The heat-affected zone is the region of a material whose microstructure or properties are changed by nearby welding or thermal cutting without itself melting. It lies outside the fusion region and can contain different subregions corresponding to different thermal histories.^[ref]

In a steel, heating and cooling can change phases, grain size and hardness; in another alloy the thermal cycle may soften a previously strengthened region. Heat input, travel speed, geometry and thermal transport determine the affected region together. Measuring a weld therefore requires examining both the joint and adjacent material. Surface heat-tint colours alone do not uniquely identify the extent of the metallurgically altered region.^[ref]

A WebAssembly module is the unit of deployment, loading and compilation in WebAssembly. It collects declarations such as functions, types and memories, can request imports, and can expose exports. Optional initialization includes data segments and a start function. A WebAssembly module is represented in a binary encoding or the readable WebAssembly text format.^[ref]

Definitions use zero-based indices, with a separate index space for each kind of definition. Imported functions precede locally defined functions in the function index space. For example, with one imported function and two local functions, their function indices are respectively 0, 1 and 2; memory index 0 belongs to a different space. This explains why inserting a function import can change a numeric call target.^[ref]

A WebAssembly module instance is the runtime realization of these declarations, with references to the allocated or imported objects. Keeping a WebAssembly module distinct from a WebAssembly module instance separates a code description from the state used while executing it. Imported state can be shared; the mere use of the same module does not specify all its runtime resources.^[ref]

The WebAssembly text format is the readable textual encoding of WebAssembly modules. Parenthesized declarations describe the same structured program that can be encoded in binary. Symbolic identifiers such as $add help readers refer to definitions; numeric indices determine references in the underlying module.^[ref]

For example, (module (func (export "add") (param i32 i32) (result i32) local.get 0 local.get 1 i32.add)) declares and exports a function accepting two 32-bit integers. The local.get instructions push its arguments onto the operand stack; i32.add consumes them and produces the result. Inputs 3 and 4 give 7. The folded instruction (i32.add (local.get 0) (local.get 1)) is an alternative notation for the same instruction sequence.^[ref]

The parentheses of a folded instruction make operand grouping visible, but do not introduce a different arithmetic operation. Decoding either spelling yields the same instruction structure. Textual readability and runtime behavior are separate: indentation can help a human follow the WebAssembly module, while the declared types and instruction order govern execution.

WebAssembly linear memory is an indexed sequence of bytes accessed by WebAssembly memory instructions. Its runtime representation records its type and current byte sequence. The length is a multiple of the WebAssembly page size, $65536$ bytes, or $64\mathrm{KiB}$. Declared limits give an initial minimum and optionally a maximum size in pages.^[ref]

A load must fit its entire access into the current WebAssembly linear memory. For an instruction offset of zero, a four-byte i32.load in a one-page memory succeeds at address $65532$: its last accessed byte has address $65535$. At address $65533$, it would need a byte at address $65536$ and therefore traps. In general, effective address $a$ and width $w$ must satisfy $a+w\le L$, where $L$ is the current length; the check includes the instruction's immediate offset.^[ref]

This boundary example also distinguishes alignment from bounds. Ordinary unaligned loads are permitted; their alignment annotation is an optimization hint. A successful bounds check establishes that the accessed bytes lie inside this memory. It does not establish that the program selected its intended object.^[ref]

A WebAssembly module instance is the runtime realization of a WebAssembly module. It collects references to the functions, memories and other entities that the WebAssembly module imports, defines or exports. The WebAssembly store contains the allocated runtime objects; a module's local indices are resolved through the corresponding WebAssembly module instance.^[ref]

For example, (module (memory (export "memory") 1)) declares and exports one page of WebAssembly linear memory. Instantiate this WebAssembly module twice. Each instantiation allocates its own declared memory, so writing17 into byte0 through the first WebAssembly module instance leaves byte0 in the second at zero. Code can be reused while mutable state remains distinct.^[ref]

