Microsoft to retire Skype Desktop API | Hacker News Reader

Microsoft to retire Skype Desktop API

networkworld.com

104 pointsnezza-_-72 comments

"The Desktop API was created in 2004 and it doesn’t support mobile application development." So what? Desktop users built Skype with their Skype-to-phone and phone-to-Skype (Skype Number) subscription money. We want what we paid for: Skype API, and whatever apps and devices are supported by it. Mobile will necessarily have different apps and devices with some (perhaps) overlap. Developers should be courted to redeploy their apps to a Mobile API if they like, and if their customers want. Microsoft should leave Desktop alone, rather than dumbing it down and damaging ongoing Desktop product value. Microsoft has been perennially wrong about product value. This is just one more bold misstep.

That's just the thing though, the customer base no-longer require the desktop API, because there is a new more attractive (mobile) market. Whatever amounts of cash we threw at them in earlier years matters nothing, nothing at all.

>These changes will significantly improve the call quality and speed of delivery of instant messages, while retaining excellent battery life of mobile devices.

How does scrapping a desktop API affect mobile in this way?

If the mobile skype app can call desktops, then maybe they interface with the desktop API.

Because Microsoft can update the Skype architecture so that it is more centralized like other IM applications. Right now a lot of Skype is peer-to-peer, which is not good on mobile. I, for one, am really looking forwards to this change.

I don't see how that affects the API.

    [API] --> [Skype] --> [Other Skype Clients]

Rather than:

    [API] --> [Skype] --> [Microsoft Servers]

Changing how the network is built shouldn't affect the API in this way since there's an abstraction layer (The Skype program). It would really only affect those who have reverse engineered the Skype protocol and built their own clients.

As for making Skype more centralized, yes, that will improve the experience for mobile users but it can be done without scrapping the API.

IMO, this is an excuse to scrap it and not a legitimate reason.

Skype moved away from peer to peer for supernodes. Here's the article: http://arstechnica.com/business/2012/05/skype-replaces-p2p-s...

I thought that MS had already re-engineered Skype, and most of the if the peer to peer operation was removed in favor of central MS server control. Which also had the side effect of making it possible/simpler for the NSA/FBI to gather information from skype conversations.

Windows powered laptops and tablets are and always were mobile devices for Microsoft.

All I can find on their developer webpage that they appear to support is Skype URIs. It's basically HTML tags that register the skype: service to send commands to skype. The functionality appears to be completely one way. The app sends a message to skype and skype takes over from there.

Skype API appears to be the same things as Skype Desktop API which is the ops focus.

SkypeKit which allowed developers to create a headless skype client is no longer accepting new developer registrations. https://support.skype.com/en/faq/FA12322/what-is-skypekit

I can't see anything at all about hardware that supports skype. This is probably through some enterprise BD channel. Hopefully they aren't being abandoned too.

A company I used to work for is panicking over this. One of their core products had an integration with Skype using Skype4Com for various reasons. They came to me recently asking if I had any recommendations for a better platform. (They knew Skype was a temporary hack but didn't expect it to die this fast)

Basically they need a white-labeled, web-based video conference solution. They'd like centralized recording with the recordings accessible via an API and support 1-to-1 and 1-to-N broadcasts. They've played around with TokBox and UStream but they're not the greatest fit.

If anyone knows of a company or is working on a startup in this space I'll gladly point them that direction. They're a global e-learning company with huge government/education contracts.

I know you want a service, and this is a product, but I'll mention it anyway: https://jitsi.org/

Not sure how integration would work.

I can highly recommend Jitsi over Skype. Jitsi also is snooping free and recommended for Video conferencing & VoIP on https://prism-break.org/

Unfortunately what I've found is that anything that claims "high security" or P2P technology means that they don't support the ability to record all broadcasts on a central server. I understand and appreciate where those products are coming from, but education institutions have policies that require all communications to be reviewed and audited.

Then they're probably going to have to roll their own.

