feat(courses): other courses under every course and post, footer advert, utm on shares, sign-in after Telegram

apps/website/i18n/course-ladder.ru.json

@@ -1,5 +1,5 @@
 {
-  "SAY_LADDER": "Лестница: первые три курса бесплатные, четвёртый и все следующие открываются с клубом. Какие три — выбираете вы.",
+  "SAY_LADDER": "Лестница: первые три курса бесплатные и без регистрации, четвёртый и все следующие открываются с клубом. Какие три — выбираете вы.",
   "SAY_RUNG": "Ступень {0} вашей лестницы",
   "SAY_FREE_TITLE": "Этот курс для вас бесплатный",
   "SAY_FREE_LINE": "Откройте одним нажатием. Ничего не спрашиваем, из браузера ничего не уходит.",
@@ -19,7 +19,7 @@
   "SAY_SHARED": "Ваши пройденные уроки видит автор курса.",
   "SAY_COURSES": "Все курсы",
   "SAY_CATALOG_TITLE": "Курсы",
-  "SAY_CATALOG_LEAD": "Каждый курс — 27 коротких уроков в браузере, в каждом живой виджет или спека, которую можно запустить. Любые три — бесплатно, остальные открывает клуб.",
+  "SAY_CATALOG_LEAD": "Каждый курс — 27 коротких уроков в браузере, в каждом живой виджет или спека, которую можно запустить. Любые три — бесплатно, без регистрации и почты; остальные открывает клуб.",
   "SAY_FREE_LEFT": "Осталось бесплатных курсов: {0} из 3",
   "SAY_FREE_NONE": "Три бесплатных курса выбраны. Все остальные открываются с клубом.",
   "SAY_IN_CLUB": "Вы в клубе: открыты все курсы.",
@@ -41,5 +41,11 @@
   "SAY_PAY_WAITING": "Ждём подтверждения оплаты...",
   "SAY_PAY_SLOW": "Оплата ещё не подтверждена. Если вы оплатили, страница продолжает проверять; можно обновить её позже.",
   "SAY_PAY_DONE": "Оплата получена. Добро пожаловать в клуб.",
-  "SAY_PAY_OR_BOT": "Или оплатить звёздами Telegram в нашем боте"
+  "SAY_PAY_OR_BOT": "Или оплатить звёздами Telegram в нашем боте",
+  "SAY_MORE_TITLE": "Другие курсы",
+  "SAY_PROMO_TITLE": "Три курса бесплатно, без регистрации",
+  "SAY_PROMO_LINE": "Каждый урок идёт прямо в браузере, с живым виджетом или спецификацией, которую можно запустить. Любые три курса бесплатно, без регистрации.",
+  "SAY_PROMO_PRESS": "Смотреть курсы",
+  "SAY_SIGN_IN": "Уже в клубе? Войти через Telegram",
+  "SAY_SIGNED_IN": "Вы вошли: ключ клуба сохранён в этом браузере."
 }

