Stdio(3) change: FILE is now opaque | Hacker News Reader

Stdio(3) change: FILE is now opaque

undeadly.org

167 pointsgslin90 comments

I think FreeBSD tried to opaque FILE[1], but it was reverted[2] and still non-opaque in main[3].

[1]: https://github.com/freebsd/freebsd-src/commit/c17bf9a9a5a3b5...

[2]: https://github.com/freebsd/freebsd-src/commit/19e03ca8038019...

[3]: https://github.com/freebsd/freebsd-src/blob/main/include/std...

OpenBSD tends to commit to breaking changes much more aggressively than others. Something tells me they're not reverting.

I think FreeBSD is also more concerned with performance regression than OpenBSD is.

FreeBSD's implementation of FILE is a nice object-oriented structure which anyone could derive from. Super-easy to make FILE point to a memory buffer or some other user code. I used that a bunch a long time ago.

Obviously making FILE opaque completely breaks every program that used this feature, so no surprise it was reverted.

fopencookie, fmemopen, you don't need transparency.

fopencookie seems glibc-specific, so unavailable on BSD.

BSD has had funopen(3) since 4.4, so it has an alternative. FreeBSD has implemented fopencookie(3) since v11, but FreeBSD is the BSD most willing to implement Linux interfaces for various reasons.

In the FreeBSD case, far from breaking "every program" it breaks very little at all. In fact it broke 1 thing at the time. Unfortunately, that 1 thing happened to be sysinstall(8).

stdin, stdout, and stderr were already pointers rather than array element addresses, and the external symbol references to __stdinp, __stdoutp, and __stderrp did not change; compiled code using the old macros continued to work as the actual structure layout was not changed; compiled code using FILE* would have continued to work as the pointer implementation didn't change; compiled C++ code with C++ function parameter overloading would have continued to link as the underlying struct type did not change; source code using the ferror_unlocked() and suchlike function-like macros would have not needed changing as there were already ferror_unlocked() and suchlike functions and those remained.

Looking at things like https://reviews.freebsd.org/D4488 from 2015 there was definitely stuff in the ports tree that would have broken back in 2008. But that won't break now should this change be made again, and that's not base.

What actually broke was libftpio, a library that was in base up until 2011, and definitely won't break now, nearly 14 years after being removed for being orphaned after sysinstall(8) itself has gone away.

* https://cgit.freebsd.org/src/commit/lib/libftpio?id=430f2c87...

If you've ever done this to a C library, the first thing that you'll look at when someone else does it is not the FILE type, but how stdin, stdout, and stderr have changed.

The big breaking change is usually the historical implementation of the standard streams as addresses of elements of an array rather than as named pointers. (Plauger's example implementation had them as elements 0, 1, and 2 of a _Files[] array, for example.) It's possible to retain binary compatibility with unrecompiled code that uses the old getc/putc/feof/ferror/fclearerr/&c. macros by preserving structure layouts, but changing stdin, stdout, and stderr can make things not link.

And indeed that has happened here.

openbsd has never, ever, ever, even once maintained binary compatibility; this is just a warning about source compatibility breaking

You have this backwards. Source compatibility is not broken by such a change, with zero source changes required in order to accommodate it for almost all applications; and far from OpenBSD "never, ever, ever" maintaining binary compatibility Masahiko Yasuoka and Philip Guenther took deliberate steps in this very case to ensure as much binary compatibility as they could, retaining structure layouts as-is and retaining several symbols for library internals that macros used to reference, even those that won't be used by freshly recompiled applications.

The warning, and the bumping of several shared library major version numbers, is most definitely about the standard streams breaking binary, not source as you have it, compatibility. Any newly compiled binary that is using the C standard streams won't run on old shared libraries because of the new symbol references for __stdin, __stdout, and __stderr.

The best comments always say "why" and that's missing here.

Does anyone know why this change was done? Security reasons? Preparing for future changes?

Can someone elaborate? I always treated FILE as opaque, but never imagined people could poke into it?

People use reflection for monkey patching and complain when using compiled languages less supportive of such approaches.

So it wouldn't surprise me, that a few folks would do some tricks with FILE internals.

The standard doesn't specify any serviceable parts, and I don't think there are any internals of the struct defined in musl libc on Linux (glibc may be a different story). However, on OpenBSD, it did seem to have some user-visible bits:

https://github.com/openbsd/src/commit/b7f6c2eb760a2da367dd51...