The import case supplies a useful counterexample. If a WebAssembly module imports memory and two instantiations receive the same memory object, writes through one are visible through the other. The WebAssembly module instances are distinct but that resource is shared. Instantiation therefore links imported entities as well as allocating local definitions; it also initializes active segments and invokes the start function when one is present.^[ref]

A WebAssembly store is the abstract runtime state containing allocated entities such as function, memory, table and global instances. A WebAssembly module instance refers to these entities through WebAssembly abstract addresses. A store memory address selects a memory instance; it is distinct from a byte offset within WebAssembly linear memory.^[ref]

For an original toy model, let the store contain two memory instances at abstract addresses 0 and 1. Module instances A and B both use local memory index 0, but A maps it to address 0 and B maps it to address 1. A third instance C imports the first memory and maps its own local index 0 to address 0.

Initially, let byte offset 0 in both memories contain zero. Writing 17 through A changes the first memory. Reading offset 0 through B gives zero, while reading it through C gives 17. There are three distinct coordinates: the module-local index, the store address selecting a memory and the offset selecting a byte inside it. Equal local index numbers do not establish sharing.

These invented addresses illustrate the semantics; they are not exposed machine addresses measured from an engine. The WebAssembly store also is not one linear memory or a persistent file store. The abstract model records allocations, while implementations may reclaim unreachable entities when doing so is not semantically observable.^[ref]

The HITS algorithm assigns two related scores to pages in a selected link graph: an authority score for being linked to by strong hubs, and a hub score for linking to strong authorities. Jon Kleinberg formulated this mutual reinforcement to identify useful sources for broad-topic web queries.^[ref]

Let the adjacency matrix satisfy $A_{ij}=1$ when page $i$ links to page $j$. With authority vector $a$ and hub vector $h$, alternatenormalizing after each update. Repeated authority updates involve $A^{\mathsf{T}}A$, connecting the procedure to power iteration and eigenvectors. If pages one and two each link only to page three, the third page becomes the authority and the first two equally strong hubs. A high score describes this link structure; it is not independent verification of the page's claims.

A web crawler automatically discovers and fetches web resources, often to supply a web search engine. Starting from known addresses, it can extract links and schedule newly discovered addresses for later fetching. Discovery, fetching and indexing are distinct: a fetched page need not be accepted into the searchable collection.^[ref]

In a toy web, page A links to B and C, while B links back to A. Recording previously encountered addresses prevents that cycle from causing endless rediscovery. A pending-work list determines what to fetch next; revisits are still useful when pages change. Real web crawlers also control request rates and handle unavailable resources. This makes crawling a continuing resource-allocation task rather than a single traversal that permanently captures the entire web.

A water-equivalent material is a material whose relevant interaction with ionizing radiation approximates that of water under specified conditions. For a water phantom, the comparison concerns energy absorption and scattering in the measurement geometry. It is not a statement that the material merely has a density near that of water, or that equivalence at one energy applies at every other energy.^[ref]

For a narrow monoenergetic photon beam, the linear attenuation coefficient satisfieswhere $\rho$ is the mass density and $\mu /\rho$ the mass attenuation coefficient. Equal densities leave the mass interaction coefficients unconstrained. Predicting absorbed dose also involves how transferred energy is deposited, for which the mass energy-absorption coefficient is relevant; the attenuation of primary photons alone is not the complete scattering and deposition problem. For charged particles, the corresponding comparison also involves stopping power.

For an original toy calculation, stipulate equal densities but linear attenuation coefficients ${\mu }_w=0.05\mathrm{c}{\mathrm{m}}^{-1}$ and ${\mu }_m=0.08\mathrm{c}{\mathrm{m}}^{-1}$ at one chosen photon energy. For $t=10\mathrm{cm}$, the unscattered fractions in the Beer-Lambert law areThe substitute transmits only $e^{-0.3}\approx 0.74082$ times the unscattered intensity of the reference, despite identical density. These coefficients are stipulated model inputs, not tabulated data for real water or a named plastic. Matching even this transmission at one energy would not establish equal absorbed dose throughout a larger scattering object. Energy dependence of water equivalence is therefore part of any meaningful equivalence specification.

