Microsoft's reusable component framework for Yammer.com | Hacker News Reader

Microsoft's reusable component framework for Yammer.com

github.com

78 pointsluisrudge31 comments

Regardless of what you might think about Yammer's UI, a repo full of tested React components written in TypeScript is very valuable as a reference.

You might also want to look at the very comprehensive office 365 react components from MS which is build on. https://github.com/OfficeDev/office-ui-fabric-react

Meant for developing Sharepoint/Office Apps but released under MIT License (Only the custom Microsoft Fonts are restricted but easily replaced)

i remember when you'd get a fully featured ui library when you bought a compiler.

Nowadays the compilers are free. I guess you get what you pay for.

As a reference, sure.

I've always liked the hover cards that they use on Yammer, the examples of those are pretty sweet.

It's pretty amazing that even this has generated negative comments. This is one of the reasons why I don't feel like I should ever open source any code. Anything I ever release would be something that was of use to me and maybe a few other people. Would I have to actively use and support it for the rest of my life to be "right"? I wonder how many of you complaining actually never open source anything while using lots of it.

I guess it's lame to complain about negative comments on the internet, but hacker news is supposed to be different. Open source from Google? That's fantastic. Open Source from Intel, Facebook, Yahoo, or HP? Those are OK too. Open source from Microsoft? Better make sure you've said something negative to make sure that your fan boy card doesn't get pulled.

Office 365 is awesome and getting better every day.

Full motherfuckin’ stop.

If they can get teams to not have so much whitespace it might even be a feasible replacement for slack.

Have you ever tried to contribute to or use a Microsoft backed open source project before? I have. I've also worked as a partner, SQL certified DBA for the best part of 20 years of my life. Considering we can't even get anything fixed when we're paying for partner support, you're SOL most of the time. I had an issue open for 8 years that we paid to get fixed and all they did was ship a lousy registry fix from first line support. So we have to deploy that fix to about 2000 people because two teams won't take ownership internally. Nice job.

It's not about fan boy cards, it's about the fact that MSFT visibly doesn't give a fuck about contributors or clients and will quite happily steamroll over everything whilst their marketing department fanfares repeatedly about how they're embracing open source and they love Linux.

You'll find that Open Source to Microsoft means business as usual, just on github, with added marketing fluff.

So because you have experience with Microsoft (a huge company with many products and projects) that means that anything that they do in open source is automatically BS?

I'm actually in close contact with the Microsoft product teams that I depend on, and while I wish they only listened to me, they don't...but it's not the end of the world and it doesn't color every perception I have of them.

Even when things are quite bad and they don't have an answer I want to hear...that doesn't mean that I can use that experience as the only criteria for all of my future interactions. What about the stuff that they get right? Does that have any value?

People here go crazy about privacy and Intel's ME. They don't flip a crazy bit when Intel releases something open source or judge every product against their hatred of the management engine. Does Intel support every single project that they open source forever? How about IBM?

If you go to McDonalds and someone serves you a turd instead of a burger three times in a row, do you go back again?

If IBM or Intel served me a turd, I wouldn't go back again. In fact IBM served me a helping of WebSphere once and I didn't go back.

They don't only sell one thing. They are the definition of a complex company and environment. Not everything they do is bad. Some things they do really well.

Seriously, you had a bad experience with one IBM product. Would that keep you from looking at other things they do? They do lots of amazing things and make some platforms that define the sector for solving certain kinds of problems.

That was one of several IBM products that I had a bad experience with. Lotus Notes too for example.

(In which I forgot about Notes and beg forgiveness....Sheesh. I had a bad experience with it too.)

Yeah it wasn't much fun was it :)

So you want to be the angry guy in the corner holding a grudge, that's cool. The Microsoft of yore was really yucky, and I'm happy I'll probably never have to consider whether SQLServer is the right DB for a job and that "a real webbrowser" has caught on as an accepted casual term for any browser that isn't IE.

But this isn't that, and the rest of us here kinda just want to get on with it and build cool stuff and this helps us do that.

I prefer pragmatist. I think we have the same goal really which is avoiding anything that is a commercial risk, money and time sink and that’s the status quo with this particular organisation even to this day.

Ok, great. This isn't a commercial risk (MIT license) nor a money sink (free). So time sink remains, and whether it is a such should be reasonably quick to work out (give it a spin, see what happens).

No, I don't think we have the same goal, and you certainly do not come off very pragmatically.

The beauty of actually free open source licences is that you don't need Microsoft to give a fuck. All that stands between you and fixing something is one click on "Fork".

Sure, it would be better if they were interested in building and engaging a community (and who says they're not? It's a huge company and a lot has changed in the past few years, but fair, you don't want to take anything for granted), but releasing code, tests and build instructions under one of the most liberal licenses around is so close to the mark that no, complaining isn't really warranted.

Let’s be realistic. How many forks, other than a personal fork for patches, are actually viable? This is one of those attractive fallacies like owning a Tesla service manual.

Complaining is warranted when you are a paying customer and you are told to post a GitHub ticket which is ignored...

It depends on what you mean by viable. Plenty of projects have fixes and features that live in forks and solve real issues for real people. Plenty more have bits and pieces copy-pasted into other repos, forming the basis of a new project. Rather than focusing on the expected viability of a fork, the real value is that it's possible at all to have one.

You can't make other people do stuff for you for free. Sometimes, as you've relayed your experiences, you can't make them do it even if you pay. At least open source gives you the option of fixing it yourself, or paying a third party of your own choice to do so.

Hey, I work on Yammer, and I'm one of the contributors to this project. YamUI is a cross-team effort between our Design and Frontend teams to create drop-in components for Yammer.com that have strictly-enforced style and behavior. As @gregod said, it is based on Office's awesome Fabric project: https://github.com/OfficeDev/office-ui-fabric-react/

The project is just us working in the open. We're not trying to push this as a new component framework for others to use, and we probably don't have the bandwidth to support external consumers, but we're happy if people can learn from it.

fantastic work, man. great to see this out in the light of day :)

Did you guys notice that upon returning from a link to yammer to the github page via the back button, a bunch of cards are visible until you scroll?

Yup, that's because those hovercards render on `componentDidMount`, and the styleguide renders all components on page load. It's a bit janky, but it only affects the styleguide, so it's low priority.

Disclaimer: I work on Yammer, and I've contributed to this project.

Yammer.com? Now there’s an unhelpful endorsement. It’s been a while since I used Yammer, but I remember its UI as pretty damn awful.

Plus like all MSFT frameworks, it'll be abandoned about 3 minutes after it is released.

Citations for the downvoters: silverlight, lightswitch, Velocity, AppFabric, Windows Workflow, half of WCF, ...

half of WCF

Why just half ? :)

It's one of the slowest UIs I know about. We also have a Yammer Embed on our Intranet and it's slow and buggy as well.

So reusable components from a crappy UI! Thank you Microsoft :~D

Well i guess it's nice they open source it but in reality nobody will bother with it. Bootstrap is where the action is these days.

Reply on news.ycombinator.com