apps/website/public/learn/a-bar-detector/index.html

@@ -123,5 +123,6 @@ <h1>A bar detector</h1>
 <section><h2>All lessons</h2><ul class="outline"><li><h3>Module 1 · Three values</h3><p>Why three, how balanced ternary writes every number without a sign, and what flipping and cutting trits do.</p><ol><li><a href="/learn/why-three/">01 Why three</a></li><li><a href="/learn/balanced-ternary/">02 Balanced ternary</a></li><li><a href="/learn/flip-and-cut/">03 Flip to negate, cut to round</a></li></ol></li><li><h3>Module 2 · Logic with unknown</h3><p>Kleene's three-valued gates, two gates binary has no twin for, and addition as a pair of tables.</p><ol><li><a href="/learn/kleene-logic/">04 Three-valued logic</a></li><li><a href="/learn/consensus-and-any/">05 Consensus and any</a></li><li><a href="/learn/sum-and-carry/">06 Sum and carry</a></li></ol></li><li><h3>Module 3 · Adding trits</h3><p>The half adder, the full adder and a carry that ripples left, one place at a time.</p><ol><li><a href="/learn/half-adder/">07 The half adder</a></li><li><a href="/learn/full-adder/">08 The full adder</a></li><li><a href="/learn/ripple-carry/">09 Ripple carry</a></li></ol></li><li><h3>Module 4 · Multiplying without a multiplier</h3><p>Copy, drop or flip: a product by one trit, long multiplication, and a MAC that only adds.</p><ol><li><a href="/learn/times-a-trit/">10 Times one trit</a></li><li><a href="/learn/long-multiplication/">11 Long multiplication</a></li><li><a href="/learn/ternary-mac/">12 A ternary MAC</a></li></ol></li><li><h3>Module 5 · Trits in binary memory</h3><p>Two bits per trit, five trits per byte, and the bits a 27-trit word needs.</p><ol><li><a href="/learn/two-bits-per-trit/">13 Two bits per trit</a></li><li><a href="/learn/five-trits-a-byte/">14 Five trits in a byte</a></li><li><a href="/learn/a-word-of-27/">15 A word of 27 trits</a></li></ol></li><li><h3>Module 6 · The TRI-27 instruction word</h3><p>The 32-bit word the Trinity emulator decodes: its fields, its 47 opcodes and its 15-bit immediate.</p><ol><li><a href="/learn/the-instruction-word/">16 The instruction word</a></li><li><a href="/learn/the-opcode-map/">17 The opcode map</a></li><li><a href="/learn/the-immediate/">18 The immediate</a></li></ol></li><li><h3>Module 7 · A ternary machine</h3><p>An ALU built from this course's adders, a three-way jump, and a program you can step.</p><ol><li><a href="/learn/a-ternary-alu/">19 A ternary ALU</a></li><li><a href="/learn/three-way-jump/">20 A three-way jump</a></li><li><a href="/learn/run-a-program/">21 Run a program</a></li></ol></li><li><h3>Module 8 · Three answers</h3><p>Compare at the highest differing trit, find a number in thirds, sort with three-way compares.</p><ol><li><a href="/learn/compare-at-the-top/">22 Compare at the top</a></li><li><a href="/learn/questions-in-thirds/">23 Questions in thirds</a></li><li><a href="/learn/sort-by-three/">24 Sort with three-way compares</a></li></ol></li><li><h3>Module 9 · Ternary neurons</h3><p>Weights of -1, 0, +1: one neuron, a detector and a small layer, with no multiplier anywhere.</p><ol><li><a href="/learn/one-ternary-neuron/">25 One ternary neuron</a></li><li class="on" aria-current="page"><a href="/learn/a-bar-detector/">26 A bar detector</a></li><li><a href="/learn/a-tiny-layer/">27 A tiny layer</a></li></ol></li></ul></section>
 <footer><p><a href="/learn/ternary-computing.t27">This course is itself a t27 spec: read it</a></p><p>Widgets keep their own English words: each one lives in its own t27 spec.</p></footer>
 </main>
+<script>(function(){var q=location.search;if(!/[?&]utm_source=/.test(q))return;document.querySelectorAll('a[href^="/#/"]').forEach(function(a){a.setAttribute('href','/'+q+a.getAttribute('href').slice(1))})})()</script>
 </body>
 </html>

