight from the README, but maybe there are other limitations:

"This project may contain Microsoft trademarks or logos for Microsoft projects, products, or services. Use of these trademarks or logos must follow Microsoft’s Trademark & Brand Guidelines. Use of Microsoft trademarks or logos in modified versions of this project must not cause confusion or imply Microsoft sponsorship."

(Disclosure: Work at Microsoft, but I work in Azure and some open source stuff, not on or directly with Fluid/Office/etc.)

That's just a trademark clause for Microsoft logos and brands. The Fluid Framework itself is MIT licensed [0] and doesn't require exposing any of those logos/brands when you use it, so the framework itself is fairly open for usage.

I think the main thing that would slow down adoption for Fluid is that the only "production" backend is an Azure service, which isn't part of the open source Fluid Framework. Other open source backends[1] aren't recommended for productions. Until there are some open source ones, I'd assume adoption will be limited to folks in the Azure ecosystem.

[0]: https://github.com/microsoft/FluidFramework/blob/main/LICENS...

[1]: https://fluidframework.com/docs/deployment/service-options/

I don't get your point? Doesn't that just mean if you fork the project you can't imply that it is sponsored by Microsoft? I have seen similar clauses in most open-source projects. This seems even more generous in that they don't demand you remove all the trademarks and logos just that you can't use the logo in a way that implies endorsement.

I am not saying whether you should or should not use this product. My point is that before using such products, you need to read really carefully where and how you can use it this product because of the included resources or intended purpose for using this. It is not sufficient to only read the README file and decide based on that.

The key part here is the second sentence "Use of these trademarks or logos must follow Microsoft’s Trademark & Brand Guidelines.". But there can be other such limitations also, just hidden somewhere else.

I'm stunned that techcrunch doesn't mention Notion instead of wave.

Collaboration technology startups are a ton of fun, but the problem is that eventually Microsoft always shows up to rain on your parade.

Notion, Airtable, and Coda always knew they'd have to deal with this at some point.

I was a heavy user of Wave for the brief period of time it existed. Although I struggle to think of how I will integrate this into my workflow today, I am excited to give it a try.

Wave was perfect for summer specific processes, like development of a tech specifications or making a complex contracts. Collapsing branches were much more usable than comments in Google Docs: you can't have a lengthy discussion in Docs, but in Wave it was natural.

Still, the team has to be disciplined, or else Wave quickly became chaotic.

Seems like a pretty direct rip-off of Notion. It's almost as if it was deliberate.

Shouts to all in here that think they were the only ones who liked Wave. I tire of the forgetful historians in here or just rehashing the old news/its place in the history of the time.

Sure this is wave. Sure this is notion.(which isn't the grandiose lightbulk product fans of it seem to think it is, not to mention its significance way overblown... you think Microsoft cares what hipster-designer/dev doc collab party you're in?)

The collaboration on any kind of docs/work is just natural and expected progression now. Anyone else just find this not that exciting?

furthermore, all these kinds of productivity tool 'advancements' and evolutions come down to people. Wave didn't fly because people didn't get it or wouldn't adapt. Sure, there were some really great parts of it and so they were extracted and moved into other apps where people could warm up to them and see some advantages.

But it's people. And it's tools. And it takes time for them to adapt and grow and old habits die hard.

Microsoft has taken 10 years to get to the point where its userbase may be ready for realtime collab apps that bring together multiple types of docs. We're talking about a userbase, and an age, where there is still tons of doc emailing and filename-ver-2-Copy.doc going on. Yes, lots of us have moved on from that thanks to perhaps more niche openness to new editing and sharing methods, Notion-heads, Figma jams, etc. But MS is just getting there now and who knows if customers will even end up using it since they've got other ways to collab in and out of the ecosystem on different scales.

But again, meh -- not that exciting. Just tools.

Microsoft can't make a quality clone of slack so I can't believe this will be a quality clone of Notion

Are you talking about Microsoft Teams? Because I hate Microsoft but Teams blows Slack out of the water by a country mile in terms of usability, reliability, and resource usage. I can't get a basic voice call working in Slack for longer than 45 seconds Teams conversely has smooth, reliable video delivery.

