In a world that is now gradually adopting HTTP/3 (which, as you know, is implemented over QUIC), the problem with the missing API for QUIC is still a key problem.
There are a number of existing QUIC library implementation now since a few years back, and they are slowly maturing. The QUIC protocol became RFC 9000 and friends, but the most popular TLS libraries still don’t provide the necessary APIs to make QUIC libraries possible to use them.
Example that makes people want HTTP/3
For a long time, many people and projects (including yours truly) in the QUIC community were eagerly following the OpenSSL Pull Request 8797, which introduced the necessary QUIC APIs into OpenSSL. This change brought the same API to OpenSSL that BoringSSL already provides and as such the API has already been used and tested out by several independent implementations.
Implementations have a problem to ship to the world based on BoringSSL since that’s a TLS library without versions and proper releases, so it is not a good choice for the big wide world. OpenSSL is already the most widely used TLS library out there and lots of applications are already made to use that.
Delays made quictls happen
The OpenSSL PR8797 was delayed back in February 2020 on when the OpenSSL management committee (OMC) decreed that they would not deal with that PR until after their pending 3.0.0 release had shipped.
“It is our expectation that once the 3.0 release is done, QUIC will become a significant focus of our effort.”
OpenSSL then proceeded and their 3.0.0 release was delayed significantly compared to their initial time schedule.
In March 2021, Microsoft and Akamai announcedquictls, an OpenSSL fork with the express idea to ship OpenSSL + the QUIC API. They didn’t want to wait for OpenSSL to do it.
Several QUIC libraries can now use quictls. quictls has kept their fork up to date and now offers the equivalent of OpenSSL 3.0.0 + the QUIC API.
While we’ve been waiting for OpenSSL to adopt the API.
OpenSSL makes a turn instead
Then came the next blow to everyone’s expectations. An autumn surprise. On October 13, the OpenSSL OMC announces:
The focus for the next releases is QUIC, with the objective of providing a fully functional QUIC implementation over a series of releases (2-3).
OpenSSL has decided to implement a complete QUIC stack on their own and with the given time line it sounds like it will take them a few years (?) to ship. And instead of providing the API lots of implementers have been been waiting for so long, they explicitly say that it is a non-goal at the start:
The MVP will not contain a library API for an HTTP/3 implementation (it is a non-goal of the initial release).
I didn’t write my own QUIC implementation but I’ve followed the work of several of the implementations fairly closely and it is fairly complicated journey they set out for themselves – for very unclear reasons. There already exist several high quality QUIC libraries, why does OpenSSL think they need to make yet another one? They seem to be overloaded with work already before, which the long delays of the 3.0.0 release seemed to show, how are they going to be able to add a complete new stack implementation of top of this? The future will tell.
On October 20 2021, the pull request that was created in April 2019, is finally closed for real as a “won’t fix”.
Where are we now?
The lack of a QUIC API in OpenSSL has held us back and with this move from OpenSSL, it will continue to hold us back for an uncertain amount of time going forward.
QUIC stacks will have to stick to using or switching to other libraries.
James Snell, one of the key contributors on the QUIC and HTTP/3 work in nodejs tweeted:
Back in February of this year, I received one of the most chilling emails in my life when someone threatened my life: “I will slaughter you“.
That email penetrated something deep into my soul and heart and I was not quite myself for a few days.
Almost six month later (a few days ago), I received another email from a person who claims to be this “Al Nocai” from that first threat. I read the beginning of this email several times before I actually could make myself continue past the opening paragraphs. Slightly worried what to find, with the memories from the last time coming back.
This email however, says “an apology” in the subject. I believe this email is written by the same guy.
The long and full apology is inserted below. The gist of it is: he claims to have been the victim of all sorts of bad stuff by several people named “Dan” and I got lumped in there because of my first name. He (?) also says he suffers from schizophrenia.
I’m publishing this partly because I got a lot of attention when I made the initial threat public so I figure it could be interesting for some to learn about this development. I can of course not verify that this is in fact the same person nor will I even attempt to verify any of his many claims of wrongdoings against him.
