Down to the bare metal

[BigV]

I'm soooooooo tired of the performance of my laptop. It takes about ten minutes from power on to cruising altitude.

*Yaaawn...*

Just thinking about it makes me really yawn.

So, I intend to take it down to the bare metal. I'm going to re-fdisk the hard disk, reinstall xp, all the service packs, office, and about 30 different applications I wrote down. I hope I don't reinstall the crap that has accumulated from the many months since the last time I went through this.

I will be incommunicado in the meantime, so I have a strong motivation to get myself back online. I don't reckon you'll all miss me much but I'll miss you. Who knows, maybe I'll be back online today. Unlikely, but possible. Hmm. Maybe I'll reinstall Firefox first. But then I may never finish the rest of the system. Oh well.

See y'all soon!

[Rock Steady]

Originally Posted by BigV

I'm soooooooo tired of the performance of my laptop. It takes about ten minutes from power on to cruising altitude.

*Yaaawn...*

Just thinking about it makes me really yawn.

So, I intend to take it down to the bare metal. I'm going to re-fdisk the hard disk, reinstall xp, all the service packs, office, and about 30 different applications I wrote down. I hope I don't reinstall the crap that has accumulated from the many months since the last time I went through this.

I will be incommunicado in the meantime, so I have a strong motivation to get myself back online. I don't reckon you'll all miss me much but I'll miss you. Who knows, maybe I'll be back online today. Unlikely, but possible. Hmm. Maybe I'll reinstall Firefox first. But then I may never finish the rest of the system. Oh well.

See y'all soon!

Yes, I do this on all my winboxes every so often. Reformat/reinstall. In the NT days it was every 9 months, with XP its more like every 2 years.

Break a leg, dude.

[barefoot serpent]

you did a defrag first?

[BigV]

I'm only back online long enough to confirm that the first 20 dingdang restarts and failed reinstalls aren't because I hosed something. Just getting the stupid thing to boot from cd is a challenge.

Oh, and my Dell supplied silkscreened labeled cd that says "Reinstallation CD for Microsoft Windows XP Professional Including Service Pack 2" is really XP Home. No sh*t. I did it twice. And that was also one of the things I wanted to double check on this restart. Yep. True. So, it appears I don't have a XPSP2 disc, and can't install usiing the gui. Whatever. I'll beat this *%@!^)^ into submission. I may find myself down to the local Fry's for a new 100GB 2.5" hd and I'll really be starting fresh.

Defrag? Waaay past the time when a defrag can be efficacious. I think I'm going to be depending more on outboard storage and less on *only* using the internal hd.

Or, maybe I'll just get so fed up with this process and get a new laptop. That's an (remote) option.

Ok, back to the grindstone.

[Beestie]

Good luck with all that.

[tw]

Originally Posted by BigV

I'm only back online long enough to confirm that the first 20 dingdang restarts and failed reinstalls aren't because I hosed something. Just getting the stupid thing to boot from cd is a challenge.

I suspect that laptop has been nuked before this warning will be read. Responsible manufacturers provide comprehensive diagnostics with their machines. Diagnostics (BTW) that should be loaded on the hard drive BEFORE formatting and installing the XP partition.

Unfortunately, the shotgun method is being used. Everything performed only to fix the problem. Very wrong procedure. First collect facts. First establish WHY the problems exist. Do not change anything until first the facts have been collected.

So many sources. For example, what did those comprehensive diagnostics report? What does the system (event) log report - valuable information destroyed by the 'nuke and pave' option. What is ongoing in Task Manager?

A system that takes minutes to boot will either be contaminated by promiscuous users (ie worms) or have hardware problems. What makes the problem even harder to solve? The human who tries to fix a hardware problem by reloading Windows

Shotgunning is why some problems are not solved. Its a well proven concept that applies to everything - not just computers. Break the problem down into parts. Then solve those parts one step at a time. In this case, we can draw a big line right down between hardware and software. One does not fix software when hardware integrity is unknown. At least does not IF one wants to fix it the first time.

Dell provided comprehensive diagnostics AND a program to load those comprehensive diagnostics on a new drive. The event logs could have long ago been warning of the failing hard drive - providing one with plenty of opportunity to transfer to a new drive without data failure and without wild speculation. Just another example of why problems are solved first by breaking down a problem into parts.

[Rock Steady]

Well, we'll just have to have a gentlemen's disagreement here. Typically, the Windows Registry gets so screwed up it's hard to repair and optimize. Then you have all the malicious stuff on top of that. I don't bother finding problems. I create a pristine environment from first principles.

I always build my own desktop computers so I don't even have any Dell tools anyway.

Sometimes I can avoid the clean rebuild by using System Restore, which actually works on XP.

[tw]

Originally Posted by Rock Steady

Well, we'll just have to have a gentlemen's disagreement here. Typically, the Windows Registry gets so screwed up it's hard to repair and optimize. Then you have all the malicious stuff on top of that. I don't bother finding problems.

