Slashdot Mirror


User: Ben+Hutchings

Ben+Hutchings's activity in the archive.

Stories
0
Comments
1,450
First seen
Last seen
Profile
(view on slashdot.org)

Comments · 1,450

  1. Re:Why? on Writing an End to the Bio of BIOS? · · Score: 1

    How about fixing the boot process? Historically, the BIOS would only boot from the first floppy drive (drive 0x00) or the first hard drive (drive 0x80). There was no need for the BIOS to pass the device number to the boot loader, since a bootable floppy would only ever be booted as drive 0x00 and a bootable hard drive would only ever be booted as drive 0x80. Of course people do want to boot off other devices, and BIOSes now support that, but they have to change the drive number mapping to do so. This makes drives inaccessible and can make it hard for an OS installer to work out what the drive number mapping will be when the OS boots off hard drive.

  2. Re:OF? on Writing an End to the Bio of BIOS? · · Score: 1

    A PC BIOS is not "8-bit"; some parts of it run in "16-bit" real mode and others in "32-bit" protected mode. Besides which, the archaic interface between OS and BIOS doesn't restrict what the BIOS can do internally. A modern PC BIOS is smart enough to configure pretty much everything automatically. The BIOS setup program lets you fiddle with bus timings because some people like to push their hardware, but there's no need to do that.

    I agree that the boot order configuration is crappy and there's no good reason for that. It would be simpler to have a setting for a default boot device and the option to override that on a one-time basis, as on the Mac (or the Amiga).

    Boot device selection is really quite weird on the PC because the BIOS spec says the BIOS will only boot from the first floppy or hard drive. If the BIOS boots from any other device it has to change the drive number mapping.

  3. Re:Of course it isn't the end of the world! on Old School Data Mining, Maritime Style? · · Score: 1
    The thing that the "we're all going to die!" extremists miss, is that the changes will happen over a very long period of time.

    The thing that the head-in-the-sand extremists miss, is that climate changes could happen very quickly - extreme inputs could push chaotic atmospheric and ocean current systems over to another attractor. For example, there is the possibility that the Gulf Stream will be stopped by meltwater from the arctic (see this thread).

  4. Re:Of course it isn't the end of the world! on Old School Data Mining, Maritime Style? · · Score: 1

    You might want to think about how the costs are going to be shared out in those cases. Implementing the Kyoto Protocol is more expensive for developed countries that use a lot of fossil fuels than for developing countries that don't. Dealing with floods is likely to be very expensive for small island nations and not so bad for large landlocked ones. The developed countries tend to fall somewhere in between those extremes which means it might be cheaper for them not to implement Kyoto. However the costs on both sides are, I suspect, largely a matter of guesswork.

  5. Re:Debian still down on Savannah Back Online With Extra Security · · Score: 4, Interesting

    Debian has gradually been bringing services back online as the relevant files are verified and new passwords and keys generated. They are also tightening security in some ways, e.g. dropping pserver access to CVS servers. Alioth and www.debian.org are the latest services to be restored.

  6. Re:No offense on Fingers Crossed for Beagle · · Score: 3, Informative

    The reason for this? They accidentally uploaded the Ariane 4 software to the Ariane 5 before launch. Needless to say, the rockets didn't work exactly the same. :-)

    This math bug caused both the primary and backup computers to hang.

    They didn't use the Ariane 4 software by accident. They intentionally re-used the software (presumably with some constants changed) and tried to save money by omitting thorough re-testing.

    See section 2.1 of the report on the Ariane 5 failure for a full explanation of how it happened.

  7. Re:so what? on Open eBook Forum Courts Controversy Over Formats · · Score: 1
    ...or (as with the PDF version of the ANSI C++ Standard) I can't copy at least small portions of to quote.

    The restriction on copying text was a publication error. The current PDF version permits it. I don't know whether there's any possibility of a free "upgrade".

  8. Re:rash of naughty dates coming on Time's Up: 2^30 Seconds Since 1970 · · Score: 1

    MS-DOS uses a 7-bit year counter and I can't think why it would be signed (all the other time fields are unsigned) so I would expect it to work until 2108.

    AmigaOS uses a 32-bit second counter which is definitely unsigned so it should be OK until 2114 (the epoch is 1978). The battery-backed clock driver would probably need to be patched before 2078 though.

