Category Archives: Technology

Really everything related to technology

cURL and libcurl, Network, Open Source, Web

http2 in curl

March 9, 2014 Daniel Stenberg

While the first traces of http2 support in curl was added already back in September 2013 it hasn’t been until recently it actually was made useful. There’s been a lot of http2 related activities in the curl team recently and in the late January 2014 we could run our first command line inter-op tests against public http2 (draft-09) servers on the Internet.

There’s a lot to be said about http2 for those not into its nitty gritty details, but I’ll focus on the curl side of this universe in this blog post. I’ll do separate posts and presentations on http2 “internals” later.

A quick http2 overview

http2 (without the minor version, as per what the IETF work group has decided on) is a binary protocol that allows many logical streams multiplexed over the same physical TCP connection, it features compressed headers in both directions and it has stream priorities and more. It is being designed to maintain the user concepts and paradigms from HTTP 1.1 so web sites don’t have to change contents and web authors won’t need to relearn a lot. The web will not break because of http2, it will just magically work a little better, a little smoother and a little faster.

In libcurl we build http2 support with the help of the excellent library called nghttp2, which takes care of all the binary protocol details for us. You’ll also have to build it with a new enough version of the SSL library of your choice, as http2 over TLS will require use of some fairly recent TLS extensions that not many older releases have and several TLS libraries still completely lack!

The need for an extension is because with speaking TLS over port 443 which HTTPS implies, the current and former web infrastructure assumes that we will speak HTTP 1.1 over that, while we now want to be able to instead say we want to talk http2. When Google introduced SPDY then pushed for a new extension called NPN to do this, which when taken through the standardization in IETF has been forked, changed and renamed to ALPN with roughly the same characteristics (I don’t know the specific internals so I’ll stick to how they appear from the outside).

So, NPN and especially ALPN are fairly recent TLS extensions so you need a modern enough SSL library to get that support. OpenSSL and NSS both support NPN and ALPN with a recent enough version, while GnuTLS only supports ALPN. You can build libcurl to use any of these threes libraries to get it to talk http2 over TLS.

http2 using libcurl

(This still describes what’s in curl’s git repository, the first release to have this level of http2 support is the upcoming 7.36.0 release.)

Users of libcurl who want to enable http2 support will only have to set CURLOPT_HTTP_VERSION to CURL_HTTP_VERSION_2_0 and that’s it. It will make libcurl try to use http2 for the HTTP requests you do with that handle.

For HTTP URLs, this will make libcurl send a normal HTTP 1.1 request with an offer to the server to upgrade the connection to version 2 instead. If it does, libcurl will continue using http2 in the clear on the connection and if it doesn’t, it’ll continue using HTTP 1.1 on it. This mode is what Firefox and Chrome will not support.

For HTTPS URLs, libcurl will use NPN and ALPN as explained above and offer to speak http2 and if the server supports it. there will be http2 sweetness from than point onwards. Or it selects HTTP 1.1 and then that’s what will be used. The latter is also what will be picked if the server doesn’t support ALPN and NPN.

Alt-Svc and ALTSVC are new things planned to show up in time for http2 draft-11 so we haven’t really thought through how to best support them and provide their features in the libcurl API. Suggestions (and patches!) are of course welcome!

http2 with curl

Hardly surprising, the curl command line tool also has this power. You use the –http2 command line option to switch on the libcurl behavior as described above.

Translated into old-style

To reduce transition pains and problems and to work with the rest of the world to the highest possible degree, libcurl will (decompress and) translate received http2 headers into http 1.1 style headers so that applications and users will get a stream of headers that look very much the way you’re used to and it will produce an initial response line that says HTTP 2.0 blabla.

Building (lib)curl to support http2

See the README.http2 file in the lib/ directory.

This is still a draft version of http2!

I just want to make this perfectly clear: http2 is not out “for real” yet. We have tried our http2 support somewhat at the draft-09 level and Tatsuhiro has worked on the draft-10 support in nghttp2. I expect there to be at least one more draft, but perhaps even more, before http2 becomes an official RFC. We hope to be able to stay on the frontier of http2 and deliver support for the most recent draft going forward.

PS. If you try any of this and experience any sort of problems, please speak to us on the curl-library mailing list and help us smoothen out whatever problem you got!

cURL and libcurl, Mozilla, Network

HTTPbis design team meeting London

March 8, 2014 Daniel Stenberg

I’m writing this just hours after the HTTPbis design team meeting in London 2014 has ended.

