TODO: <Prodcut name>



  •  I have been reading TDWTF for quite a while, and I recently came across something possibly worth a submission.  I have recently built an entirely new computer (for gaming) and have suffered through a number of minor WTFs when trying to find up-to-date drivers for some of the hardware. Some of Gigabyte's software was especially odd.  They include @BIOS, which is a program run from within Windows to update the BIOS.  Said program has a "LiveUpdate" button that appears to have nothing to do with the BIOS but is designed to update the bios auto-updater. 

    [IMG]http://i710.photobucket.com/albums/ww108/bawkbawkboo1/screenshotBIOS.png[/IMG]

    Clicking it made the program temporarily unresponsive while it attempted and (minutes later) failed to access Gigabyte's server.  The following is the info for the autoupdate "feature" (through Windows Defender).

     [IMG]http://i710.photobucket.com/albums/ww108/bawkbawkboo1/ScreenshotWindowsDefender.png[/IMG]

     

    Not sure if this is worth a submission.

    - Jack L.



  • I beleive "TODO: <Product name>" is the default product name for MFC templates.  You can often see it in applications with the default MFC blocks icon.



  •  I wonder how many people know what "Clear DMI data pool" means?



  • @bawkbawkboo1 said:

    This is a product purchased in mid-2009, it has flash memory installed on the motherboard, but only 1M Byte of it?  Or is it 8M Bytes, and they just mis-labeled it?  Either way, it's a WTF.

    (Yes, I know I've posted things which one could construe to me not being a fan of motherboard flash memory.  It's true that in certain circumstances I think it's a bad idea (although, those circumstances mostly revolve around either not having a read-only/read-write jumper or selling the product to an idiot.)  However, if you're going to do it, flash memory's cheap enough these days there should be *at least* 4MB if you're going to do at all.)

    @bawkbawkboo1 said:

    Not sure if this is worth a submission.

    Several, IMHO.  I suspect strongly that we have not explored the depths of the WTFs here.



  •  That seems daft to me, we don't know the size of the data in the flash apart from that its obviously less than 1MB. Lets assume it 600 - 800KB. Why would you put 4MB of it in there just because its cheap enough? If your never going to exceed the 1MB barrier, than 2MB should be absolute maximum (and if it's something like 600KB, then I'd say 1MB is ample). 

    It's not like the user can use the flash to storing his MP3's in after all, it's just the BIOS code and settings. 

    Secondly, why would the user care about the flash memory type or size when they can't access it anyway? All they'll care about is if there's an update available, and if so, use a program to update it (with good reason, I don't know anyone that upgrades there bios just because theres a newer version.They subscribe to the "If it works, don't f___ with it" rule). 



  • @Mole said:

    I don't know anyone that upgrades there bios just because theres a newer version.They subscribe to the "If it works, don't f___ with it" rule
     

    I did once on a UMPC, but because the update promised "improved performance" (which I later learned meant it runs the fan more often, so less heat, so better performance, right?). Oh well, I figure it'll improve my CPU's lifetime as this computer runs pretty hot...



  • The RTWTF is that nobody worries about spelling prodcut right, including me.



  • @Mole said:

    I don't know anyone that upgrades there bios just because theres a newer version.

     

    I know several... this included me until I upgraded the BIOS on my laptop hoping for a new option... and ended with my laptop regularly exiting sleep on its own, of course always tightly packed in its bag, with me finding a barely touchable 70°C block with empty battery when taking it out to use it. And it required several WTFy workarounds to revert, as the manufacturer never thought they would screw up and the updater program didn't allow flashing of an older version than the installed one...



  •  So the new feature was to turn your laptop into a radiator?



  • TRWTF is that the BIOS flasher runs under Windows. That's wrong for so many reasons.

    Also, you might put more memory in than you need if you expect to add features later. Not really likely for a BIOS though. Is 4MB really necessary for what is essentially a boot program? Super Mario RPG fits in 4MB, and I don't think your average BIOS has nearly as much in the way of 16-bit pseudo-3D graphics.


