Follow Slashdot blog updates by subscribing to our blog RSS feed


Forgot your password?

Comment Re:I can guess why IBM was pushing for IEEE 754r (Score 4, Informative) 158

(...) I'm guessing that at best, 128 bit IEEE 754r performs about half the speed of 64bit IEEE 754, purely because of the data width.

According to Douglas Crockford "'s literally hundreds of times slower than the current format.".

I don't doubt that the software implementations are "hundreds of times slower". I've had my hands deep into several implementations of decimal arithmetic and none of them are even remotely close to IEEE 754 in hardware. IEEE 754r is better than some of the predecessors because a software implementation can map the internal representation to integer maths. However, IEEE 754r does exist in hardware and I was guessing that the hardware IEEE 754r is still half the speed of hardware IEEE 754.

One other thing that IEEE 754 has going for it is the emerging GPU-as-co-processor field. The latest GPUs can do full 64bit IEEE 754 in the stream processors, making massive parallel floating point processing incredibly speedy.

Toby Haynes

Comment Re:I can guess why IBM was pushing for IEEE 754r (Score 4, Informative) 158

Why do processors need decimal number support? 10 is just an arbitrary number humans picked because they happen to have 10 fingers. There's no connection between that and computers.

Clearly you've never dealt with an irate customer who has spent $$$ on your software product, has created a table using "REAL" (4-byte floating point) types and then wonders why the sums are screwing up. IEEE754 can't accurately represent most fractions in the way that humans do and this means that computers using IEEE 754 floating point give different answers to a human sitting down with pen and pencil and doing the same sums. As humans are often the consumer of the information that the computer spits out, making computers produce the correct results is important.

There are plenty of infinite precision computing libraries out there for software developers to use. However, they are all a lot slower than the 4, 8 or 10 byte floating point IEEE 754 calculations which are supported directly by the hardware. Implementing the IEEE 754r calculations directly on the CPU means that you can get close to the same performance levels. I'm guessing that at best, 128 bit IEEE 754r performs about half the speed of 64bit IEEE 754, purely because of the data width.

Toby Haynes

Comment I can guess why IBM was pushing for IEEE 754r (Score 4, Interesting) 158

The debate over floating point numbers in ECMAScript is interesting. IEEE 754 has plenty of pitfalls for the unwary but it has one big advantage - it is directly supported by the Intel-compatible hardware that 99+% of desktop users are running. Switching to the IEEE 754r in ECMA Script would have meant a speed hit to the language on the Intel platform until Intel supports it in hardware. This is an area where IBM already has a hardware implementation of IEEE 754r - its available on the POWER6 platform and I believe that the z-Series also has a hardware implementation. I suspect that IBM will continue to push for IEEE 754r in ECMAScript, I wonder whether Intel is considering adding IEEE 754r support to its processors in the future.

Disclaimer: I have no contact with the IBM ECMAScript folks.

Toby Haynes

Comment Perl is Elegant (Score 4, Interesting) 160

Perl is one of those languages that most people meet in passing because someone else has hacked up a script to get something out of some file. Which is sad, because understanding what makes Perl different from other languages and why it is often a better choice for wrangling data isn't going to be obvious in one lousy foreach search-and-replace hack. And most people exposed to perl scripts in this manner fall over on the difference between scalar and list context and never discover why perl expressions like $lookup{$term}++ will save them years of work, while making their analysis scripts go faster.

I write Perl modules day in, day out to cope with processing DB2 internals in an attempt to model and improve them. Object-oriented Perl makes this easy, fast and effective. Closures (which I'm sure aren't understood by 90% of the Slashdot community) back this up being able to create anonymous subroutines with data attached which can be processed later. Perl is also effective for parallel task analysis - I have modules for jobserving many tasks across multiple machines and Perl threads make it easy to fire a task off while something else is done.

Perl is an essential part of data analysis for any serious volume of unstructured data. However, I'm not unhappy that it is little understood. Perl makes in the impossible merely hard. If everyone knew how to leverage Perl, I wouldn't have so much fun.

Toby Haynes

Comment Repeat after me - man pages are not help pages (Score 1) 769

Man pages are not tutorials. They are not help pages. They are reference material, designed to quickly and briefly inform the experienced user about the usage of a command and any related files.

Info pages are the next step up. They are often more detailed (witness the Emacs pages) but they are generally written for a technical audience and quite frankly, the average user isn't going to find them anyway or be able to do anything with them if they can find them,

Man pages and Info pages will, if you are very very lucky, be written by the authors of the software who speak the same language you do and are used to writing technical information. They stand some chance of being updated as the software is updated.

Next up on the list of information resources is wikis, hosted at the site where the development is done. These will generally provide more accessible information for new users, will be community driven and hopefully will reflect the state of the code in the last two years. Keeping wikis up to date is a huge task for large projects.

Beyond that, proper documentation requires constant community involvement. Most projects don't have someone looking after the docs - the software is the focus. Even teams like GNOME and KDE struggle to keep their docs even vaguely up to date - witness the dearth of information on the latest release of gdm (2.28.1) for example. And these teams do have volunteers who are trying to keep up with the changes.

Forums actually give the best location for problem solving because they quickly acquire a list of problem reports that are the things that 90% of people hit. These show up clearly in google and will, hopefully, get solutions posted or FAQs written.

