Doctors, lawyers and other high earning professionals are often members of “limited liability partnerships”. They’re not employees – and so there’s no 15% employer national insurance. This creates a big tax saving. The Times is reporting that Rachel Reeves is considering changing this – and that it could raise £2bn.

Taxing people differently just because of their choice of legal vehicle is irrational – and there’s certainly a principled justification for equalising the position. It also achieves the political aim of mostly affecting only high earners – around 0.1 % of taxpayers receive 46% of all LLP income.

But it’s not without political cost – more reasonably paid professionals (like GPs) would also be affected: the average GP who’s a member of an LLP earns £118k, and would see their take-home pay fall by about £7k (although some of the tax revenues raised by the new measure could be used to fund an increase in GP pay).

The response of those affected, and the impact on tax revenues and the wider economy, is hard to predict. There are also practical problems, and fairness issues around where precisely the line would be drawn.

This is certainly something any Chancellor should consider – and there may be ways of squaring the circle, and raising revenue without hitting GPs or creating a series of unfair new anomalies.

The think tank/academic group CenTax published a detailed report in September analysing HMRC data around LLP/partnership taxation. The £2bn figure comes from their report – which I highly recommend. Note that their data is from 2020 – so realistically all the figures should be uprated by around 15-20% for inflation/wage growth.

I’ve tried to put together some illustrative data on the impact of the change for individual taxpayers – it’s all in this spreadsheet. Please don’t take this as definitive – it was thrown together quite quickly.

The current situation – the doctors

When someone is employed, their employer applies employer national insurance to their pay packet. So, for example, if a hospital has £118k to pay its doctors.¹ About £18k comes out immediately as employer national insurance. The doctor only ever sees the remaining £100k – and of course pays income tax and employee national insurance on it. He takes home about £69,000.

The doctor never sees that missing £18k, and might be completely unaware of it – but in the long term, evidence shows that he’s paying it (because it reduces his wage).

Now imagine a doctor who’s a “locum”. They’re often (but not always) taxed as self-employed. There’s no employer’s national insurance. So the doctor is paid the whole £118k. She’s paying more income tax and national insurance (because of the higher gross pay), but ends up taking home around £76k. Our locum is £7k better off than an employed doctor.²

Let’s take a third category – a GP. The £118k figure I’ve been using comes from the CenTax report – they estimate it’s the average earnings of a GP who’s a member of an LLP.

Most GP practices are set up as partnerships. A traditional partnership is just people working together in business, but many GPs use a more modern entity, a “limited liability partnership” which behaves like a company in most respects but is taxed like a partnership.

A member of a partnership isn’t an employee and (usually) is taxed in the same way as someone who’s self employed. So a GP will be taxed in the same way as the locum. No employer, and overall she’s £7k better off than an employed doctor.

This is a very irrational result.

It looks more irrational when we get to very highly paid professionals.

Highly paid partners

Most of the £2bn revenue comes from people earning far higher amounts than the £118k received by the average GP. Around 0.1 % of taxpayers receive 46% of all LLP income.

This is from the CenTax report:

Not shown on this table are much less profitable partnerships such as farm partnerships. CenTax proposes an allowance or exemption that prevents them being affected.

The greatest number (but not the highest earners) are solicitors. CenTax reckons the average income of solicitors who are partners/members of LLPs is £316,000.³

A solicitor whose gross income is £316k currently takes home about £180k. If his income was subject to employer national insurance, he’d take home £155k.

This is a very big difference. His effective tax rate (i.e. overall tax divided by overall income) has gone up from 43% to 51%. His marginal tax (i.e. the % tax they pay on the next pound he earns) has gone up from 47% to 55%.

We see more dramatic effects if we go to the largest law firms, where many partners earning well into seven figures.

A partner earning £2m currently takes home £1,072k. If employer NICs applied, she’d take home £914k – meaning £159k more tax. Her effective tax rate has gone up from 46% to 54% and her marginal tax rate is now also 55%.⁴

This puts our £2m partner in the same position as (say) a trader at a bank where their salary and bonus pot are together £2m. Previously she paid less tax; now she pays the same.

An important point: the reason law firms are usually structured as partnerships is history rather than tax. Until relatively recently, solicitors were required to practice as partners or sole practitioners. Firms weren’t able to become companies until 1985. Even today, most of the big firms aren’t in practice able to incorporate because, whilst it would be permissible for their English lawyers, it’s not permitted for many of the foreign lawyers they practice with.

Similarly, auditors (and thus many accountants) historically had to structure as partnerships, and still do in some countries.

However many professionals absolutely do structure as partnerships for tax purposes. Most fund management businesses – private equity and hedge funds – are structured as LLPs rather than companies. The main, and perhaps only, reason for this is tax.

According to CenTax, the average member of a financial services LLP earns £675,000. There will be some earning ten or twenty times this figure.

The arguments for and against

There are two obvious arguments in favour:

• If the Chancellor is to stick to her fiscal rules then, absent very large spending cuts, she needs to find additional tax revenue. This is a relatively easy way of taxing high earners.
• It’s in principle correct that everyone who makes their living from work should be taxed the same way.

CenTax estimated that imposing employer national insurance on LLP members’ pay would raise around £2bn. Their analysis seems sensible to me – although it’s based on 2020 numbers so the figure today would be around 15-20% higher.

