Posts by Alan Brown

Not that it matters

Congress has subpoenaed the unredacted version.

It WILL come out eventually. The only question is how long it takes for it to come out.

Don't forget that for all the huff-n-puff, Mueller is a republican.

"It is when Spamhaus won't tell you why you've been blacklisted."

Claim lacks credibiilty.

Please cite your evidence and your IP.

Re: Self-appointed but

"They may be self-appointed"

but so is every organisation that runs a mail server and Spamhaus didn't sneak out one night and install use of their lists into those mailservers.

It's not overly surprising that the loudest whiners in the "Spamhauze blocked me" crowd have tie ins to the "alt right" and fake news generating crowd. The same mentalities tend to apply.

"The ISP and Spamhaus should be able to work together to resolve the issue."

In general ISPs as described respond to being listed by screaming their tits off about damage to their users, and their "RIGHT TO SEND EMAIL"(*) instead of actually removing the spammer.

In cases where ISP servers end up repeatedly listed due to spamming events, ISP customers should take a hint and vote with their wallets.

(*) There is no "right" to send email. It's a privilege, extended by receiving networks and it can be revoked at any time for any reason with no negotiation entered into whatsoever. As soon as some dolt starts ranting about his/her "rights" in this area, you know you have a fucktard on your hands and the best solution is to hang up on the conversation and walk away until they learn a few of life's lessons.

"The damage caused by each side to the other is grossly disproportionate."

'My network, my fucking rules' - aka behave yourself, or you can fuck right off.

If Packet.tel is claiming "damage" because they're being _boycotted_ (ie, outfits across the network are refusing to do business with them) then perhaps they should have thought their business model through before they started that behaviour.

- Your IP connection is supplied by your ISP to you under a contract.

- They buy from their supplier under another contract.

- Even if those contracts contain a clause that you have guaranteed access to MY network, it has zero legal standing if they don't have a contract with me where I agree to it.

- Companies which fail to comprehend this and have tried to litigate to force access to networks that they have no contractual agreements with (aka, forced acceptance without a contract) have found themselves in very _very_ hot water.

Re: For the love of..

"I use Spamhaus at the moment but it's catching virtually nothing these days. "

Way back when, that was why other blacklists started popping up.

When they got "too effective" is when they started getting heavily targetted by abuse-friendly networks for affecting their bottom lines. Burnout becomes a major issue when faced with those kinds of attacks and that has a lot to do with the high turnover of lists

If Spamhaus lists something(*) there's invariably a bloody good reason for it and when the whiners start actually giving out the affected IPs(**) you'll usually find 20 or 30 people popping up with their own evidence of abuse - or you'll find that the whiner in question is trying to send mail from dialup/DSL enduser IPs and being smacked down by someone who refuses to accept mail from such hosts. There's a lot of cases of throwing shit at a wall and hoping it'll stick.

(*) - Spamhaus entries are always listed along with the evidence trail.

- There are multiple lists, comprising various classes of problem.

- Spamhaus LISTS suspect IPs. It's up to admins to make the final decision. Private property rights and all that guff - attempting to force a network to accept connections from any other network is an extremely dangerous path to go down and can backfire spectacularly on the litigant.

(**) - Complainants seldom if ever disclose their IPs, or their background.

- Fatuous accusations of "Waaaaah!!! Spamhaus blocked me. My Frea Speach is being oppressed" without any actual backup are commonplace. (see point 2 above).

- In virtually all such cases it's because the whiner is attempting to send mail from a designated enduser IP (DSL or dialup) and the receiving side says NO. Most of the rest are customers of egrariously spam-friendly ISPs (such as certain Ukranian/Russian/Chinese networks).

Re: Banning IP addresses is probably not a long term solution.

"Criminals and pro-criminal hosting engage in legal and technical attacks against blacklist maintainers."

And this happens a lot more than you may think. Think "Userfriendly" and "Steph" - at least one set of legal (and media) actions has been launched by company MARKETING departments, rather than actually deal with their abuse-friendly policies.

Re: Why no ARM servers?

