Does it mean that EU actually wants to have their national services to have taps into the each IM platform disguised among other vendors?
The European Parliament's new Digital Markets Act, adopted as a draft law this week, could compel big platforms owned by large firms including Apple, Google, and Facebook to make their tech interoperable. Among other things, this might mean forcing the tech vendors' messaging apps to allow communication with other services. …
No, that's why KKR is trying to buy TIM and get control of Tier 1 internet operator Sparkle.
I understand El Reg is mostly an English news outlet - but the takeover bid launched by KKR, being a large EU telco involved and its Tier 1 network should have been interesting, especially because it has CLOUD Act side which is not irrelevant.
No it means it is time for the differing IM platforms to become a unified network, just like the PSTN and computer networks and email networks before them.
There really is no reason why someone shouldn't be able to use WhatsApp to send messages to some using whatever service and client they want, totally unknown to the message sender.
Next step unification and interop of the social media networks...
Re: "no reason why someone shouldn't be able to use WhatsApp to send messages to some using whatever service"
Only WhatsApp (and Telegram and Signal) are not concerned by this proposed regulation, as it applies to "number-independent interpersonal communication services" - as per the article.
So Teams will speak to Skype and Google Chat? How revolutionary and world-changing.
No, it wouldn't break end-to-end encryption. The message can be encrypted on the sender's device using a public key looked up from the recipient's ID, and decrypted on the recipient's device using the recipient's own private key which resides there. None of this depends on what channel the message is sent through -- it's still encrypted end-to-end.
What potentially would be broken by a measure that makes E2E encrypted messaging providers interchangeable, though, is the ability for the messaging platform to control the private keys -- and therefore, to benefit from information gleaned by decrypting messages in transit without making any of this information available to anyone else through whose equipment it may have incidentally passed.
But of course, technology companies would never, ever display such brazen contempt for their own paying customers .....
"There really is no reason why someone shouldn't be able to use WhatsApp to send messages to some using whatever service and client they want, totally unknown to the message sender."
I can think of a reason. Because it aint built that way. There have been open communication standards before so it already exists for text chat. Why doesnt the EU bugger off and make such a wonderful system? If its any good people will move (unlikely).
"Next step unification and interop of the social media networks..."
Until the EU makes itself an iron curtain where they can remain static while the world progresses.
Why doesnt the EU bugger off and make such a wonderful system? If its any good people will move (unlikely).
Which completely ignores the network problem.
It really, REALLY doesn't matter one bit how good or otherwise any system is - many users are forced to use a particular system simply because the group they need to communicate with are using that system.
So, for example, you send you child to school and the school informs you that they use WhatsApp for messaging and you'll need to have WA installed. Personally, my response would be "in that case, I'll be having a conversation with the information commissioner's office about them fining you out of existence" - but it seems few parents take that route. So the parent is forced to use WA otherwise they can't be involved in various activities.
The idea behind these rules is that the school might use WA, but the parent use Facetime, or ... That would break the network effect that the larger players exploit to lock out new competition, and in theory, alternatives could appear and work. Which is exactly why none of the services use and support open protocols - it's not in their interest to do so, it's very much in their interests to be as closed and proprietary as possible.
The same problem used to happen with mobile networks a couple of decades ago - before the days of "truck loads of bundled minutes" etc. Each network operator would charge a lot more for calls/texts to mobiles on other networks. So a teenager with limited funds would find it expensive to be on a different network to all their mates. So a significant financial incentive for groups of friends to all be on the same network - which was, of course, a benefit to those networks with the largest groups to start with, and a significant barrier to growth for any new entrants.
"Which completely ignores the network problem."
I am not sure how there would be a network problem. Calls and texts used to be cheaper to the same network (as you mention) until people just used apps which caused a drop in calls and text. If the EU made some open standard then it would be down to developers to implement something. With the abundance of apps there could be a wonderful explosion of apps. Assuming any of this is any good.
"So, for example, you send you child to school and the school informs you that they use WhatsApp for messaging and you'll need to have WA installed."
To which the answer is 'I dont have a smart phone'. Why this would be a poor excuse to force existing successful systems to do something entirely different doesnt make sense.
"That would break the network effect that the larger players exploit to lock out new competition"
Are larger players locking out the competition or are these separate systems from different groups that different people use based on preference? Forcing them to then have to cooperate because its such a pointless idea of no (or negative) value that nobody is using an open standard already to do this. If its such a good idea the EU can go make such a standard and see if anyone has any actual interest in it.