Around 30 people attended the meeting i Mozilla’s central London office. The fridge was filled up with drinks, the shelves were full of snacks and goodies. The day could begin. This is the Saturday after the IETF89 week so most people attending had already spent the whole or parts of the week before here in London doing other HTTP and network related work. The HTTPbis sessions at the IETF itself were productive and had already pushed us forward.

We started at 9:30 and we quickly got to work. Mark Nottingham guided us through the day with usual efficiency.

We all basically hang out in a huge room, some in chairs, some in sofas and a bunch of people on the floor or just standing up. We had mikes passed around and the http2 discussions were flowing back and forth depending on the topics and what people felt about them. Some of the issues that were nailed down this time and will end up detailed in the upcoming draft-11 are (strictly speaking, we only discussed the things and formed opinions, as by IETF guidelines we can’t decide things on an offline meeting like this):

Priorities of streams will have a dependency graph or direction, making individualÂ streams less or more important than other
A client can send headers without compression and tell the proxy that the header shouldn’t be compressed – used a way to mitigate some of the compression security problems
There will be no TLS renegotiation allowed mid-session. Basically a client will have to tear down the connection and negotiate again if suddenly a need to use a client certificate arises.
Alt-Svc is the way forward so ALTSVC will appear a new frame in draft-11. This is the way to signal to an application that there is another “route” to the same content on the same server. This will allow for what is popularly known as “opportunistic encryption” or at least one sort of that. In short, you can do “plain-text” HTTP over a TLS connection using this…
We decided that a server should support gzip contents from clients

There were some other things too handled, but I believe those are the main changes. When the afternoon started to turn long, beers and other beverages were brought out and we did enjoy a relaxing social finale of the day before we split up in smaller groups and headed out in the busy London night to get dinner…

Thanks everyone for a great day. I also appreciated meeting several people in real-life I never met before, only discussed with and read emails from online and of course some old friends I hadn’t seen in a long time!

Oh, there’s also a new rough time frame for http2 going forward. Nearest in time would be the draft-11 at the end of March and another interim in the beginning of June (Boston?).

As a reminder, here’s what was happened for draft-10, and here is http2 draft-10.

Out of all people present today, I believe Mozilla was the company with the largest team (8 attendees) – funnily enough none of us Mozillians there actually work in this office or even in this country.

Technology

A payment on a flight story

February 3, 2014 Daniel Stenberg

“Three beers and three chicken sandwiches, thanks” – I said, graciously handing over my VISA card to the flight stewardess to pay for everything for me and my two Haxx friends BjÃ¶rn and Linus. 20 something Euros. Neither of us were carrying any cash.

We were happy that we had seats in the 7th row on the way back to Stockholm since when we flew down to FOSDEM Brussels just two days earlier we were in the 18th row and by then they had ran out of sandwiches. Brussels Airlines on a direct flight.

The lady inserted the card into her handheld card-reader, messed around a while since it wouldn’t “take” at the first attempt and then she handed it to me to enter my PIN. So I did.

credit cards Ms stewardess pressed some buttons or something and then she said, “no it doesn’t work” and continued to try inserting my card in basically every variation you can (especially having the chip side turned out has to be a very clever way) while repeating to me that it doesn’t work. This is the same card I’ve used numerous times during the weekend and I used it several times at the airport less than an hour ago. I know it works.

“OK let’s say you’re right”, I sighed as I really didn’t think I nor my card were to blame but I also didn’t feel like just causing trouble. I handed her my second VISA card. “Here, try this instead then”.

“No it doesn’t work either”. This certainly wasn’t right.

Linus reached over and offered his MasterCard. Perhaps there was a VISA problem with the device but “Sorry sir, it doesn’t work” came back. We’re now at farce level. BjÃ¶rn joins the fun and offers a 4th card, another MasterCard.

By this time we’ve eaten most of the sandwiches and were enjoying the beers.

The lady continued to struggle and it still didn’t work. She was starting to act a bit troubled about this and I asked her if she really insisted that all our cards are broken and she admitted that she suspected the machine wasn’t working quite as it should. Then she leaned towards the male steward who was serving drinks a couple of steps away. They fiddled a bit more and then she came back to us.

“That worked, I swiped it” she said without much further explanation and returned with BjÃ¶rn’s card and the receipt for the purchase. I felt quite done with this by now so I didn’t ask nor pushed her why she didn’t do that earlier. I was glad it finally was fixed. So much for me paying, now BjÃ¶rn did it instead. Oh well, truly fascinating that they would do like this on an airline where people have to do purchases with credit cards all the time.