"For example, I've seen benchmarks where Intel's compiler produces significant improvements on x86 over gcc."

Almost all of this is down to the Intel compiler using undocumented cheats vs the GCC one playing strictly by the rules. Remember when MS used to pull that with undocumented APIs?

Re: Looks like we're still relying on Bruce

"if we want timely redirection of incoming rocks"

In general, we only ever see said rocks once they've gone past us on the way OUT from the sun and have a large spotlight shining on their matte, coal black rears which are in close proximity to us.

When the large spotlight is between us and said matte black objects, they're a little hard to see. Ditto trying to see them against a background which is predominantly black. Finding a silent black cat at the bottom of an unlit coal mine with no torch (or can of tuna), at midnight is probably easier.

Re: And thus the problem

"There is no downside to using dry copper pairs between the street and the premises when you're talking about distances of a couple hundred meters and frequencies of a few MHz"

Except we're long past "a few MHz" and a hell of a long way past "a couple hundred metres"

And that's only half the story anyway - since it was installed, my VDSL speed has dropped by 25% on a 150m connection simply as a result of interference from _other_ VDSL circuits.

This isn't a new phenomenon - Back in the early 1980s when I was working in Telco side stuff, we found that you could only put 8 2MB/s trunk circuits in a cable before they interfered with each other. The alternative was substantially reducing the distance between regenerators (as in less than half) - and this was a major driver of deploying fibre, as fibre links would go the entire distance without a repeater and without relying on the electricity supply to each of those repeaters going titsup.

Re: "If you are getting a gigabit"

"G.fast can provide a gigabit over phone wiring"

On a good day

Downhill

With a tail wind.

And a bulletnosed driver.

More to the point. G.fast is the embodiment of the rather pointed fact that the equipment at each end of such copper links ends up costing FAR more than that needed for fibre - AND the fibre in between.

The reason telcos like xDSL is simple:

If they lay new fibre it's an infrastructure project and they have to make the cost of the fibre back over 10-20 years.

If they whack in xDSL equipment they can charge for it up front.

Re: BT's new 'up to' Fibre Product - Two baked bean tins and a wet piece of string.

"Openreach are not going to be restricted"

Oh yes they are. They DO NOT own the infrastructure, ARE NOT a separate company and are beholden to decisions made by BT head office.

This ongoing and well -documented market abuse is exactly why New Zealand took the step of forcing the lines and dialtone sides to be completely split up when Telecom NZ attempted to sell the BT model as a way of staving off regulatory intervention.

The result was a change in the market there (in terms of opening up) which would make your head spin when compared to the market here - The day I see Openreach freely selling duct space and maintenance contracts to Virgin and _actively_ selling/supporting services to 3rd parties - as now happens in New Zealand - is when I'll believe that the UK has a truely open market.

Re: Utterly Shameful decision.

"The ASA is there to protect consumers "

WRONG!!!!

The ASA is there to protect marketers - from the government.

Re: Utterly Shameful decision.

"But if FTTC can deliver similar latency to FTTP (and yes, it certainly can) - "

Only if your FTTP has _really_ shitty latency and your upstream is _really_ badly overloaded

256QAM has 9 bits per symbol (8 if you use trellis coding, but 9 for encoding purposes) - that means your BAUD rate is 1/9 the BIT rate - and latency is determined by the baud rate.

xDSL is effectively a shitload of 2400 baud 64/128/256 QAM acoustic modems running side by side at different carrier frequencies spaced at 3.5kHz intervals ranging from just over 64kHz up to 33MHz. You simply aren't going to get better latency than that (in fact the best latency is about 140ms)

xDSL also have a _very_ limited reach and is subject to crosschannel interference which limits the number you can have on one set of cable pairs, whereas I can run 10GB/s 80km on a single fibre and if I use DWDM can easily put 1TB/s down the same fibre. The tranceivers to do that are only about £250 a pair and for typical GPON shortrange 1GB/s use are more like £3-6

Re: "I'm a judge, and I can't understand the difference, so nobody can"

"We 'ad teletypes wit' 110baud acoustic couplers... "

Sheer luxury. Some of us had to make do with 45.5 baud and Baudot code.