I’m happy for “Al” that he’s getting help and tries to move on. For me, this apology at least finally proves that this threat is over and in fact never was intended literally. I hope I will never receive anything close to that again.
The apology in full
I am Al Nocai. When I contacted you initially, I believed you to be a Dan E., from texas, or a Dan S from delaware or a Dan from Minneapolis. I didn’t do my research, and when I found it was actually you and you had nothing to do with my situation, I became indignant and even more of an asshole. You had every right to be mad, and publish as you did. I’m not trying to justify what I did, there is none, I should have been a lot more cordial. I just want to provide context around what was happening, I believe I at least owe you why.
I had to retire from my career do to schizophrenia. Again, I should have not let my delusions go to the point they did nor should I have acted the way it does. My illness doesn’t detract from the rashness of my actions.
I, at the time, lost my defense project after getting hacked through the Department of Veterans Affairs in the US. It was a few years into development, and it was meant to be a pathway to get homeless veterans off the streets. I was trying to develop a “trade-route” in tech. At the time I was very lucid. I was volunteering regularly in my community and I had a great family life. In October 2020, there was an attempt on my life which left me bludgeoned and near death with 2 orbital fractures. These event led me to uncover a massive money laundering operation and then had ties elsewhere. I then got hacked. When I say hacked, I lost every device. They rooted my charge arbitrator, I was bios bonded and I basically lost every document and all my software.
The people who did it gloated to me about it to me through linkedin. Trying to take my computers back, I ended up QAing a lot of their malware. This led to be being whaled for months. People attempted to blackmail me, they stole my identity and I was lied to about what I was doing as the people approaching me hid their relations to the people who initially hacked me.
These stressors where the predicates to my psychotic break. My federal correspondence about the hacking was re-routed. I had people impersonating microsoft employees. I even had a $50 billion dollar mortgage servicer trying to sue me over tweets. I also had firms like outpost 24 doing MiTM attacks on me, which is also a factor as I didnt realize at first where all my EU Certs were coming from in my Active directory.
Last, just to give you a clearer picture at how much I was monitored: I ended up not being able to take anything back in Windows. They were in my Windows Registry directly from install. Finding what I did, I believe I misattributed a good amount of sloppy programming to malicious behavior due to stress and paranoia.
So why did I believe you to be someone else? All the people responsible for these actions, well most of them were named Dan. They were mostly ex-google employees. Which is why my results changed and I ended up with about 19k various porn accounts in my name. This, they were also posing as random people named Dan (i.e. when IO tried telling Convercent, Microsoft’s auditor) about the issues. They were named Dan, and they were fraudulent.
So when I found what I did, and as well maintained as I did, I jumped the gun and assumed you were, again, one of the individuals I had whaling me. And at that point, after going to every agency I could, state and local, about the issue, I was a vicious dog backed into a corner. And I also at the time didn’t mean I’d physically harm you. I meant I was going to keep taking out your site.
Again, this doesn’t excuse my behavior. And in the end after all I tried to do to report everything that was happening. I ended up making the people who whaled me a lot of money. And I ended up still losing everything, but I also hurt and was a real asshole to a lot of people that had nothing to do with anything.
In the end, I had a hard time realizing what I did and the curl reason was a piss poor one. Again, I was just being an indignant ass. Especially to someone who actually reported security issues.
Sorry it took me a long time to write this. I should have apologized right after I knew I was wrong. I apologize for that mistake as well.
This amends is also a part of my getting better. So I apologize if this angered you. I just needed to make sure I tell you that I was wrong and I should have had betterjudgement.
When you use the name localhost in a URL, what does it mean? Where does the network traffic go when you ask curl to download http://localhost ?
Is “localhost” just a name like any other or do you think it infers speaking to your local host on a loopback address?
The name was “resolved” using the standard resolver mechanism into one or more IP addresses and then curl connected to the first one that works and gets the data from there.
The (default) resolving phase there involves asking the getaddrinfo() function about the name. In many systems, it will return the IP address(es) specified in /etc/hosts for the name. In some systems things are a bit more unusually setup and causes a DNS query get sent out over the network to answer the question.
