Test Equipment Anonymous (TEA) group therapy thread

[med6753]

I do have one PC with a spinny drive and a CD in it other than that the others have SSD's and somewhere I have a plug in CD reader but where  Apart from some ancient backups and photos I should add to my NAS at some time CD's time is about passed.

I thought everyone still had CD/DVD drives in their computers?? Mine is in the bay below the 5.25in floppy drive... Oops.

McBryce.

Seriously, all of my computers do indeed have a CD/DVD drive. It still comes in handy every once in a while. 

[beanflying]

My largish Coolermaster box didn't come with any external drive bay slots or options to make one. I think it would have been over 6 months since I used the one on my Old I3/HP PC on the Laser to copy a CD for a pendrive. Good reminder to get all the older stuff onto the NAS as a third backup point instead of CD's

[McBryce]

I do have one PC with a spinny drive and a CD in it other than that the others have SSD's and somewhere I have a plug in CD reader but where  Apart from some ancient backups and photos I should add to my NAS at some time CD's time is about passed.

I thought everyone still had CD/DVD drives in their computers?? Mine is in the bay below the 5.25in floppy drive... Oops.

McBryce.

Seriously, all of my computers do indeed have a CD/DVD drive. It still comes in handy every once in a while. 

I was serious about the 5.25 drive too! As I do a lot of classic computer repairs (Commodore 64, Atari XL, Amstrad CPC etc) I still need 3.5in and 5.25in drives on occasions. I have a separate WinXP computer for the "retro work" which still supports direct read/write to these devices. My main computer with Win10 couldn't access them properly. The XP also has the option of connecting a rather rare 3in floppy drive too if needed.

McBryce.

[bd139]

From the  WERE THEY THINKING?!? Files...

https://www.zdnet.com/article/red-hat-resets-centos-linux-and-users-are-angry/

mnem
Jeebus... even OS is going monopolistic.

I find it completely unsurprising. Red Hat has become the favoured corporate flavour of Linux because the PHBs think "You can get support" and thus shell out for Red Hat licences. However, anyone who has actually tried to get the support from Red Hat that they're paying for will find that it's useless, and furthermore that Red Hat is pathetically weak at keeping up to date with things that need to be patched. With the weight of a huge Red Hat customer behind me I have previously failed to get them to do anything useful for me, and worse still they took a long time to do nothing. However, most of the application software, middleware and the like that corporate types pay for is built to run on Red Hat, so corporate types still need Red Hat to run stuff on, even after they've discovered that Red Hat's support isn't worth paying for. Oh, but wait, there's Centos - that is release identical to Red Hat but for the branding, and it's free.

So, I suspect that a lot of paid Red Hat licences have been migrating to free Centos licences. Cue Red Hat discovering that their business model is crumbling under them. Instead of fixing the problem, making the support worth paying for, they are trying to change the playing field so that people who need to run software that is only officially supported on a Red Hat release have to use Red Hat rather than Centos. Reason being that corporate change control teams will have a blue fit if you try to run a configuration that updates on a rolling basis, and the security team will have a blue fit that you want to punch a hole in their firewall to let it install patches that they haven't approved (Both reasons that a lot of corporate systems run software that can be years behind current patch levels).

Oh don’t get me started on this. I just inherited a three figure sum of mixed CentOS 7 and 8 nodes haphazardly managed with spacewalk and ansible on top of VMware. At least it’s not on HPE crap this time. One reason I haven’t been posting much recently is working out how to approach the problem. Literally as we landed on a half decent solution they pulled this shit so it’s back to the drawing board.

And yes this is going in Amazon.

Edit: on a positive note there’s only one windows node in the entire cluster which is used to wrap some piece of shit COM in a service.

Didn't IBM buy Red Hat last year? Yep, they did. And didn't IBM jump on the "Open Source" bandwagon years ago when it comes to selling their servers? Yep, they did. So while I don't completely understand what is going on now it seems to this peon that they are now talking out of both sides of their mouth. The hypocrisy of it all and not surprising from my former employer. 

Yeah they did indeed. They're talking out of their asshole not their mouth

It's a bad move. Firstly RedHat wrestled control over the entire Linux ecosystem through freedesktop/systemd/gnome/wayland etc, bought up all the developers and contributors, now they force a paid model on to people by turning the only stable and consistent enterprisey linux into a testing branch for their commercial product. This has all been at the cost of somewhat resource constrained Linux distributions such as Debian who have found it hard to keep up with the rate of change nor do they want to because some of the design decisions are pretty awful.

