Steve Troughton-SmithHow about Apple rebrand itself with a new CEO?
| github | https://github.com/michaelengel |
| web | https://multicores.org |
Michael Engel
@me_@sueden.social
- 489 Followers
- 217 Following
- 9.1K Posts
34 years of Unix experience, it's still fun!
Quinn “The Eskimo!”Apple’s API Review team is hiring.
https://jobs.apple.com/en-us/details/200605934/api-review-design-engineer
The job description says “you’ll have a unique opportunity to help shape the technologies that inspire developers” and that’s no exaggeration.
Werawelt@tagesschau
"...hat Dobrindt wörtlich gesagt: "Wir haben heute Entscheidungen getroffen zur Reduzierung der illegalen Migration." Und das stimmt nicht. Was das Kabinett heute abgeschafft hat, ist eins der wenigen kontrollierten Instrumente regulärer Migration, die Deutschland überhaupt hat....Ein Innenminister, der diese handverlesenen 1.000 Menschen im Monat mal eben zu illegalen Migranten macht, trägt damit verbal zur Spaltung der Gesellschaft bei."
Ein Kommentar von Bianca Schwarz, ARD-Hauptstadtstudio
Danke Frau Schwarz.
Humphrey Archer@sarahjeong.bsky.social
Indeed
Techbro billionaires have an astonishing ability to invent things that already exist, solve problems that don't, and redo things we did before, just at more waste and with less quality than what we did before..
Indeed
Techbro billionaires have an astonishing ability to invent things that already exist, solve problems that don't, and redo things we did before, just at more waste and with less quality than what we did before..
xlengineer1For USA:
Tim Carroll - VP Digital Business, DigiKey" In light of recent tariffs, our team has adapted the website, adding a filter that allows customers to search for tariffed and non-tariffed products"
David Chisnall (*Now with 50% more sarcasm!*)The first major brand to use the slogan 'No AI. No bullshit.' is going to make so much money.
CGdoppelpunktHabeck war fleißig. Er hat niemanden als faul bezeichnet.
Bei Merz und Linnemann ist das anders.
Nick Stevens GraphicsSo, lets see if I got this right.
Saturn V, 13 launches, 100% success rate.
SpaceX Starship, 9 launches, 0% success rate.
Seems to me that Nazi rocket engineers are not what they used to be...
Adrian SampsonThe final-project blog posts for my grad compilers course, #cs6120, are trickling out, and some of them are pretty fun! A couple of examples:
Simon Bertron turned quantifier elimination into a source-to-source optimization for Python programs with `any` and `all`: https://www.cs.cornell.edu/courses/cs6120/2025sp/blog/imperative-quantifier-elimination/
Ethan Uppal & Zihan Li got some speedups by parallelizing a dataflow solver: https://www.cs.cornell.edu/courses/cs6120/2025sp/blog/zihan-ethan-project/
Quantifier Elimination For Imperative Programs
The Background When people want to write mathematical statments formally, they make use of the quantifier symbols $\exists$ and $\forall$. Of course, many formally distinct mathematical statements are actually logically equivalent. For example, the following two logical formulas are equivalent (if $x$ is a real number): $$ \exists x: a x^2 + b x + c = 0 $$ $$ b^2 - 4ac \geq 0 $$ It is worth noting briefly that the two formulas are only equivalent when $x$ is restricted to be a real number. If we allowed $x$ to be any complex number (or further restricted $x$ to be an integer!) the two formulas would no longer be equivalent. We say that the formulas are equivalent in the theory of real numbers. For our purposes, a theory is a formal language together with some axioms defining which formulas constructed from that language are true. When a formula can be written with and without quantifiers, the quantifier-free version has some advantages from a computational perspective. The main advantage in the context of formal mathematical statements is that in most theories we care about, it is possible to describe an algorithm that decides the truth of any quantifier-free formula in that theory. This is not generally true of quantifier-ful