apps/website/public/learn/a-bend-at-zero/index.html

@@ -94,5 +94,6 @@ <h1>A bend at zero</h1>
 <section><h2>All lessons</h2><ul class="outline"><li><h3>Module 1 · Lab: our own research</h3><p>A number format of our own, an honest scoreboard, and a model's tables multiplied on the board.</p><ol><li><a href="/learn/t27-number-format/">01 T27: a number format of our own</a></li><li><a href="/learn/an-honest-scoreboard/">02 An honest scoreboard</a></li><li><a href="/learn/a-model-on-the-board/">03 A model's tables on the board</a></li></ol></li><li><h3>Module 2 · AI numbers: the MX block</h3><p>How AI chips keep weights in a few bits: one shared scale per block, the scale byte itself, and what one outlier does to its neighbours.</p><ol><li><a href="/learn/one-scale-per-block/">04 One scale for a block of weights</a></li><li><a href="/learn/scale-byte-native/">05 The scale byte on a real machine</a></li><li><a href="/learn/one-outlier/">06 One outlier, many zeros</a></li></ol></li><li><h3>Module 3 · Ternary weights</h3><p>Weights that are only minus, zero or plus a scale, the five rules a ternary alphabet must pass, and a test pass that checked nothing.</p><ol><li><a href="/learn/phi-scaled-ternary/">07 Weights of minus phi, zero and plus phi</a></li><li><a href="/learn/a-pass-that-checked-nothing/">08 A pass that checked nothing</a></li><li><a href="/learn/five-rules-for-weights/">09 Five rules for a weight alphabet</a></li></ol></li><li><h3>Module 4 · The Ternary Network Float</h3><p>A rule the compiler enforces before any test runs, and a 17-bit float whose exponent is four balanced trits.</p><ol><li><a href="/learn/blocked-is-not-failed/">10 Blocked is not failed</a></li><li><a href="/learn/four-trit-exponent/">11 A float whose exponent is four trits</a></li><li><a href="/learn/negate-twice/">12 Negate twice, get the number back</a></li></ol></li><li><h3>Module 5 · Arithmetic on signed numbers</h3><p>Multiply two signed numbers, add them when their signs differ, and do both at once in a multiply-accumulate.</p><ol><li><a href="/learn/a-sign-is-one-bit/">13 A sign is one bit</a></li><li><a href="/learn/opposite-signs-subtract/">14 Opposite signs subtract</a></li><li><a href="/learn/multiply-then-add/">15 Multiply, then add</a></li></ol></li><li><h3>Module 6 · Parts of a neuron</h3><p>A ReLU that bends at zero, a power of two for softmax, and an argmax that names the answer.</p><ol><li class="on" aria-current="page"><a href="/learn/a-bend-at-zero/">16 A bend at zero</a></li><li><a href="/learn/two-to-the-x/">17 Two to the x</a></li><li><a href="/learn/pick-the-largest/">18 Pick the largest</a></li></ol></li><li><h3>Module 7 · Learning from a mistake</h3><p>A loss that prices a wrong guess in bits, one step that moves a weight against its gradient, and the hidden layer that XOR needs.</p><ol><li><a href="/learn/a-loss-in-bits/">19 A loss in bits</a></li><li><a href="/learn/one-step-downhill/">20 One step downhill</a></li><li><a href="/learn/xor-needs-a-bend/">21 XOR needs a bend</a></li></ol></li><li><h3>Module 8 · BitNet: ternary networks</h3><p>A threshold that squeezes a sum back to three values, one neuron that becomes a different function when its weights change, and a neuron that reads its inputs 27 trits at a time.</p><ol><li><a href="/learn/back-to-three-values/">22 Back to three values</a></li><li><a href="/learn/same-neuron-new-weights/">23 Same neuron, new weights</a></li><li><a href="/learn/a-neuron-in-chunks/">24 A neuron in chunks</a></li></ol></li><li><h3>Module 9 · The ternary MAC as a chip</h3><p>The 27-trit dot product as wires with no register, the same sum added into a register on every clock, and a small whole network to close the course.</p><ol><li><a href="/learn/a-dot-product-in-wires/">25 A dot product made of wires</a></li><li><a href="/learn/add-it-up-every-clock/">26 Add it up on every clock</a></li><li><a href="/learn/a-whole-network/">27 A whole network</a></li></ol></li></ul></section>
 <footer><p><a href="/learn/ai-numbers.t27">This course is itself a t27 spec: read it</a></p><p>Widgets keep their own English words: each one lives in its own t27 spec.</p></footer>
 </main>
+<script>(function(){var q=location.search;if(!/[?&]utm_source=/.test(q))return;document.querySelectorAll('a[href^="/#/"]').forEach(function(a){a.setAttribute('href','/'+q+a.getAttribute('href').slice(1))})})()</script>
 </body>
 </html>