In other words: localhost was not really special and using this name in a URL worked just like any other name in curl. In most cases in most systems it would resolve to 127.0.0.1 and ::1 just fine, but in some cases it would mean something completely different. Often as a complete surprise to the user…
Starting in commit 1a0ebf6632f8, to be released in curl 7.78.0, curl now treats the host name “localhost” specially and will use an internal “hard-coded” set of addresses for it – the ones we typically use for the loopback device: 127.0.0.1 and ::1. It cannot be modified by /etc/hosts and it cannot be accidentally or deliberately tricked by DNS resolves. localhost will now always resolve to a local address!
Does that kind of mistakes or modifications really happen? Yes they do. We’ve seen it and you can find other projects report it as well.
Who knows, it might even be a few microseconds faster than doing the “full” resolve call.
(You can still build curl without IPv6 support at will and on systems without support, for which the ::1 address of course will not be provided for localhost.)
Specs say we can
The RFC 6761 is titled Special-Use Domain Names and in its section 6.3 it especially allows or even encourages this:
Users are free to use localhost names as they would any other domain names. Users may assume that IPv4 and IPv6 address queries for localhost names will always resolve to the respective IP loopback address.
Name resolution APIs and libraries SHOULD recognize localhost names as special and SHOULD always return the IP loopback address for address queries and negative responses for all other query types. Name resolution APIs SHOULD NOT send queries for localhost names to their configured caching DNS server(s).
Mike West at Google also once filed an I-D with even stronger wording suggesting we should always let localhost be local. That wasn’t ever turned into an RFC though but shows a mindset.
(Some) Browsers do it
Chrome has been special-casing localhost this way since 2017, as can be seen in this commit and I think we can safely assume that the other browsers built on their foundation also do this.
Firefox landed their corresponding change during the fall of 2020, as recorded in this bugzilla entry.
Safari (on macOS at least) does however not do this. It rather follows what /etc/hosts says (and presumably DNS of not present in there). I’ve not found any official position on the matter, but I found this source code comment indicating that localhost resolving might change at some point:
Since some time back, Windows already resolves “localhost” internally and it is not present in their /etc/hosts alternative. I believe it is more of a hybrid solution though as I believe you can put localhost into that file and then have that custom address get used for the name.
Secure over http://localhost
When we know for sure that http://localhost is indeed a secure context (that’s a browser term I’m borrowing, sorry), we can follow the example of the browsers and for example curl should be able to start considering cookies with the “secure” property to be dealt with over this host even when done over plain HTTP. Previously, secure in that regard has always just meant HTTPS.
This change in cookie handling has not happened in curl yet, but with localhost being truly local, it seems like an improvement we can proceed with.
Can you still trick curl?
When I mentioned this change proposal on twitter two of the most common questions in response were
can’t you still trick curl by routing 127.0.0.1 somewhere else
can you still use --resolve to “move” localhost?
The answers to both questions are yes.
You can of course commit the most hideous hacks to your system and reroute traffic to 127.0.0.1 somewhere else if you really wanted to. But I’ve never seen or heard of anyone doing it, and it certainly will not be done by mistake. But then you can also just rebuild your curl/libcurl and insert another address than the default as “hardcoded” and it’ll behave even weirder. It’s all just software, we can make it do anything.
The --resolve option is this magic thing to redirect curl operations from the given host to another custom address. It also works for localhost, since curl will check the cache before the internal resolve and --resolve populates the DNS cache with the given entries. (Provided to applications via the CURLOPT_RESOLVE option.)
What will break?
With enough number of users, every single little modification or even improvement is likely to trigger something unexpected and undesired on at least one system somewhere. I don’t think this change is an exception. I fully expect this to cause someone to shake their fist in the sky.
However, I believe there are fairly good ways to make to restore even the most complicated use cases even after this change, even if it might take some hands on to update the script or application. I still believe this change is a general improvement for the vast majority of use cases and users. That’s also why I haven’t provided any knob or option to toggle off this behavior.
The top photo was taken by me (the symbolism being that there’s a path to take somewhere but we don’t really know where it leads or which one is the right to take…). This curl change was written by me. Mike West provided me the Chrome localhost change URL. Valentin Gosu gave me the Firefox bugzilla link.
