Replies

In 1975 I was a university undergrad, learning BASIC (or FORTRAN – my memory that far back is a bit fuzzy)  programming on a PDP 11/75 mini computer.  I was actually an Art major at that time – but became fascinated with computers.

Around 1979 I got an Apple II+ and my course was set for an inevitable career path in computing.  A few years later, when I got my first job offer for a programmer position, I also had a hard won offer (which I had been waiting 3 years for) in hand for an electrician apprenticeship.  It was tough choosing between the two – but the chance for a computing career won out.  No complaints on that decision!

In one of the final servicing stack updates, MS included updated code that would trigger if Windows Update detected a processor that was not considered a supported CPU. The result would be a nagging warning popups and no more updates. No, I don’t know which servicing update that was. I had been running Win 7 on a 5900X system for a while before that without issue. I do know that when I saw the unsupported CPU popups, I just rolled it back to a previous system backup and called it a day. However, I could not install any newer updates successfully after that. Wasn’t a problem for me as I had Win 7 off the internet anyway.

As for your question about rendering times:

For me, the >=3% faster rendering speed is not considered insignificant. My 3D scenes can take a few minutes per frame. The app displays the render times in both total minutes and total seconds. So it’s easy to measure and note differences.

Modest to large 3D animation projects can be rendered a number of ways: frames distributed to multiple computers (locally or remote), frames distributed to multiple GPU cards on the local machine (or multiple networked machines) or via a cloud service (that uses what I described previously). In any case, for a large animation project with high computational demands, a 3% improvement can be the difference between hitting the deadline, making a profit – or both.

Thanks for the questions, I enjoyed answering them.

I ‘dunno, I remember the things MS did to kneecap Win 7 during its final months of life. Regardless, I managed to get Win 7 installed on a system with a newer (at the time) Ryzen 5900X cpu. It ran fine & was stable. It actually rendered 3D images at least 3% faster than on Win 10 Pro (installed on the same system). Kind of makes me wonder if THAT was one of the real reasons MS wanted to kill Win 7.

Will they kneecap Win 10?? Well, I’m not rolling the dice…

During 2025, I’m going to skip Win 10 updates unless it’s something super critical AND I’m absolutely sure MS is not trying to ‘bork my system.

I generally keep my Win 10 system off the internet anyway & use Linux as my primary platform for doing business. I “curate” any Win apps that I plan to install: googling for malware hits for the software, running the files through a few virus scanners plus something like Virustotal.

My advice to those still on Win 10: ALWAYS do a full backup of the system before you do any degree of 10 updates. I stuck with this habit during the final months of Win 7 & it saved my bacon a few times.

We have had an Epson ET-2750 for years. We live in a low humidity region, so we have to do a print head cleaning cycle more frequently than folks in areas with higher humidity. Like the printer you mentioned, ours has a Maintenance Box that’s easy to get to.

Recently, we got an alert on the printer that indicated the Waste Ink Tank (aka: Maintenance Box) was nearing being full. I looked-up info on getting a replacement Waste Ink Tank, how the tank works & how to replace it…

First, the tank is basically full of absorptive sponges. There is no level sensor for the sponges – that would surely be too expensive for Epson to engineer. So, in reality, the Epson firmware just puts a limit on the total number of cleanings it performs. When you approach the maximum allowed number of cleanings you get warning notices on the printer’s display.

From what I learned, when you hit the max allowed number of total cleaning cycles, the printer effectively bricks itself (it won’t do anymore cleanings & therefore it won’t print). You have three choices at this point: 1) pay the $$$ & send the printer to Epson to get serviced & re-activated, 2) attempt do-it-yourself or 3) buy a new printer.

I learned that if I chose option 2 (DIY), just swapping in a new tank is NOT a complete fix. I still would need a “key” & app to set the firmware to re-activate the printer. I believe this amounts to just resetting the cleaning counter maintained by the firmware. Epson won’t supply the key or app, so I would have to go on a 3rd party website (you’ll have to google for this, I won’t post any examples), buy a key and use their tool to reset the counter. I wasn’t sure any of these websites were 100% trustworthy (let alone legit). I found myself quite disappointed with Epson.

So, I chose option 3: get a new printer. However, this time around I got a simple Brother B&W laser printer. A better fit for our drier climate. Oh, and the laser printer is NOT an Epson, it’s a Brother model.

Honestly, what Epson is doing here is such a “dark pattern”. They should just sell the replacement Waste Ink Tanks & let you reset the cleaning counter via the printer’s settings directly (no special top secret app required).

“But it can’t be all bad. After all, if we decided not to embrace change, we’d be calling ourselves the AskWoody XP Newsletter.”

Hah hah!  If XP 64 was still maintained and constantly updated to support new hardware (including new motherboards, CPUs, GPUs, storage, busses, etc), I’d surely be using it.  Heck, I’d gladly PAY a (reasonable) annual fee to use it.  Seriously.