apps/website/public/learn/a-build-receipt/index.html

@@ -170,5 +170,6 @@ <h1>A build, itemised</h1>
 <section><h2>All lessons</h2><ul class="outline"><li><h3>Module 1 · The chip</h3><p>What an FPGA is, which chip we use, and how its pins meet the board.</p><ol><li><a href="/learn/fpga-is-a-grid/">01 An FPGA is a grid you configure</a></li><li><a href="/learn/which-chip/">02 Which chip is this?</a></li><li><a href="/learn/pins-and-package/">03 Pins, balls and the package</a></li></ol></li><li><h3>Module 2 · Numbers in hardware</h3><p>Bits, trits and number formats, and what arithmetic costs in logic.</p><ol><li><a href="/learn/bits-and-trits/">04 Bits, trits and GF16</a></li><li><a href="/learn/what-arithmetic-costs/">05 What arithmetic costs</a></li><li><a href="/learn/how-much-fits/">06 How much fits on the chip</a></li></ol></li><li><h3>Module 3 · Your t27 program</h3><p>Write a spec, test it, and see why compiling is not the same as being right.</p><ol><li><a href="/learn/hello-t27/">07 Hello, t27</a></li><li><a href="/learn/tests-are-the-spec/">08 Tests are part of the spec</a></li><li><a href="/learn/compiles-is-not-correct/">09 Compiles is not the same as correct</a></li></ol></li><li><h3>Module 4 · Inside t27c</h3><p>How the compiler reads a spec and what it writes, including the native t27b.</p><ol><li><a href="/learn/compiler-stages/">10 The stages of t27c</a></li><li><a href="/learn/t27c-on-a-real-machine/">11 t27c on a real machine</a></li><li><a href="/learn/seven-backends-and-t27b/">12 Seven backends, and t27b</a></li></ol></li><li><h3>Module 5 · From spec to hardware</h3><p>The Verilog t27c writes, the cells it becomes, and what one LUT does.</p><ol><li><a href="/learn/spec-to-verilog/">13 From a spec to Verilog</a></li><li><a href="/learn/count-the-cells/">14 Count the cells</a></li><li><a href="/learn/one-lut-up-close/">15 One LUT, up close</a></li></ol></li><li><h3>Module 6 · Reading synthesis</h3><p>What yosys reports about your design, and which warnings matter.</p><ol><li><a href="/learn/synthesis-warnings/">16 Synthesis warnings that matter</a></li><li><a href="/learn/state-machines/">17 State machines</a></li><li><a href="/learn/where-the-luts-go/">18 Where the LUTs go</a></li></ol></li><li><h3>Module 7 · Place, route, timing</h3><p>Where the cells land on the die, and whether the clock is met.</p><ol><li><a href="/learn/placement/">19 Placement: where the cells land</a></li><li><a href="/learn/logic-depth/">20 Logic depth</a></li><li><a href="/learn/timing-slack/">21 Timing and slack</a></li></ol></li><li><h3>Module 8 · The bitstream</h3><p>How a routed design becomes the bits the chip loads, with no Vivado.</p><ol><li><a href="/learn/fasm-to-frames/">22 From FASM to frames</a></li><li class="on" aria-current="page"><a href="/learn/a-build-receipt/">23 A build, itemised</a></li><li><a href="/learn/read-a-bitstream/">24 Read a bitstream</a></li></ol></li><li><h3>Module 9 · On the board</h3><p>Ask the chip who it is, load the bits, and check them.</p><ol><li><a href="/learn/jtag-asks-the-chip/">25 JTAG asks the chip</a></li><li><a href="/learn/same-bits-on-the-board/">26 The same bits, on the board</a></li><li><a href="/learn/capstone/">27 Capstone: out-guess yosys</a></li></ol></li></ul></section>
 <footer><p><a href="/learn/course.t27">This course is itself a t27 spec: read it</a></p><p>Widgets keep their own English words: each one lives in its own t27 spec.</p></footer>
 </main>
+<script>(function(){var q=location.search;if(!/[?&]utm_source=/.test(q))return;document.querySelectorAll('a[href^="/#/"]').forEach(function(a){a.setAttribute('href','/'+q+a.getAttribute('href').slice(1))})})()</script>
 </body>
 </html>