Re: "How many people are piping their gigabit FTTH into enterprise class routers"

"For the matter the EU regulation 015/2120 bars ISP forcing their own equipment onto customers"

But that doesn't stop them bloody well trying - I've found a surprising number of ISPs who _will not_ provide "router free" connections

"Judges in the UK are famously impartial."

The Judge even acknowledged the issue - the problem is that the ASA "has a procedure" and then followed it meticulously.

What this decision underscores is that the ASA IS NOT A FUCKING REGULATOR - and as such it is not bound by the kinds of rules that a regulator would be - they tend to make it up as they go along.

At heart is that the challenge wasn't about truth in advertising, it was about the technicalities of the way the ASA came to its decision about truth in advertising - and bear in mind that the ASA is a boys' club, not a regulator - it exists specifically in order to stave off having government intervention, so its rules are written to ensure that everything favours itself and its internal decisions.

A judicial review against following procedures and rules will fail - even if the procedure happens to be completely WRONG - the challenge would need to be against the rules or the procedures themselves.

The court system IS NOT and never has been about "justice", it's about "law"(*) - and as long as the ASA kept within the confines of their rules then they couldn't lose this case.

Of course a few legal rulings upholding obviously boneheaded ASA decisions might well result in it being replaced with a real regulator with proper regulatory oversight, but in the meantime, it acts as a highly effective way of "the industry" running interference between advertisers and Trading Standards.

In the meantime: The Register and others need to stop referring to the ASA as a "regulator" and give it its _real_ designation of "industry-self-appointed internal watch poodle"

(*) Remember all the things which have been "legal" in the past - Apartheid, Concentration camps, interment, Jim Crow laws, Enclosure laws, etc.

Re: 1/3:3/1

" ITU-designated band of frequencies from 0.3 to 3 terahertz (THz)"

Yes....... now check the wavelengths in cm or mm.

There's a reason the bands are in 3-30 ranges and it has something to do with the speed of light

Re: When Urban Myths Come True

"could have built your server room underneath a swimming pool,"

Or under a 500 gallon water tank in the ceiling above.... (BTDT, finally got a new room built, everything out - and then some cretin decided to put a pile of unauthorised computers "in all that spare space in your server room"(*) whilst the door was unlocked. We let him stay there and cackled....)

(*) A converted cloakroom

Re: When Urban Myths Come True

"a water pipe broke above our offices and hundreds of gallons of water per minute were released over the ceiling tiles." "Can't say that this was the result of human error though"

I'm still waiting for this to happen in one location. The human error was that although we specifically stated "no water piping is to run through this area _at all_, it was left off the spec by the architect without bothering to consult with us and when the builders needed to run pipework, guess what they did? (We noticed and complained, but the builders didn't give a shit because "you aren't the customer, mate, now fuck off our site!")

Re: When Urban Myths Come True

"one of our mainframes losing network connectivity at the same time every night,"

We were copping abuse from staff in one building for the networking being intermittently dead on friday afternoons.

It turned out the network cabinet had been rewired into the nonessential power circuit of the workshops in the building and the staff there would knock off early if all jobs were complete, turning off the power as they left.....

This (and a few other Furphy) are why I insist on paperwork for all proposed electrical jobs - making sure that what's going to happen is what was actually asked for and that what was done actually meets the spec. There are a shitload of cowboy sparkies around and no paperwork == "not me mate" or "I just did what i was told" (even when speciifically told to NOT do something)

Re: So...

"Fortunately, this "cleaner uplugged the life support" story is a myth."

"Cleaner unplugged the server" is not. I've had it happen on customer sites.

Re: So...

"And if the boss really is too mean for that, a filing cabinet in front of the socket outlet."

I usually add a "DO NOT UNPLUG" label on the plug and "DO NOT TURN OFF" on the socket - it does seem to work.

Unfortunately that's no protection against other outlets on the same circuit being overloaded.