A focused physical comparison places the same characterized detector in reference water and in a candidate slab assembly, with controlled material thickness, irradiation geometry and detector position. Record density and composition as well as detector response over the intended radiation conditions. Slab gaps and changes in surrounding scattering material can confound the comparison; the primary IAEA discussion explicitly treats material variability and the approximate nature of plastic-to-water scaling.^[ref]

Placental matrotrophy is nourishment of developing offspring by transfer of maternal material across a placenta. It supplements or replaces reliance on reserves packaged in the ovum, such as yolk. Viviparity specifies retention of developing offspring and live birth; placental matrotrophy additionally specifies a route of nutrient supply.

A striking example comes from the lizard described as Mabuya heathi in a 1984 study. Its very small ovum could not account for the large increase in offspring dry mass; the authors estimated that placental transfer supplied more than 99 percent of the near-term fetus's dry mass. Examination of the placenta linked the growth measurements with an anatomical route of exchange.^[ref]

Dry mass matters because uptake of water alone can greatly increase wet mass. In a synthetic comparison, an initial 1 mg of dry mass and a final 10 mg imply a net addition of 9 mg of dry material. Measuring that increase supports transfer beyond the original reserves; determining the particular nutrients and transport mechanisms requires further measurements. The lizard example also makes placental matrotrophy a comparative character across vertebrates, rather than a synonym for being a mammal.

Vitamin B3 is the nutritional category including nicotinic acid and nicotinamide, compounds the body can use to make nicotinamide adenine dinucleotide and nicotinamide adenine dinucleotide phosphate. These coenzymes participate in metabolism, including transfer of reducing equivalents between reactions. Adequate dietary supply supports a reusable biochemical system rather than providing the bulk of the body's fuel.^[ref]

For example, an oxidized coenzyme can accept reducing equivalents in one reaction and transfer them in another. The coenzyme is regenerated during the cycle, so its importance is not measured by how much of it is burned for energy. Humans can also derive niacin equivalents from tryptophan; severe inadequacy is associated with pellagra.^[ref]

Vitamin K is a family of fat-soluble compounds with a shared quinone framework. It includes phylloquinone, conventionally called vitaminK1, and menaquinones, grouped as vitaminK2. The family supports an enzymatic modification required for several biologically active proteins, including proteins involved in coagulation.^[ref]

An important role is supporting conversion of selected glutamate residues in proteins into gamma-carboxyglutamate residues. The protein sequence and the completed chemical modification are different kinds of information. Producing a polypeptide of the expected length does not establish that all its required modifications have occurred.^[ref]

The chemistry became concrete in a 1974 prothrombin study: researchers isolated a peptide from its calcium-binding region and used nuclear magnetic resonance and mass spectrometry to identify modified glutamic acid. This linked vitamin dependence to a molecular structure that could be measured.^[ref]

For a mechanistic example, consider two preparations containing the same amount of a vitamin-K-dependent protein. If one is insufficiently modified, equal protein abundance need not mean equal activity in an assay requiring the completed binding region. The example distinguishes nutrient-supported chemical maturation from simply supplying more structural protein. VitaminK is involved in several physiological systems; the name should not be reduced to a synonym for a single clotting factor.

Vitamin B6 is the collective name for six related compounds with vitamin-B6 activity. They comprise pyridoxine, pyridoxal and pyridoxamine, together with their corresponding 5′-phosphate forms. The phosphorylated forms include pyridoxal phosphate, an active coenzyme. A family name and a particular molecular species therefore describe different levels of chemical identity.^[ref]

An enzymatic example makes the distinction concrete. 1-aminocyclopropane-1-carboxylate synthase requires pyridoxal phosphate when converting S-adenosylmethionine into the precursor of the plant signal ethene. Naming the broader Vitamin B6 family does not establish that the required coenzyme form is present at an enzyme's active site.^[ref]

For an analytical comparison, measuring total members of the family and measuring only pyridoxal phosphate answer different questions. Equal total chemical amounts can conceal different distributions among forms, and equal coenzyme concentration does not establish equal enzyme abundance. Separating those quantities helps interpret a biochemical assay without reducing the vitamin's role to the amount of one named precursor.