Recently found the defective part, and replaced that IC in a Gateway power supply. Why? Clearly not to save money. Clearly not faster. But I learned. I learned why power supplies fail - and even that Gateway power supplies are missing essential functions. A problem also found in so many clone computers. Those who are considering a new machine - Gateway power supply was found missing essential functions.

Been using Windows NT on systems even before Windows 98 was marketed. Never once saw problems traceable to registry - except viruses. And yes, I have even cleaned a virus by deleting and modifying files - to learn the hows and whys. A rare case where the registry was part of a problem because the registry kept reloading the virus. More often, the registry is too often blamed for problems elsewhere.

If a system contains no value, then a 'nuke and pave' may work. But defined is a problem more often directly traceable to hardware - which means the 'nuke and pave' would only make things more difficult.

Meanwhile, what does a clone computer builder not get? Comprehensive diagnostics. A first thing executed when hardware integrity is suspect. Therefore clone builders rarely recommend diagnostics since such was never available. It explains why shotgunning is so often recommended - a technique that even auto mechanics no longer call acceptable.

Always start by not trying to fix the problem. Fixing comes later. First collect facts. NT based OSes (unlike the unreliable Win9x systems) make that easy. If problem was a failing hard drive, then that too would be obvious in system (event) logs. (his symptoms are consistent with a failing hard drive but facts are not yet provided.) Then step two - the repair - is done right the first time.

[Rock Steady]

No, the Win Registry is a well known single point of failure in desktop winboxes.

Registry defragmentation performs physical defragmentation of the Windows registry file. After defragmentation your registry will acquire linear structure which will reduce application response time and registry access time. Do not worry if sometimes after defragmentation the registry will become smaller only by 1-5%, the key is not in size but in its linear structure, hence it determines access time.

http://www.majorgeeks.com/Registry_D...ion_d4004.html

I built around 20 PCs and had about three hardware problems in five years. I have had a lot of success renewing systems. I have to debug my complex code at work, I don't need to be screwing around with OS problems.

[busterb]

TW I'm with you all the way about MBAs, But I don't agree with this statement. Do you by chance work for Dell? "Meanwhile, what does a clone computer builder not get? Comprehensive diagnostics." Are there not 3rd party programs to do this? To get this Comprehensive diagnostics. I have to talk to someone from who knows where.
A hypo problem. I use Yankeeclipper for copying & paste. What's wrong is some times it doesn't copy? If it screws the pouch I just exit and restart, but sometimes that does work, I reboot. I know that I'm not the sharpest knife in the drawer, But I've been playing around a little while and have built a few. bb

[dar512]

In general, tw, I agree. When debugging, I always look for the root cause of the bug and not the symptoms. But in this case, I gotta go with Rock Steady and buster. Sometimes it's just too much trouble to pick the fly shit out of the pepper.

[Undertoad]

In other words, Dell or Gateway will sell you a crappy machine, with a few bundled debugging tools to help you figure out what is so crappy about it. Did the tool help find the bad power supply tw?

[tw]

Originally Posted by dar512

But in this case, I gotta go with Rock Steady and buster. Sometimes it's just too much trouble to pick the fly shit out of the pepper.

5 minutes to collect information is too long? Rock Steady will spend hours rebuilding a system without any reasons for that work to be successful. Without even knowing why he has a failure. He is assuming a software problem when the problem is probably a hardware one.

For example, let's say the failure is a bad disk. He has spent how long reloading that bad disk only to have it still fail. Meanwhile, in five minutes, the system (event) log from the original system could have reported the problem. Information he destroyed because he did what so many do - shotgun.

The Dell diagnostic (available to everyone without wasteful human intervention on a phone, also located on Dell disks, and also provided with my system on a CD) may have also reported the problem. Both information sources in five minutes would save Rock Steady hours of wasted effort.

Responsible manufacturers provide comprehensive diagnostics. Such diagnostics are not provided by Gateway and not provided by clone computer manufacturers. Yes one can download numerous, individual diagnostics from component manufacturers. That takes longer.

Meanwhile, the 'nuke and pave' option takes longer, destroyed user data on the drive only on speculation, destroyed any hope of learning from XPs diagnostic information, and then still may not solve the problem.

[tw]

Originally Posted by Undertoad

In other words, Dell or Gateway will sell you a crappy machine, with a few bundled debugging tools to help you figure out what is so crappy about it. Did the tool help find the bad power supply tw?

The highest failure rate computers were, far and away, clones. I basically undermined an alternative source of income by getting these victims to replace those disasters with brand name systems. In at least one case, the adult was so stubborn that the kids finally had to demand change. The adult considered those failures as normal.

Meanwhile, you tell me. The computer does not power. So why would I execute diagnostics? Obviously, one might waste time shotgunning the power switch, power supply controller, and power supply. Or again first collects facts. In but under two minutes, the power supply was identified as reason for that problem by first using a standard diagnostic tool. It took longer to remove the problem.

[Undertoad]

And what was that standard diagnostic tool?