If you expose it, someone will probably sooner or later use it, but probably not in any sane / portable code. On the face of it, it doesn't seem like a consequential change, but maybe they're mopping up after some vulnerability in that one weird package that did touch this.

In gnulib, there is code that patches FILE internals for various platforms to modify behavior of <stdio.h> functions, or implement new functionality.

https://cgit.git.savannah.gnu.org/cgit/gnulib.git/tree/lib/s...

Yes, it's not a good idea to do this. There are more questionable pieces in gnulib, like closing stdin/stdout/stderr (because fflush and fsync is deemed too slow, and regular close reports some errors on NFS on some systems that would otherwise go unreported).

Yes, that part of Gnulib has caused some problems previously. It is mostly used to implement <stdio_ext.h> functions on non-glibc systems. However, it is also needed for some buggy implementations of ftello, fseeko, and fflush.

P.S. Hi Florian :)

And now updated for this change.

https://git.savannah.gnu.org/cgit/gnulib.git/commit/?id=69a0...

> Yes, it's not a good idea to do this. There are more questionable pieces in gnulib, like closing stdin/stdout/stderr (because fflush and fsync is deemed too slow, and regular close reports some errors on NFS on some systems that would otherwise go unreported).

Hyrum's law strikes again. People cast dl_info and poke at internal bits all the time too.

glibc and others should be using kernel-style compiler-driven struct layout randomization to fight it.

> Hyrum's law strikes again.

Is there a name for APIs that are drawn directly from some subset of observed behaviors?

Like Crockford going, "Hey, there's a nice little data format buried in these JS objects. Schloink"

> Is there a name for APIs that are drawn directly from some subset of observed behaviors?

Desire paths. https://en.wikipedia.org/wiki/Desire_path

*BSD stdio.h used to include macro versions of some stdio functions (feof, ferror, clearerr, fileno, getc, putc) so they would be inlined.

    /*
     * This has been tuned to generate reasonable code on the vax using pcc.
     */*

The MH and nmh mail clients used to directly look into FILE internals. If you look for LINUX_STDIO in this old version of the relevant file you can see the kind of ugliness that resulted:

https://cgit.git.savannah.gnu.org/cgit/nmh.git/tree/sbr/m_ge...

It's basically searching an email file to find the contents of either a given header or the mail body. These days there is no need to go under the hood of libc for this (and this code got ripped out over a decade ago), but back when the mail client was running on elderly VAXen this ate up significant time. Sneaking in and reading directly from the internal stdio buffer lets you avoid copying all the data the way an fread would. The same function also used to have a bit of inline vax assembly for string searching...

The only reason this "works" is that traditionally the FILE struct is declared in a public header so libc can have some of its own functions implemented as macros for speed, and that there was not (when this hack was originally put in in the 1980s) yet much divergence in libc implementations.

Historically some FILE designs exposed the structure somewhere so that some of the f* methods could be implemented as macros or inline functions (e.g., `fileno()`).

I guess these things usually boil down to performance concerns right?

Hyrum's Law applies: the API of any software component is the entire exposed surface, not just what you've documented. Hence, if you have FILE well-defined somewhere in a programmer-accessible header, somebody somewhere can and will poke at the internal bits in order to achieve some hack or optimization.

OTOH, yes.

OTOH, when coding, I consider FILE to be effectively opaque in the sense that it probably is not portable, and that the implementers might change it at any time.

I am reminded of this fine article by Raymond Chen, which covers a similar situation on Windows way back when: https://devblogs.microsoft.com/oldnewthing/20031015-00/?p=42...

Yes, it would not be sane to depend on implementation details of something like this.

But the sad reality is that many developers (myself included earlier in my career) will do insane things to fix a critical bug or performance problem when faced with a tight deadline.

Such is life, yes.

> to achieve some hack or optimization.

Or functionality. Happens to me all the time I have some Java class that's marked Final, so instead of just extending the class and moving on, I have to copy/paste the entire class wholesale to accomplish my goal.

Personally I hate "nanny" languages that block you from accessing things. It's my computer, and my code, and my compiler. Please don't do things "for my own good", I can decide that for myself.

(And yes, I am aware of the argument that this lets the original programmer change the internals, in practice it's not such a big problem. Or the cure is worse than the problem - for example my copy/paste example.)