Our options are:

1. Move to Debian. Likely the short term outcome so far.
2. Move to FreeBSD. Would like to but Java on FreeBSD is problematic and there is no reliable valgrind port.
3. Move to Amazon Linux. Possibility but this is based on CentOS so YMMV there and also has more vendor lock in!
4. Move to OpenSUSE. Buggy.
5. Stick it all in EKS/kubernetes where we don't have to give a shit about the underlying platform any more. Long term outcome.

Note the lack of Ubuntu - it's a shit show.

Someone did this: https://centos.rip

[med6753]

Yeah they did indeed. They're talking out of their asshole not their mouth

As long as they send me that pension check every month they can be flaming assholes for all I care. 

[bd139]

Hey at least our RHEL license money is going into a good cause: old scope retirement home 

[med6753]

Hey at least our RHEL license money is going into a good cause: old scope retirement home 

Ha ha. Speaking of which....the Type 503 is at the UPS Poughkeepsie distribution center as I speak. If it gets to the Kingston office this morning, and it should, I will get it today but delivery date still sez Monday. We'll see.

[bd139]

Fingers crossed for you 

[Cerebus]

Oh don’t get me started on this. I just inherited a three figure sum of mixed CentOS 7 and 8 nodes haphazardly managed with spacewalk and ansible on top of VMware. At least it’s not on HPE crap this time. One reason I haven’t been posting much recently is working out how to approach the problem. Literally as we landed on a half decent solution they pulled this shit so it’s back to the drawing board.

And yes this is going in Amazon.

Edit: on a positive note there’s only one windows node in the entire cluster which is used to wrap some piece of shit COM in a service.

He, he. Count yourself lucky. I've had to try and deal with network management based off of Ansible Tower. Why Tower? Politics - because it's got paid support. Ha! After that, using Ansible to actually manage servers would seem like bliss. (Ansible's a bit clunky, but the core philosophy that, unlike every other equivalent remote management product,  it doesn't need any specialised Ansible components available or installed on the managed machines is, IMHO, the right one. Tower however, is a huge crock of shit.)

Didn't spacewalk get canned this year? [Checks] Yup, on May 31st.

[bd139]

They bought Tower here apparently but never got around to using it  . What I do have is stuff which was written for some ancient version of Ansible so depending on which playbook you run it shits out a hundred warnings first and some of the roles just don't work because they were cut and paste from some drunkard on github about 5 years ago and never run

Tower has to be better than Harness + Terraform + EKS 

[mansaxel]

Our options are:

1. Move to Debian. Likely the short term outcome so far.
2. Move to FreeBSD. Would like to but Java on FreeBSD is problematic and there is no reliable valgrind port.
3. Move to Amazon Linux. Possibility but this is based on CentOS so YMMV there and also has more vendor lock in!
4. Move to OpenSUSE. Buggy.
5. Stick it all in EKS/kubernetes where we don't have to give a shit about the underlying platform any more. Long term outcome.

Note the lack of Ubuntu - it's a shit show.

Someone did this: https://centos.rip

(Hilarious!)

I'm going to do a deploy of a bunch of bare metal machines at work, for stuff that is the stuff that you need to get the stuff that lies under the stuff where the developers can do things with Frameworks and Concepts and other neologism shit, to actually run. (Which is why I'm not even considering fads like containers (run someone elses badly cobbled "it compiles, ship it" bug-infested insecure shit, opaquely) or similar systems.

They're Centos today. A decision I grudgingly made based on the fact that I wanted someone else to be able to look after them.

I'm actively considering a BSD now, either OpenBSD (fastest route; all I need is in the base OS, 0 packages required but no GSSAPI-ssh) or FreeBSD, which requires like 5 ports built but has native GSSAPI support.

Yes, DeadRat are the scum of the Linux continent. The good thing is that there's no sign of systemd on BSD, so time to emigrate overseas.

[bd139]

In agreement with containers. I rather like the original containers that we came up with i.e. processes. Throw a static linked binary in a directory from some semi managed language (Go), add ulimit, pledge and proper permissions and you're done. chroot if you're feeling fancy. And no fucking overlayfs bollocks.

Talking of which I'm actually planning to spend some time this weekend fixing some of the go runtime issues on FreeBSD as no one else appears to have taken much of an interest in it. There's a nasty one that is breaking reading directory entry syscall results and truncating things which is blocking the whole Go test suite.

[med6753]

Listening to you software engineers makes me want to go shock myself on anything analog. 

[bd139]

It does that to us too 

[med6753]

Oh good...the Type 503 is indeed out for delivery. 

[mansaxel]

Listening to you software engineers makes me want to go shock myself on anything analog. 