If the boss _IS_ too mean to pay for things to be done properly, then document everything and when the shit hits the fan, make sure _HIS_ boss is made aware of the paper trail. You'd be surprised how often a "There's no money for this" becomes "How much do you need to fix this?"

Re: So...

"So that the fuse had a near 100 percent load."

Fuses don't blow at 100% load. They run indefinitely at 100% load.

They "blow" almost immediately at _200%_ load and will last about 8-12 hours at 150% load or several weeks at 120% load.

Re: Historical inaccuracy ...

" Except, the size of those floppies has surely been overstated?"

Nope.

2.44MB 5.25" floppy drives did exist - they were rare and eyewateringly expensive - mostly punted by IBM as an improvement on 1.2MB 5.25inch ones. I think I've handled two such disks in my entire life.

3.5 inch were available in 720kb, then 1.44MB, then (rarely) in 2.88Mb (there were also 120MB 3.5" floppies but that's a story for another day and a ATAPI interface - they were blisteringly fast to read and write 1.44MB floppies. I timed mine at about 7 seconds for an entire disk)

Re: back in the day..

"storeman who kept a window sash weight in stock"

And an extra long one made up out of bar stock for the inevitable twat who would send the apprentice back because it wasn't long enough... (a lot of storemen didn't like the game of "pick on the apprentice")

Re: Magnets!

"We have an industrial magnet currently in the office to "scramble" the backup tapes from 2005!"

Very few data tape formats from 2005 will even notice a "large industrial magnet" - which is causing me headaches disposing of a store of LTO2 tapes that a user has dredged up from that era just after I'd dumped the LTO2 drives.

Re: On a side note

"The read/write heads were not always calibrated identically on different machines."

That was a common problem - the fact that you can _buy_ audio calibration cassettes escaped most people's notice. The fact they're about $100 apiece and you need at least two - 1kHz and 10kHz - is offputting for many more. (you need to calibrate both cassette spindle/drive motor speed AND head alignment - with the speed being surprisingly out of whack on a lot of mechanisms)

For floppy drives it was a bit trickier but calibration disks did (and do) exist if you're willing to stump up for them AND can read an oscilloscope. The demise of stepper motors driving linear positioning strips and uncontrolled spindle motors(*) made absolute track calibration less of an issue anyway

(*) These had a rudimentary trimpot speed control in the end of the motor just like cassette mechanism but no feedback mechanism telling the drive what speed the disk was rotating at. That coupled with "rubber band drive" between the spindle motor and spindle (the motors didn't have enough torque for direct drive) meant "wow and flutter" caused by varying friction between the floppy disk and its case was a serious issue. The original head stepper motors had no positional feedback to locate the first track and would just seek N times until they "should be" at the end of travel. Newer drives use a worm gear to slide the head, have an optical sensor for the end of travel (not just the hole in the floppy) and can hunt more precisely to locate the index track in the case of misaligned tracks.

Re: More urban myths busted

" an over zealous cleaner used a scouring pad to try and 'clean' the screen protector."

That happened more times than I like to think about. Most times people had to live with the results

Re: Then there is the "send me a copy"

"When asked for a screenshot from site, we got one, all right."

In some cases (BIOS issues) you actually _need_ to do this.

In all other cases I pretend I can't open the attachment and ask them to d it as requested.

Re: Well if the US ships want the Chinese to keep out of the way

" But an electrical engineer who doesn't understand that colour CRTs have magnetic fields?"

The magnetic fiield around a CRT isn't enough to affect a floppy.

The magnetic field around the DEGAUSSING COIL is another matter.

How often was this monitor powered down/up?

Re: Well if the US ships want the Chinese to keep out of the way

"Even now you can still visit offices where there are separate circuits for thew pc's with marked 'cleaner' and 'kettle' sockets."

We still do this.

The alternative was people unplugging things to insert their variety of equipment in said receptacle. Cleaners are informed that plugging into any outlet not marked for their use is a sacking offence.

It's compounded by shitty consumer equipment that breaks down or otherwise shorts out and takes out circuit breakers, You don't want them on the same spur as your computers or comms equipment - not because of filtering, but because you don't want the power going off without a good reason.