Diagnostics

Generated from dk.simongray.podcast.diagnostics by dev/reference.clj, so don't edit it by hand.

A diagnostic is a map with an :id, a :level and a :message, where the message is the meaning below unless the code gives a more specific one. Other keys give details, such as a :path into the data or the :tag of an element. Use report for the diagnostics of a parsed feed, transcript or chapters document, validate for those of a feed, and summarize to group diagnostics by kind.

The level says how much a diagnostic matters:

• error: a rule of a spec or of the chosen requirements is broken, so apps or a directory can misread the feed, or refuse it
• warning: the feed works, but apps are likely to show or treat something worse than the podcaster means
• info: something worth knowing that needs no change, e.g. a repair that the parse made

The requirements of a directory can raise a level, e.g. a repeated enclosure URL is an error for Apple. An error under Apple's requirements breaks a rule of its guide, though Apple Podcasts Connect doesn't say which rules it refuses a feed for. To tell whether a feed passes, give the diagnostics of validate/report to diagnostics/verdict.

The namespace of an id says which part of podcast-clj found the problem:

• fetch: the fetch of a feed or a list
• discovery: finding the feeds of a show from an address
• repair: the repairs of malformed XML
• xml: the parse and namespace resolution
• quirk: the normalisations of the tree
• feed: the model
• transcript and chapters: the companion formats
• mp3, mp4 and file: the media files and the images
• id3: the ID3v2 tags of media files, through id3-clj
• validate: the validator
• check: the checks over the network
• sync: the sync between devices, through like-minded
• gpodder: a sync through the gpodder API
• library: a saved library
• podping: a poll of the Hive chain
• comments: the threads of comments
• value: the records of payments, the remote items of time splits, the Lightning addresses and the wallets
• analytics: the list of user agents, and the downloads of OP3

A function that sends requests also returns its failures as diagnostics, under the :error of its result.