"it's very much in their interests to be as closed and proprietary as possible."
Lets assume this is true. Then it should remain closed and proprietary. Its theirs, they made it and its being made freely available to use. Anyone can go and make their own so go for it. Why must the existing and successful be stolen and corrupted by the incompetent and stupid who refuse to go make something themselves?
"The same problem used to happen with mobile networks a couple of decades ago - before the days of"
Aka market solution works.
Why must the existing and successful be stolen and corrupted by the incompetent and stupid who refuse to go make something themselves?
No-one is stealing.
But consider this. Go off and design a new phone system - make it really great, it'll do stuff no-one knew they wanted to do, absolutely fabulous. Only one glitch - it's not interoperable with any phone system currently in existence. So who's going to buy one of your phones & the service behind it when they won't be able to use it to communicate with anyone they know ? THAT is the network problem which the small number of big players use to lock out competition. Basically, until enough of your mates use a system, you aren't going to find it useful - and to get to that point, a lot of people have to take a punt on something that effectively "doesn't work" (where "doesn't work" equates to "can communicate with "some group").
The phone system, where you can pick up almost any phone anywhere in the world, and call another phone almost anywhere else in teh world works ... drumroll ... because of open standards. Those standards don't stop providers (for example) competing on other features (for example, great voicemail) - but they do allow a user on one network to call a user on another network.
As to the example of schools and whatsapp - AIUI that does happen, and the school's response is "tough - we use whatsapp, either you do or you miss out", safe in the knowledge that parents without a smartphone really are rare, and ones willing to (and knowledgeable enough) to stand up are rarer still.
"No-one is stealing."
These systems are private property. This topic is about forcing them to work as other people want (EU law makers) not the owners. So stealing.
"Only one glitch - it's not interoperable with any phone system currently in existence."
I thought this was about making them all work together so there must be some base standard for that so the EU could go and make one. And they could churn some apps like everyone else does. Put in the leg work to do something themselves instead of stealing someone elses work because success is hard.
"THAT is the network problem which the small number of big players use to lock out competition"
So why on earth would people use facebook messenger as its locked out of the other systems. And same with whatsapp. The who argument here is none of this is interoperable, yet they exist and are successful. If the network problem was a true obstacle then we wouldnt be having this discussion because they wouldnt all exist. But they do.
"Basically, until enough of your mates use a system, you aren't going to find it useful - and to get to that point, a lot of people have to take a punt on something that effectively "doesn't work""
Aka the legwork to be successful which the EU is not willing to make and so wants to steal.
"As to the example of schools and whatsapp"
I still dont see an issue with that and for the hold outs I know they are quite happy not to be bothered by their phone pinging all the time.
I'm in two minds about this.
On the one hand it would be nice if interoperability were the norm rather than the exception in the world of software development. It would make the lives of users infinitely easier, and wouldn't (in theory) make much difference to how developers work.
But on the other hand, protocols and file formats designed by committee are usually unwieldy and bloated to the point of being unusable. And said committees usually work at a snail's pace so technological advancements and improvements take far longer than they should. Of course fast moving committees are no better either, with iterations of specifications released with very little actual thought being put into them. (Hello, C++ committee.)
On reflection, I think what the EU and other governments should really aim for is for any broad protocol or file format to be made available either as an open specification and reference library, or licensable under frand like terms to encourage adoption at a wider level. I think that would strike a balance between interoperability and technological advancement. Maybe that's what they are trying to do - the article wasn't exactly clear on what their goal is.
Of course all of the above is irrelevant while companies continue to build services around the concept of data hoarding to sell to anyone with a cheque book. Maybe web3 service providers will have more incentives to be more open. (But I'm not holding my breath there.)
What's the point of it doesn't apply to Whatsapp et al?
That said, maybe the simplest would be to force all chat applications to support a common standard like RCS, and accept RCS messages to/from other services. They can still offer bells and whistles on top if they want, knowing that those will not be transferred to users on another service. Pretty much like iMessage supports SMS to a degree.
They might consider forging alliances with certain competitors to share some exclusive features that they have in common, as an advantage on other competitors that they don't like — pretty much like countries sign trade deals. RCS would be WTO, and everybody could choose whether to ally or Brexit.
You can do the same with Facebook, and I'm not sure it's even possible to stop iMessage from working that way. The problem isn't with the technical side, it's that most of the companies handling messaging have a vested interest in not allowing compatibility with anyone else. Apple have been very explicit that they use iMessage as a way to lock groups into the Apple ecosystem by making it a worse experience to communicate with anyone outside it; Facebook certainly do the same even if they've generally avoided saying it out loud. Google probably would if they could manage to go more than a week without killing off their latest chat app. It's only the smaller players, like Signal, that benefit from better compatibility, since the network effect tends to work against them rather than for them.