apps/website/public/learn/a-correlator-finds-the-frame/index.html

@@ -284,5 +284,6 @@ <h1>A correlator finds the frame</h1>
 <section><h2>All lessons</h2><ul class="outline"><li><h3>Module 1 · Adding numbers</h3><p>Why addition is slow: the carry that walks up the word, the dedicated chain in every slice, prefix networks and carry-save trees.</p><ol><li><a href="/learn/the-carry-chain/">01 The carry chain</a></li><li><a href="/learn/prefix-adders/">02 Looking ahead: prefix adders</a></li><li><a href="/learn/carry-save/">03 Three numbers at once</a></li></ol></li><li><h3>Module 2 · Multiplying</h3><p>A product is a sum of shifted copies; Booth recoding halves them, and the DSP48E1 slice does the rest in one block.</p><ol><li><a href="/learn/shift-and-add/">04 Shift and add</a></li><li><a href="/learn/booths-trick/">05 Booth's trick</a></li><li><a href="/learn/the-dsp48e1-slice/">06 The DSP48E1 slice</a></li></ol></li><li><h3>Module 3 · Fixed point</h3><p>Where the binary point sits, what rounding does to a value and to its average, and what each bit of a quantizer buys.</p><ol><li><a href="/learn/q-formats/">07 Q formats</a></li><li><a href="/learn/rounding-and-saturation/">08 Rounding and saturation</a></li><li><a href="/learn/quantization-noise/">09 Quantization noise</a></li></ol></li><li><h3>Module 4 · Functions in hardware</h3><p>Sine, cosine, angle and length from shifts and adds, and when a table is the better answer.</p><ol><li><a href="/learn/rotating-by-shifts/">10 Rotating by shifts</a></li><li><a href="/learn/angle-and-magnitude/">11 Angle and magnitude</a></li><li><a href="/learn/tables-and-interpolation/">12 Tables and interpolation</a></li></ol></li><li><h3>Module 5 · Signals and sampling</h3><p>What sampling does to a frequency, how an oscillator is built from an adder, and what a DFT bin measures.</p><ol><li><a href="/learn/sampling-and-aliasing/">13 Sampling and aliasing</a></li><li><a href="/learn/the-nco/">14 A numerically controlled oscillator</a></li><li><a href="/learn/bins-of-a-spectrum/">15 Bins of a spectrum</a></li></ol></li><li><h3>Module 6 · FIR filters</h3><p>The moving average, a windowed-sinc design with integer taps, and folding the taps onto DSP slices.</p><ol><li><a href="/learn/the-moving-average/">16 The moving average</a></li><li><a href="/learn/designing-taps/">17 Designing taps</a></li><li><a href="/learn/folding-taps-onto-slices/">18 Folding taps onto slices</a></li></ol></li><li><h3>Module 7 · Multirate</h3><p>Lowering the sample rate without folding noise in: decimation, the CIC filter and the polyphase form.</p><ol><li><a href="/learn/lowering-the-rate/">19 Lowering the rate</a></li><li><a href="/learn/the-cic-filter/">20 The CIC filter</a></li><li><a href="/learn/polyphase/">21 Polyphase</a></li></ol></li><li><h3>Module 8 · The FFT</h3><p>N log N instead of N^2: butterflies, the bit-reversed input order and the bits each stage adds.</p><ol><li><a href="/learn/the-butterfly/">22 The butterfly</a></li><li><a href="/learn/twiddles-and-bit-reversal/">23 Twiddles and bit reversal</a></li><li><a href="/learn/growth-through-the-stages/">24 Growth through the stages</a></li></ol></li><li><h3>Module 9 · On the bench</h3><p>From the arithmetic to the board: a correlator from the modem spec, a budget of slices and timing, and the capstone filter.</p><ol><li class="on" aria-current="page"><a href="/learn/a-correlator-finds-the-frame/">25 A correlator finds the frame</a></li><li><a href="/learn/does-it-fit-and-close-timing/">26 Does it fit and close timing?</a></li><li><a href="/learn/capstone-a-filter-on-the-board/">27 Capstone: a filter on the board</a></li></ol></li></ul></section>
 <footer><p><a href="/learn/arithmetic-and-dsp.t27">This course is itself a t27 spec: read it</a></p><p>Widgets keep their own English words: each one lives in its own t27 spec.</p></footer>
 </main>
+<script>(function(){var q=location.search;if(!/[?&]utm_source=/.test(q))return;document.querySelectorAll('a[href^="/#/"]').forEach(function(a){a.setAttribute('href','/'+q+a.getAttribute('href').slice(1))})})()</script>
 </body>
 </html>