Id	Level	Meaning
:fetch/unreachable	warning	The feed or list couldn't be fetched: the name of its host wasn't found, the host couldn't be reached, the secure connection failed, the request took too long, the answer was too large, or the URL wasn't allowed. The :type of wary-fetch says which, when the platform tells.
:refresh/offline	info	The device seems to be offline: :online? said so, or every fetch of the refresh failed before a server answered, at more than one host. No feed counts it as a failure, and each stays due, so the next refresh fetches it.
:refresh/stopped	info	The :signal of the refresh stopped it, e.g. as the listener cancelled it. The feeds whose fetch it stopped or kept from starting count no failure, and stay due.
:fetch/malformed	warning	The document isn't an RSS feed, or for a list, an OPML document: it isn't XML, or its root element isn't rss or opml.
:fetch/unsupported-format	warning	The document is an Atom or RSS 1.0 feed rather than RSS 2.0, so it can't be read. The :format names which.
:fetch/page-failed	warning	A page of a paged feed couldn't be fetched or read, so the feed lacks the items of that page and the pages after it.
:fetch/status	warning	The server answered with a status that isn't a success, such as 404. The feed stays as it was, and the next fetch waits longer after each failure in a row.
:fetch/empty-feed	warning	The feed came back without items, though it had some before, as some hosts answer during an outage. The feed stays as it was, and the next fetch waits as after a failure.
:fetch/gone	warning	The feed is gone: its server answered 410, or it failed many times in a row, as the :failures say. It isn't checked any more, unless a Podping or a refresh of that feed asks for it, and a fetch that succeeds brings it back.
:fetch/move-unverified	warning	The feed declares that it moved to the :new-url, but no feed of the show answered there, as the :reason says. So the feed stays at its old URL, and the move is tried again at the next fetch.
:fetch/guid-taken	warning	The feed declares the podcast:guid of another subscription, at another URL, and would take that subscription over, so it isn't added. To replace the other subscription, unsubscribe from it first.
:fetch/guid-changed	info	The feed gives itself another guid than before, from :old to :new, e.g. a new podcast:guid, or a new URL of a feed that declares none. Its subscription keeps its key, the :feed-guid, and notes the new guid.
:discovery/unreadable	warning	The text isn't a URL, a domain or a link of an app, so there's nothing to fetch. It may be words to search a directory with.
:discovery/not-found	warning	No feed was found: the page names none that could be read, and none of the other places that were tried has one.
:directory/source-failed	warning	A directory gave no answer to a search, a chart, a lookup or a call for the names of its categories, so what it knows is missing from the results. The :source names it, and the :reason says why.
:repair/leading-junk	warning	Text before the XML declaration or the rss element, such as whitespace, a byte order mark or a PHP warning, was removed.
:repair/trailing-junk	warning	Text after the end of the rss element, such as a PHP warning, was removed.
:repair/control-characters	warning	Control characters that XML forbids, and character references to them, were removed.
:repair/undeclared-prefixes	warning	Known namespace prefixes that were used without a declaration were declared.
:repair/html-entities	warning	HTML entities that XML lacks, such as  , were replaced by character references.
:repair/bare-ampersands	warning	Bare ampersands were escaped.
:repair/lenient-parser	warning	The document was still malformed after the repairs, so it was parsed leniently, with the parser that the app installed or with jsoup.
:xml/skipped-entity	warning	An entity that only an external DTD could declare was left out of the text, since no external DTD is read. The :entity names it.
:xml/default-namespace	info	A default namespace isn't known, so the elements in it have no alias, and its declaration stays as an xmlns attribute.
:xml/unknown-namespace	info	The namespace URI isn't known, so its prefix is used as the alias.
:xml/undeclared-prefix	warning	A prefix was used without a declaration, which only the JVM parser accepts, so the prefix is used as the alias.
:quirk/namespace-variant	info	A namespace was bound to a variant of its URI, such as http for https, another case, a trailing slash or an old URI, and was recognised anyway.
:quirk/rss-default-namespace	info	The rss element declared a default namespace, such as that of Atom, though RSS 2.0 has none, so it was ignored.
:quirk/itunes-image-text	info	An itunes:image gave its URL as text or in a url attribute instead of href.
:quirk/explicit-value	info	An itunes:explicit said yes, no, explicit or clean, which was read as true or false.
:quirk/yes-no-value	info	A yes/no element said true or false, which was read as yes or no.
:quirk/transcript-type	info	A transcript type was written loosely, as srt, vtt, json or text, or with a parameter such as a charset, and was corrected to its MIME type.
:quirk/attribute-name	info	An href attribute was used where the spec asks for url, and was renamed.
:quirk/enclosure-length	info	An enclosure length that wasn't a number was dropped.
:quirk/empty-guid	info	An empty guid was dropped, so the identity of the item is its enclosure URL.
:quirk/deprecated-images	info	A deprecated podcast:images srcset was turned into podcast:image elements.
:quirk/item-image	info	An item had an image element with only attributes, which RSS 2.0 lacks, and it was read as the itunes:image that it stands for.
:quirk/podcast-category	info	A channel had the non-standard podcast:category elements that Wavlake writes, and no itunes:category, so they were read as itunes:category, whose shape they copy.