The official publication date of the relevant QUIC specifications is: May 27, 2021.
I’ve done many presentations about HTTP and related technologies over the years. HTTP/2 had only just shipped when the QUIC working group had been formed in the IETF and I started to mention and describe what was being done there.
I’ve explained HTTP/3
I started writing the document HTTP/3 explained in February 2018 before the protocol was even called HTTP/3 (and yeah the document itself was also called something else at first). The HTTP protocol for QUIC was just called “HTTP over QUIC” in the beginning and it took until November 2018 before it got the name HTTP/3. I did my first presentation using HTTP/3 in the title and on slides in early December 2018, My first recorded HTTP/3 presentation was in January 2019 (in Stockholm, Sweden).
In that talk I mentioned that the protocol would be “live” by the summer of 2019, which was an optimistic estimate based on the then current milestones set out by the IETF working group.
I think my optimism regarding the release schedule has kept up but as time progressed I’ve updated that estimation many times…
HTTP/3 – not yet
The first four RFC documentations to be ratified and published only concern QUIC, the transport protocol, and not the HTTP/3 parts. The two HTTP/3 documents are also in queue but are slightly delayed as they await some other prerequisite (“generic” HTTP update) documents to ship first, then the HTTP/3 ones can ship and refer to those other documents.
QUIC is a new transport protocol. It is done over UDP and can be described as being something of a TCP + TLS replacement, merged into a single protocol.
Okay, the title of this blog is misleading. QUIC is actually documented in four different RFCs:
RFC 9002 – QUIC Loss Detection and Congestion Control
My role: I’m just a bystander
I initially wanted to keep up closely with the working group and follow what happened and participate on the meetings and interims etc. It turned out to be too difficult for me to do that so I had to lower my ambitions and I’ve mostly had a casual observing role. I just couldn’t muster the energy and spend the time necessary to do it properly.
I’ve participated in many of the meetings, I’ve been present in the QUIC implementers slack, I’ve followed lots of design and architectural discussions on the mailing list and in GitHub issues. I’ve worked on implementing support for QUIC and h3 in curl and thanks to that helped out iron issues and glitches in various implementations, but the now published RFCs have virtually no traces of me or my feedback in them.
In the curl project we make great efforts to store a lot of meta data about each and every vulnerability that we have fixed over the years – and curl is over 23 years old. This data set includes CVE id, first vulnerable version, last vulnerable version, name, announce date, report to the project date, CWE, reward amount, code area and “C mistake kind”.
We also keep detailed data about releases, making it easy to look up for example release dates for specific versions.
All this, combined with my fascination (some would call it obsession) of graphs is what pushed me into creating the curl project dashboard, with an ever-growing number of daily updated graphs showing various data about the curl projects in visual ways. (All scripts for that are of course also freely available.)
What to show is interesting but of course it is sometimes even more important how to show particular data. I don’t want the graphs just to show off the project. I want the graphs to help us view the data and make it possible for us to draw conclusions based on what the data tells us.
The worst bugs possible in a project are the ones that are found to be security vulnerabilities. Those are the kind we want to work really hard to never introduce – but we basically cannot reach that point. This special status makes us focus a lot on these particular flaws and we of course treat them special.
For a while we’ve had two particular vulnerability graphs in the dashboard. One showed the number of fixed issues over time and another one showed how long each reported vulnerability had existed in released source code until a fix for it shipped.
CVE age in code until report
The CVE age in code until report graph shows that in general, reported vulnerabilities were introduced into the code base many years before they are found and fixed. In fact, the all time average time suggests they are present for more than 2,700 – more than seven years. Looking at the reports from the last 12 months, the average is even almost 1000 days more!
It takes a very long time for vulnerabilities to get found and reported.
When were the vulnerabilities introduced
Just the other day it struck me that even though I had a lot of graphs already showing in the dashboard, there was none that actually showed me in any nice way at what dates we created the vulnerabilities we spent so much time and effort hunting down, documenting and talking about.
I decided to use the meta data we already have and add a second plot line to the already existing graph. Now we have the previous line (shown in green) that shows the number of fixed vulnerabilities bumped at the date when a fix was released.