Really? I would of figured anything that claims to be "high security" was just making delusional marketing points and was the same bog standard security you would get elsewhere.

You are less cynical than I.

GhotiFish,

Why do you give snarky trolls like kevinxucs advice on how to avoid/circumvent the hellban? If you took the time to read his post history, you'd see he's exactly the kind of 1 line, trollish, thoughtless commenter that we don't want to see in HN.

...

are you... looking at the same user I am?

https://news.ycombinator.com/threads?id=kevinxucs

I can make the following deductions about him.

1. This user is young, likely male, opinionated. None of these are problems.

2. This user tried take a shot at IE to posture socially, and did not expect that tact to fail.

3. When he was down voted, he changed his behaviour.

I told him exactly because of his post history. It tells a very clear story of what happened. A very clear and unfair story.

edit: Oh wait, you have show dead off, don't you? Yah turn that on, it paints a different picture.

I guess it would be possible to build a client that always connect to a recording peer (and don't allow to disconnect it)?

Since they seem to build a product around whatever solution they will pick, they could as well do that with jitsi. What am I missing?

At vLine, we provide a WebRTC-based cloud solution: https://vline.com. We're still working on some of the features you mentioned (recording), but would be happy to chat with them to see what we can do.

Thanks! I'll send them your way. A lot of service proiders seem to be in the same boat. They're all "working" on recording and say they get many requests for it. This company has reviewed dozens of services which were all very close to being what they want minus the recording feature. If you were able to deliver on that you would have a significant advantage over many other competitors.

Hey! I work on the BD team at TokBox - we're planning on having an archiving beta on WebRTC by the end of the month - if you want to know more feel free to drop me a line at edward [at] tokbox [dot] com!

Also would love to know more about why TokBox doesn't fit your/their needs!

TokBox is one of the leaders in this space: http://tokbox.com/

My word, that's an annoying home page.

I'd checkout https://layer.com

I haven't tried it out personally, but I've heard good things.

There's other, newer Skype APIs, I believe. CallBurner already announced they would still work, using another API.

Now that's interesting. Is there any link to the new API or at least to the announcement?

My company sells solutions to intercept any application playback and recording at the Windows OS level and we have a specific product targeted for Skype and other VoIP applications: http://www.nektra.com/products/audio-recorder-api/ we also sells binary instrumentation tools for companies who want to speedup their own implementation http://www.nektra.com/products/deviare-api-hook-windows/

Our technology is being used in the security field in companies such as BAE Systems stratsec, CyberPoint International, and for other uses like recording YouTube videos in http://www.jaksta.com/

Feel free to contact me, even if you just need some guidance.

Sales on HN, right before my eyes.

Well wadyaknow

It is help and completely focused on this topic, there are no many companies focusing on bringing APIs for closed desktop/server applications.

I used VSee on a project a couple of years ago: http://vsee.com/api

Looks like you are searching for www.dozeo.com. You can find the API at api.dozeo.com.

Who still uses Skype after it was confirmed[1] that Microsoft sells our personal text, voice and video conversations to dubious agencies?

Edit: Personally, I've switched to OTR (Adium/Pidgin) for text and ZRTP (Jitsi) for voice/video.

[1] http://www.theguardian.com/world/2013/jul/11/microsoft-nsa-c...

Yet another example of what happens when you depend upon proprietary software. The developers that are upset should have known better.

It's a little ridiculous to assert that no one should rely on proprietary software. Are all the folks writing Windows applications or PlayStation games idiots? A cautious approach toward relying on an API that's not central to another company's business is the key.

As a counter-point, the Facebook APIs are pretty crucial to Facebook's business (third party developers and application functionality for users), but they've seen plenty of instability and breaking changes in the past, many of which cause real problems for businesses.

If you're aiming for the user-facing side of your business to be self-reliant, under your own governance, and at the mercy of your market rather than of your partners, then it can and does make sense to look at self-hosting open source services and components; you pay a premium for the hosting, but you know that you opt-in to upstream changes on your own timescale, and you can fix issues yourself.