Then BjÃ¶rn looked at the receipt he got back:

Pringles and a bottle of water, 5 Euro.

It made all three of us burst out in uncontrolled laughter. It was then followed by some contemplation what it actually meant. What exactly did BjÃ¶rn pay for? Did he only get the wrong receipt or did he pay for it? If so, who paid for our food and drinks?

It is 2014 and we haven’t come further than this.

(Update: BjÃ¶rn reminded me that we did verify the last 4 digits of the card number on the receipt and it didn’t match any of our cards…)

Network, Web

HTTP2 – the next step

January 24, 2014 Daniel Stenberg

The HTTPbis working group of the IETF had an interim meeting in Zurich January 22nd to 24th. I participated from remote and I listened in on the discussions over webex and followed the jabber room while the meetings were going on, addressing HTTP2 protocol issues one by one ironing out quirks and progressing forward.

I won’t bore you with details why I wasn’t present in Zurich.

Here’s a couple of quick and brief reflections from my perspective:

Listening in from remote like this is not at all adequately compensating for not being there. A room full of people discussing something is really hard to follow from remote and completely impossible to interact with. It is better than not being able to listen in at all, but it is certainly not a replacement for being there.

It is amazing how much faster people can come to conclusions and fix issues when being in the same room. Issues that have been lingering in the tracker for a very long time could be dealt with and closed within minutes. Things like what to call the protocol in ALPN or to remove the ability to switched off flow control. Not all issues of course…

HTTP2 draft-09 that soon will become draft-10 to reflect the updates from this meeting and more, is from my perspective quite far in its process. It is clearly at a point that seems to be OK with most people and the discussions are now just about details. Of course the devil is in the details and I’m not saying it can’t take a long time to settle on them, but the structure and main concepts of the protocol are probably now established.

There were not very many proxy or server people at the interim. Most of the people seemed to be client-side oriented and some service oriented. I’m personally client-side biased myself but I hope this doesn’t lead to us deciding on things that the “other side” will have problems with down the line.

Firefox nightly supports HTTP2 draft-09 (for https:// URLs) and twitter supports it server side. Enable it in the browser by entering “about:config” in the URL bar and change the config entry called “network.http.spdy.enabled.http2draft” to true. Done.

Some of the biggest HTTP2 changes brought up compared to what draft-09 says include:

no more ability to switch off flow control
the prioritization field/concept “weighted dependency tree approach”
>= TLS 1.2 with ephemeral ciphers
MUST not use TLS compression
tolerant to TLS false start or at least must accept/buffer application layer data
padding

There was also a whole lot of discussions about TLS for http:// urls, proxys, MITM for SSL, opportunistic encryption and more but I believe most of those issues remained at the same position as before – I missed out on parts of the last afternoon so I may have missed some details. It’ll all be revealed in draft-10 and the mailing list I’m sure!

Update: the minutes from the interim meeting.

cURL and libcurl, Development, Mozilla, Network

My first Mozilla week

January 17, 2014 Daniel Stenberg 1 Comment

Working from home

I get up in the morning, shave, eat breakfast and make sure all family members get off as they should. Most days I walk my son to school (some 800 meters) and then back again. When they’re all gone, the house is quiet and then me and my cup of coffee go upstairs and my work day begins.

Systems and accounts

I have spent time this week to setup accounts and sign up for various lists and services. Created profiles, uploaded pictures, confirmed passwords. I’ve submitted stuff and I’ve signed things. There’s quite a lot of systems in use.

My colleagues

I’ve met a few. The Necko team isn’t very big but the entire company is huge and there are just so many people and names. I haven’t yet had any pressing reason to meet a lot of people nor learn a lot of names. I feel like I’m starting out this really slowly and gradually.

Code base

Firefox is a large chunk of code. It takes some 20 minutes to rebuild on my 3.5GHz quad-core Core-i7 with SSD. I try to pull code and rebuild every morning now so that I can dogfood and live on the edge. I also have a bunch of local patches now, some of them which I want to have stewing in my own browser for a while so that I know they at least don’t have any major negative impact!

Figuring out the threading, XPCOM, the JavaScript stuff and everything is a massive task. I really cannot claim to have done more than just scratched the surface so far, but at least I am scratching and I’ve “etagged” the whole lot and I’ve spent some time reading and reviewing code. Attaching a gdb to a running Firefox and checking out behavior and how it looks has also helped.

Netwerk code size

“Netwerk” is the directory name of the source tree where most of the network code is located. It is actually not so ridiculously large as one could fear. Counting only C++ and header files, it sums up to about 220K lines of code. Of course not everything interesting is in this tree, but still. Not mindbogglingly large.

