Eric S. Raymond

Announcing pytogo release 1.13 Fast, crude, incomplete, but useful Python to Go source translator. Intended to assist a human translator rather than fully automate the process, but often produces something closer to idiomatic handwritten Go than fully automated tools can. New in this release: Forward-port to recent versions of Python that choke on \s in strings. Corrected sume buggy translations of string replace methods, Corrected some bugs in library-function translation. gitlab.com/esr/pytogo

RE: The Way of War of Our Enemies In every hot war the United States has become involved in since the Korean War, we have enjoyed absolute tactical and operational dominance over our enemies. We win every tactical engagement, overwhelmingly. Operationally we can and do dominate any theater of our choosing. No one—and I mean NO ONE—can stand toe to toe with the US military. This has been true for decades. We’ve talked before about the elements of national power—the “DIME” (Diplomacy, Informational, Military, Economic). Our military power is unsurpassed. We are masters of diplomacy. We have the world’s strongest economy. So how do we lose? The INFORMATIONAL component. Our military opponents, from Ho Chi Minh to Osama bin Ladin, knew that the only way to defeat the USA is to demoralize the American populace such that it demands withdrawal and throws the then current Commander-in-Chief out of office. The ONLY way to defeat America militarily is to convince the American people that a war is unwinnable. The slow dribble of IED deaths in OIF was not actually targeting soldiers and Marines—it was targeting YOU, the American people. And CNN eagerly complied with death counts running across the bottom of the screen. The Tet Offensive? It was a decisive US victory that could have ended the Vietnam War in our favor. But Walter Cronkite instead declared the war lost, protests erupted nationwide, and the war was lost. The Highway of Death in Kuwait? We could have taken out Saddam Hussein in 1991 and never needed to go back in 2003, but international media made the attack on retreating Iraqis look “too cruel,” so we halted just short of the finish line. The strategic imperative of every one of America’s military enemies is to break the will of the American people with skewed information, propaganda, and extreme emphasis on America’s minor losses amidst overwhelming military victory. But the Ho Chi Minhs and Osama bin Ladins can’t do that by themselves. They need willing partners in the American media and government. And for Operation Epic Fury, boy oh boy do the Iranian mullahs have an over abundance of American morale killers to draw from in order to defeat America through the informational instrument of national power. Tucker Carlson. Senator Mark Kelly and the rest of the Seditious Six. CNN. ABC. NBC. CBS. NYT, WaPo. Pakistani bot armies on social media. X “influencers” like Cerno, Candace, MartyrMade and Ian Carroll. Every idiot claiming we are fighting “Israel’s war." There is an entire Army of American politicians and media figures who are willingly fighting Iran’s informational war on its behalf (and in some cases, at its behest). America is DECISIVELY WINNING the war on Iran in every measurable respect. Yet there are so many influential Americans who are desperately determined to make you believe otherwise. In days of old in non-US countries, such people would have been strung up for treason. Thankfully it’s 2026 and we have a First Amendment, so no one fear being treated in such a medieval manner. But we can still ostracize and ridicule such people and sources for the irreparable harm they are wreaking upon the USA as they do the bidding (intentionally or unintentionally) of Theo-fascist mullahs who are determined to set off a nuclear bomb so that the Twelfth Imam will arise from a well in Qom and precipitate the global apocalypse. We all need to choose sides. Are you with America, or are you with theologic-inspired, deliberate Armageddon? And anyone who chooses the latter needs to be the target of mockery, derision and clearly-stated facts disproving their lies. And if YOU are an American Patriot, you can fight that informational war on America’s behalf, right now, right here on social media, right there in your own living room. Your voice matters, and your voice is actually a part of the war. FIGHT FIGHT FIGHT.

@Efroymson Alas, even AI is powerless in the face of my typos. Dammit.

@esrtweet Hadened and Repqaired! I see you are trusting your AI tools well, sir

Announcing mad release 1.3 An implementation of Michigan Algorithmic Decoder, and its manual. This program is a work of reconstructive archeology, a compiler for the Michigan Algorithmic Decoder (MAD) language from 1959-1962. It first translates MAD code into C, then uses the local C compiler to make an executable binary. It includes both a complete transcription of the original MAD manual and 100dpi scans of the original. New in this release: Hadened the code with ChatGPT 5.2. Repqaired the PLY loading procedure Added PEP484 type hints. gitlab.com/esr/mad

@JackSla07094857 @WilliamWolfe Not intentional. PHP is a sieve.