I just bodged together a higher voltage AC source, from a function generator, and a 220V:9V mains transformer backwards; then spent some time looking at wobbly pointers on my newer VTVM (the UK MoD Marconi one that's only 52yo) in the higher ranges. It comes in to about 2% accuracy worst case, entirely uncalibrated, at 1KHz. This as compared to my Gossen MetraHit 25S, that actually has a calibration from this century. I didn't shock myself, mostly because I was careful around the transformer secondary.

[BU508A]

Interesting milivoltmeter from Grundig:



Never saw before a Meter with an integrated scope like this for the Lissajous.

Available here (NAWTS): https://www.ebay.com/itm/293889182317

[AVGresponding]

Fook, that's expensive. Nice though.

[med6753]

Interesting milivoltmeter from Grundig:



Never saw before a Meter with an integrated scope like this for the Lissajous.

Available here (NAWTS): https://www.ebay.com/itm/293889182317

That is cool. Never saw that combination either. But not sure if it justifies that price. 

[mnementh]

Someone did this: https://centos.rip

Hey at least our RHEL license money is going into a good cause: old scope retirement home 

Your linky made me *snerk* coffee out my nose... 

I really can't think of a better use for that money...

mnem
It doesn't matter what socio-economic system you subscribe to: The corporate business model eats its customers, and when it runs out of customers it eats itself.

[mnementh]

I do have one PC with a spinny drive and a CD in it other than that the others have SSD's and somewhere I have a plug in CD reader but where  Apart from some ancient backups and photos I should add to my NAS at some time CD's time is about passed.

I thought everyone still had CD/DVD drives in their computers?? Mine is in the bay below the 5.25in floppy drive... Oops.

McBryce.

Seriously, all of my computers do indeed have a CD/DVD drive. It still comes in handy every once in a while. 

No optical in my last two three builds, because the thing got used so little the loading mechanism belt turned to marshmallow. 

Only laptop I've had with optical in the last 5 years (not counting the old CF-29 Toughbook I've kept over a decade for nuking CISCOs and Automotive diag) has been my wife's, which is a .edu fleet model Lenovo.

As often as it is needed, I've had for over a decade one old HP sub-note accessory USB DVD-ROM with ATAPI interface which is BIOS-compatible back to Windows 95-era machines. I have a similar 3.5" floppy drive from the same era, and I keep both of them in a box with install media and blank discs for DOS3.3/Win 3.1, Win98 and Server2K. I think that's enough legacy media for any sane person; not that I necessarily qualify as such.

I think the last time I opened the box for anything was 3 years ago, and that was just to burn a PW-recovery disc.

Fuck optical drives... they've been a lame cupholder joke for nearly a decade.   

mnem
*still using 4GB Sandisk Memory Stick Pro and SD+ USB cards scavenged from Suckage City in like 2009*

[med6753]

The Type 503 arrived, in one piece. It was not packed all that well IMHO. Only 1 inch foam rubber on all sides. But no damage.

And it powers up. This scope is absolutely cherry. Just a little dust. The covers have some scuffs but that's it. I'll just clean them. No re-paint required. Even the handle is in good shape.




Applying the calibrator there's 2 issues. Severe roll off but worse yet it won't trigger. No biggie. We'll fix.



View inside. Very clean. Just a little dust.



And here's why it won't trigger. That white wire hanging in the breeze. It comes from the Trigger Source switch. Need to figure out where it goes.



View of the other side. Again, just minor dust.



And any old timer like me recognizes a beam power tube. It's a 6DQ6A. Used in the Horizontal Output of analog TV's. But why does this 503 have one? Weird. I'll investigate and report back.

[Cerebus]

Gosh, that does look clean (especially for its age).

[med6753]

So why does the Type 503 use a 6DQ6A beam power tube? I have the answer. It utilizes a vacuum tube SMS supply.  Those clever Tek engineers. See the schematics. Transformer T601 supplies the tube heater voltages plus the +500V plate voltage to the 6DQ6A. That tube is fed from the grid with an approx 30kHz pulse from the oscillator. That amplified 30kHz pulse is fed to transformer T620 which in turn supplies all the PSU voltages plus the HV. The only disadvantage I see to this is that the 6DQ6A is passing a lot of current and is going to run hot. And this scope has no fan. We'll have to see how warm it gets.

And also, check out C682 and C684. Paralleled capacitors. As I mentioned before I've see this often on Tek equipment. Why it's done I don't know. But I will replace them with a single capacitor of 22uF and 82uF.


 

[mnementh]

That's a delicious example of Tek hollow-state pulchritude. Excellent find. 

mnem
   I wanna lick it all over!