In reality it is indeed tricky to be so disciplined, and if some suppliers provide very high-uptime services with reasonable failure modes (e.g. Google Analytics), then it seems fair to go that route - but on the flip-side, data really is money nowadays, and user privacy may be important if you care about such things, so even these decisions can be questionable.

How is depending on proprietary software any different than consuming an API such as Stripe which could disappear tomorrow too?

If you move from Stripe to another payment gateway, you will not lose all your users. Microsoft has monopoly access control to the other users, who are needed to make your skype extension useful to the user using your extension. That is why Microsoft control over such user directory is unacceptable.

Anyone know is this will effect the Skype4COM library?

I built a Skype bot using it for work just a few months ago. Really annoying if it is.

According to this it will stop working: http://stackoverflow.com/questions/18822634/will-skype4com-s...

Oh well. That's so daft. Will be either abandoning that project, or switching to another alternative that does support such an API.

Maybe a new API that works well with mobile too will be released in time for us to switch to it. Wouldn't take much to improve on the API, it is (was) horrible.

Last time I tried to use the Skype API, it was horribly broken, and the documentation was downright incorrect in several places. I guess it's not had any love for a while.

The current leadership at microsoft seems to actively trying to destroy the company.

Compatibility is a word that could explain microsoft's success. Destroy that and their competitive advantage is none existent.

I dont want to say more... But a lot of decisions these days are "change for change sake."

I agree with you, and I hope they continue their glorious plan to follow current trends rather than be market leaders, all the way to the end-goal of being an also-ran or footnote in history.

After watching the last 20 years of Microsoft's rise, domination, and fall, I've concluded that no tech corporation should be the 800lb gorilla for more than 5 years. At that point, it should be considered an area ripe for disruption. I'm seeing Apple, Google, and Canonical (!) falling into that category now. The sooner they're disrupted/replaced, the better off the rest of us will be.

Why is canonical on that list? Well Apple seems to have hit their limit and are experiencing gravity.

I'm heavily vested in microsoft and just wish they'll stop trying to commit suicide.

Canonical is now starting to throw its weight around on things like Graphic interfaces (Semi-proprietary Mir vs. Wayland), User Privacy (Selling Unity Lens search results to Amazon), and then there's the Reality-Distortion Field; Shuttleworth is starting to become a little too much like Steve Jobs (harrassing people like the KWin Author), but without all the legal/business chops Jobs had. People are starting to become Ubuntu zealots. Not linux zealots or GNU zealots, or even those interested in Free Software. just zealots of the Ubuntu distribution. And if you say something against it, they'll downvote and flame away in response.

I moved to Arch and Gentoo, personally.

I wish you luck with your microsoft vestment. I'd advise you to diversify, however; Skills only relevant in a (rapidly decreasing) fraction of the computer market will only provide limited rewards in the future.

And this is why building a great product with the intention of selling it to the highest bidder is a fundamental betrayal of your customers.

>And this is why building a great product with the intention of selling it to the highest bidder is a fundamental betrayal of your customers.

Maybe you might furnish us with a list of business models you do approve of? How about:

- Building a crap product and selling it to the highest bidder?

- Building a great product and selling it to the lowest bidder?

Building a great product and selling it (when the offer is) made to the highest bidder is the basis a of large number of business models. Are you against selling companies at all?

How about:

-Building a great product, retaining ownership, and maintaining it for your user base to the best of your ability.

-Building a great product and selling it to someone who you're confident is committing to maintaining the quality of your product, even if they're not the highest bidder.

-Building a great product and selling it to the highest bidder, under pre-agreed terms that let you keep enough control to maintain quality.

People snark at me every time I suggest this. I really don't understand why. Hacker News is brimming with stories of "I'm sad because Google/Microsoft/Apple/Amazon/whatever bought a product/service I love and immediately ruined it/shut it down." I don't think it's unreasonable to wish that a product you use regularly and pay for (whether directly or through advertising views) be maintained in good condition and not gutted. Maybe I'm not cynical enough.