But: that’s assuming it would not be the ugly bloated legalized malware/spyware platform Windows is today.  Pity.

2 users thanked author for this post.

Cybertooth, Mothy

I use a dedicated firewall/router appliance (opensource IPFire installed on a low power PC) to protect my home net.  IPFire is configured to also serve as the DHCP server for my home net.  Some aspects of my DHCP server configuration:

• I establish static (permanent) DHCP leases for all hosts (PCs, printers, servers, etc.).  The host’s ethernet (MAC) address is required to create the static address.
• I only allow hosts with a pre-configured DHCP static/permanent address to have access to the home net.

This allows me to always have all our home net hosts setup as DHCP clients – rather than as a non-DHCP static IP address.  This simplifies managing IP address assignment on my home net because – in effect – it’s all done via static/permanent DHCP leases on the firewall/router alone.  This also makes things really convenient for mobile devices as they can remain always configured for DHCP.

For family & friends that are visiting, I’ll add their host info to the DHCP server config – as static/permanent leases.  No big deal: it’s a one-time deal for each and IPFire’s UI makes it easy to do.

For anyone else: if they request internet access at my home I’ll tell them to use  HotSpot (or the equivalent) on their cell phone… or I’ll just roll my eyes and change the subject.  I never let strangers on my net.

 

I’ve administered Linux and Windows systems for many years.  For server administration, I often rely on the command line/shell to get things done.

But when I’m using my personal laptop/PC to do non-administrative stuff I rarely have to resort to using the command line on Linux.

Heck, now that I think of it, I think I resort to Powershell on Windows more than using command line on Linux because so much stuff is tucked away deep into convoluted GUIs in Windows 10 & 11 now.

1 user thanked author for this post.

rc primak

For the installing Linux on an Acer C710, this does a good job of outlining the steps:

https://www.linux.com/topic/desktop/how-install-linux-acer-c720-chromebook/

It’s focused on the C720 but the C720 & C710 are quite similar.

 

For installing Mint on the ZA3, the graphics drivers are apparently the biggest challenge.  This gives some key info:

https://community.linuxmint.com/hardware/view/3924

===

Personally, I wouldn’t waste time on old laptops/chromebooks that have less than 4gb of RAM and are incapable of upgrading memory & disk.

On the other hand, I’ve had great success taking old laptops, upgrading the RAM & storage (if needed) and then turning them into nice Linux laptops. But I’ll only do this with laptops that have a CPU with at least 4 physical cores (that can do at least 2.6GHZ) and can physically support up to at least 16GB of memory.

1 user thanked author for this post.

rc primak

When visiting some (not all) family members I sometimes am asked to look into issues with their computer or home network.

For the computers, I always start off doing what most here have already mentioned: check backups are being done, make sure AV & anti-malware is working & updating, check logs, looking for persistent errors, etc.  I added “BIOS check” to this checklist after visiting a relative a last year.

During that visit I discovered they had a fairly new but virtually unused high end custom workstation (from a local mom&pop computer shop) that they bought for their CAD work.  I asked about it and they said it would always crash after a few minutes of any kind of use, so they just didn’t use it.  The vendor had apparently been of no real help & the relative was just too embarrassed to pursue the issue any further.

I looked over the system & confirmed the issues  – but none of my typical checks uncovered an obvious cause.  I spent hours on this and then – as an afterthought – I checked the BIOS.

OMG!  The vendor had set one of those lame auto-overclock options… which I quickly confirmed had mis-configured memory and made the memory very unstable.  Disabling the auto-overclock & setting proper explicit parameters for CPU, power and memory resulted in a stable awesome system.

So now when I debug a relative’s (or friend’s) computer/laptop I always look over the BIOS settings first.

1 user thanked author for this post.

windbg

My ISP blocks some ports (closed, not stealth) related to SMB and NetBIOS.  So GRC’s portscans always show those particular ports as closed but not stealth – despite my home firewall configured to stealth those ports.

 

[ek]

Ascaris wrote:

What have you had to turn off? The stuff I have seen is opt-in when it comes to the OS itself.

After I do an Ubuntu install, there are a number of services I disable or uninstall.  One of these is the “whoopsie” service that “phones home” to share data on any app/service/system crashes.  The other “phone home” stuff I just opt out.

I actually uninstall the whoopsie service.  That’s because – years past – some old versions of Ubuntu would sometimes re-enable whoopsie due to a heavy-handed update.

The other thing I always disable/de-install on newer Ubuntu installs is the snap packages subsystem (ie:snapd).  This is due to how snap erodes Linux security and (to a lesser degree) reliability.

Due to similar concerns, Linux Mint has taken a stand against snapd & does not install it in its more recent releases.

It appears Canonical’s ‘devs are eventually going to make snap stuff absolutely essential for Ubuntu desktop releases.  This is why I’m starting to completely abandon Ubuntu and go with Mint for desktops.