There are, inevitably, several arguments against.

Consistency

If the Government only taxed LLPs, and not traditional partnerships, then some GP practices would pay more tax and others would not. That seems hard to justify.

The issue is even more stark when it comes to law firms. One of the most profitable law firms in the country is structured as a traditional partnership, not an LLP. Can it be right they pay less tax because of this historical accident?

And some lawyers practice as individuals. They wouldn’t pay employer NICs. That seems odd. (And presumably we’d see more like that.)

What about barristers? Junior barristers at leading commercial barristers’ chambers can earn up to £360,000 in their first year. Some senior KCs earn ten times that. Barristers aren’t (usually) members of partnerships; but it’s hard to see why a barrister who earns £2m should pay less tax than a solicitor who earns the same.

This becomes quite hard to fix unless employer national insurance (or something equivalent) is applied to all the self-employed (and see further below).

Behavioural response

It’s easy to calculate the “static” revenue from a tax change – it’s just multiplying numbers together.

Estimating the actual revenue is much more difficult, because you have to take into account the “behavioural response”.

Here there will be several:

• Some people will move from LLPs to become self-employed consultants (and escape the new tax).
• Large law firms practice all over the world. In many cases it’s possible to do much the way work in Dubai as in London. So (at the margins) we will see some members of these firms move from London to Dubai to escape the tax. And not just Dubai – for various reasons, lawyers in many European countries pay lower tax than lawyers in the UK.
• Some people will work less, because they are less motivated. Conversely, others will work more, because they need to work more hours/years to earn the same amount.
• Some firms will restructure into companies. The LLP partners/members will become shareholders. On the fact of it this saves just a small amount of tax – my calculations suggest an average GP could save £3k, and even a £2m law firm partner would save only £13k. However in practice it may save more than this, as the companies could retain and reinvest profit. That may even have business and economic advantages.⁵

CenTax used historical “elasticity” data to estimate that imposing NICs on LLPs would cause a loss of tax revenue equal to about 20% of the “static” estimate. That feels in the right range.

The question is whether there would be a wider impact on UK law firms, fund managers etc, beyond just the loss of tax revenue, and perhaps a wider impact on the City and the economy as a whole. I don’t know the answer to that.

The doctors

I am not very good at politics, and try not to make political predictions.

That said: it seems to me there are likely to be few people opposed to the idea of increasing the tax of millionaire lawyers. There may be rather more people opposed to the idea of increasing the tax on GPs. A cut in take-home-pay is likely to go down badly with GPs, particularly when compared with increase of £7k idea of taxing millionaire lawyers the (net) £2,000 increase they received from the most recent pay deal.

That raises obvious political questions: but exempting doctors from any new rule would be unprincipled.

One answer, suggested by CenTax, is for central Government to increase GP pay, funded by the new tax measure.

A slightly less unprincipled approach would be to create a per-partner/member exempt amount, set at a level so doctors pay little or no additional tax.

If the exempt amount were set at the average GP partner pay of £118,000, I estimate this would reduce the yield from about £2bn to about £1bn (calculation on the second tab of the spreadsheet).

But this unprincipled compromise could actually prepare the way for the most principled change of all: applying employer national insurance to all forms of work, employed and self employed. That’s clearly out of the question if we’re talking about the moderately paid self-employed (e.g. tradespeople). But if it’s done with an exempt amount, then suddenly it seems more realistic.

And that has the laudable side-effect of dealing with the consistency problems identified above.

Obvious disclosure: I was a partner in a large law firm. I have no economic interest in any law firms today, but it goes without saying I am going to be influenced by my background.

We’ve investigated many tax avoidance schemes. None of them had any technical merit – indeed many were closer to tax evasion than tax avoidance. All of the schemes, without exception, should have been disclosed to HMRC under DOTAS – the rules requiring up-front disclosure of tax avoidance schemes. None of them were.

The whole point of DOTAS was to enable HMRC to catch tax avoidance schemes early, and then either change the law (e.g. if someone found a loophole) or challenge the scheme (if it was just hopeless). In the last fifteen years, the landscape has changed dramatically. Almost all tax avoidance schemes are of the “hopeless” rather than “loophole” variety. And so the only way the promoters can stay in business is by breaking the law and avoiding DOTAS.

My view is that many promoters are criminals, but criminals who are very hard to prosecute under current law. My view was (and is) that the law should be changed to make prosecution a real possibility. Only then would we be able to deter the rogue promoters. My proposal was that failure to disclose under DOTAS should be a criminal offence, and I was delighted to see this adopted by the Government in the package it published back in the Spring.

Except I’ve changed my mind – following a series of discussions with HMRC and advisers.

I changed my mind partly because of two serious problems with the proposal – one that can be fixed, and one that can’t. And partly because two of the other elements in the Government’s package render it unnecessary.

I discussed these issues at the House of Lords Finance Bill Sub-Committee on Monday (starting at 07:45):

The problem we can fix

Many representative bodies have complained that the proposal as it stands is likely to create a chilling effect on ordinary tax advice, or swamp HMRC with unnecessary disclosures – or both.