Added is the new line (in red) that instead is bumped for every date we know a vulnerability was first shipped in a release. We know the version number from the vulnerability meta data, we know the release date of that version from the release meta data.
This all new graph helps us see that out of the current 100 reported vulnerabilities, half of them were introduced into the code before 2010.
Using this graph it also very clear to me that the increased CVE reporting that we can spot in the green line started to accelerate in the project in 2016 was not because the bugs were introduced then. The creation of vulnerabilities rather seem to be fairly evenly distributed over time – with occasional bumps but I think that’s more related to those being particular releases that introduced a larger amount of features and code.
As the average vulnerability takes 2700 days to get reported, it could indicate that flaws landed since 2014 are too young to have gotten reported yet. Or it could mean that we’ve improved over time so that new code is better than old and thus when we find flaws, they’re more likely to be in old code paths… I don’t think the red graph suggests any particular notable improvement over time though. Possibly it does if we take into account the massive code growth we’ve also had over this time.
The green “fixed” line at least has a much better trend and growth angle.
Present in which releases
As we have the range of vulnerable releases stored in the meta data file for each CVE, we can then add up the number of the flaws that are present in every past release.
Together with the release dates of the versions, we can make a graph that shows the number of reported vulnerabilities that are present in each past release over time, in a graph.
You can see that some labels end up overwriting each other somewhat for the occasions when we’ve done two releases very close in time.
tldr: the level of HTTP/3 support in servers is surprisingly high.
The specifications are all done. They’re now waiting in queues to get their final edits and approvals before they will get assigned RFC numbers and get published as such – they will not change any further. That’s a set of RFCs (six I believe) for various aspects of this new stack. The HTTP/3 spec is just one of those. Remember: HTTP/3 is the application protocol done over the new transport QUIC. (See http3 explained for a high-level description.)
The HTTP/3 spec was written to refer to, and thus depend on, two other HTTP specs that are in the works: httpbis-cache and https-semantics. Those two are mostly clarifications and cleanups of older HTTP specs, but this forces the HTTP/3 spec to have to get published after the other two, which might introduce a small delay compared to the other QUIC documents.
The working group has started to take on work on new specifications for extensions and improvements beyond QUIC version 1.
In early April 2021, the usage of QUIC and HTTP/3 in the world is measured by a few different companies.
netray.io scans the IPv4 address space weekly and checks how many hosts that speak QUIC. Their latest scan found 2.1 million such hosts.
Arguably, the netray number doesn’t say much. Those two million hosts could be very well used or barely used machines.
HTTP/3 by w3techs
w3techs.com has been in the game of scanning web sites for stats purposes for a long time. They scan the top ten million sites and count how large share that runs/supports what technologies and they also check for HTTP/3. In their data they call the old Google QUIC for just “QUIC” which is confusing but that should be seen as the precursor to HTTP/3.
What stands out to me in this data except that the HTTP/3 usage seems very high: the top one-million sites are claimed to have a higher share of HTTP/3 support (16.4%) than the top one-thousand (11.9%)! That’s the reversed for HTTP/2 and not how stats like this tend to look.
It has been suggested that the growth starting at Feb 2021 might be explained by Cloudflare’s enabling of HTTP/3 for users also in their free plan.
HTTP/3 by Cloudflare
On radar.cloudflare.com we can see Cloudflare’s view of a lot of Internet and protocol trends over the world.
This HTTP/3 number is significantly lower than w3techs’. Presumably because of the differences in how they measure.
All the major browsers have HTTP/3 implementations and most of them allow you to manually enable it if it isn’t already done so. Chrome and Edge have it enabled by default and Firefox will so very soon. The caniuse.com site shows it like this (updated on April 4):
(Earlier versions of this blog post showed the previous and inaccurate data from caniuse.com. Not anymore.)
curl supports HTTP/3 since a while back, but you need to explicitly enable it at build-time. It needs to use third party libraries for the HTTP/3 layer and it needs a QUIC capable TLS library. The QUIC/h3 libraries are still beta versions. See below for the TLS library situation.
curl’s HTTP/3 support is not even complete. There are still unsupported areas and it’s not considered stable yet.
curl supports 14 different TLS libraries at this time. Two of them have QUIC support landed: BoringSSL and GnuTLS. And a third would be the quictls OpenSSL fork. (There are also a few other smaller TLS libraries that support QUIC.)
