I've personally witnessed a number of T12 and T8 ballast failures.  The T8s were always the electronic.  I've turned on fixtures and POOF- the fuse inside blew.  Heard it.  I've also seen a few that failed when I had them on.  Have had a few Advance T8s and early generation Sylvania T8 ballasts that would smoke and get very hot when they failed. 

As for T-12 Magnetic.. oh man do those things go!  fire hazard.  Those tar burners would get very hot and a few would flash arc when they failed, the others just got super hot and dripped tar out the ends and when I would go to change them, there was heat damage to the fixture from how how they got.  And they smelled HORRIBLE. 

I just disposed of about 400 3000K fluorescent lamps, both T12 and T8.  Offered them for free on Craigslist and Facebook Marketplace, no takers.  Got a few people messaged me if I had any 3500 or 4100 though.  I am not getting rid of those. 

Adding on to me finding my old computer stuff, I found my two old 160gb hard drives.
Even though one of them still works fine, the other had a broken sata data connector (it was sitting in improper storage long before I got it), I had to balance the cable on what's left of the pins to get it to work.
Unfortunately, it's now making a repeated clicking and screeching sound (only started doing it recently), probably its last few hours of life (it was my main drive) and the other was wiped.


It booted up instantly when I found it (this morning), but now it takes ages to boot up.

Could you upload a picture?

I hope you would enjoy the car show there

I went to the Category:May 2016 Coffee and Chrome in Hapeville, Georgia, USA back then.

Cost being the most decisive factor, I guess. Now, cost aside, what would prevent say a LED edge lit panel to have a rectifier to run it effectively on DC ?

They do effectively run on DC, or on high frequency PWM when dimmable. At the very least the bargain basement LED panels used at my school all run from switching power supplies.

The thing is, if you are a paying customer for the weather service, you will get more reliable access than someone trying to look at it for free, mainly when the server load approaches its limits...

The problem also is, the only way to eliminate flicker is by using rather large (for the power) electrolytic capacitors and these are rather problematic components mainly from reliability perspective. Not that it wouldn't be possible to make them reliable enough, but it costs money and mainly building space within the ballast cavity (which uses to be extremely small with modern LED products), so much it forces the other components to operate in not that favorable conditions, increasing the cost there (to compensate with "better" components) as well.
Skipping this capacitor allows you to achieve significantly better reliability for the same budget, power and efficacy.

Why the cheepeese are skipping it is clear - as it allows to cut cost significantly even without giving up the reliability of the rest, along with compromising on the quality of the rest it really saves a lot.
But with "quality brands" I see the market rather went into eliminating flicker with mediocre reliability, I'm missing the uncompromised reliability with tolerating some flicker - as to me many applications don't benefit that much from true flicker free operation, but do benefit from the reliability.

"getter" is a chemical intended to absorb atoms and molecules contaminating fill of sealed structures.
On incandescents a red phosphorus was used to absorb any remaining traces of moisture from the lamp internal atmosphere.

A barium is used to absorb mainly oxygen mainly in vacuum devices - in vacuum tubes it use to have the form of a coating on the glass surface, applied by evaporating the pellet containing Ba after the tube was evacuated (either heating the cup with Ba pellet with laser, or earlier filling thin tube by the BA, welding this tube into a loop and then firing it by an induction heater so thethin tube bursts releasing the Ba) and letting it condense on the tube glass wall. During this it reacts with most of the remaining air in the tube, but then the coating on the glass was absorbing the gasses that got released from the internal materials over the lifetime, so helps to maintain the good vacuum there. It became the characteristic mirror-like finish on the top of the tubes. You may find the same form also in some Na lamp outer bulbs - usually the mirory like patches around the bulb base.

On discharge outer bulbs the vacuum does not have to be that good, but the "mirror like" form was unwanted (as it obstructs light, or even reacts with the IR reflector coating on SOX), it is just put into a cup welded on some stem wire and let it act from there.

Eric, even more interesting when such anomalies as yours begin to turn up!  Its seems there are no hard rules when it comes to documenting lamp designs over time.  Engineers are always playing with ideas for improvements (which sometimes accidentally turn out to be a backward step during the later phases of lifetest, like when they prematurely introduced full-coated bulbs).  So it is not unexpected that they may have made some runs according to the earlier fully clear-dome design to check whether that made any real improvement, and we end up with lamps like yours whose date seems not correct at first sight.