DOTAS is complex, and honest tax advisers – who currently “take views” that uncontroversial arrangements aren’t disclosable – won’t be able to be as relaxed when breach of DOTAS is a criminal offence. So some firms might move out of tax advice altogether – a bad thing for business and HMRC alike. Others could respond by disclosing everything out of prudence.

This is an important criticism, but one that can be addressed without too much difficulty. The answer is to create an “bona fide firm” defence to the criminal offence so that mainstream firms need have no fear that a one-off accident could result in criminal liability.

Mainstream firms look really different from the promoter firms. Mainstream firms advise varied clients doing varied things, ,and it would be exceptional for any of those things to be caught by DOTAS. The promoter firms typically run a very small number of schemes (often just one), marketed in high volumes.

So we can create a defence that applies to any firm that can demonstrate that at least 90% of its tax-related business (measured by fee income) relates to bona fide tax advice.¹ Tax advice would be “bona fide” for this purpose if either it relates to an arrangement which is not properly disclosable under DOTAS, or it is disclosable but was properly disclosed.

I expect every single mainstream firm that provides tax advice, large and small, would be comfortable the defence applied to them, and so would not suffer the chilling effects that concern the CIOT.

The “chilling effect” problem is therefore not insurmountable.

The problem that can’t be fixed

The more challenging issue is that the same complexity that scares normal tax advisers will, perversely, make the actual criminal tax advisers more relaxed.

Take a look at the judgment in the Hyrax case. This was a low quality tax avoidance scheme which in my view had no reasonable prospect of success. It failed to disclose under DOTAS by running extremely poor arguments that in my view also had no reasonable prospect of success. It still took the Tribunal 56 pages to throw out Hyrax’s appeal.

The Hyrax judgment shows a KC (who I expect designed the scheme) running a long series of complex but meritless arguments. The Tribunal had no difficulty dismissing them, but arguments like these would be far harder for a jury to evaluate, and could easily create reasonable doubt. That’s particularly the case when promoters have an opinion from a KC who is willing to bless the most far-fetched arguments. Until the Bar gets its house in order, it’s going to be far too easy for a promoter to manufacture an excuse for its actions.

In the course of the recent consultation, I and other advisers spoke to knowledgeable people at HMRC, and we came away with the distinct impression it might be challenging to present any DOTAS case to a jury. We then spoke to retired HMRC inspectors with experience of tax prosecutions, and then to barristers specialising in “white collar” crime (doing both prosecution and defence work). All of them thought that prosecutions would be problematic – one barrister said the offence might be “unprosecutable”.

Our team has explored ways that the offence could be made simpler, but unfortunately none seem very workable. The various forms of simplicity we’ve considered all either risk criminalising innocent firms, or present too many loopholes for criminal firms.

So I’m disappointed to have to conclude that my basic concept of criminalising DOTAS breaches is not viable.

I’d love to be wrong, and for there to be a way to design an offence that neither criminalises the innocent or lets the guilty off the hook, but I’m not currently seeing it.

The alternative

My original thought process was that rogue promoters were breaking the law without consequence, and severe sanctions were required to stop this.

That’s still my view, but I think the same aim may now be achieved by two other measures in the Spring package:

Universal stop notices

HMRC currently has a power to issue a “stop notice” to the promoter of a tax scheme, making further promotion of that tax scheme a criminal offence.² The problem is that the people ultimately behind these schemes hide themselves behind trusts, nominee directors and nominee shareholders. HMRC will issue a stop notice to one entity, and the promoter will simply move its business to another (very possibly run by different frontmen).³ This example in the HMRC consultation document is an accurate reflection of what’s been happening:

The Government is proposing to end this game of “whack a mole” with a “universal stop notice“. This would empower HMRC to issue a notice specifying a scheme; no person could then promote or enable that or any similar scheme. Anyone who did would commit a criminal offence, as would any person controlling or influencing them.

The key question is whether HMRC is able to speedily identify schemes, and HMRC and the Parliamentary drafting team are then able to craft USNs which encompass all the variations of the schemes that the promoters (who are smart and devious) will be able to come up with. If they can, then the USNs should do the job of criminalising the most significant areas of tax avoidance.

DOTAS penalties

One of the reasons promoters ignore DOTAS is that it’s hard for HMRC to apply penalties.

Under the current rules, failing to disclose a tax avoidance scheme under DOTAS can lead to daily penalties of up to £600. These penalties must be imposed by the Tax Tribunal on application by HMRC. If the failure continues, HMRC can apply again for higher “continuing” daily penalties of up to £1 million in total. HMRC cannot impose these penalties itself – each one requires a formal tribunal process. There are also smaller fixed penalties, which HMRC can issue directly, for users who fail to include a scheme reference number on their returns (typically £5,000–£10,000), with a right of appeal to the tribunal.

This regime has proved weak. The tribunal process is slow and expensive, and promoters can often dissolve or re-incorporate before penalties are determined. Even when penalties are imposed, they are rarely paid – as in Hyrax, where the tribunal approved a £1 million DOTAS penalty but nothing was ever recovered. The result is that promoters routinely breach DOTAS with little real consequence.

The Government’s Spring 2025 package proposes to change this by allowing HMRC to impose DOTAS penalties directly, bringing the rules into line with other modern anti-avoidance regimes. Penalties would still be appealable to the tribunal, but HMRC could issue them immediately, without needing a prior tribunal order.