  • Discourse touched me in a no-no place

     @Kilrah said:

    @Mole said:

    I don't know anyone that upgrades there bios just because theres a newer version.

     

    I know several... this included me until I upgraded the BIOS on my laptop hoping for a new option... and ended with my laptop regularly exiting sleep on its own, of course always tightly packed in its bag, with me finding a barely touchable 70°C block with empty battery when taking it out to use it. And it required several WTFy workarounds to revert, as the manufacturer never thought they would screw up and the updater program didn't allow flashing of an older version than the installed one...

    The solution is easy. Get the REAL flasher for the BIOS - the one from the BIOS vendor, not the system vendor. Now call me when you've done something truly impressive with a laptop BIOS - like hacked it so it can run unauthorized PCI cards in the internal socket. My HP currently identifies itself as a "Hackett Peckerd"  with BIOS revision A69 (or something like that) and wears a VASTLY overpowered (108mw, the stock nic was 22mw) internal wifi NIC using an Atheros chipset - not the silly Broadcom garbage HP sold with it stock - nevermind the beefy upgraded antennas. The BIOS is even signed.

     

    And I did all of this with nothing but BALLS OF STEEL and a hex editor (and a misappropriated BIOS authoring toolkit that I used to sign the damn thing) - no documentation, no tutorials, and only a vague idea that it might be possible. 



  • @Weng said:

    The solution is easy. Get the REAL flasher for the BIOS - the one from the BIOS vendor, not the system vendor. Now call me when you've done something truly impressive with a laptop BIOS - like hacked it so it can run unauthorized PCI cards in the internal socket. My HP currently identifies itself as a "Hackett Peckerd"  with BIOS revision A69 (or something like that) and wears a VASTLY overpowered (108mw, the stock nic was 22mw) internal wifi NIC using an Atheros chipset - not the silly Broadcom garbage HP sold with it stock - nevermind the beefy upgraded antennas. The BIOS is even signed.

     

    And I did all of this with nothing but BALLS OF STEEL and a hex editor (and a misappropriated BIOS authoring toolkit that I used to sign the damn thing) - no documentation, no tutorials, and only a vague idea that it might be possible.

    You bumped up the wifi transmit power in the BIOS?  I didn't even know that was possible.  The wifi cards I've seen require the firmware to be monitored by a "regulatory daemon" to ensure the power levels are within FCC spec.  I always figured it was possible to hack the firmware and regulatory daemon to circumvent this (although I really never have needed to) but I was unaware the BIOS could even control the wifi transmit power...



  • Not sure, but IIRC Atheros' cards have unofficial unlocked firmware available.

    What he did with BIOS was removal of the PCI-lock that accepted only OEM cards with HP as vendor ID in the slot... or even not removal but change of the one accepted ID, perhaps?



  • @Weng said:

    And I did all of this with nothing but BALLS OF STEEL and a hex editor (and a misappropriated BIOS authoring toolkit that I used to sign the damn thing) - no documentation, no tutorials, and only a vague idea that it might be possible.
    Reminds me of the times when I was changing the bootup logos on various motherboards (and removing useless [to me] ROMs to make more space for nicer images). Of course, the first motherboard where such BIOS failed to boot completely had the chip soldered.



  • I wonder how much it cost them to make that fancy UI for what is essentially just a window with 4 text fields, 4 push buttons and 2 checkboxes? Oh, and a help button. Is the help button usually next to the minimize button on Windows?



  •  I personally adore the "current status of the motherboard"...like it's going to change suddenly.



  • @Benanov said:

    I personally adore the "current status of the motherboard"...like it's going to change suddenly.

    and where does it say that?



  • @Kilrah said:

    @Mole said:

    I don't know anyone that upgrades there bios just because theres a newer version.