Until the non-technical users of Free software step up to volunteer to write and maintain documentation for software projects, there will always be a lack of current, complete and easily accessible documentation. However, good documentation is hard to write and I often find that commercial software projects are distinctly lacking when it comes to good writing, so I don't think this is an issue which is restricted to open source software.

Toby Haynes

Open Source

Linux Kernel 2.6.32 Released 195

diegocg writes "Linus Torvalds has officially released the version 2.6.32 of the Linux kernel. New features include virtualization memory de-duplication, a rewrite of the writeback code faster and more scalable, many important Btrfs improvements and speedups, ATI R600/R700 3D and KMS support and other graphic improvements, a CFQ low latency mode, tracing improvements including a 'perf timechart' tool that tries to be a better bootchart, soft limits in the memory controller, support for the S+Core architecture, support for Intel Moorestown and its new firmware interface, run-time power management support, and many other improvements and new drivers. See the full changelog for more details."
The Internet

Are Ad Servers Bogging Down the Web? 387

blackbearnh writes "The work of making high-volume web sites perform well is an ongoing challenge, and one that continues to evolve as the nature of web content changes. According to Google Performance Guru Steve Souders, fat JavaScript libraries and rich content are creating new problems for web site tuning, but one of the biggest problems lies outside the control of web site administrators — ad servers. In an interview previewing the upcoming Velocity Online conference run by O'Reilly, Souders talks at length about the real causes of poor web performance today, and in particular, the effect that poorly performing ad servers are creating. 'We adopted a framework of inserting ads, of creating ads, that's pretty simple. And because it's pretty simple, it's not highly tuned. That's one reason why we shouldn't be too surprised that we see performance issues in third party ads. The other reason is that ad services are not focused on technology. Certainly companies like Yahoo and Google and Microsoft, we're technology companies. We focus on technology. So it's not surprising that our web developers are on the leading edge of adopting these performance best practices. And it's also not surprising that ad services might lag two, three or four years behind where these web technology companies are.'"

Comment Re:Great work! (Score 1) 236

That is hardly "Which you could 'install' by just sticking them in ~/.fonts ." and while upon closer inspection it turns out i wasn't compiling running rpmbuild, i still think that is far more work than anybody should have to do to get a some essential fonts.

No - what was pointed to you was a way you could get all the Microsoft Core Web fonts installed under the control of the package management system. That is one way.

If you happened to have all the MS fonts on a USB key, you could have just plugged it into your Linux box and copied it into your ~/.fonts directory. This too would have worked for your user, but would not have allowed another user on the same machine to use the same fonts.

Toby Haynes

Comment Plug fuses are per-appliance (Score 1) 1174

Then the electrician fucked up. He is supposed to know that breaker should not allow higher currents than the line and outlet can handle.

The electrician has no control on how many devices you hook up to the circuit. A circuit probably has 5 - 12 outlets/lights on it. The circuit breaker is protecting the entire circuit against current exceeding the safe range for the wiring. Per-plug fuses allow the manufacturer to place a lower limit per appliance - no reason why a TV would need more than a 5A fuse, whereas a kettle or microwave is probably pushing 10A+. Also, a light appliance (say a radio) probably has wiring that only copes with Cheers,
Toby Haynes

Comment Re:If LotusLive iNotes is in any way based on (Score 2, Insightful) 171

Lotus Notes isn't an email program. It's an application development platform with multiple backend databases, networking interfaces plus a scripting language, a plugin system and a ton of other stuff. That it reads and writes email is just proof of JWZ's assertion.

Toby Haynes

Comment Re:If LotusLive iNotes is in any way based on (Score 1) 171

If you've moved up from Notes 7 to Notes 8.5, you've just changed from the native Windows client to the Eclipse client. Folks like me who got our first taste of the Eclipse client at the Notes 7 native Linux release still shudder with horror (although it did allow for an extended coffee break at the start of the day). Certainly Linux Notes 7 -> 8 -> 8.5 has been a journey of performance improvements.

Now I hear colleagues in the corridors moving from Notes 7 to Notes 8 on Windows XP. They aren't too happy right now.

Toby Haynes

Comment Re:If LotusLive iNotes is in any way based on (Score 4, Informative) 171

Two simple examples: - we just "upgraded" to v8.5. It takes 127 seconds to start up. It takes 38 seconds more to show me my inbox. It takes 47 seconds to bring up the editor to reply to the first mail of the day. This is all on a fairly new Dell D630 laptop.

You're kidding me, right? Or you're making up numbers. Or you are running the Windows version amid the antivirus scans...

Linux box, Fedora 11, T60p, 5400rpm drive - hardly a world beater laptop these days. Times are all intervals.

  • Time to password screen - 3 seconds
  • Time to interactivity with the welcome screen (the one that displays all the mail, calendar and to-dos) - 10 seconds
  • Time to display Mail window after clicking mail tab - 2 seconds
  • Time to show contents of the top mail message - 1 second.

Release 8.5 Revision 20081211.1925 (Release 8.5) Standard Configuration

So either you've got the CPU clocked down or something is eating your cycles. I hardly hold Lotus Notes in high regard but its improved performance significantly in recent releases.

Toby Haynes

Slashdot Top Deals

I go on working for the same reason a hen goes on laying eggs. -- H.L. Mencken