There is another practicality to consider.
If these comms are going between systems how is the integrity of the message going to be maintained?
Signal, WhatsApp, iMessage all use end to end encryption which is fine when they are on the same system, how is this going to work when they pass between systems? As presumably the keys aren’t going to be known or trusted by the receivers application.
Or is there going to be some gateway that will decode the message, code it to a new standard and then pass it on (great for snooping)
Also would signal users WhatsApp knowning who they are talking to?
SMS and email at least are based on standards everyone adheres to.
Mind you have a look at week or two back for the story about problems with getting a calendar to work between applications and that is static (mostly) data.
>If these comms are going between systems how is the integrity of the message going to be maintained?
PKI, Standards and unification - solved problems.
>Also would signal users WhatsApp knowning who they are talking to?
If the sender has your phone number, they know who they are sending a message to; just that they don't know you are using Signal.
>SMS and email at least are based on standards everyone adheres to.
And there is no reason whatsoever why the same shouldn't apply to other communications mediums: IM, social media etc.
The only issue is what is baked into every device, which determines the base level of interop..
>Mind you have a look at week or two back for the story about problems with getting a calendar to work between applications and that is static (mostly) data.
Given the incentive that also will be a solved problem.
Basically, we are just waking up to what was recognised as a problem in the 1980's and decided to "leave it to the market", and thus contributed to the rise of propriety standards such as the shifting MS office document formats...
It has always seemed so ridiculous to me, that I need to open one of several different apps to communicate depending on who I am communicating with. It's almost like having to have different SIM cards so I can talk to other mobile users on different networks. The way it was in the beginning, smtp/pop3 worked regardless of client software, and you could make your own. SMS worked on any phone, any network. Any browser could talk to any web server, mostly. These things remain today, but the newer messaging and social media platforms have setup for members only in the fight to dominate and this is the antithesis of what the internet original best intentions were.
I believe iMessage is nowadays built around XMPP and WhatsApp used to be based around XMPP. What I don't know is if the Signal protocol (that WhatsApp is supposedly now using internally but not in an interoperable fashion) is to some extent built around XMPP.
The reason XMPP is so often-used is because it's just simply a very well working protocol that is easily adaptable and extensible to new situations. That extensibility allowed encryption to be bolted on, and (with quite a bit of headache if I recall correctly) allowed it to be adapted to the mobile environment that did not exist when XMPP was developed. For instance, in the olden days people used to have one device with one IP address and their chat client on that device was either online or not ("presence"). Nowadays people are logged into the same account with multiple devices, some of which roam and may change their IP mid-session, some of which are data or battery constrained so sending keep-alive packets 24/7 is not ideal.
I am not in favour of governments requiring the use of XMPP for interoperability but I would have no issue with a requirement for the industry to create interoperability in a way that works for them. "This is the goal, you work out the best way to reach it".
If so many existing services are essentially built around XMPP then that should make the creation of bridges between the various networks easier.
but I would have no issue with a requirement for the industry to create interoperability in a way that works for them.
"a way that works for them" would probably mean a way that allows them to extract more private information. So need to take care that the industry players are prevented from collecting data from messagers that are not registered with their service. Or indeed requiring a native account before accepting cross messaging.
There was a time, around 15 years ago, when instant messaging *was* all XMPP based. You could chat with GTalk or Facebook users from your ISP's or free mail provider's XMPP service or from your own server for that matter.
Then the MBAs came along and shouted "monetisation!" and everyone rushed to disable federation, creating artificial scarcity.
> that should make the creation of bridges between the various networks easier.
In XMPP a bridge (or transport) is a very specific concept which essentially describes proxies to integrate non-XMPP services (other chat protocols, SMS, email, home automation, notification services, etc., Etc) with XMPP.
To connect two XMPP services together what you want is federation, which is a core part of the protocol and enabled by default.
I need to read the proposal, but I wonder if it contemplates the use of open protocols (XMPP) or merely closed interconnection amongst those "significant players". Because if it's the latter, all it's doing is perpetuating the current cartel situation and not improving the competition landscape in any way.
Thank you for the reminder to look into this.
I think a key premise of this article may be wrong.
The European Electronics Communication Code (EECC) came into effect across the EU only one year ago. Its implications for OTT providers are intended to be profound.