The by far most popular TLS library to use with curl, OpenSSL, has postponed their QUIC work:
At the same time they have delayed the OpenSSL 3.0 release significantly. Their release schedule page still today speaks of a planned release of 3.0.0 in “early Q4 2020”. That plan expects a few months from the beta to final release and we have not yet seen a beta release, only alphas.
Realistically, this makes QUIC in OpenSSL many months off until it can appear even in a first alpha. Maybe even 2022 material?
The Google powered OpenSSL fork BoringSSL has supported QUIC for a long time and provides the OpenSSL API, but they don’t do releases and mostly focus on getting a library done for Google. People outside the company are generally reluctant to use and depend on this library for those reasons.
The quiche QUIC/h3 library from Cloudflare uses BoringSSL and curl can be built to use quiche (as well as BoringSSL).
Microsoft and Akamai have made a fork of OpenSSL available that is based on OpenSSL 1.1.1 and has the QUIC pull-request applied in order to offer a QUIC capable OpenSSL flavor to the world before the official OpenSSL gets their act together. This fork is called quictls. This should be compatible with OpenSSL in all other regards and provide QUIC with an API that is similar to BoringSSL’s.
The ngtcp2 QUIC library uses quictls. curl can be built to use ngtcp2 as well as with quictls,
Is HTTP/3 faster?
I realize I can’t blog about this topic without at least touching this question. The main reason for adding support for HTTP/3 on your site is probably that it makes it faster for users, so does it?
We’ve seen other numbers say h3 is faster shown before but it’s hard to find up-to-date performance measurements published for the current version of HTTP/3 vs HTTP/2 in real world scenarios. Partly of course because people have hesitated to compare before there are proper implementations to compare with, and not just development versions not really made and tweaked to perform optimally.
I think there are reasons to expect h3 to be faster in several situations, but for people with high bandwidth low latency connections in the western world, maybe the difference won’t be noticeable?
I’ve previously shown the slide below to illustrate what needs to be done for curl to ship with HTTP/3 support enabled in distros and “widely” and I think the same works for a lot of other projects and clients who don’t control their TLS implementation and don’t write their own QUIC/h3 layer code.
This house of cards of h3 is slowly getting some stable components, but there are still too many moving parts for most of us to ship.
I assume that the rest of the browsers will also enable HTTP/3 by default soon, and the specs will be released not too long into the future. That will make HTTP/3 traffic on the web increase significantly.
The QUIC and h3 libraries will ship their first non-beta versions once the specs are out.
The TLS library situation will continue to hamper wider adoption among non-browsers and smaller players.
The big players already deploy HTTP/3.
I’ve updated this post after the initial publication, and the biggest corrections are in the Chrome/Edge details. Thanks to immediate feedback from Eric Lawrence. Remaining errors are still all mine! Thanks also to Barry Pollard who filed the PR to update the previously flawed caniuse.com data.
curl is an internet transfer engine. A rather modular one too. Parts of curl’s functionality is provided by selectable alternative implementations that we call backends. You select what backends to enable at build-time and in many cases the backends are enabled and powered by different 3rd party libraries.
curl has a range of such alternative backends for various features:
International Domain Names
HTTP content encoding
Stable API and ABI
Maintaining a stable API and ABI is key to libcurl. As long as those promises are kept, changing internals such as switching between backends is perfectly fine.
The API is the armored front door that we don’t change. The backends is the garden on the back of the house that we can dig up and replant every year if we want, without us having to change the front door.
Already back in 2005 we added support for using an alternative TLS library in curl when we added support for GnuTLS in addition to OpenSSL, and since then we’ve added many more. We do this by having an internal API through which we do all the TLS related things and for each third party library we support we have code that does the necessary logic to connect the internal API with the corresponding TLS library.
Today, we merged support for yet another TLS library: rustls. This is a TLS library written in rust and it has a C API provided in a separate project called crustls. Strictly speaking, curl is built to use crustls.