    Some more dates:

    06/07/2079 - SQL Server SMALLDATETIMEs

    01/01/10000 - ANSI SQL DATEs and TIMESTAMPs

  9. Re:Prepare for the Y10K Bug! on Time's Up: 2^30 Seconds Since 1970 · · Score: 1

    atoi is not your friend. Its behaviour when given invalid or out-of-range input is undefined. What you really want is strtoul.

  10. Re:Prepare for the Y10K Bug! on Time's Up: 2^30 Seconds Since 1970 · · Score: 1

    Have you tried looking on eBay?

  11. Re:$4000! on Former Netscape Executive gives $4000 to AmiZilla · · Score: 1

    I think the problem is that Mozilla would need at least 64 MB of RAM and a 200 MHz processor to be at all usable. That means it's going to require a PowerPC board, which limits the potential user-base further. (It may also complicate the programming due to the need for coordination between the PowerPC and 680x0 side of things and the various different PowerPC boards out there but maybe that has become easier in the past few years.)

  12. Re:$4000! on Former Netscape Executive gives $4000 to AmiZilla · · Score: 1

    The only one worth using is GNU C++. The others are buggy as shit and no longer maintained, so far as I can tell.

  13. Re:Intriguing, but likely impossible on Intel To Produce Cheap LCoS Chips · · Score: 1

    Well I wasn't entirely serious.

  14. Re:"TrustedTV(tm) on Intel To Produce Cheap LCoS Chips · · Score: 5, Interesting

    I expect that in the future video cameras will detect watermarked images and refuse to record them.

  15. Re:Still a problem? on Blockbuster Chief: End DVD Region Codes · · Score: 1

    There are a few exemptions, but that's approximately right. Here are the details.

  16. Re:Examples? on Blockbuster Chief: End DVD Region Codes · · Score: 1

    Oddly, Blockbuster in the UK doesn't seem to do anything like this. I rented the original version of Y Tu Mama Tambien from there, which was unrated (aka X-rated) in the US.

  17. Re:Sql Server CE? on Server CE Database Development with .NET · · Score: 1

    If you read the small print, you'll find that SQL Server is only claimed to support SQL-92 Entry, which is a tiny subset of SQL-92 that doesn't even include VARCHAR. Yet it still fails to conform to that; it follows SQL-89 in some cases where SQL-92 differs. If you care about ANSI conformance I think you'll be better off with PostgreSQL.

  18. Re:you're missing a lot on 64-bit Linux On The Opteron · · Score: 1

    Many people at Microsoft seem to have adopted a form of Hungarian notation without really understanding it. As you say, prefixes like "dw" or "ul" are quite useless; the prefix on a variable name should convey the purpose of the variable and not its concrete type. File positions should have a prefix like "pos". Handles should have a prefix of "h" plus something indicating what type of thing they are handles to.

  19. Re:I dont' get it... on Microsoft: Patches, Patches Everywhere! · · Score: 1

    OK, I won't mention the XDA or the Orange SPV.

    Oops.

  20. Re:Cool. But get your own postal system. on ICANN Troubles At UN Summit On Internet · · Score: 1

    Only stamps on international mail need to have the country name on them. For domestic mail it's obviously unnecessary.

  21. Re:Screw em all. Use OpenNIC on ICANN Troubles At UN Summit On Internet · · Score: 1

    This is about the ITU, not the UN General Assembly. Don't think that just because one part of the UN is impotent that its agencies have no influence. The ITU has real power. If you ever make international phone calls you're relying on ITU standards such as country codes. The ITU is dominated by current and former state telephone monopolies that like smart networks and dumb nodes and who feel threatened by the Internet. It does seem to me to pose a real danger.

  22. Re:Pentax K-1000 on Best 35mm SLR Camera for Beginners? · · Score: 1

    There should be no need to log the camera settings if you're shooting on digital. EXIF files can hold all that information in the header and I expect that the uncompressed file formats do the same.

  23. Re:Truly Sad..... on Microsoft Retires Windows 98 · · Score: 1

    Huh? AmigaOS has (and always had) preemptive multitasking. Of course it doesn't have memory protection.

  24. Re:The fair vote initiative on Gerrymandering by Computer · · Score: 1

    Many European countries use PR, including Italy and Germany. Germany has a 5% threshold for national elections (i.e. parties with less than 5% of the total vote get no seats) which keeps the far right out. It currently has a fairly stable red-green coalition.

  25. Re:In the UK on Gerrymandering by Computer · · Score: 1

    It's not really gerrymandering - she used the alternate scheme of moving voters (council tenants) instead of boundaries, which is what cost so much public money.