     

    I know several... this included me until I upgraded the BIOS on my laptop hoping for a new option... and ended with my laptop regularly exiting sleep on its own, of course always tightly packed in its bag, with me finding a barely touchable 70°C block with empty battery when taking it out to use it. And it required several WTFy workarounds to revert, as the manufacturer never thought they would screw up and the updater program didn't allow flashing of an older version than the installed one...

    And right here's the number one reason for more memory in the BIOS flash: space for backups.  Someone else already mentioned the room for expansion - which is why I suggested 4M instead of 2M.

    For what it's worth, the LinuxBIOS people apparently found great performance benefits by replacing the BIOS with a real OS.  I'd have been more interested, except that the benefits are entirely at boot time - an event I would prefer to eliminate rather than optimize.  (That is, boot once (preferably at the factory), and then suspend for short downtimes and hibernate for long ones.  OS upgrades get performed live with something like Ksplice when kernel updates are needed.)



  • Backups are of no use if you can't use them, as is the case of a typical bios upgrade failure or virus. 

    A more reliable version of your suggestion already exists:

    "Since the procedure to recover erased or corrupted BIOS involves the
    removal of the BIOS chip with the computer turned on, many motherboards
    manufacturers launched motherboards equipped with two BIOS chips: a
    main BIOS and a backup BIOS. Therefore if the main BIOS is erased -
    either accidentally by means of an unsuccessful BIOS upgrade or
    intentionally by a virus - you can turn on the machine using the backup
    BIOS and recover the contents of the main BIOS without the need of
    any fancy procedure to reprogram it. You can find motherboards from
    Gigabyte, Albatron and Chaintech with this feature."


  • Discourse touched me in a no-no place

    @bannedfromcoding said:

    Not sure, but IIRC Atheros' cards have unofficial unlocked firmware available.

    What he did with BIOS was removal of the PCI-lock that accepted only OEM cards with HP as vendor ID in the slot... or even not removal but change of the one accepted ID, perhaps?
    I identified a whitelist of about 7 PCI ID's for the internal MiniPCI socket - inserting anything other than one of those would cause the BIOS to throw up a message accusing you of being gay and refuse to boot. Funnily enough, one was the Intel centrino wireless board - and this was an AMD laptop. Code reuse anybody?

     At any rate, vendor IDs were stored in a list (little endian bytes of course), and then the device IDs (little endian) in another parallel list (so vendor IDs were duplicated in the previous list), and then valid subsys IDs (big endian, oddly enough) in yet another parallel list (so device IDs were duplicated in the previous list) - all scattered about the file in totally random places.

     

    Amusingly, figuring that out was the easy part. Actually getting it flashed was rather harder. The windows-based HP flasher would only take HP-signed (not Phoenix-signed, as I had created) BIOSes. The Phoenix flasher required a floppy boot - and even once I'd procured a USB floppy drive and a working floppy disk, the buggy-ass HP BIOS wouldn't let me boot to it.

     Eventually, I discovered that the board contained a rescue flasher (the Phoenix version) that would automatically load if the installed BIOS failed. Using my BALLS OF STEEL, I flashed the HP BIOS over with an ancient version and pulled battery in the middle of it. Next boot, I was staring at a BIOS flash screen. It happily read my BIOS off the floppy and all was well.



  • @Weng said:

    It read my BIOS off the floppy
     

    Between this and the molten microwave, you are one of the more awesome people to have entered my onlife.



  • @Mole said:

    Backups are of no use if you can't use them, as is the case of a typical bios upgrade failure or virus

    In a typical motherboard.

    In a well-designed motherboard, the BIOS nvram starts off with a 'BIOS selector' routine which isn't included in the upgrade process, that activates the newest code that hasn't failed recently - and thus, the backup is accessible, so long as it hasn't been overwritten with a virus.  The backup hasn't been overwritten by a virus, because it's only writable when the BIOS write switch is in the 'allow' position, and the user only puts it in that position for a BIOS update, and immediately resets it after he's done with the update.  And he really does this, because this is important to him - otherwise he wouldn't have managed to go through all the hoops you have to in order to get such a motherboard, due to the fact that idiots prefer the motherboards that let viruses infect them whenever, and idiots greatly outnumber the rest of us.