Another example is a private constant. Instead of allowing me to reference it, I have to copy it. How is that any better? If the programmer has to change how the constant works then they can do so, and at that point my code will break and I'll .... copy the constant. But until then I can just use the constant.

Typical "early-in-carrier" thinking. Copying implementation is totally correct move here.

All projects mentioned should have forked stdio and added their hacks/optimisations/functionality to that.

They were just too lazy. Can't blame them though. Writing C code is torture after all. One should cut all the corners they could.

> Typical "early-in-carrier" thinking.

Other way around. When I first started I thought these access restrictions were a great idea. Make sure, I, and other, only program "correctly".

When I got more mature I found how often they are impediments, and how little they actually help.

I don't want someone else protecting me from myself. I just don't want that, no matter how well intentioned.

What I meant are "avoid copy paste at all cost" and "I'm center of the world. What I want is the only thing that matters"-thinking.

About access restrictions. We have two nice examples here

1. Stdio devs can't freely make modifications because someone's code depends on private implementation details

2. And yours example. You tried to do wrong thing and those access restriction made you suffer and do the right thing.

Works as intended I'd say.

Access modifiers exist so ppl dont shoot themselves

I never asked anyone for such protection, and I don't want it.

Fortunately as time showed such things are over all beneficial for software engineers worldwide

You can usually use other langs

The OpenBSD answer to this is: fuck them they should've known better. The few pieces of software that do this and have an active port maintainer will get patched. The rest will stay broken until somebody cares to deal with the change.

And for everything that people say about backwards compatibility, and for all the times that something has broken on me... man, I'm still glad that there are people who will stand up and defend that attitude. It isn't breaking things if you point out where they were already broken. And moving fast is defensible when the only alternative is standing still.

One time, when we were discussing a potential internal change that threatened to break ill-behaved software, Steve Summit told a story like this:

"Once upon a time, pointers on the Macintosh had 24 bits. The upper 8 bits were reserved for flags. Apple warned developers not to look directly at the flags in the upper 8 bits, but to use the macros that were supplied as part of the API -- but third-party developers looked directly at the upper 8 bits anyway. When System 7 came out with full 32-bit pointers, a lot of old applications broke because of this!"

Of course, what he didn't mention at the time was that System 7 provided a toggle that allowed these programs to run with old-school 24-bit pointers -- the equivalent concession is something I don't think OpenBSD is willing to make.

Nevertheless, vendors can and have broken full backward compatibility in cases where the developers "should've known better". Hyrum's Law just states that there will be a few that don't get the message and will watch their software break when these changes are made...

I always assumed that people could poke into it, but shuddered at the thought.

I've seen old code do this over the years. When you consider for example that snprintf() didn't used to be standardized until the late 1990s. People would mock up a fake FILE* and use fprintf.

One issue – ISO C defines setvbuf (configure buffer mode and size of a stdio stream) but not getvbuf (get current buffer mode and size). On many platforms with non-opaque FILE, you can implement your own getvbuf by peering inside its undocumented fields.

I guess part of why it is not in the standard is that it is rarely requested functionality, but there are rare use cases where it may have value. And I think it is an unfortunate lack of orthogonality to have a setter but no corresponding getter.

stdio_ext.h offers some functionality like a "getvbuf", but not quite – e.g. __fbufsize tells you a stream's buffer size, and __flbf whether it is line-buffered – but it isn't clear how to distinguish fully buffered and unbuffered streams. And stdio_ext.h has never been standardised, it is an extension invented on Solaris and copied by Linux (and a few other platforms too, e.g. IBM z/OS).

In addition to "some code frobs internals", non-opaque FILE also allows for compatibility with code which puts FILE into a structure, since an opaque FILE doesn't have a size.

It happened to microsoft a while ago, they changed something in FILE and something else broke, so they went for opaque FILE.

But code outside the standard library can’t do that, can it? fopen returns a pointer to a FILE, and you can’t know how a struct FILE should be copied.

You can’t just memcpy the bits and then mix calls to fread using pointers to the old and the new FILE struct, for example. I think the standard library need not even support calls using a pointer to a FILE struct it didn’t create.

You can’t just memcpy the bits...

You certainly shouldn't, but sadly this is something which people do.

