Preliminary diagnosis for a VAIO PCV-J15

I’ve made a few attempts on building complete systems on the hand-me-down VAIO I mentioned a day or two ago. For my money (which was nothing) the machine is working almost perfectly, with no visible defects aside from fading on the case, or perhaps an odd scratch or two.

The best improvement I made, at this point, was to clean the screen properly, with a solution of one part white vinegar with one part distilled water. That’s usually what I prefer for laptop screens, and it took off a nice layer of gunk without any trouble. I buffed the glass with a microfiber cloth and now it’s crystal clear.

The monitor, although it is a giant rock, actually has a very clear image and will display an extraordinary resolution. I actually knocked down the screen dimensions because the detail it was showing was making the text hard to see, if you can believe that. It’s labeled as a Trinitron, and so I suppose I shouldn’t be surprised that it has a good picture. Too bad the physical size of the screen isn’t a little bigger; it would make that detail a little more comfortable on the eyes.

Internally, this has an SIS730 graphics system and chipset, as I mentioned earlier. This particular setup shares memory with the rest of the system and is configurable (to a degree) through the BIOS. I went ahead and gave it the full 32Mb it was allowed, and 3D acceleration — through the xorg-video-sis driver — was smooth in both Ubuntu and Arch, using a resolution of 1024x768x24. It’s definitely no contender for a modern graphics prize, but most of the GL xscreensavers in Arch were quite fluid.

That same architecture handles audio and a few other subsystems on this machine; everything seems to be working well in that category too.

Memory requirements for the desktops I put together — one using Openbox with LXDE, and one using IceWM — were low enough that I never noticed any paging or stutter, even though there is only 192Mb in the machine. In fact, the only time I was ever cautioned about memory at all was by mtpaint of all things, when I tried to resize a giant TIFF image. Adjusting mtpaint’s preference made that error go away.

Network speeds are about what I expected through the RealTek RTL-8059 — around 70-80kbps, with occasional spikes depending on the server. That card needs the ne2k-pci module in order to work; it took me a little time to find the relationship between the two, but once I had that magic combination, everyone was happy.

From what I can tell thus far — and I still haven’t cracked the case yet — there’s a 54X CDRW in this machine that would not have come standard. And I believe there’s a high-speed USB card in place, which would explain an additional row of USB jacks in one riser panel. Aside from that I’ll need to open it all up to see if there’s more fun stuff inside. And no, I’m not going to play with the modem.

The hard drive is running strong, with no errors yet reported. I’ve blanked the drive once, reinstalled two or three times and even gone so far as to rerun the original recovery CDs and put Windows Me back on there. The only disappointment was, of course, WindowsMe. I forgot how rotten that was. …

Anyway, boot times for Linux are pretty strong for an 800Mhz machine; Ubuntu takes its normal 1-minute-plus for a console-based system; Arch is lagging at around 44 seconds, with almost 18 of those seconds lost to the udev-modules-setup sequence. Taking out module autoloading chops another 10 seconds off that boot time; I’ll save my rants against the Arch loading procedure for another time. 👿

(I didn’t mention Crux because the only Crux system I managed to build was somehow out of whack, with the Grub bootloader spitting out an infinite trail of GRUB GRUB GRUB GRUB GRUB messages. I blame myself for that one; I wasn’t paying attention while I was installing. 😳 )

And for comparison WinMe appears to be up in around 56 seconds, but of course, you can’t really do anything until about 1:12. Or at least, that’s when the hard drive stops churning.

So the preliminary diagnosis for this machine is rather strong: No failed components (or at least no failed crucial parts), a speedy system and a clean display. All the cables and tertiary components are here, along with the system recovery software. I’m confident it will perform quite well in both Windows and Linux, so I’m comfortable with setting up a dual boot system with a lightweight but comprehensive alternative to Windows Me, and passing it on to its next life. If you have suggestions on a distro to suit it, I’d be happy to hear them. It’ll need full multinational support for both keyboard input and text display, because it’s likely to go on to a Japanese-speaking user. Beyond that, it just needs to look nice.

If I can get a stretch of clear weather in the next week or so, I’ll strip it down to the metal, scrub it vigorously and reassemble. I promise photos. 😉

2 thoughts on “Preliminary diagnosis for a VAIO PCV-J15

  1. Pingback: Fair but honest? Xubuntu 8.10 « Motho ke motho ka botho

  2. Pingback: Returning the animal to its natural habitat « Motho ke motho ka botho

Leave a comment