Skype's pretty much the worst thing ever anyway

Really, I don't get why people care about it. It's closed source, shady, loaded with spam bots and the company treats customers terribly.

Network effect. All my friends and family use Skype and Facebook. It's difficult to avoid using them when people have picked them as their sole method of communication.

Very much so. I use it at work because we use it for interviews, since most people have it, and their voice and screen-share are more reliable (on nearly every system) than alternatives (which people never have installed anyway).

We universally hate it. We haven't found anything better :(

Google Voice/Talk/Hangouts does about the same stuff and has similar market penetration, as far as people's familiarity with the company and likeliness that they already have an account.

Which is why we've been testing Hangouts for such things, yeah :) I have hopes. It's pretty nice for some things. That said, it isn't there yet.

It has the ability to share windows (awesome!), but there's a persistent watermark that's annoying and covers stuff up. Quality is often worse than Skype, which means waiting a few seconds while it clears up the text.

It has crashed and/or disconnected more often (never thought I'd say that, it's kind of hard to beat Skype here, but it has). Acceptably stable still, but it's still an annoyance.

For communication purposes, it doesn't have real persistent chatrooms, so we lose history if someone doesn't check in for a day or so. Same with searching the history.

All of which means we still have to resort to Skype for some things, at which point the annoyance of having multiple simultaneous communication systems outweighs our annoyance with Skype.

None of the alternatives meet our needs (google talk etc). The needs are: ability to set up rooms or group chats that you can favorite, ability to make calls and video chat, desktop based and always on at all times. If anyone knows of an alternative that can do all of this let me know.

I am a freelancer and literally all of my clients use Skype, most of them primarily so.

This is really screwing us up and will force us to cancel paid subscriptions prematurely. There was no outreach to the developer community to even find out what interesting and/or thriving applications are relying on the API - just an expiration date. I understand a new architecture for mobile is most important right now, but it's at the expense of pissing off a loyal group of Skype developers who could help further drive the platform's success.

who cares, webrtc is here!

Deployment in usable apps timeline? I'm in.

http://talky.io, http://vmux.co, http://vline.com

Don't forget http://tokbox.com !

Anyone know if you can still use an older client to use the api? I guess if you can it's just a waiting game till the client becomes unsupported.

I just got a forced update with no way to cancel. But I haven't gone to OldApps to see how far back to go so that A) forced updates don't happen and B) the API and apps still work reliably. Don't know which version at which those two criteria are satisfied.

> I just got a forced update

Heh, Linux baby ;)

Quite possibly.

I've taken a copy of the most recent standalone installer which precedes the api-warning-nag version. It appears to be able to install over the top of the more recent version, and when my desktop API is disabled I will install it and see what happens.

The change-log of the api-warning-nag version does not mention the warning popups and just lists a higher contrast UI as an accessibility improvement. I think I can do without those kinds of updates if it means keeping the API. Fluxon mentions a forced update, hopefully that was also a featured added with the api-warning-nag update.

The standalone installer has already been signed by Skype, so it can be passed around without worries about whether it's genuine.

(I would post all the version info but am on a different computer)

Hello, Google Hangouts!

Did anyone mention closed source? They just retired their XMPP interface... If there is any way to go, it's not something that is almost equally closed source.

Also Google+ and anything related has about the same reputation as Nokia in the Netherlands. I don't know how it is in the US, but here I'm quite sure I can't even find a single friend willing to use that service.

Nokia has a poor reputation in the netherlands? I don't think most people in the US realize that.

Well not poor so much as 'old'. Carrying Nokias by people under 30 is almost frowned upon and buying them is like buying a computer that has Windows XP on it.

Show me something that is a sufficiently functional open-source equivalent of Skype or Hangouts, and I'll use it. But in the mean time, I have bigger fish to fry than ridding my life of ideologically impure software.

Reply on news.ycombinator.com