@WilliamWolfe heh, that was predicted more or less by @esrtweet in 2006, in his essay, Gramscian Damage, based upon what was happening in Europe at the time. P.S. there are a bunch of hidden porn-links embedded in the essay at one spot, not sure if that's intentional: esr.ibiblio.org/?p=260

Woke, for as bad as it is, is just a blip on the radar of the long 20th century. The future of America will not be secular, pluralistic, or progressive. It will either be a 1) Renewed Christian nation or 2) a Muslim nation. Someone’s morality WILL rule.

@cathyr19355 @planefag Vegetable oil? Who's talking vegetable oil? The canonical pasty-faced Celto-Teutonic thing to fry with is butter, of course.

I was trained by an ethnic mother and grandmother so I can't look at meat without mentally adding carrots and turnips and onions and potatoes. Like a Valheim dev I guess. Just troll me up fam

No, it doesn't mean that. Shipper is an orchestrator; it doesn't actually do anything itself, it knows how to make other programs do things and generates a script that you can pipe to your shell. The things it requires for normal operation are mature and stable, like...scp. Igor only gets involved if you want the specific optional feature of pushing release assets to GitLab. In theory, shipper could do the asset push itself. But I prefer it to have it not directly execute commands that modify the state of the world. Instead, it gives you an executable recipe.

@esrtweet I thought the rule was "succeed quietly; fail noisily". And if shipper relies on alpha software, that makes shipper itself alpha software.

It was all Igor's fault! That's my story and I'm sticking to it! I succeeded in getting shipper to ship itself. The source of the problem turned out to be that I had delegated part of the problem to Igor, my hunchbacked assistant. No, actually, gitlab-igor is a little program @ianbruene wrote to use the gitlab API to push tarballs and other assets to a project's releases area. Why did he name it Igor? " Let's just say that one of Ian's major qualifications for becoming my apprentice was a sufficiently geeky sense of humor. "Walk this way..." Poor Igor is alpha software. Good concept, but it needs to be noisier when it fails. It had a bad interaction with one of shipper's assumptions - which is my fault, really, I should have realized that sanity check I put in last week would fire spuriously when Igor was on the job. Ian and I need to pile on and make Igor howl. When he hits an error condition, that is. Sometimes, failing too quietly is a bug.

Announcing shipper release 1.22 Automated shipping of open-source project releases shipper is a power distribution tool for developers with multiple projects who do frequent releases. It automates the tedious process of shipping a software release and (if desired) templating a project web page. It can deliver releases in correct form to SourceForge, Savannah. New in this release: Code hardening with ChatGPT 5.2. Get release-notification addresses from repology.org. Debian-specific packaging data is obsolete and removed. Also, generate packaging-status URLs into templated web pages. Added -t option for generating tweet texts. It was all Igor's fault! That's my story and I'm sticking to it! Explanatory tweet to follow... gitlab.com/esr/shipper

I have a hilarious bug this morning. Shipper won't ship. Let me explain. You may have noticed that, with the help of robot friends, I've been maintaining a pretty steady pace of a software release every day for over a month. I'm on a mission to retire all of my technical debt. I started with 61 projects; my punch list is now down to 28. If you ever ever cut a software release yourself, you will know that this involves a fair amount of process friction. You have to do validation testing, set a release tag, push a release tag to the repo, make tarballs, update a project web page, mail notifications to package maintainers who you know will be interested, and perhaps ship an announcement to a friendly IRC channel. It's all very boring, and it's fiddly, and it's easy to miss a step and embarrass yourself. I have a low tolerance for process friction, and I don't like embarrassing myself, so I wrote a tool. It's called shipper. The most recent thing I taught it to do is automatically excavate the names of the package maintainers it needs to notify by looking up the project's package name on repology.org. Before that I had to maintain those address lists by hand, which was exactly the kind of annoyance I was trying to get away from. To use shipper, you set a bunch of things in a per project-control file that it knows about. Then your release process consists of putting the new version number near the top of your NEWS file, and typing "make release" in your project directory. Usually this works like a charm. I mean, yesterday I shipped morse, and the day before that it was sng, and the day before that it was Super Star Trek, I think. This morning, I tried to ship shipper, but shipper wouldn't ship. I got a message about "no deliverables with versions", which means shipper thinks that there's something screwed up somewhere in a Makefile or the project control file. I'll figure it out. And ship. Until I do, shipper has become the barber that does not shave itself. Probably its next move will be to warn me that all Cretans are liars.