formulas. If we squint and pretend our two formulas are computer programs, the quantifier-free version looks like a few arithmetic operations while the quantifier-ful version looks an awful lot like a loop that might never halt! This suggests an algorithm for deciding the truth of a formula. First convert the formula to a quantifier-free form, then apply the universal-quantifier-free-deciding algorithm to the result. This is possible in some theories which have known procedures for removing quantifiers from any formula in that theory and, in fact, the Z3 SMT solver has a quantifier elimination tactic which attempts to apply such procedures. Of course, quantifier elimination is not performing any decidability magic. Theories which admit quantifier elimination are fully decidable even by algorithms that do not perform quantifier elimination. So one reason Z3 has a quantifier elimination tactic (and not some other deciding algorithm) is because it can make the deciding cheaper. Even if our “loopy” pseudo-program above eventually terminates, it might take a long time. It would be easier to do the arithmetic instead. The Project This brings us to the task at hand. If quantifier elimination can speed up SMT solving, maybe it can speed up regular old computer programs. As we saw above, true quantifiers generally translate to loops that may never terminate. In most programming circles, writing loops that you know ahead of time might never terminate is generally frowned upon, so we will need more limited substitutes. We will use the commonly available any and all functions as our restricted quantifiers. any returns true if and only if any of its inputs is true while all returns true if and only if all of its inputs are true (get it?). This makes any an existential quantifier ($\exists$) and all a universal quantifier ($\forall$). We will also restrict ourselves to a single theory: Presburger Integer Arithmetic. Formulas in Presburger Arithmetic are built from the following language: $[ \neg, \wedge, \vee, \exists, \forall ] \cup \mathbb{N} \cup [ +,-,<,= ]$. Where the first three symbols are the standard first order predicate logic not, and, and or. We will operate on Python programs because Python has a few nice properties:
Erik Uden 🍑EU plant neues Massenüberwachungsgesetz mit Vorratsdatenspeicherung & Zwangs-Backdoors – jetzt ist noch Widerspruch möglich!
Hey Leute,
die EU plant gerade ein neues Gesetz zur Vorratsdatenspeicherung, das echt krass werden könnte. Es geht nicht nur um ein paar Verbindungsdaten, sondern darum, jeden Online-Dienst zur Überwachung zu verpflichten – also auch Messenger, Hosting-Anbieter, Webseiten usw.
Das Ganze läuft unter dem Titel:
„Retention of data by service providers for criminal proceedings“
Hier kann man bis zum 18. Juni 2025 Feedback abgeben: Have Your Say
Was ist geplant?
- Pflicht zur Datenspeicherung mit Identitätsbindung – also alles, was du online machst, muss auf dich zurückführbar sein.
- Sanktionen für Dienste, die keine Nutzerüberwachung einbauen – darunter könnten auch VPNs, selbstgehostete Sachen oder Open-Source-Projekte fallen.
- Backdoors in Geräte und Software – mit Hardwareherstellern soll zusammengearbeitet werden, um „gesetzlichen Zugriff“ zu ermöglichen.
- Auch kleine Anbieter betroffen – es geht ausdrücklich nicht nur um Meta, Google & Co.
Das Ganze basiert auf Empfehlungen einer „High Level Group“, deren Mitglieder komplett geheim gehalten werden. Patrick Breyer (Piraten/MEP) @echo_pbreyer hat nachgefragt – die EU hat ihm eine geschwärzte Liste geschickt.
Laut EDRi wurde die Zivilgesellschaft explizit ausgeschlossen. Lobbyismus deluxe.
Was kann man tun?
Einfach Feedback abgeben, geht in 2 Minuten.
Kurz schreiben, dass man gegen anlasslose Vorratsdatenspeicherung und Überwachung ist, reicht schon. Jeder Kommentar zählt.
Deadline ist der 18. Juni 2025, Mitternacht (Brüsseler Zeit).
Wäre gut, wenn wir aus der IT-Szene da nicht still bleiben. Das betrifft wirklich alle – Entwickler, Admins, SysOps, Hoster, ganz normale Nutzer.