Are there any POSIX or ISO guarantees on "FILE"? I think it's safe to assume that it isn't an incomplete type, but all functions that use it operate on pointers anyway. Storing a copy of a "FILE" object might result in each copy pointing to the same underlying file handle but having different internal state.

However, who should really rely on internals of FILE? Isn't this a bad practice?

In general, it is a bad practice. However, it can be useful for some low-level libraries. For example, https://github.com/fmtlib/fmt provides a type-safe replacement for `printf` that can write directly to the FILE buffer providing comparable or better performance to native stdio.

Doesn't fwrite more or less write directly to the FILE buffer, if buffering is enabled?

I'm curious to take a closer look at fmtlib/fmt, which APIs treat FILE as non-opaque?

Edit: ah, found some of the magic, I think: https://github.com/fmtlib/fmt/blob/35dcc58263d6b55419a5932bd...

I'm curious how much speedup is gained from this.

With fwrite that would be another level of buffering in addition to FILE's buffer. If you are interested in what {fmt} is doing, a good starting point is https://github.com/fmtlib/fmt/blob/35dcc58263d6b55419a5932bd.... It is also possible to bypass stdio completely and get even faster output (https://vitaut.net/posts/2020/optimal-file-buffer-size/) and while it is great for files, it may introduce interleaving problems with things like stdout.

In SunOS 4.x `FILE` was not opaque, and `int fileno(FILE *)` was a macro, not a funciton, and the field of the struct that held the fd number was a `char`. Yeah, that sucked for ages, especially since it bled into the Solaris 2.x 32-bit ABI.

Indeed, that was the way it originally worked in all UNIXes: https://github.com/dspinellis/unix-history-repo/blob/Researc...

It was a then-important optimization to do the most common operations with macros since calling a function for every getc()/putc() would have slowed I/O down too much.

That's why there is also fgetc()/fputc() -- they're the same as getc()/putc() but they're always defined as functions so calling them generated less code at the callsite at the expense of always requiring a function call. A classic speed-vs-space tradeoff.

But, yeah, it was a mistake that it originally used a "char" to store the file descriptor. Back then it was typical to limit processes to 20 open files ( https://github.com/dspinellis/unix-history-repo/blob/Researc... ) so a "char" I'm sure felt like plenty.

I don't know if I agree, but this is one shining example of what makes *bsd's great, not being afraid of change. Linux should take note. So much of Windows' headaches stem from not wanting to break things, and needing to support old client code.

There isn't really much of "Linux" here - this code is in libc, so glibc, but that was built from portability, it isn't very Linux specific. Linux doesn't have an all encpmpassing community for userspace.

I see. I thought OpenBSD maintained their own downstream fork of glibc or something since the title/link are for their site/lists.

It may not be all encompassing,but I was referring to GNU/Linux. you can swap out bits and pieces, but what mainstream distros include by default, that's what I meant.

What you are looking at is not a GNU C library at all. It is a BSD C library.

I think you are looking for the Linux Standard Base. It started out with a great idea, but the LSB grew so large most popular distros publicly stated they would no longer pursue compliance, so the effort kinda fizzled out.

> So much of Windows' headaches stem from not wanting to break things

Quite acceptable for not having the headache for things breaking.

I am not disagreeing, but it is a design fault if you continually end up having to choose between breaking or improving things.

When you have major platform updates like windows NT rewrite back in 2000, windows 8 and now windows 11, they're opportunities to shed legacy things. The choice should have been to keep supporting a long-term-stable version of windows for security fixes (like XP or Win7) and get rid of tech designed to support old software.

Their problem now is they want everyone to be on win10 and then win11 and then whatever they come up with, or else.

You can carry legacy dependencies with you to new major versions or you can support old versions for security fixes long-term.

I empathize with the sentiment, but I think Microsoft isn't that conservative. Since XP, they have continuously dropped hundreds of features and components with varying significance. https://en.wikipedia.org/wiki/List_of_features_removed_in_Wi...

That said, there isn't a one to one correlation between the gains of dropping a legacy component and the increased capability to innovate either. Windows is highly modular, and maintaining legacy compatibility rarely is a blocker for a line of innovation, nor is a high budget workload.

I wonder though, what kind of innovation do you want to see but think that it's held back by legacy components?