This is still early days for the rustls backend and it is not yet feature complete. There’s more work to do and polish to apply before we can think of it as a proper competitor to the already established and well-used TLS backends, but with this merge it makes it much easier for more people to help out and test it out. Feel free and encouraged to join in!
We count this addition as the 14th concurrently supported TLS library in curl. I’m not aware of any other project, anywhere, that supports more or even this many TLS libraries.
The TLS library named mesalink is actually already using rustls, but under an OpenSSL API disguise and we support that since a few years back…
The TLS backend code for rustls was written and contributed by Jacob Hoffman-Andrews.
On February 11th, 2021 18:00 UTC (10am Pacific time, 19:00 Central Europe) we invite you to participate in a webinar we call “curl, Hyper and Rust”. To join us at the live event, please register via the link below:
What is the project about, how will this improve curl and Hyper, how was it done, what lessons can be learned, what more can we expect in the future and how can newcomers join in and help?
Participating speakers in this webinar are:
Daniel Stenberg. Founder of and lead developer of curl.
Josh Aas, Executive Director at ISRG / Let’s Encrypt.
The event went on for 60 minutes, including the Q&A session at the end.
If you already have a question you want to ask, please let us know ahead of time. Either in a reply here on the blog, or as a reply on one of the many tweets that you will see about about this event from me and my fellow “webinarees”.
You might recall that my Twitter account was hijacked and then again just two weeks later.
The first: brute-force
The first take-over was most likely a case of brute-forcing my weak password while not having 2FA enabled. I have no excuse for either of those lapses. I had convinced myself I had 2fa enabled which made me take a (too) lax attitude to my short 8-character password that was possible to remember. Clearly, 2fa was not enabled and then the only remaining wall against the evil world was that weak password.
The second time
After that first hijack, I immediately changed password to a strong many-character one and I made really sure I enabled 2fa with an authenticator app and I felt safe again. Yet it would only take seventeen days until I again was locked out from my account. This second time, I could see how someone had managed to change the email address associated with my account (displayed when I wanted to reset my password). With the password not working and the account not having the correct email address anymore, I could not reset the password, and my 2fa status had no effect. I was locked out. Again.
It felt related to the first case because I’ve had my Twitter account since May 2008. I had never lost it before and then suddenly after 12+ years, within a period of three weeks, it happens twice?
Why and how
How this happened was a complete mystery to me. The account was restored fairly swiftly but I learned nothing from that.
Then someone at Twitter contacted me. After they investigated what had happened and how, I had a chat with a responsible person there and he explained for me exactly how this went down.
Had Twitter been hacked? Is there a way to circumvent 2FA? Were my local computer or phone compromised? No, no and no.
Apparently, an agent at Twitter who were going through the backlog of issues, where my previous hijack issue was still present, accidentally changed the email on my account by mistake, probably confusing it with another account in another browser tab.
There was no outside intruder, it was just a user error.
Okay, the cynics will say, this is what he told me and there is no evidence to back it up. That’s right, I’m taking his words as truth here but I also think the description matches my observations. There’s just no way for me or any outsider to verify or fact-check this.
A brighter future
They seem to already have identified things to improve to reduce the risk of this happening again and Michael also mentioned a few other items on their agenda that should make hijacks harder to do and help them detect suspicious behavior earlier and faster going forward. I was also happy to provide my feedback on how I think they could’ve made my lost-account experience a little better.
I’m relieved that the second time at least wasn’t my fault and neither of my systems are breached or hacked (as far as I know).
I’ve also now properly and thoroughly gone over all my accounts on practically all online services I use and made really sure that I have 2fa enabled on them. On some of them I’ve also changed my registered email address to one with 30 random letters to make it truly impossible for any outsider to guess what I use.
(I’m also positively surprised by this extra level of customer care Twitter showed for me and my case.)
Am I a target?
I don’t think I am. I think maybe my Twitter account could be interesting to scammers since I have almost 25K followers and I have a verified account. Me personally, I work primarily with open source and most of my works is already made public. I don’t deal in business secrets. I don’t think my personal stuff attracts attackers more than anyone else does.