This https://assets.gov.ie/95410/a688502b-84d4-411d-9f62-80d4c29aa35d.pdf is a slidedeck presented directly by the Irish government to Facebook, Microsoft, Apple et al in November 2020, makes clear the intention to regulate them in the same manner as telcos (obligation to interconnect being a key remedy against dominance).
Libertarian commentators often hone in on the (equally key) obligation to provide for “lawful intercept”, and our politics will likely decide which of the two we citizens consider more important. The EU, being broadly left-of-centre in a global context, is enthusiastic about public networks.
This Digital Markets act may be designed to eliminate the potential greyness around whether a service is number based or not. Whatsapp is the 600lbs gorilla in this case - the genius innovation of Whatsapp was the way it harvests the ubiquitous public unique identifier of phone number to link to a proprietary client identify.
All of this has caused the PSTN to be progressively subsumed into private walled gardens. MS Teams is well on the way to doing “a Whatsapp” to the Business comms market. Many readers here will have taken refuge in Signal/Telegram, but these are short-term havens that are no substitute for a true public network.
Register readers will have watched the walled gardens spread over the years, and I imagine the majority will be wishing the EU commission success on this one.
The field of instant messaging is still remarkably diverse. Just off the top of my head, I can think of WhatsApp, Facebook Messenger, iMessage, Signal, Telegram, and Discord, most of which can fall back to SMS as a common interconnection protocol. The legislation in question is meant to open up interconnection between non-number-identified platforms, which seems oddly targeted insofar as I can't think of too many of those that people willingly use (Teams and whatever Google Chat is calling itself these days are the main ones that leap to mind), so do we really need legislation? From a technical perspective, I think we can all agree that better intercommunication would be nice to have, but it doesn't seem to be essential to consumer choice, and chat services can (mostly) at least still talk to one another in some fashion, so communication is not totally Balkanized. If I want to chat with someone on Facebook (which I don't, but stay with me here), it makes sense to use Facebook Messenger as an immediate chat mechanism, but when I'm not on Facebook, literally anything else makes more sense. As others point out upthread, technical solutions by committee, especially by government committee, are rarely ideal themselves, so we ought to ask which is the least bad solution, the imperfect market condition currently in action or a stifling government mandate, keeping in mind, of course, that this is not a problem on a scale with solving Global Warming but rather one of mild consumer inconvenience.
The devil is in the details. If you say "it needs to be able to send to users of other platforms", if everything supports SMS (there are already gateways that support SMS to/from the internet) they're covered. If it needs to support the same features for all platforms like presence detection, end to end encryption and so forth, then it would require dumbing down everything to the lowest common denominator, and everyone loses.
I don't see this as a problem that needs to be solved. If I have friends chatting with WhatsApp and I can't join in that group chat because I refuse to use that product, that's my problem. Why should everything be forced to be interoperable to lower that barrier? What's next, require apps to be supported on all platforms, so I can run Windows apps on my iPhone, an Android can run macOS apps, and so on?
Requiring app interoperability would arguably be much more meaningful to consumers in the world at large, if it wasn't hopelessly unrealistic. But since when has that stopped the EU do gooders from trying to enforce their will on the tech world?
"So how do you feel about Vodafone, O2, 3 and EE only allowing calls on their own networks? I mean if your friends had Vodafone SIM and you were on EE?"
That would be fantastic, try it. How long will it last do you think? And why do you think it would only last that long?
> That is how it worked originally for sms. e.g. You could not send messages (out of the box) between Vodafone, Cellnet, Orange and One2One.
Yeah I remember getting involved in SMS infrastructure at the tail end of the introduction of interoperability within the UK (1996-97) when there still was not full international interoperability (i.e. on Vodafone UK you could SMS Vodefone users in (some) other countries but not other OpCos' users in other countries).
At that time Cellnet in UK and MTN in South Africa were the biggest/busiest SMS operators in the world. Then in late 1997/early 1998 Japan introduced SMS for the 1st time (on "J-Phone" network) and woosh! they passed Cellnet and MTN traffic levels in a matter of months. SMS in Japan was a kludge as, unlike GSM and CDMA, the Japanese mobile standard PDC didn't have a separate control/data channel for out-of-band sending/delivery (i.e. during a voice call) so a data "call" had to be made by the phone or SMSC to send/receive SMSes. Still got a J-Phone advertising leaflet somewhere with Quentin Tarrantino on the cover. Fun times!
Biting the hand that feeds IT © 1998–2022