Video conferencing

I’ll admit I’ve not participated in this sort of large scale video conferences before this. With Vidyo and all the different people and offices signed up at once – it is a quite impressive setup actually. My only annoyance so far is that I didn’t get the sound for Vidyo to work for me in Linux with my headphones. The other end could hear me but I couldn’t hear them! I had to defer to using Vidyo on a windows laptop instead.

Doing the video conferencing on a laptop instead of on my desktop machine has its advantages when I do them during the evenings when the rest of the family is at home since then I can move my machine somewhere and sit down somewhere where they won’t disturb me and I won’t disturb them.

Bugzilla

The bug tracker is really in the center for this project, or at least for how I view it and work with it right now. During my first week I’ve so far filed two bug reports and I’ve submitted a suggested patch for a third bug. One of my bugs (Bug 959100 – ParseChunkRemaining doesn’t detect chunk size overflow) has been reviewed fine and is now hopefully about to be committed.

I’ve requested commit access (#961018) as a “level 1” and I’ve signed the committer’s agreement. Level 1 is entry level and only lets me push to the Try server but still, I fully accept that there’s a process to follow and I’m in no hurry. I’ll get to level 3 soon enough I’m sure.

Mercurial

What can I say. After having used it a bit this week without any particularly fancy operations, I prefer git so much more. Of course I’m also much more used to git, but I find that for a lot of the stuff where both have similar concepts I prefer to git way. Oh well, its just a tool. I’ll get around. Possibly I’ll try out the git mirror soon and see if that provides a more convenient environment for me.

curl

What impact did all this new protocol and network code stuff during my work days have on my curl activities?

I got inspired to fix both the chunked encoding parser and the cookie parser’s handling of max-age in libcurl.

What didn’t happen

I feel behind in the implementing-http2 department. I didn’t get my new work laptop yet.

Next week

More of the same, land more patches and figure out more code. Grab more smallish bugs others have filed and work on fixing them as more practice.

Also, there’s a HTTPbis meeting in Zürich on Wednesday to Friday that I won’t go to (I’ll spare you the explanation why) but I’ll try to participate remotely.

cURL and libcurl, Haxx, Network, Open Source, Web

I go Mozilla

December 20, 2013 Daniel Stenberg 4 Comments

In January 2014, I start working for Mozilla

I’ve worked in open source projects for some 20 years and I’ve maintained curl and libcurl for over 15 years. I’m an internet protocol geek at heart and Mozilla seems like a perfect place for me to continue to explore this interest of mine and combine it with real open source in its purest form.

I plan to use my experiences from all my years of protocol fiddling and making stuff work on different platforms against random server implementations into the networking team at Mozilla and work on improving Firefox and more.

I’m putting my current embedded Linux focus to the side and I plunge into a worldwide known company with worldwide known brands to do open source within the internet protocols I enjoy so much. I’ll be working out of my home, just outside Stockholm Sweden. Mozilla has no office in my country and I have no immediate plans of moving anywhere (with a family, kids and all established here).

I intend to bring my mindset on protocols and how to do things well into the Mozilla networking stack and world and I hope and expect that I will get inspiration and input from Mozilla and take that back and further improve curl over time. My agreement with Mozilla also gives me a perfect opportunity to increase my commitment to curl and curl development. I want to maintain and possibly increase my involvement in IETF and the httpbis work with http2 and related stuff. With one foot in Firefox and one in curl going forward, I think I may have a somewhat unique position and attitude toward especially HTTP.

I’ve not yet met another Swedish Mozillian but I know I’m not the only one located in Sweden. I guess I now have a reason to look them up and say hello when suitable.

Björn and Linus will continue to drive and run Haxx with me taking a step back into the shadows (Haxx-wise). I’ll still be part of the collective Haxx just as I was for many years before I started working full-time for Haxx in 2009. My email address, my sites etc will remain on haxx.se.

I’m looking forward to 2014!

Mail

Subject: pharma hack

October 16, 2013 Daniel Stenberg

Time to submit some more strange emails I’ve received recently. Here’s one I suspect may be someone who spots that curl is being abused against some host. I really wouldn’t even know how to begin to answer this…

Someone is using your code to continually hack small businesses I work
with. How on earth do I stop them?!

[name]

Mail, Open Source

Don’t email me

October 8, 2013 Daniel Stenberg 12 Comments

Why I insist on people to keep issues on the mailing list(s)

A recent twitter discussion I had with Andrei Neculau contributed to his blog post on this subject, basically arguing that I’m wrong but with many words and explanations.