All in all I think Discord is better than either of them but I'll take Teams over Slack any day.

I admit the voice/video chat in teams is really well done. But that's because they are using Skype's technology for that. But the actual interface, messaging, and overall experience of Teams is really bad. Just yesterday I wasn't able to do a cmd+a to select all the text in my message box. It's a clunky and annoying daily experience

Hopefully it doesnt slow down as much though

This looks like a potential competitor to Notion more than Wave (as others have pointed out).

It's worth pointing out that Microsoft has let its OneNote product wither on the vine in recent years, ceding use cases and market share to competitors like Evernote, Bear, Craft, and even Obsidian. People have been asking for Markdown support and the ability to insert code blocks in OneNote for years; maybe Loop will end up being the de facto replacement and contain those features.

I'd be happy with Find&Replace in OneNote, but that's not on the table. They even have a handy help page [0] that needs to be seen to be believed: explaining how to copy the text you want to replace, Ctrl+F, input your search, and just repeat Ctrl+F, Ctrl+V!

Someone actually took the time to write that doc page instead of implementing a basic editor feature...

[0] https://support.microsoft.com/en-us/office/find-and-replace-...

Did they really cede market share to Evernote much? Isn’t Evernote ceding market share? I have always assumed Bear and Craft are pretty niche. Aren’t they Apple ecosystem only?

I was one of the (evidently very few) users of Wave. While it had many features, the one I remember now very fondly was proper message threading. It wasn't until years later that I used my first email program (mutt, of all things) with simple, correct threading.

I'm still somewhat annoyed that major email providers don't support proper threading. I wonder why. Is it because email is supposed to be "easy" and threads are cognitively "hard"?

Yep... that's the thing that killed it... everything was constrained to a threaded chain of replies, instead of being a proper multi-user realtime editor.

It was trying to be Reddit, Slashdot, HN, etc... a threaded forum post, instead of hypertext you could markup, annotate, or edit.

Define proper threading?

It wasn't flattened like in many consumer and business targeted email readers these days. Which was a sharp contrast with Google's other main communication product, Gmail, which flattened threads into conversations.

This should be called OneNotion

…for Business!

So this is same thing Microsoft did to Slack with Teams but with Loop directly towards Notion?

The fact that it is completely free and open source, tells you that it is a straight 'extinguish' which was rare but now they have gone and done just that (Again).

This is what 'extinguish' looks like with the new Microsoft.

Wasn't Wave the cloud service that promised to be a federated application, but delivered the open source server very late (like after the failure/shutdown was announced)?

Comparing anything that is not federated with Wave sounds odd to me.

Wave was federated? First i've heard of this. Where can i find this OSS server?

Also, federated seems like the least important wave feature that people talk about.

At the time federation was not that special. E-Mail was federated and Jabber/XMPP was federated too. So it was expected, that the successor (as which Wave was marketed) of e-mail was federated too.

The real time aspects, on the other hand, were pretty amazing at the time. From todays perspective the reverse might be true ;-)

The underlying architecture was federated; basically an XMPP extension. When Wave shut down, they handed the federated server code to the Apache Foundation, who briefly developed it under the name "Apache Wave". Apache Wave was shut down in 2018, but you can still download the code.

That product doesn't look like Wave.

It's a shockingly shameless clone of Notion.

Can we stop putting any of these product ideas on pedestals, as if they were acts of technological genius? They are all incremental interesting improvements, often thought of around the same years with unique insights, but not some revolutionary inventions no one has tried to build before. They are what they are because of their execution.

I'm referring to design.

And Notion is a better clone of Quip.

And both are a shameless clone of Microsoft Sharepoint - it all comes full circle

Which was built to combat Lotus Notes

Which is a clone of Confluence, which is a clone of any wiki software.

can you really clone a broad category? Wouldn't we just call that an alternative implementation?

Wouldnt be so funny, would it?

Would parent's statement be more accurate if they said MediaWiki?

Isn't SharePoint a clone of lotus notes (without any of the cool parts)?