This means HMRC should be able to issue DOTAS penalties as a simple procedural step, whenever it identifies a tax avoidance scheme that is being marketed but has not been disclosed under DOTAS. The process should be much faster, making it harder for promoters to evade penalties by delay or phoenixing.

If HMRC wants to change the avoidance landscape then it will have to act aggressively, and be willing to use the joint and several liability rules to pursue the individuals behind the schemes if/when the companies fold. I also feel that maximum penalties of £1m isn’t enough – the HMRC example of a promoter making £1m in fees per month is not fictional. I would make the maximum penalty the higher of (1) £1m, and (2) 150% of fees charged.

The next step

I still think more is needed. It’s offensive to me and many other tax advisers that flagrantly doomed tax avoidance schemes are still sold at scale. It’s become a mis-selling problem as much as it is a tax problem – lives are ruined when these schemes are sold to people (often people on modest incomes).

Part of the answer is regulation (and regulation that goes further than the current Government proposal, which won’t apply to most scheme promoters). But it’s not clear that existing regulation is working. The Bar in particular has tolerated bad actors for too long. If the professions won’t regulate themselves then Government should step in.

One man has set up a series of fake banks on Companies House – with real bank names, no checks, and zero consequences. When he’s caught, he does it again. And he’s been facilitated by a legitimate incorporation agent that, for reasons we don’t understand, incorporated an obviously fraudulent fake bank for him.

This report identifies a serial fraudster, the incorporation agent that enabled him, and how we can stop these frauds in the future.

Meet the fake banker

Meet Barbarat Giuseppe, the world’s most prolific banker. He’s run most of the world’s largest banks: JPMorgan, Soc Gen, SEB, Citi, Westpac, Natixis, Goldman Sachs, and many more:

Mr Giuseppe’s spectacular career is spoilt only by the small detail that it’s all fraudulent.

And it’s also spoilt by the other small detail that Giuseppe might not exist, or could be an innocent person whose name has been stolen (although why someone would invent or steal such an unusual name is unclear).¹

“Giuseppe” sometimes gets caught. Perhaps an investigator like Graham Barrow spots one of the fake banks and informs the real bank. Then the real bank has to apply to the UK’s most obscure court, the Company Names Tribunal, for an order changing the company’s name. Here’s an example, brought by the real UBS against “UBS Group AG Ltd”. The fraudsters didn’t bother to defend their position, and the company was then renamed to “14695023 LTD” (after its company number).

The problem is that then nothing happens. No investigation by the police or any regulator.

In one case, the bank made a report to the police and was told that it was a civil matter. In another, a bank was told that the police couldn’t take any action unless a victim of fraud came forward.²

The fraudsters therefore act with total impunity.

So what happened after the fake “UBS Group AG Ltd” was forcibly renamed by the Tribunal and Companies House to “14695023 Ltd”? The fraudsters renamed it again – to “Goldman Sachs Finance Ltd“, and kept going. That wasn’t a one-off.

What is the scam? It’s hard to say. One knowledgable insider told me he suspects it’s money laundering – Giuseppe opens up a bank account in a country with slack KYC procedures and provides ID showing he’s the director of Goldman Sachs. But it could be almost anything – investor fraud, money laundering, sanctions-busting…

And Giuseppe is only one of many playing the same game. In 2024 there were nine defended Company Names Tribunal cases (which appear to have been bona fide disputes). There were 339 undefended orders – most of which look like plain fraud.

Who helped him do it?

The fraud is obvious on the face of the companies’ incorporation documents: a company called “UBS Group AG Ltd”, with the declared beneficial owner an individual living in France. It’s obviously nothing to do with the real UBS. The stated (paid) share capital of EUR1bn makes reasonably clear this isn’t someone’s weird game, but simple fraud.

So who incorporated the company? The incorporation document contains a statement of compliance:

The declared agent is “Registered Office (UK) Ltd”, which³ trades as The London Office and MYCO Formations.

We see a statement of compliance from Registered Office (UK) Ltd on all the Barbarat fake bank companies.

The precise status of Registered Office (UK) Ltd isn’t clear to us. Generally speaking, companies providing incorporation or registered office services are treated as a TCSP (trust or company services provider) and have to be regulated for anti-money laundering (AML) purposes, usually by HMRC. However we can’t see Registered Office (UK) Ltd on the HMRC list.⁴ A previous story suggested that Registered Office (UK) Ltd believed all they had to do to comply with AML rules was to check clients’ passports. This is not correct.

The first response from Registered Office (UK) Ltd

The Barbarat network was first spotted by Graham Barrow two years ago, and he asked MYCO about this.

MYCO Works said they’d investigate:

There’s been no response since. Nor was there any response when we wrote to Registered Office (UK) Ltd back in February asking for comment.

Registered Office (UK)’s defence

We wrote to Registered Office (UK) Ltd again this week – and MYCO Works (their trading name) replied. They first claimed the fraudsters “just used our address”. After we pressed, the story changed – blaming their software and, finally, Companies House.

Here’s the complete chain.

This is Registered Office (UK) Ltd telling us that they didn’t incorporate the UBS Group AG Ltd, but just provided a registered office:

We asked how that could be. The company was incorporated stating that Registered Office (UK) Ltd was the incorporation agent. Even if that could be forged (and I wasn’t sure it could be), surely they would have spotted the forgery, and the other obvious signs of fraud, and not accepted it as a client?

MYCO Works then escalated the matter internally, and we received a completely different response, which doesn’t make much sense:

We replied saying that we didn’t understand how they could let unknown people use their software to incorporate companies, and received another response blaming software:

This again makes no sense. The reference to Companies House’s lack of verification looks like a distraction. Legally, an incorporation agent is responsible for companies they incorporate. That’s clear from the statement of compliance filed when UBS Group AG Ltd was incorporated

We don’t know what Registered Office (UK) Ltd is doing, or why it thinks it’s acceptable to let third parties incorporate companies in its name. We can’t imagine they knew that a fraud was being committed, but they seem to have taken the business decision to let absolutely anyone incorporate absolutely anything in their name.

The bottom line: Registered Office (UK) Ltd provided a certificate of compliance for a string of fake banks, each incorporated with obviously fake share capital of hundreds of millions or billions of pounds. Those certificates of compliance were false – and that should have consequences.

How to shut this down

There are three obvious steps that could be taken.

First, we need to see prosecutions. Skip the hard stuff of finding victims of fraud. Do an “Al Capone” and prosecute the easy offences instead. Anyone filing false information with Companies House has committed a criminal offence.

If (as in Giuseppe’s case) they do knowingly, it’s punishable by up to two years in jail. And Giuseppe has claimed to have paid £100m for shares in the fake UBS/Goldman Sachs (almost certainly false), and provided Companies House with a series of false addresses.

Second, Companies House should introduce address verification. Right now anyone can incorporate a company giving Buckingham Palace as their address – there are no checks whatsoever. But if someone falsely uses your address for their company, you have to provide full identification details to Companies House to get it fixed. That doesn’t feel right.

There are many ways to solve this: the simplest would be for Companies House to send automated letters with a security code to the registered office address, and only finalise incorporations once the security code has been entered.

Third: the authorities should go after the facilitators: legitimate businesses who share responsibility for the fraud. Not because they are criminals, but because (we believe) they’ve decided that conducting even small checks is inconvenient for their business. It appears in this case, providing fraudsters (unwittingly but, in our view, unethically) with the means to incorporate companies with zero oversight.

When that business decision facilitates fraud, it should have consequences. The lack of any due diligence suggests a breach of anti-money laundering rules. However the more serious consequence is the basic criminal offence (an unlimited fine but no jail time), which applies to any person who, without reasonable excuse, causes a false document to be delivered to Companies House.

It seems likely that this criminal offence was committed by the company formation agent, Registered Office (UK) Ltd.⁵ We will continue to see company formation agents facilitating fraud until there is a strong deterrent – meaning well-publicised prosecutions.

Many thanks to N for the original tip, and L, O and C for their input. And, most of all, to Graham Barrow for spotting this network years ago, and for freely providing us with his invaluable expertise.

A sentiment just crossed Fediverse recently, which is in the vein of "RSS was peak social media, change my mind". The original post was from https://hachyderm.io/@Daojoan@mastodon.social and is quoted below:

RSS never tracked you.
Email never throttled you.
Blogs never begged for dopamine.
The old web wasn’t perfect.
But it was yours.

https://mastodon.social/@Daojoan/114587431688413845M

I was there for the rise and fall of blogging, so the rest of this post is me over thinking this particular post.

RSS never tracked you

RSS never tracked you the way HTTP or HTML never tracked you. What tracks you with HTTP/HTML are rendering engines for HTML and Javascript. I can say with absolute certainty that tracking-tech during blogging's height was used to track audience, click-through rates, and other site-engagement metrics. RSS was the loss-leader for clicks on sites. This particular post uses one of those tricks; an opening paragraph promising more, which if you're reading this through RSS you will have to click through to read. Tracking RSS feeds was done through tracking pixels back in the day because you couldn't trust Javascript to run in the RSS readers but HTML rendering generally worked.

Email never throttled you.

Email absolutely did, absolutely. The problem with spinning up a self-hosted newsletter service in 2025 is getting your IP reputation good enough that the big mail-box vendors (Google, Microsoft, Yahoo/AOL) let you deliver. This issue was in its infancy in 2005, but was an emerging problem as IP reputation became clear as a low-cost first-pass anti-spam technique. Mailing list operators ran into this all the time back in 2005 as various subscribers moved behind security appliances doing IP reputation.

Blogs never begged for dopamine.

On a factual basis, this is false. Sole operators like myself lived for comments, that dopamine hit from people liking what I wrote. If I couldn't get that, I'd enjoy reshare statistics (see the first point about RSS tracking). Many commercial blogging platforms even created RSS feeds for comments on specific articles to make it easier to keep up on discussions. If that isn't dopamine, I don't know what is.

The old web wasn’t perfect.
But it was yours.

True, to a point. I tagged this article blogger because that's what I hosted this blog through for most of its first decade.  You know, a centralized blogging platform akin to LiveJournal or Dreamwidth back then, or Medium and Substack today. I didn't own the platform. After I moved to my own domain and Movable Type, this was true. And yeah, it wasn't perfect but it was most definitely mine.

There is another layer to this post beyond the simply factual, and that's a critique of platforms. Blogging in its heyday was perhaps the last major gasp of the Old Internet, where a bunch of hobbyists create something beautiful and widely adopted, which then gets enclosed by commercial interests. Blogging was absolutely not centralized. That lack of centralization means there was no unitary profit motive looking to drive engagement to increase sales, those were limited to individual sites.

Blogging's decline is traceable to two big trends:

• Google killed Google Reader, which was the dominant RSS reader by a long shot. This death forced a bunch of folks to look for alternate platforms. My stats show my readership plummeted over 50% on death-day.
• Twitter and other early social media provided a much shorter dopamine feedback loop than then blog-publish-comments loop.

The fact that Google Reader's death functionally killed RSS-based distribution is proof that blogging was already beginning to centralize. Google couldn't control production and distribution, only engagement; and the real money was in controlling all three. Google wanted what Twitter had, which was the entire production, distribution, and engagement framework on the same site with the same owners and tracking infrastructure. Once they had that, start engineering dopamine feedback loops to improve stickiness and engagement; and we have all the algorithmic and dark pattern pathologies we know and loathe today.

Modern internet users have been trained for a decade in a half to expect social media to involve a single, or small number of sites to do everything, and individual posts are short enough to deal with while waiting for a bus or the kids to walk from school to car. The old blog+rss model simply can't compete with this, and better fits as attention-competition to a given user's news media consumption. Medium and Substack both offering paid subscription options for individual blogs is proof that news media is the competition to blogging, not social media.

The nostalgic exhortation to set up a blog and distribute through RSS is in the same vein as calls to return to IRC and dump Slack/Teams. Some people/groups can do that, but the platform features are different enough that the old tools don't feel feature-complete and that lack nudges folk back into the commercial walled gardens.

(Like this article? Read more Wednesday Wisdom! No time to read? No worries! This article is also available as a podcast).

People who are getting on a bit in years (like me), like to discuss how nothing is like it was anymore. In that line of conversation, there is a fine distinction between outright complaining and a potentially useful analysis of changes in society and how these changes impact the things that are happening. Last week, during a dinner with my former manager at a big tech company (who is also past the target age for an AARP membership), we started discussing how we got into the field, how that was different from the situation people find themselves in today, and what the consequences of that are.

When I started getting interested in computers, everything was an uphill battle in all directions: Computers were scarce, they were underpowered, and they were expensive. There was little information available and what was available was difficult to get and, again, expensive. I regularly bought computer magazines and tried to get my hands on any and all books that discussed computers and programming. It was also a very lonely endeavor, as I knew almost nobody who knew anything about computers.

Then, when I went to college, access to computers, books, and people improved significantly, but it was still not a panacea. For instance, during my first two years at college, we had to reserve time slots to use the terminals of our school’s mini-computer. Bookings were for 30 minutes time slots you could have at most three slots outstanding. Yes, you read that right, we could have a max of 1 1⁄2 hours of terminal time reserved up front! When that time was spent, you had to go find the reservation terminal and then you could book another 1 1⁄2 hours, subject to availability. In these circumstances, when you get to sit down in front of the terminal, you better know what you are going to type in.

Additionally, the Pascal and COBOL compilers ran on a batch system overnight, so you had to carefully plan a ½ hour slot for the next day to see if your program had compiled, and if it hadn’t, make any corrections and resubmit, so that you could get another attempt at maybe running it the day thereafter. As you can imagine, this system did not do wonders for velocity and there was a bonus for getting your work in early in the trimester.

To make matters worse, whenever the college IT department was faced with a low free disk space situation, they ran a program called “DailyReport” which removed “temporary files” from the disk. Unfortunately, “DailyReport” considered student’s binaries temporary files, because they could be recreated from the sources. On at least one occasion, I logged on in the morning to find that my compilation had succeeded at 2am but that DailyReport had come round at 4am to remove my binary. Try submitting your code for a software engineering class on time under these circumstances!

All of that hassle was to get a professional qualification for a job that didn’t even pay that well. It wasn’t bad, but it was not “doctor or lawyer” good. I have written about this before, but when I graduated, I could maybe look forward to a comfortable middle class existence. As a matter of fact, in my first job, my yearly income was so low that I qualified for healthcare under the state’s social health fund plans. Nothing wrong with that, I guess about 75% of the country earned under the threshold for the funds, but it just goes to show that nobody, not even me, thought I was heading for a cushy tech job.

There are some consequences to this state of affairs. First of all, in that environment, only people who really (and I mean really really) like toying with computers get into the field. It’s not just that coding wasn’t cool yet; it was actually difficult and cumbersome. Only people with a deep interest in the subject matter and with lots of passion studied computer science and they had to struggle mightily to become any good at it. But, good they typically became, because: Pressure makes diamonds.

Another consequence of this story is that, tech skills being rare, it was very easy to let it go to your head.

In the small village I grew up in, I knew literally nobody who was into technology. That’s quite lonely as it means that you have nobody to talk to about your passion and it doesn’t make you very popular either. But it also makes you think you might be a wizard. Myths and legends are full of wizards and they are always solitary figures that are holed up somewhere and doing things that nobody understands while speaking strange languages. I was a solitary figure that was holed up in my room all day doing things that nobody understood in strange languages, so I thought I was a wizard. Stands to reason.

This feeling persisted into college. Sure, there were more people there, but in the meantime computer science had become somewhat hip and so I had a fair amount of fellow students (especially at the start of the first year) who did not have my deep pre-college experience hacking with computers. Fortunately, a handful of them did and we hung out together, because it is expected for wizards to form a conventicle, especially in wizard school (forming our own little pre-HP Ravenclaw).

Most of the rest did not make it very far in college. We started the first year with 220 students and four years later, about 60 would graduate. During the introductory program, the professors helpfully said: “Look left and right of you, only one of you will be here next year. Even then, of the people who graduated in the same year as me, I would trust approximately a quarter of them to touch a computer that I care about.

The problem with feeling you are a wizard is that it can lead to arrogance. Being in the possession of secret knowledge and arcane spells, I would look down on the muggles who were struggling to set their VCR’s clock to adapt to daylight savings time or who were trying to tame the dragon of WordPerfect without knowing the incantations and wand movements required to do so.

One day, a couple of guests came in to dine in our restaurant. They were obviously exhausted and my mother asked what was going on. They explained that they ran a small bookkeeping firm and they had just bought their first computer. For the whole day, they had been trying to set it up, install some software, and make it print. My mother graciously offered the services of the local wizard (me). Next day, I went over there and did the needful, showing off my mastery of the arcane “MODE 9600,N,8,1” and the contents of CONFIG.SYS.

Until I started working, I really did not meet anyone who knew more about computers than I did. Not even the college professors. However, when I did start working, I soon found myself surrounded by real wizards who knew much more than I did. This was a humbling experience. It’s not just that they knew and understood things I didn’t know and understand, they knew and understood so much more that I feared I might never catch up…

This is of course the young person’s underestimation of the value of time. Time heals all wounds and it affords compound interest (in both money and knowledge). Steadily grinding over a period of time does wonders.

After my first job, I kept finding people who were much smarter and more knowledgeable than me and I started to thoroughly enjoy their presence. At almost every company or contract, I found people, often older than me, who had knowledge and experience that I didn’t have. Fortunately, I had never become so full of myself that I resented meeting these people. It also didn’t break any fundamental feelings of superiority because, being Dutch, I really didn’t have these to begin with; all things considered, feeling like a wizard was a really thin layer of veneer.

Holland is the country where people say: “If you behave like a normal person, you are already crazy enough.” Dutch parents tell their kids to “Doe normaal!” (Act normal!) It is one of the two Dutch sentences the lovely Mrs Wednesday Wisdom can pronounce. The other one being: “Hand voor je mond!” (Hand in front of your mouth!), which is what we say if someone yawns while providing access to the light source at the other end of the tube that goes through your body all the way to the other end.

When I joined Google, I finally found my place. Here was a company where everyone was a wizard. When people asked me what it was like, I used to say: “Well, everyone’s a wizard and I am the dumbest person in the building”. Being in that exalted company robbed me from any last vestiges of feeling special that might still have been lingering. My colleagues had written books on a range of topics, had authored open source software that everyone in the world used, patented new stuff left and right, invented new (and useful) programming languages, and together we did things that had never been seen before. I cannot lay claim to any significant contributions to any of that, but I am really good at taking other people’s great ideas and running with them, which is a skill in itself. And also, I can grind, which in a world where every success is 1% inspiration and 99% perspiration is also quite useful.

Being the dumbest person in the building means that you experience a lot of pressure from everyone above you. But, you have to realize that the easiest way to get better at something is if you have the examples right in front of your eyes to copy and learn from. It might be a lot of pressure, but pressure makes diamonds!

I kept running into people that looked like the younger me, but “gone wrong”. I once interviewed a young gentleman from South America who had applied for an SRE role at Google. I started the interview with a lowball Linux system administration question, which he answered satisfactorily. After giving the answer he looked at me proudly and said: “Have you ever met anyone aged 26 who knows as much about Linux as I do?” “Yes,” I answered, “everyone in this building.”

He did not make the hiring bar, but he became legendary for asking our lovely receptionist out on a date on the way out of the building.

In the hiring committees we regularly rejected candidates who we felt were probably smart enough, but who had spent too much time being a big fish in small ponds. It is not hard to shine in a dim room and these people had grown complacent and not continued to develop as engineers. I’ll say it again: Pressure makes diamonds, without that pressure you just have some carbon…

The times have really changed a lot. Compared to 40 years ago, there is an abundance of everything you need to get going in the field: Powerful computers are cheap and widely available; there’s the Internet, and, until recently, getting into technology was a surefire way to get a high paying job. Consequently, it appears to me that many people choose a career in tech not because of a passion for the subject matter, but for purely monetary reasons. I do not begrudge anyone that choice though and I totally understand parents who give their children three career options: Doctor, lawyer, software engineer. But I do regularly see that lack of passion and without that passion you do not have the intrinsic motivation to grind and without that grinding it is hard to become good.

I regularly told people who asked how to land a job at Google that the only people who could do that were the ones who had thoroughly misspent their childhood.