"Windows" did this 11 years ago:

>FILE Encapsulation: In previous versions, the FILE type was completely defined in <stdio.h>, so it was possible for user code to reach into a FILE and muck with its internals. We have refactored the stdio library to improve encapsulation of the library implementation details. As part of this, FILE as defined in <stdio.h> is now an opaque type and its members are inaccessible from outside of the CRT itself.

https://devblogs.microsoft.com/cppblog/c-runtime-crt-feature...

Windows has kept FILE opaque for as long as I can remember. Granted, that's not very long, only 10 or so years.

> Linux should take note

Ugh, no, it should not. As a user i prefer my existing programs to keep working whenever i update my OS and as a developer i prefer to work on new code than playing nanny with existing previously working code (working code here means the code did the task it was supposed to do) because some dependency broke itself.

it's already not working that way in Linux, it's just that there is no rhyme or reason to it. Even on stable Debian I have to have pipx/pyenv/venv for python and juggle multiple versions of rust and go because of mismatches between software and distro versions.

Python, rust and Go aren’t beacons of software stability though.

This is a Python, Rust and Go issue, not a Linux issue. With "Linux" i meant the Linux kernel which does not break backwards compatibility.

Sadly, the userland does not have the same mindset as very few libraries and languages care about backwards compatibility.

However at least on Linux it is possible to do that if you care by choosing languages and libraries that do not break compatibility without worrying that the OS will break things. On OpenBSD you may write your program in C (which has strong backwards compatibility as a language) and use only libraries such as curl, cairo, opengl, openal and x11 that in general do not break backwards compatibility and yet have the OS throw all that effort out of the window.

Even on windows, it isn't usually the NT kernel that's the issue, it's the user space api's. For example, why hasn't ripgrep (and similar tools) replaced grep as a default? they're much better for performance right? (i mean on stable/mainstream distros). i get supporting old stuff, that's why there are LTS releases. But I think it is the fear of not being able to shed enough old things in new releases to displace the maintenance burden of LTS type releases that's the big issue.

Features shouldn't be ridden of just because either, there needs to be a good reason, like a better alternative.

> Even on windows, it isn't usually the NT kernel that's the issue, it's the user space api's.

On Windows the user space APIs do not change though - and Windows has a MUCH larger API exposed compared to Linux. Of course stuff does break occasionally, but considering how much software is released constantly for Windows, that breakage is incredibly minimal (and Microsoft doesn't seem to care as much about it as they used to in the past).

> For example, why hasn't ripgrep (and similar tools) replaced grep as a default?

I do not think there is a case of replacing anything here, you can use ripgrep alongside grep just fine, they are two different programs.

However there are way more people knowing grep and a ton of scripts, applications, etc, assume grep is there so it makes sense to both avoid breaking existing working stuff and take advantage of people's existing knowledge. If someone needs ripgrep's performance they can use that instead.

> i get supporting old stuff, that's why there are LTS releases.

LTS releases exist largely exactly because stuff breaks way more than it should.

> But I think it is the fear of not being able to shed enough old things in new releases to displace the maintenance burden of LTS type releases that's the big issue.

No, there is nothing about "fear" here, it is purely practical: breaking stuff that works is not a good thing. If there is some better way to do things then that is fine but 99.9999% of the cases it can be implemented alongside the existing stuff (sometimes even having the existing stuff use the new thing) instead of breaking people's code (and, even worse, forcing them to waste time learning how to use a new tool just so they can do the exact same thing they were already doing previously).

> Features shouldn't be ridden of just because either, there needs to be a good reason, like a better alternative.

No, a "better alternative" is not a good reason by itself, to get rid something you must make sure that practically nobody uses it. For example, if there was a bug on a feature or program that made it unusable and yet nobody bothered to report the bug in years, that is a good indicator nobody is using that feature / program so that it can be removed. But as long as there are people who rely on things, you should never break them.

So many words in the commit message and the announcement article, yet not a single mention of the rationale? I have a bad feeling about their practice.

Out of curiosity. If we have this inside `bits/types/FILE.h` in Linux/GNU, does it mean the type is opaque?

  ...
  struct _IO_FILE;
  
  /* The opaque type of streams.  This is the definition used elsewhere.  */
  typedef struct _IO_FILE FILE;
  ...

Unless struct _IO_FILE is expanded elsewhere, yes, it's opaque and can only be interacted through a pointer, as it would otherwise be an unsized object.