A new era of productivity tools by Microsoft is needed. All of their current tentpole products are just web versions of their old apps, and most of them are very annoying or slow or just weird. I'm intrigued to see where this leads as they double down on their web-first strategy.

As for the Wave comparisons, it's more of a technical comparison of the Fluid framework and Wave's real-time collaboration data model/editor, right?

> and most of them are very annoying or slow or just weird

It's kinda mindblowing that in $CURRENT_YEAR we're able to do run beautiful-looking ray-traced video games on a home console that costs a few hundred dollars, but when I launch Microsoft Word on a several-thousand-dollar MacBook Pro it takes over 8 seconds from a cold start (just timed it right now).

We've all just accepted that 8 seconds is an acceptable time to wait to type some words into a computer.

My crappy HP laptop running Windows 10 starts Word in less than 2 seconds... you do have an SSD, don't you?

> Loop tracks your cursor position in real time. That’s the current state of the metaverse for you right there. Nothing says I’m present and active in a meeting like moving your cursor around, after all.

Great, can't wait to automate my cursor to look awake in meetings while being resting away from the computer.

Google Wave was amazing technology stuck with the worlds worst user interface. I recall that each person's edits were stuck in their own boxes, it wasn't a fluid multi-editor document system that it should have been.

Does Notion host their source code on github ?

I realize that you comment is likely a joke but... hmm, would having Copilot v100 complete a near clone of a some competitor's software be considered as good as clean room implementation [0], legally?

[0] https://en.wikipedia.org/wiki/Clean_room_design

For the record, Notion lives up to the hype.

Serious question, has anyone ever used a collab product in the way described (3 cursors and several people typing)?

Yes, in my company some of the fastest proposal writing has been done during lockdown as 3-4 people co-author the same document online via Word. It works very well.

We use it for meeting agenda / notes.

The note taker takes the note, while other people queue up the topics to talk about next.

I'm not suggesting this is the best way to take the meeting note and I'm happy to hear about the alternatives, but it is reasonably usable.

It doesn't look gorgeous either. People wouldn't like to distract the meeting through the doc, which is projected to the large screen in front of everyone.

NCIS has been doing this for years: https://www.youtube.com/watch?v=u8qgehH3kEQ

Often in Google Docs and Notion, however usually it's in separate sections of a document. I've never seen three people writing different points on the same bullet list - sounds like it would result in an inconsistent writing style or conflicting information.

We have a brief weekly status meeting where we have a word document that's used as a reference - where one person owns each section. Rather than share screen, everyone just opens that file, and see the live changes.

It works OK.

No, but as others have said I've seen it with multiple people in different parts of the same document. Demoing 3 cursors on screen at the same time expresses the capability of the product in a couple seconds, explaining how you should actual use it would have taken longer and killed flow.

We use Miro realtime with live cursors etc. in online workshops. It's ok.

Only for vacation planning where people are researching and inputting suggestions on a spreadsheet, or where people are brainstorming on a docs sheet.

I did that kind of thing often enough when i was in school, usually while in a room together with the other people editing the document.

I think I experienced it with Google Docs once. So much context switching happens until you cannot switch contexts anymore. Then you are just numb.

So they announced it but didn't launch anything real yet... like there's no where to go to demo it?

This seems like a carbon copy of Notion which seems to have a lot of users so may be a better comparison

I don’t see the notion in using Notion’s UI here albeit with Miro/Figma flying cursors and pointers

Looks like the evolution of Microsoft OneNote which I like and use daily.

This looks promising compared with collaborating in word

It is a HUGE improvement.

I used the publicly available fluid preview for a while and it is miles better using word for any kind of collaboration. I have since gone back to using Word for collaboration and I despise it. I feel like pulling hairs off my scalp.

I would say it is a competent new tool that gives you everything that notion did, while leveraging the kind of integration microsoft already goes well. The improvement is particularly noticeable not just because the product is good, but because what it replaces (word online) was so incredibly bad.

Disclosure - I would put a disclosure about my employer here, but don't want to be seen publicly bad mouthing another actively maintained product by my company.

On what, if you don't mind the question?

Where can i try this?


So... Notion?

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact