If you’ve ever looked at the bottom of a bank check, you probably glanced over some strangely formed numbers? If you’re a fan of science fiction or retro computers, you’ve probably spotted the same figures on any number of books from the 1980s. They’re mostly readable, but they’re chunky and thin in places you don’t expect.

Those oddball numerals didn’t come from just anywhere—they were a very carefully crafted invention to speed processing in the banking system. These special fonts were created to be readable both by humans and machines—us with our eyes, and the computers with magnetic sensors. Let’s explore the enigmatic characters built for Magnetic Ink Character Recognition (MICR).

Machines Will Do The Work

Early examples of machine-readable magnetic fonts from the Department of Commerce—Automatic Character Recognition, A State-Of-The-Art Report, May 1961.
These days, much of the money in the world is sent and received via digital transfers. Once upon a time, though, paper was king when it came to moving money. The almighty check was how you got money out of one account and into another one.

Sadly, as populations grew and economic activity skyrocketed, the status quo couldn’t hold. By the mid-1940s, the problem was already apparent, with the Federal Reserve dealing with 2 billion checks a year in 1946. While mechanical adding machines and various other techniques helped, fundamentally, bankers and clerks were processing millions of checks daily, all by hand.

The financial world needed a way to speed handling of checks as much as possible. The solution was to enable machines to read as much of the information on a check as possible, so they could handle the basic sorting and processing steps at speed. This would eliminate much of the manual reading and handling by humans, and greatly improve throughput.

The problem was that in the middle of the 20th century, technologies like optical character recognition, or even digital cameras, were decades away. Instead, the key innovation that saved banking was MICR—short for Magnetic Ink Character Recognition. It involved printing certain characters on checks with an iron oxide-based ink. The combination of the ink’s magnetic content and the unique shape of each number meant machines could read the checks easily and unambiguously—even in the case they were physically damaged. Meanwhile, the MICR characters were also designed to remain human readable, so they could be readily understood by the humans using them, too. This was an important backup in the event a check failed machine reading for whatever reason.
An example US-style check with the MICR line along the bottom—printed with the E-13B font. Credit: Federal Reserve Bank of Philadelphia
With MICR, checks could be pre-printed with a bank’s routing number and the customer account to draw from, leaving just the payment amount to be read from the check user’s handwriting. Alternatively, even the amount could be printed in MICR characters if the check was fully machine-issued, speeding processing further. With the aid of magnetic ink, processing speeds went up prodigiously. In 1950, mechanical aids had allowed one clerk to process 1,300 checks in an hour. Fast forward to the magnetic ink era just a few years later, and clerks were able to handle 33,000 checks or more in the same amount of time.

As is so often the way, the world did not agree on one standard for MICR purposes. Developments across the banking world occurred during the 1950s, with two major magnetic fonts being developed in parallel.

If you’re based in the United States, Canada, the UK, Australia, or much of the rest of the English-speaking world, you’re probably most familiar with a font called E-13B. This is the one with the gloopy letters and the worst ‘1’ numeral ever committed to print. It was developed by General Electric and the Stanford Research Institute. Its designation was entirely pragmatic—E denoted that it was the fifth font considered, and B denoted the second revision. 13 referred to the fact it was designed for use on an 0013-inch grid.

The font was designed to create a unique magnetic signal pattern when each numeral or symbol was scanned by a magnetic reader. The shapes were specifically engineered to avoid any possible confusion – that’s why the 0 has those straight sides, and the 8 is so hefty at the bottom, for example. Each number generates a waveform that’s distinct from the others, making it easy to process the signal and read the check accurately. E-13B wasn’t perfect, with 2s and 5s putting out rather similar signals in some cases that could cause confusion, but it proved itself more than reliable enough to do the job.
The 14 characters of the E-14B MICR font—the last four are for control purposes.This book cover from the 1980s was typical of the era – leaning on E-13B tropes to convey a technical aesthetic.
The standard was trialled in 1956 and was adopted by the American Bankers Association by 1958. By 1963, the American National Standards Institute (ANSI) designated that E-13B would be the standard, and by 1967, the Federal Reserve mandated the use of magnetic ink on checks. E-13B went on to become a graphical motif commonly associated with computers and modernity, with artists commonly creating lookalike characters for the whole Latin alphabet. However, the official E-13B standard only ever had 14 characters—numerals 0 to 9, plus four additional control characters for check processing—”transit,” “on-us,” “amount,” and “dash.”
The CMC-7 font, designed by Groupe Bull.
At roughly the same time, French computer company Groupe Bull was working on its own standard. In 1957, it developed the CMC-7 font, which used an entirely different approach to E-13B. Rather than relying on the varying the intensity of magnetism by the amount of ink in a character, CMC-7 instead relied on characters made up of vertical bars. The spacing between the bars could be read by machine to determine the numerals. The design gave CMC-7 characters more of a barcode-like appearance. Notably, CMC-7 also featured a full alphanumerical character set—41 glyphs, including A-Z, 0-9, and five control characters.
An Italian check signed by Enzo Ferrari – note the CMC-7 font along the bottom. Credit: Morio, CC BY-SA 3.0
Thanks to the geopolitics of the mid-20th century, each MICR standard ended up with its own stomping ground. While E-13B dominated in the Anglophone world, CMC-7 ended up being used in France, Spain, and much of Europe and South America. At heart, both standards did the same thing—they enabled machines to read most of the data on a check with a minimum of fuss.

Banks might feel mostly digital these days, but MICR fonts are still an important standard in the financial world. If you’re issuing checks, you might end up running into some problems if you’re not printing them with the appropriate MICR font and magnetic ink. For most of us, checks are a simple tool of the past, but it turns out a great deal of engineering went into perfecting them before the computer came along.

hackaday.com/2025/02/10/how-ma…

Un presunto leak di documenti classificati della National Security Agency (NSA) è stato pubblicato su Breach Forums, uno dei più noti marketplace underground per la compravendita di dati rubati.

L’utente “HumanError”, con il titolo di “GOD User” all’interno del forum, ha annunciato il caricamento di un set di documenti appartenenti alla Five Eyes Intelligence Group, l’alleanza di intelligence che comprende Stati Uniti, Regno Unito, Canada, Australia e Nuova Zelanda.

Secondo il post, il leak sarebbe avvenuto attraverso l’intrusione nei sistemi di Acuity Inc., un’azienda che collabora direttamente con il governo degli Stati Uniti e i suoi alleati.

Cosa è stato compromesso?

Il post pubblica una serie di samples e fornisce alcune informazioni sul contenuto dei dati sottratti, tra cui:

• Nomi completi, email, numeri di telefono (personali e aziendali)
• Indirizzi email governativi, militari e del Pentagono
• Comunicazioni classificate tra Five Eyes, 14 Eyes e i loro alleati

L’autore del leak menziona anche il coinvolgimento di altri membri con gli alias @IntelBroker, @Sangsiero e @EnergyWeaponUser, suggerendo che l’operazione possa essere stata condotta da un gruppo piuttosto che da un singolo attore.

Tali informazioni sono interconnesse ad una precedente fuoriuscita di informazioni di Aprile del 2024, quando IntelBroker pubblicò un post analogo riportando dei dati trapelati dalle infrastrutture della NSA. Infatti i samples pubblicati in entrambi i post risultano gli stessi.

L'articolo NSA Data Leak: Documenti classificati trapelano su Breach Forums per la seconda volta proviene da il blog della sicurezza informatica.

The global status quo on everything from artificial intelligence governance to platform accountability to digital trade is being rewritten.

digitalpolitics.co/newsletter0…

The BASIC language may be considered old-hat here in 2025, and the days when a computer came as a matter of course with a BASIC interpreter are far behind us, but it can still provide many hours of challenge and fun. Even with our love of all things 8-bit, though, we’re still somewhat blown away by [Matthew Begg]’s BASIC interpreter written in 10 lines of BASIC. It’s an entry in the BASIC 10-liner competition, and it’s written to run on a Sinclair ZX Spectrum.

The listing can be viewed as a PNG file on the linked page. It is enough to cause even the most seasoned retrocomputer enthusiasts a headache because, as you might expect, it pushes the limits of the language and the Sinclair interpreter. It implements Tiny Basic as a subset of the more full-featured BASICs, and he’s the first to admit it’s not fast by any means. He gives a line-by-line explanation, and yes, it’s about as far away from the simple Frogger clones we remember bashing in on our Sinclairs as it’s possible to get.

We love it that there are still boundaries to be pushed, even on machines over four decades old, and especially that this one exceeds what we thought was a pretty good knowledge of Sinclair BASIC. Does this language still have a place in the world? We always look forward to the BASIC 10-liner competition.

Header: background by Bill Bertram, CC BY-SA 2.5.

hackaday.com/2025/02/10/basica…

Era una mattina qualunque per i dipendenti di un’importante azienda italiana del settore retail. L’aria nei corridoi profumava ancora di caffè e il tintinnio delle tastiere riempiva gli uffici. Nessuno avrebbe mai immaginato che, in quello stesso momento, i loro dati fossero in vendita nel dark web.

Un post comparso su un noto forum di cybercriminalità informatica chiuso (accesso su presentazione o attraverso pagamento) aveva appena messo in vendita l’accesso a un database contenente più di un milione di clienti: nomi, email, numeri di telefono, indirizzi, persino informazioni sui pagamenti.

Il venditore, noto con lo pseudonimo Panigale, offriva un accesso privilegiato a un sistema SAP compromesso, permettendo a chiunque fosse disposto a pagare 10.000 dollari di effettuare ordini fraudolenti e accedere alle email aziendali.

Ma chi era la vittima? Il post non la menzionava esplicitamente. Solo pochi dettagli lasciavano intendere che fosse un’azienda italiana, forse un grande rivenditore con un’enorme base di clienti.

Initial Access broker: i mercanti del crimine digitale

Quello che stava avvenendo non era un caso isolato. I broker di accesso, come Panigale, sono figure chiave nell’ecosistema del cybercrime. Non sono necessariamente gli autori dell’attacco iniziale, ma fungono da intermediari: ottengono credenziali compromesse, accessi a database o interi sistemi aziendali e li rivendono a gruppi ransomware o altri cybercriminali.

Nel caso specifico, l’offerta riguardava un accesso a SAP, il cuore pulsante della gestione aziendale di molte multinazionali. Controllando questo sistema, un attaccante potrebbe manipolare ordini, visualizzare informazioni finanziarie o persino sabotare la catena di approvvigionamento.

Ma non tutti possono vedere i dettagli dell’inserzione: per visualizzare i sample, il forum richiede più di 100 reazioni, una misura per limitare l’accesso ai soli membri fidati e mantenere la community al sicuro da infiltrazioni delle forze dell’ordine.

La reputazione del venditore: un fattore chiave

Nel dark web, la reputazione è tutto. Panigale non è un novellino: con 150 messaggi e 37 reazioni ricevute, è considerato un utente affidabile. Inoltre, ha già utilizzato il servizio di “garante automatico”, una funzione del forum che permette di effettuare transazioni sicure con il supporto di un escrow. Questo dettaglio rafforza la sua credibilità, attirando potenziali acquirenti pronti a sfruttare i dati rubati.

Ma la vera domanda rimane: quale azienda italiana ha subito questa violazione? E, soprattutto, si è già accorta dell’intrusione?

L'articolo Il Giallo Dell’Attacco Hacker All’azienda italiana! Chi ha perso un milione Accessi SAP? proviene da il blog della sicurezza informatica.

Mechanical switches are pretty easy to understand—the contacts touch, the current flows, and Bob is, presumably, your uncle. But what about soft switches? Well, they’re not that difficult to understand either, as explained by [EDN].
You can build a touch switch quite easily with old-school chips.
The traditional softswitch takes input from a momentary single-pole pushbutton and lets you press to toggle power on and off. This operation is easy to achieve with a simple flip-flop constructed with old-school logic to create a “bistable” circuit. That means it will happily remain stable in one of two states unless you do something to make it switch.

So far, so simple. However, you’ll need to consider that a simple mechanical pushbutton tends to have an issue with the contacts bouncing as they come into contact. If ignored, this would see your softswitch rapidly flicking on and off at times, which is no good at all. To avoid this, you simply need hook up an RC network to smooth out or “debounce” the button input.

Read the post for the full circuit dynamics, as well as how to make the system work with a touchpad instead of a pushbutton. It’s rare to construct such elements from raw logic these days, what with microcontrollers making everything so easy. Still, if you want or need to do it, the old techniques still work just fine! There’s more than one way to solve the problem, of course.

hackaday.com/2025/02/10/flip-f…