 

• This reply was modified 4 years, 2 months ago by ek.
• This reply was modified 4 years, 2 months ago by ek.
• This reply was modified 4 years, 2 months ago by ek.

2 users thanked author for this post.

rc primak, Ascaris

Well, in fact, various Linux distros have had degrees of  “telemetry” for many many years.  Most of this was/is “phone home” stuff to give the distros some idea about the number & kind of installs.  Then there’s the “phone home” to report details of a software/system crash.

A decade ago, this kind of voluntary “phone home” stuff wasn’t that big of a deal.  But the world we live in now makes folks (understandably) more sensitive to anything that can be associated with the word “telemetry”.  The old IT security adage “trust no one” really does ring true in today’s world.

Unlike Windows, this stuff is pretty easy to turn off, disable or de-install.  I always turn off telemetry when I install Linux.

 

2 users thanked author for this post.

rc primak, Cybertooth

I purchased a new laptop 1.5 years ago from Dell.  It was a nice Precision model with Ubuntu (18.04) exclusively preinstalled.  It has worked great since I got it.

When ordering it online from Dell, I found that choosing Ubuntu instead of Windows for the laptop’s config saved me some bucks & made it a little easier to choose 16GB of RAM instead of 8GB.

Dell & Lenovo both offer desktops & laptops with Linux pre-installed.

Dell, Lenovo, HP and others, work with Ubuntu’s foundation/company – Canonical – to certify some of their desktops/laptops for use with Ubuntu releases.

https://certification.ubuntu.com/

I have indeed installed Ubuntu & Linux Mint on a number of desktops, servers and laptops with minimal-to-no issues.  The key to this is to – up front – do adequate research to find specific desktop/laptop products that are already well documented to work with Linux.

Sometimes I’ve been lucky and old (3+ years old) laptops I already had on hand worked fine with Ubuntu or Mint.  My experience has been that the older the equipment the more likely it will work with the latest Linux release.

All my Windows systems are in fact dual boot Windows / Linux.  I install Windows first & then Linux after.  Some motherboard UEFI implementations can make dual boot Linux installs a bit more complicated than necessary; but this is tempered by the various Linux distributions getting better at UEFI installs, doing a better job documenting how to do dual boot installs AND – in their forums – how to deal with particular manufacturer UEFI quirks.

Nowadays, when I plan new desktop/laptop purchases I select products I can confirm will work well with Linux as well as Windows.  I won’t consider a server/desktop/laptop that has firmware that would effectively physically lock me into an OS exclusively – Windows or otherwise.

3 users thanked author for this post.

OscarCP, rc primak, Ascaris

[ek]

In my book, the list of “improvements” is pathetic for a 5 year scope for a OS with a relatively rapid release cadence.

It just confirms my suspicions of the Win 10 effort: that the product/dev teams are probably spending most of their time mired in project churn & wheel spinning.  Much of this probably due to being stuck in a reactive-mode chaos caused by the relatively rapid release schedule.  They just can’t handle it effectively, so actual measurable productivity goes out the window.

The product (and consumers) would fare so much better if the release cadence was dialed way way back to give the product teams and devs room enough to do the good work they are capable of.

It’s clear to me that the current release cadence has not added any value to Win10 and it is surely costing MS a lot of $$$ (e.g.: wasted productivity, PR impacts) by ignoring the reality and sticking with it at all costs.

• This reply was modified 5 years, 3 months ago by ek.

7 users thanked author for this post.

Kranium, AlexEiffel, wavy, AngryJohnny75, Cybertooth, OscarCP

anonymous wrote:

No, he’s right. If you can’t point to any evidence it has occurred, you shouldn’t be reporting it as a risk. Being able to hack an IoT device and being able to use it to actually stalk you are different things.

Plus, unless you modified your ISP’s modem, you will be unreachable from the outside Internet without some way to get the TV to initiate the connection. If you haven’t put on any unregulated apps, it’s as safe or safer than your smartphone, which also has a camera and mic on it. But you don’t get warnings about those.

We are on a site where we are told not to worry about exploits until they actually exist in the wild. Why are so many people not applying that to TVs, and thinking a completely hypothetical hack is a valid thing to release an FBI Warning about?

Huh??

I’ve used multiple ISPs over the years for home connectivity.  Each with different modems.  In every case, the logs of my firewall/router show constant inbound probing/scanning from addresses all over the world – 7x24x365.  I haven’t ever observed the modems doing any significant form of packet filtering.

Some ISPs do block some protocols to/from home service.  Like Windows file & printer sharing, SMTP (server side) and sometimes uPNP.  But whatever blocking they do (if they do it) is quite limited.

For consumers, safe computing (“security”) is the result of the priorities they set & choices they make.  When I learn of a new vulnerability, I don’t wait until I’m a victim before I take steps to mitigate the risk(s) in a reasonable fashion.