It triggered me to write up my primary reasons for why I strongly object to handle open source issues, questions and patches privately (for free) in open source projects that I have a leading role in.

1. I spend a considerable amount of my spare time on open source projects. I devote some 15-20 unpaid hours a week for those communities. By emailing me and insisting on a PRIVATE conversation you’re suddenly yanking the mutex flag and you’re now requesting that I spend parts of this time on YOU ALONE and not the rest of the community. That’s selfish.

2. By insisting on a private conversation you FORCE me to repeat myself since ideas and questions are rarely unique or done for the first time. So you have a problem or a question that’s very similar to one I just responded to. And the next person will ask the same one tomorrow. By insisting on doing them in public already in the first email, already the second person can read it without me having to write it twice. And the third person who didn’t even realize he was interested in that topic will find out and read it as well (either now when the mail gets sent out or even years later when that user find the archived mailing list on the web). Private emails deny that ability. That’s selfish.

3. By emailing me privately and asking questions and help, you assume that I am the single best person to ask this question at this given time. What if I happen to be on vacation, be under a rough period at work or just not know the particular area of the project very good. I may be the leader or a public person of a project, but I may still not know much about feature X for operating system Z about which you ask. Ask on the list at once and you’ll reach the correct person. That’s more efficient.

4. By emailing me privately, you indirectly put a load on me to reply – or to get off as a rude person. Yes you’re friendly and you ask me nicely and yet even after you remind me after a few days I STILL DON’T RESPOND. Even if I just worked five 16-hour work days and you asked questions I don’t know the answer to… That’s inefficient and rude.

5. Yes, you can say that subscribing to an email list can be daunting and flood you with hundreds or thousands of emails per month – that’s completely true. But if you only wanted to send that single question or submit the single issue, then you can unsubscribe again quite soon and escape most of that load. Then YOU do the work instead of demanding someone else to do it for you. When you want to handle a SINGLE issue, it is much better load balancing if you do the extra work and the people who do tens or HUNDREDS of issues per month in the project do less work per issue.

6. You’re suggesting that I could forward the private question to the mailing list? Yes I can, but then I need to first ask for permission to do so (or be a jerk) and if the person who sent me the mail is going to send me another mail anyway, (s)he can just as well spend that time to send the first mail to the list instead of say YES to me and then make me do his or hers work. It’s just more efficient. Also, forwarded questions tend to end up so that replies and follow-up questions don’t find their way back to the original poster and that’s bad.

7. I propose and use different lists for different purposes to ease the problem with too many (uninteresting) emails.

Electronics

Bye bye Nexus 10 it was fun while it lasted

August 22, 2013 Daniel Stenberg 2 Comments

In the middle of my otherwise happy summer vacation, my Nexus 10 had a serious case of depression and took a nose-dive from a little over a meter above floor level and crashed into the mighty fine stone tiles. It got some serious damage and there are cracks all over the screen.

A while ago I posted a service request on Samsung’s site to get it fixed (they manufacture the Nexus-10s and the device have a product number and everything in Samsung’s systems), and they responded and directed me to my nearest “quick supportÂ center” for screen and display repairs. The nearest one happens to be located just a few hundred meters from where I work these days.

Today I stepped in there and asked to get my screen fixed.

– “It’ll cost you some 2000-3000 SEK” one of the two service guys says at once. Clearly not really wanting to fix it.

– “Eh, that’s a bit unspecific. Can you tell me with some better accuracy?” I reply. After all, I didn’t pay an awful lot more than 3000 SEK for it as new – in the US and I wanted to figure out if a repair would be worth the money.

– “Okay”, he says and leaves the room through a door and is gone for a while.

– “Do you have the serial number for it?” the guys says returning and yeah, I brought it there in its original box and the serial number is there. The guy leaves again.

– “Did you buy that device here?” He’s back. Without really specifying where “here” means but I figured he means in Sweden so I of course had to tell him

– “No, it’s purchased in the US”

– “Then we can’t fix it.” and then he explains how they can’t order the parts necessary for the device and that’s it. Nothing to do. They can’t.

Okay, so I knew there’s always a risk when “grey-importing” so I’m not really very upset. I’m mostly baffled by their response and also by the fact that there apparently is something different with the Nexus 10 display if it was bought in the US – or at least they say so.

We still have another Nexus 10 in the family and I’m ordering a 2nd gen Nexus 7 now to replace this dead thing with…