What about the risk or the temptation for bad guys in trying to backdoor curl? It is after all installed in some 10 billion systems world-wide. I’ve elaborated on that before. Summary: I think it is terribly hard for someone to actually manage to do it. Not because of the security of my personal systems perhaps, but because of the entire setup and all processes, signings, reviews, testing and scanning that are involved.
So no. I don’t think my personal systems are a valued singled out target to attackers.
Status: 00:27 in the morning of December 4 my account was restored again. No words or explanations on how it happened – yet.
This morning (December 3rd, 2020) I woke up to find myself logged out from my Twitter account on the devices where I was previously logged in. Due to “suspicious activity” on my account. I don’t know the exact time this happened. I checked my phone at around 07:30 and then it has obviously already happened. So at time time over night.
Trying to log back in, I get prompted saying I need to update my password first. Trying that, it wants to send a confirmation email to an email address that isn’t mine! Someone has managed to modify the email address associated with my account.
It has only been two weeks since someone hijacked my account the last time and abused it for scams. When I got the account back, I made very sure I both set a good, long, password and activated 2FA on my account. 2FA with auth-app, not SMS.
The last time I wasn’t really sure about how good my account security was. This time I know I did it by the book. And yet this is what happened.
I was in touch with someone at Twitter security and provided lots of details of my systems , software, IP address etc while they researched their end about what happened. I was totally transparent and gave them all info I had that could shed some light.
I was contacted by a Sr. Director from Twitter (late Dec 4 my time). We have a communication established and I’ve been promised more details and information at some point next week. Stay tuned.
Was I breached?
Many people have proposed that the attacker must have come through my local machine to pull this off. If someone did, it has been a very polished job as there is no trace at all of that left anywhere on my machine. Also, to reset my password I would imagine the attacker would need to somehow hijack my twitter session, need the 2FA or trigger a password reset and intercept the email. I don’t receive emails on my machine so the attacker would then have had to (also?) manage to get into my email machine and removed that email – and not too many others because I receive a lot of email and I’ve kept on receiving a lot of email during this period.
I’m not ruling it out. I’m just thinking it seems unlikely.
If the attacker would’ve breached my phone and installed something nefarious on that, it would not have removed any reset emails and it seems like a pretty touch challenge to hijack a “live” session from the Twitter client or get the 2FA code from the authenticator app. Not unthinkable either, just unlikely.
As I have no insights into the other end I cannot really say which way I think is the most likely that the perpetrator used for this attack, but I will maintain that I have no traces of a local attack or breach and I know of no malicious browser add-ons or twitter apps on my devices.
Firefox version 83.0 on Debian Linux with Tweetdeck in a tab – a long-lived session started over a week ago (ie no recent 2FA codes used),
Browser extensions: Cisco Webex, Facebook container, multi-account containers, HTTPS Everywhere, test pilot and ublock origin.
I only use one “authorized app” with Twitter and that’s Tweetdeck.
On the Android phone, I run an updated Android with an auto-updated Twitter client. That session also started over a week ago. I used Google Authenticator for 2fa.
While this hijack took place I was asleep at home (I don’t know the exact time of it), on my WiFi, so all my most relevant machines would’ve been seen as originating from the same “NATed” IP address. This info was also relayed to Twitter security.
The actual restoration happens like this (and it was the exact same the last time): I just suddenly receive an email on how to reset my password for my account.
The email is a standard one without any specifics for this case. Just a template press the big button and it takes you to the Twitter site where I can set a new password for my account. There is nothing in the mail that indicates a human was involved in sending it. There is no text explaining what happened. Oh, right, the mail also include a bunch of standard security advice like “use a strong password”, “don’t share your password with others” and “activate two factor” etc as if I hadn’t done all that already…
It would be prudent of Twitter to explain how this happened, at least roughly and without revealing sensitive details. If it was my fault somehow, or if I just made it easier because of something in my end, I would really like to know so that I can do better in the future.
What was done to it?
No tweets were sent. The name and profile picture remained intact. I’ve not seen any DMs sent or received from while the account was “kidnapped”. Given this, it seems possible that the attacker actually only managed to change the associated account email address.