The definition of struct _IO_FILE seems to be inside `bits/types/struct_FILE.h`, and goes like this:

  struct _IO_FILE
  {
    int _flags;
    ...

so I suppose this means it's indeed expanded somewhere and, thus, not opaque?

When we're talking about opaque it's really in relation to an individual translation unit - somewhere in the binary or its linked libraries the definition has to exist for the code that uses the opaque type.

Forgive my ignorance in this topic, but if "stdio.h" itself includes "bits/types/struct_FILE.h", is anything preventing me from accessing the individual elements of FILE as they are defined in the latter header file?

It looks like FILE is not opaque in glibc. Create a translation unit that includes <stdio.h> & declares a FILE variable and it compiles fine. For comparison, create a translation unit that declares your own struct (but does not provide a definition) and declares a variable of the same type, and you'll get a "storage size of 'x' isn't known" error when compiling.

Thanks for the explanation. In case FILE was opaque in glibc, would the same test (including <stdio.h> and declaring a variable of type FILE) also fail with the unknown storage size error? If so, would linking again some library (-l) be necessary?

EDIT: after some more thinking I assume the key is that we wouldn't be able to have a variable of type FILE, but a pointer, whose size is always known.

You'd have an error about an incomplete type - see https://godbolt.org/z/G4Gsfn7MT

> a pointer, whose size is always known

Yeah, this is exactly how it works. You work with a pointer that acts like a void* in your code, and the library with the definition is allowed to reach into the fields of that pointer. Normally you'd have a C API like

    struct Op;
    Op* init_op();
    void free_op( Op* );
    void do_something_with_op( Op* );

in the header provided by the library that you compile as part of your code, and the definition/implementation in some .a or .so/.dll that you'll link against.*

Slightly side note - I donʼt understand why C++ invents new stdio based functions like std::print but rejects to import funopen() or fopencookie() which make stdio flexible similar to iostreams. As for me it should have been done ≈20 years ago.

CHERI would defend against access to internal data structures without having to bounce between address spaces, FWIW.

Please elaborate.

fopen would hand out a FILE* without capabilities to do anything with the resulting data structure, but libc itself could work with it. Libraries would get the same kind of memory protections processes do today.

How would libc get a FILE* pointer with capabilities back from a FILE* passed by the user?

libc allocates the FILE* from an array of them or a heap of some sort. It has a private capability on the start of the array and so can recover a full-capability pointer by offsetting its private capability by the distance encoded in the user FILE*. No actual memory access required, I'd think.

See https://www.cl.cam.ac.uk/research/security/ctsrd/pdfs/202306...

This sounds about right. Under CHERI when you're returning a pointer from a function, you can choose to limit its valid dereferencable range, I imagine all the way to 0 (i.e. it can't be dereferenced).

When the pointer is passed back into libc, libc can combine the pointer with an internal capability that has the actual size/range of the structure.

This isn't _too_ different to having libc just hand out arbitrary integers as FILE; libc has to have some way to map the 'FILE' back to the real structure.

To misquote the street fighter movie: OpenBSD to Linux:

"For you the day you changed your ABI was the most important day in your life, but for me? It was Tuesday"

I enjoy the dichotomy between how bad the Linux project is at changing their ABI and how good OpenBSD is at the same task.

Where for the most part Linux just decides to live with the bad ABI forever. and if they do decide it actually needs to be changed it is a multi year drama with much crying and missteps.

I mean sure, linux has additional considerations that make breaking the ABI very scary for them. the big one is the corpus of closed source software, but being a orders of magnitude bigger project and their overall looser integration does not help any.

This has nothing to do with Linux-the-project. An equivalent change would be in glibc / musl / ...

I think the difference is just the amount of people using the technology.

> bad the Linux project is at changing their ABI and how good OpenBSD is at the same task

From my perspective as a user who wants to have his programs keep working whenever the OS updates and as a programmer who does not want to waste their time playing nanny with broken dependency upgrades for previously working code (working in the sense that it did what it was supposed to do), the Linux project is actually doing the thing the right way and OpenBSD the bad way. It is basically the #1 reason i never considered using OpenBSD.

Linux' stance on not breaking backwards compatibility is exactly what i want from an OS. Now if only the userspace libraries weren't so happy to break things too...

Hold on. CVS?

Reply on news.ycombinator.com