    I can't find motherboards with this feature, because I'm not determined enough.  But I believe they exist, because I'm like that.

    (I admit, that design would need some form of writable nvram or semi-vram (such as capacitor-backed RAM) to detect for 'recent failure's.)


  • Discourse touched me in a no-no place

    @tgape said:

    I can't find motherboards with this feature, because I'm not determined enough.  But I believe they exist, because I'm like that.

    (I admit, that design would need some form of writable nvram or semi-vram (such as capacitor-backed RAM) to detect for 'recent failure's.)

    Funnily enough, most of Gigabyte's current lines have a redundant setup like this - including the OP's. Their setup is two distinct physical chips in contrast to your soft barrier. The Gigabyte update tool in the OP will update one chip, verify it, and then update the other. Flashing "manually" with the BIOS vendor's tools will just update the primary chip.  This thread has officially come full-circle.


  •  There exists several versions. The last motherboard I bought didn't have this "Update the primary, verify, and then update the secondary". It only ever updated the primary - the secondary wasn't flashable. If your system didn't boot, a simple jumper selected the alternative bios chip to get you up and running. 

     (A failed flash also greeted you with the "Insert boot disk with BIOS to read" prompt if you didn't use the jumper, but really, how many people have floppy drives these days?)


  • Discourse touched me in a no-no place

    @Mole said:

     (A failed flash also greeted you with the "Insert boot disk with BIOS to read" prompt if you didn't use the jumper, but really, how many people have floppy drives these days?)
    I DO! Funny story - back when I bought the chassis I currently use for my primary PC, Vista was still in its early days. In fact, it wasn't even a certainty I was going to continue using it because my hardware was supported exclusively by alpha and beta drivers (this, combined with 512mb-of-RAM PCs, is actually what caused Vista to fail so miserably) - but XP didn't support my hardware without a mass storage driver, and as we all know the only way to get that into XP setup (without having to waste a small lifetime making custom Windows disks) is with a floppy disk. Therefore, I installed a floppy drive.

     

    However, I wasn't about to ruin the smooth lines of the computer with a floppy drive visible out the front panel. Instead, I installed it on its 5.25" carrier sled BACKWARDS, so 100% internal to the chassis.

     I recently took the PC offline to do a complete upgrade and the first words out of my mouth when I opened the case were "WHAT THE HELL IS THAT!?" - I'd completely forgotten about the internal floppy drive.

     I contemplated removing it, because Vista (well, Win7) is here to stay, but decided it's not hurting anything. I then contemplated not recabling it and just leaving it there disconnected - but I decided against that because the connectors were RIGHT THERE anyway.

    This is why my shiny, monolithic black Windows 7-based quad core PC, with its gigantic 12 inch long aluminum-armored video card, and huge pile of RAM makes that hella retro floppy seek sound on boot. Younger/more forgetful people who have witnessed the boot process have actually asked "What was that?"



  • @Weng said:

    This is why my shiny, monolithic black Windows 7-based quad core PC, with its gigantic 12 inch long aluminum-armored video card, and huge pile of RAM makes that hella retro floppy seek sound on boot. Younger/more forgetful people who have witnessed the boot process have actually asked "What was that?"


    I want to be like him when I grow up.



  • @bannedfromcoding said:

    I want to be like him when I grow up.
     

    We all want a little weng.



  • @dhromed said:

    @bannedfromcoding said:

    I want to be like him when I grow up.
     

    We all want a little weng.

    Not me.  I already have a little weng.  I want a huge one!



  •  @bstorer said:

    I already have a little weng.  I want a huge one!

    You're in luck. This forum is brimming with huge wengs.


Log in to reply
 

Looks like your connection to What the Daily WTF? was lost, please wait while we try to reconnect.