@mroe1492 File a bug.

@esrtweet Two languages that loccount doesn’t yet support: Bluespec System Verilog (hardware description language); HOL4 (theorem prover).

Announcing loccount release 2.17 Your system package manager probably knows this as 'loccount' Count source lines of code (SLOC) and logical lines of code (LLOC). This is a faster implementation (in Go) of David A. Wheeler's sloccount tool. It can count LLOC as well as SLOC and handles a wider spread of languages, as well. New in this release: Code hardening by ChatGPT 5.2 New -t option describes languages matching a prefix. Skip unreadable files; correctly classify files with 0 SLOC Language names can now have embedded spaces. More exclusion extensions. Ignore files with an embedded NUL in the first 1K. Improved behavior of Go files with syntax errors. Suppress following symlinks in code trees. Fixed incorrect occam file extension. 507 languages and markups supported. From 164 languages last release to 507 in this one. Thank you, robot friends. gitlab.com/esr/loccount

Your system package manager probably knows this as 'morse' A morse-code training program and QSO generator. Morse Classic is a Morse-code training program for aspiring radio hams. It can generate random tests or simulated QSOs resembling those used in the ARRL test (a QSO generator is included). There are a plethora of options to vary the training method. In one of the simpler modes, this program will take text from standard input and render it as Morse-code beeps. New in this release: Code hardering, UB/buffer-overflow prevention by ChatGPT 5.2. Documentation fully spellchecked. gitlab.com/esr/morse-clas…

@theCTwelve @CDoombeard @klara_sjo Not much to tell. I am allergic, or strongly averse to, something in fermented cheese. So I enjoy a local variant of the Philly delicacy called a mushroom steak.

@esrtweet @CDoombeard @klara_sjo Now I do find it interesting that you 1) live in/around Philly and also 2) shun the Cheese on a meat-laden sammich. I suspect your opinions of the Cheesesteak may border on heresy. Please. Tell us more.

It's just occurred to me that it's common to see bacon put in a cheeseburger, but never have I seen a sausage added into the mix. I shall be running experiments and reporting back in time. Keep your phones on and never forget the valuable work I do.

If you live in the Philly/South-Jersey area something close to this thing is normal street food - my version differs mostly by using better-quality sausage, and they wouldn't be tempted to add Chinese hot oil. The Italian sandwich tradition is really vigorous here and has many tasty offshoots. A curious thing about Italian sausage sandwiches is that though pizzerias are prone to slather cheese on almost everything else they serve (including the famous Philly steak sandwich) sausage sandwiches are somehow immune. Tradition, tradition!

@esrtweet @klara_sjo Hmmm. I feel we are very close to designing the ideal sausage sandwich. I should take this from the whiteboard to the lab. I will report results when possible.

@CDoombeard @klara_sjo Food of the gods!

@esrtweet @klara_sjo Oh my, this is some advanced fatassery, I approve Any sausage-shaped meat object can be treated this way. There are better sausages out there than bratwurst, I'll agree. How do you feel about carmelized onions?

Joke's on you. The telecoms people don't talk about it, but they went to IPv6 internally almost completely, years ago - major carriers like T-Mobile and AT &T and Verizon are more than three quarters IPv6 at this point, with a sharp gradient correlating with the age of their networking gear. They did this because IPv4 address exhaustion was a real problem and they were staring down both barrels of it. This doesn't mean Paul Ehrlich wasn't an evil idiot - he was both things. It means the two cases you're analogizing aren't nearly as parallel as you think they are.

Back in the 1990s, early Internet nerds were convinced we were running out of IPv4 addresses and that we needed to create a new standard, IPv6, to prevent this. They used the same arguments as Paul Ehrlich, exponential growth meeting a fixed supply, leading to devastating collapse. They did this even after Ehrlich was proven completely wrong in 1990. This argument is a higher level Truth that can never be disproven, no matter how many times its predictions fail to come about. Over 30 billion devices are connected to the IPv4 Internet and we sill haven't run out of IPv4 addresses. Of those are 8.5 billion mobile phones on the IPv4 Internet. Yet, when people explain IPv6 to you, the first thing they'll tell you is that it's needed to prevent the Internet from running out of addresses.