TideLog Posts Tagged “bug”

I’ve just refurbished a Thomson Lyra PDP2842. Replaced the HDD, dumped and modified the NAND firmware to iron a few “HDD Access Error” bugs out, and replaced the casing. It now seems to run OK, but hasn’t got the latest firmware yet, 3.03. I have it, but not yet installed it.

The firmware is in two parts. The NAND firmware on a chip on the board that controls the VFD LCD, bootloader, power and USB functions, and the HDD user-upgradeable software, which controls MP3, WAV, WMA playback, playlists, profiling, and radio play/record and tuner. The Hitachi hard drive firmware is separate completely from the player, and can be upgraded, but there’s no need to.

I connected the player to my Windows 7 machine, and the thing doesn’t work as a USB HDD. The computer makes the “do-dum” sound, and then pops up the dreaded “A device attached to this computer has malfunctioned, and Windows does not recognise it” message about six seconds later. THIS IS NOT A FAULT. Not an intentional one, anyhow. The USB function does work under XP and Vista. It’s happening because the motherboard of the player isn’t identifying itself and the HDD fast enough for ol’ Seven to think it’s responding. The board of the player is a bridge between the computer and the HDD, so it identifies itself as an IDE controller bridge, and then passes communication to the HDD, which the computer then detects, and installs, letting you use the HDD.

If this process doesn’t happen in a certain time, Windows 7 thinks there’s a fault, and ceases communication to prevent damage. My modifying of the NAND dump didn’t cause it, it happened before so I thought I’d see if I could fix it. My old one does the same.

I’ve had two of these blasted things. Poor firmware, poor design, and crap customer support saw an early death by Thomson and RCA of its own product. I sent bug report after bug report back in ’05, only to have them read the same tripe off a script to me in emails. I got fed up, and got Kassie to hex edit the lot. We recently figured how to dump the NAND contents using a probe, so I grabbed another Lyra and gave it another go. I sent my knackered Lyra to Her Ladyship in Tokyo (the HDD ribbon tore, making it damn more useless than ever!) and she played about with her probes, dumpers and chip programmers.

NEVER, EVER buy Thomson or RCA intentionally! Only if you wanna smash the shit up! Having said that, the Lyra Kassie has just helped me fix is almost brand new, and runs better than damn Thomson ever made it do. The software is smooth as oil! Oh, and I got rid of the stupid EU volume limit, because they said they had, and buggered the Equaliser up!

Comments 4 Comments »

This is the version on mine, and it seems really weird to me. It’s stopping me installing Mac OS x86 on it, no matter what I try, I’m getting “Still waiting for root device” after installation, regardless of what drivers I install. I’m using iPC 10.5.6.

Primary Master has the DVD-RW drive. The S.M.A.R.T option is on Primary, but Disabled and unselectable. For an optical?!?

Secondary Master is where the hard drive is. There’s NO S.M.A.R.T option. See the wrong picture?

From my days as a PC engineer, I know for sure that optical drives DO NOT even support S.M.A.R.T, let alone USE it. And the HDD should almost ALWAYS be Primary. The assignments cannot be changed, either, no matter how many times you take drives out, try just one, etc. SATA doesn’t use jumpers anyway, but I think the optical on this still uses IDE, but has no jumper (I’ve seen optical drives with a jumper switch on the back, in fact, I have one somewhere!)

I’m sure it’s a bug. My Novatech support rep, Rebecca (she’s lovely) is sending me an Update ISO image. My instincts tell me that this is either a BIOS code bug, a design fault, or just lazy engineering. I’ll update when I have more, and will share the disc image if successful.

Comments 3 Comments »

The error I posted yesterday isn’t directly a fault with the developers of the ReactOS source itself, but the Installer code for the ReactOS Build environment (RosBE) It is set to install to “%systemroot%\Program Files” by default, and it shouldn’t, as spaces in directories can cause breaks in code compilation.

I had totally overlooked this when installing RosBE, now everything seems to be fine, I’ve installed it in C:\RosBE

Comments 3 Comments »

Debugging SparkIV in SharpDevelop didn’t tell me anything, it just crashes out, without any errors in the breakpoint watcher. Whether the situation is different in VS2008 I don’t know, as I don’t have it installed on this machine at present

SparkIV debug from source was inclonclusive. I expected the breakpoint watcher to detect a bug, but it didn't.

Comments No Comments »

I discovered this while setting up my SVN environment for debugging SparkIV. TortoiseSVN is my preferred program, but it is entirely context menu driven. You HAVE to use the Explorer context menu, otherwise running the EXE link in the program group results in this not-so-helpful dialog popping up:

Tortoise's not very helpful .exe message! What when the context menu doesn't work, huh?

Tortoise's not very helpful .exe message! What when the context menu doesn't work, huh?

Trying to run the commandline doesn’t extend to creating repositories, either, you HAVE to use the context!


Comments No Comments »