Vitamin E is a family of fat-soluble vitamins with antioxidant activities, rather than the name of one chemically unique molecule. The naturally occurring forms comprise alpha-, beta-, gamma- and delta-tocopherols and the corresponding four tocotrienols. Alpha-tocopherol is the form recognized as meeting human vitamin-E requirements; the forms differ in biological activity.^[ref]

The family distinction matters analytically. Measuring total compounds grouped under the vitamin name, measuring alpha-tocopherol specifically and measuring radical-trapping antioxidant activity answer different questions. The NIH fact sheet describes preferential hepatic resecretion of alpha-tocopherol, so equal amounts of different vitamin-E forms are not automatically equivalent biological inputs.^[ref]

A useful chemical model concerns lipid peroxidation. An antioxidant hydrogen donor can intercept a chain-carrying peroxyl radical:Here $\mathrm{TOH}$ denotes a tocopherol hydrogen donor and $\mathrm{T}{\mathrm{O}}^{\bullet }$ its resulting radical. Interception competes with attack on another lipid molecule. The biological effect depends on location, reaction rates and the fate of the resulting products, not only on a bulk vitamin concentration.

A historically informative experiment compared two independent observables. In 1983, researchers measured lipid-soluble chain-breaking antioxidant concentrations in human plasma and erythrocyte ghosts using inhibited autoxidation, then compared those estimates with vitamin E measured by high-performance liquid chromatography. The primary abstract reports closely similar concentrations in those studied samples.^[ref] Functional chemical activity and separated chemical composition therefore supplied complementary evidence. The abstract does not establish a universal assay-to-health-outcome conversion.

An original idealized inhibition-time calculation explains one possible activity readout. Let radicals be introduced at constant rate $R_i$, let initial antioxidant concentration be $C_A$, and suppose each antioxidant molecule intercepts $n$ radicals before being exhausted, with negligible competing losses. ThenFor stipulated $C_A=10\mu \mathrm{mol}{\mathrm{L}}^{-1}$, $n=2$ and $R_i=2\mu \mathrm{mol}{\mathrm{L}}^{-1}\mathrm{mi}{\mathrm{n}}^{-1}$, the ideal lag is ten minutes. This is a model example, not a reconstruction of the 1983 measurements or a universal stoichiometry for every vitamin-E form.

The interpretation needs a calibrated radical-production rate, reaction controls and a specified medium. Other antioxidants or interfering chemistry can change the apparent lag. Conversely, chromatography identifies and quantifies chosen chemical species only within its separation and calibration model. Keep these chemical measurements distinct from nutritional requirements and clinical effects; none of the calculations here assigns a supplementation regimen.

A head-mounted display places a visual display and viewing optics on the user's head. In an immersive virtual reality system, head tracking supplies the viewing pose and the renderer updates the scene accordingly. Separate eye views can provide binocular depth cues; motion-dependent views also change as the head translates or rotates.

For example, when a viewer leans sideways, a nearby virtual post should shift against a distant wall by more than the wall shifts. This requires tracking position as well as orientation. The two eye views also need distinct view transforms. WebXR exposes poses and per-view projections so an application can construct these images, while the device handles its particular display arrangement.^[ref]

WebXR is the Web interface for accessing tracked views and input associated with virtual reality and augmented-reality devices. A session supplies animation frames; the application obtains the viewer pose relative to a reference space and renders the views described for that frame. A head-mounted display commonly supplies separate views for the two eyes.^[ref]

For example, a virtual cube fixed one metre in front of the initial viewpoint must be transformed differently as the viewer turns or moves. Updating only the picture's projection while ignoring the tracked position makes the cube move incorrectly. Reference spaces give these poses a defined coordinate meaning; they are part of the application's spatial model rather than an interchangeable screen size.