04-13-2018 09:08 PM
Ran into this today while checking some tracking at USPS.com
For most of you it will be a non-issue, but I'm going to have some big troubles I think. In their blind quest for "Internets Security" they are no longer going to support "outdated" browsers.
This isn't about security for the website infrastructure - it's about security on the client side and transport, and it should be my choice to take my chances, but they had to go whole nanny-state.
Windows XP users can't use any of the Firefox, Chrome, or IE versions, and neither of the listed Safari versions run on Windows at all. Opera browser isn't even an option on the list. Probable that Opera Mini on my Windows 10 phone won't work, and unknown if Edge on the phone will.
From what I see there the rest of you are going to be forced to use latest and greatest versions.
It looks like they might be going to a TLS 1.2 minimum protocol based on the browsers listed, but many older versions of those browsers also support 1.2 (but maybe not all the cipher suites), so we shall see how limited it is in a few weeks.
Also remains to be seen what "some features" means. It would be absolute idiocy to bar all least common denominator browsers from checking tracking numbers, looking up simple info, getting a rate quote, or downloading notice123.pdf and other USPS document, but makes a bit more sense to apply more stringent security when using a credit card to purchase stamps or shipping labels (but should once again be my choice - it's my credit card that might get compromised by pretty much nation state level hackers sniffing and brute force decrypting my SSL traffic)
https://www.usps.com/browser-check/
05-06-2018 12:07 AM
Well, this is going to be a problem for me and FireFox. On my computer, every Firefox update changes my tabs to black. Very unreadable and ugly. I don't know if it is just me. For that reason I refuse to update by Firefox browser. Maybe I am missing something? Is there an option that can prevent Firefox updates from changing my light theme to a dark one?
05-06-2018 02:01 AM
05-06-2018 02:49 AM
@berserkerplanet wrote:
Amazing that that incorrect power supply (1A) was able to supply double? triple rated current for that long without melting or letting all the magic smoke out
Interjecting here that a power supply does not "put (force) out" the rated amps.
Rather, it supplies whatever current is drawn, up to the rated amps --
and then the supply voltage drops.
Is that magic smoke injected and sealed into every device at the factory --
and if ever the magic smoke escapes, the device stops working?
05-06-2018 03:31 AM - edited 05-06-2018 03:33 AM
►Interjecting here that a power supply does not "put (force) out" the rated amps.
Agreed, But rated amperage/wattage is not max amperage/wattage, consumer power supplies usually don't have terribly good voltage/current regulation, and thermal response of supply components can cause supply to put out more than "rated" current and wattage (at least for a time until something melts, transformer windings short, or electrolytic cap juice leaks out.)
Yes, my observation has been that any time the magic smoke escapes from a device, it quits working. Especially if the smoke escape is accompanied by a pop! There must be a lot of magic smoke in new caps, because I've been finding that when I install new ones in some dead devices they magically start working again.
And I'm still surprised that that 1A supply that was run at or beyond rated output for 3 years hasn't cooked its caps, drifted seriously out of tolerance, and still works to power the WAP transmitter. (assuming that some of it's magic smoke escaped which is what caused it to no longer meet the power/voltage threshold required by the more power hungry Gateway device)
05-06-2018 07:30 AM
@berserkerplanet wrote:And I'm still surprised that that 1A supply that was run at or beyond rated output for 3 years hasn't cooked its caps, drifted seriously out of tolerance, and still works to power the WAP transmitter. (assuming that some of it's magic smoke escaped which is what caused it to no longer meet the power/voltage threshold required by the more power hungry Gateway device)
I know, right? One detail I left out was that when first tested (right after I pulled its plug out of the back of the Arris and before trying anything else), its voltage as measured on an analog sweep-needle meter was not only low but also actually pulsing slightly, in a regular rhythm between 5-6 VAC. (At no time was any magic smoke seen escaping, no funny smell, etc. Just way too much heat detectable through the plastic case.)
I thought, "Okay, that's not right," and started looking for others with the same specs (which I thought at the time was 1.0 amp) and a matching female plug. It wasn't until I got to the WAP power supply and noticed its 3.0 rating and its brand name sticker that I realized what had happened.
After the flaky supply had cooled for about 10 mins. or so, I tested it again and found that it was back to a solid 12 VAC, so I thought I had nothing to lose by trying it with the WAP transmitter that it had been intended for all along, and in fact it's been fine ever since. How about that.
05-06-2018 12:33 PM
Thanks for your suggestions, however this happens to me when I try to update from v57.0.4 to anything higher. In fact it happened on any version after v50, but I have been able to work around it. After v57 caused issues again, I just stopped updating Firefox. It can't be just me because when I Googled it, I noticed many people were having the same issues. So far no one seems to have come up with a solution.
05-07-2018 08:01 AM
There's no conspiracy to kick XP off the internet. Here's what usually drives these changes.
1. The company outsources to, or imports from, India. The knowlege and experience levels are extremely low. Whatever they "borrow" from some random GitHub this morning is what you get. If it doesn't work, you'll be the first to know because they won't even test before rolling out to production. But, hey, cheap, so there's always money (and time) to do the work over and over and over...
2. There's no real work for the US staff to do, so they're either laid off or drift into management positions where they don't actually do anything. Management then makes decisions based on numbers, like how Windows 2016 is 4 better than Windows 2012. Nothing else matters, not the dropped support, not the additional costs, not the totally randomized-in-unproductive-ways UI, Windows 2012 is 4 worse than 2016 and therefore must be eliminated before 2020 comes along and is twice as better.
If I thought more of you had programming skills, I'd suggest finding the TLS 1.2 code in Firefox 1634934XP++ or whatever version they're up to this morning and backporting it to the appropriate place in your favorite older version. All Mozilla's done for the last hundred versions is rearrange the UI (first to confuse, then to copy Chrome) so anybody familiar with the Netscape code can probably find it in short order.
06-14-2018 01:21 PM