Personally I miss the days of pouring over a book on Z80 assembler and trying to make heads or tails of it. I got into this field because it seemed to me that making the idiot box do something useful was one of the greatest puzzles around and it has never disappointed in that sense. Choosing my employers so that I was always the dumbest person in the building worked extremely well for me. As a strategy for selecting your next gig, I can highly recommend it.

The Department of Government Efficiency, Musk's vehicle. made news by "discovering" the General Services Administration uses tapes, and plans to save $1M by switching to something else (disks, or cloud-based storage). Long time readers of this blog may remember I used to talk a lot about storage and tape backup. Guess it's time to get my antique Storage Nerd hat out of the closet (this is my first storage post since 2013) to explain why tape is still relevant in an era of 400Gb backbone networks and 30TB SMR disks.

The SaaS revolution has utterly transformed the office automation space. The job I had in 2005, in the early years of this blog, only exists in small pockets anymore. So many office systems have been SaaSified that the old problems I used to blog about around backups and storage tech are much less pressing in the modern era. Where we have stuff like that are places that have decades of old file data, staring in the mid to late 1980s, that is still being hauled around. Even when I was still doing this in the late 2000s the needle was shifting to large arrays of cheap disks replacing tape arrays.

Where you still see tape being used here are offices with policies for "off-site" or "offline" storage of key office data. A lot of that stuff is also done on disk these days, but some offices still kept their tape libraries. I suspect a lot of what DoGE found was in this category of offices retaining tape infrastructure. Is disk cheaper here? Marginally, the true savings will be much less than the $1M headline rate.

But there is another area where tape continues to be the economical option, and it's another area DoGE is going to run into: large scientific datasets.

To explain why, I want to use a contrasting example: A vacation picture you took on an iPhone in 2011, put into Dropbox, shared twice, and haven't looked at in 14 years. That file has followed you to new laptops and phones, unseen, unloved, but available. A lot goes into making sure it's available.

All the big object-stores like S3, and file-sync-and-share services (like Dropbox, Box, MS live, Google Drive, Proton Drive, etc) use a common architecture because this architecture has been proven to be reliable at avoiding visible data-loss:

• Every uploaded file is split into 4KB blocks (the size is relevant to disk technology, which I'm not going into here)
• Each block is written between 3 and 7 times to disk in a given datacenter or region, the exact replication factor changes based on service and internal realities
• Each block is replicated to more than one geographic region as a disaster resilience move, generally at least 2, often 3 or more

The end result of the above is that the 1MB vacation picture is written to disk 6 to 14 different times. The nice thing about the above is you can lose an entire rack-row of a datacenter and not lose data; you might lose 2 of your 5 copies of a given block, but you have 3 left to rebuild, and your other region still has full copies.

But I mentioned this 1MB file has been kept online for 14 years. Assuming an average disk life-span of 5 years, each block has been migrated to new hardware 3 times in those years. Meaning each 4KB block of that file has been resident on between 24 and 42 hardrives; or more, if your provider replicates to more than 2 discrete geographic region. Those drives have been spinning and using power (and therefore requiring cooling) the entire time.

These systems need to go to all of this effort because they need to be sure that all files are available all the time, when you need it, where you need it, as fast as possible. If a person in that vacation photo retires, and you suddenly need that picture for the Retirement Montage at their going away party, you don't want to wait hours for it to come off tape. You want it now.

Contrast this to a scientific dataset. Once the data has stopped being used for Science! it can safely be archived until someone else needs to use it. This is the use-case behind AWS S3 Glacier: you pay a lot less for storing data, so long as you're willing to accept delays measurable in hours before you can access it. This is also the use-case where tape shines.

A lab gets done chewing on a dataset sized at 100TB, which is pretty chonky for 2011. They send it to cold storage. Their IT section dutifully copies the 100TB dataset onto LTO-5 drives at 1.5TB per tape, for a stack of 67 tapes, and removes the dataset from their disk-based storage arrays.

Time passes, as with the Dropbox-style data. LTO drives can read between 1 and 2 generations prior. Assuming the lab IT section keeps up on tape technology, it would be the advent of LTO-7 in 2015 that would prompt a great restore and rearchive effort of all LTO-5 and previous media. LTO-7 can do 6TB per tape, for a much smaller stack of 17 tapes.

LTO-8 changed this, with only a one version lookback. So when LTO-8 comes out in 2017 with a 9TB capacity, a read restore/rearchive effort runs again, changing our stack of tapes from 17 to 12. LTO-9 comes out in 2021 with 18TB per tape, and that stack reduces to 6 tapes to hold 100TB.

All in all, our cold dataset had to relocate to new media three times, same as the disk-based stuff. However, keeping stacks of tape in a climate controlled room is vastly cheaper than a room of powered, spinning disk. The actual reality is somewhat different, as the few data archive people I know mention they do great restore/archive runs about every 8 to 10 years, largely driven by changes in drive connectivity (SCSI, SATA, FibreChannel, Infiniband, SAS, etc), OS and software support, and corporate purchasing cycles. Keeping old drives around for as long as possible is fiscally smart, so the true recopy events for our example data is likely "1".

So another lab wants to use that dataset and puts in a request. A day later, the data is on a disk-array for usage. Done. Carrying costs for that data in the intervening 14 years are significantly lower than the always available model of S3 and Dropbox.

Tape: still quite useful in the right contexts.