:quirk/duration-format	info	An itunes:duration was written as a clock time rather than in seconds, and was read as seconds.
:quirk/date-format	info	A date wasn't in strict RFC 2822 form, but could still be read.
:feed/duplicate-element	warning	An element that the spec allows once occurred again, and only the first one is kept. The :tag names the element, and the :count is the number of the others.
:feed/empty-element	info	An element with a single value had no text, and was left out.
:feed/empty-attribute	info	An optional attribute was blank, and is left out of the model. A required one that's blank is kept for the validator.
:feed/invalid-value	warning	A value didn't read as its type, and stays a string.
:feed/string-value	info	A feed built from data held a string where the model takes another type, e.g. a date, and it was read as that type, as parse reads the text of XML. The :value is the string, and the :type the type.
:feed/markup-in-text	info	Markup appeared where the registry expects text. It's kept as markup in HTML content, and flattened to text elsewhere.
:feed/unexpected-text	info	Text inside an element that the spec leaves empty was dropped.
:feed/dropped-attributes	info	Attributes that the model has no place for were dropped: those of an element read as a single value, those of rss other than its version, and those of channel.
:transcript/speaker-prefixes	info	Speakers were taken from Name: prefixes in the cues, and carried over to the cues that follow.
:transcript/voices-carried	info	A WebVTT file gives a voice span only where the speaker changes, so the cues without one took the speaker before them. The :count is the number of such cues.
:transcript/cues-dropped	warning	Cues of a WebVTT file have timings that don't parse, e.g. 1000 milliseconds, so they were left out, as a browser leaves them out. The :count is the number of such cues.
:transcript/srt-tags	info	SRT cues held the formatting tags of HTML, e.g. for italics or a font, which SRT doesn't define, and the text was read without them. The :count is the number of such cues.
:transcript/string-times	warning	Segment times were given as strings rather than numbers, and were read as numbers.
:transcript/wrapped-srt	warning	An HTML transcript was an SRT file with every line wrapped in an element, and was read as SRT.
:transcript/unknown-type	warning	The declared type isn't one that the spec names, so the format was detected from the file instead.
:transcript/type-mismatch	warning	The declared type doesn't match the file, which was read as the format that it really is.
:transcript/not-an-object	warning	A JSON transcript isn't JSON, or isn't a JSON object, so it holds no segments.
:transcript/malformed	warning	Segments of a JSON transcript that aren't JSON objects in a list were left out.
:transcript/not-text	warning	Speakers in a JSON transcript that aren't text, such as a JSON object, were left out.
:chapters/string-times	warning	Chapter times were given as strings rather than numbers, and were read as numbers.
:chapters/not-text	warning	Chapter titles, links or place names that aren't text, such as a JSON object, were left out.
:chapters/missing-start-time	warning	Chapters without a start time can't be placed.
:chapters/unsorted	info	The chapters weren't in order of start time, which the format assumes.
:chapters/unknown-keys	info	Keys that the format doesn't define were kept as they are.
:chapters/old-version	info	The document was written for an older version of the format.
:chapters/bad-timestamps	warning	Podlove chapters whose start isn't a timestamp were dropped.
:chapters/inferred	info	The chapters were read from timestamps in the show notes rather than from a chapters file.
:chapters/not-an-object	warning	The document isn't JSON, or isn't a JSON object, so it holds no chapters.
:chapters/malformed	warning	Chapters that aren't JSON objects in a list were left out.
:mp3/junk-before-audio	info	Bytes that aren't MPEG audio frames lie between the tags and the first frame, e.g. padding that the size of a tag leaves out, and were skipped.
:mp3/estimated-duration	info	The MP3 file has no Xing, Info or VBRI header, so the duration was estimated from the bitrate of the first frames and the size of the file. The estimate is exact for a constant bitrate.
:mp3/no-audio	warning	No MPEG audio frame was found after the tags, so the file has no duration.
:mp4/zero-duration	info	The movie header of the MP4 file gives no duration, so the duration of the longest track was used.
:mp4/fragments-unread	info	The MP4 file is in fragments, with no mehd box and no index at its end that gives the duration of all of them, so the file has no duration.
:mp4/no-movie	warning	The MP4 file has no movie box, so it gives no duration and no chapters.
:mp4/chapter-images-skipped	info	The images of the chapters of the MP4 file are larger than a reading takes, so the chapters have no images.
:ogg/unknown-codec	warning	The Ogg file has no Opus or Vorbis stream, e.g. it holds FLAC or Speex, so nothing was read from it.
:ogg/no-audio	warning	The Opus or Vorbis stream has no audio after its headers, so the file has no duration.
:ogg/no-end	info	No page of the audio stream ends a packet in the last 64 KB of the file, e.g. because other streams follow it, so the file has no duration.
:ogg/chapter-start	warning	A CHAPTERxxx comment doesn't give the start of its chapter as a time, such as 00:05:00.000, so the chapter was left out.
:ogg/image-type	info	A METADATA_BLOCK_PICTURE gave the type of its picture loosely, such as JPG or image/jpg, or gave none, so the type was taken from the bytes of the image.
:aac/estimated-duration	info	The duration of the AAC file was estimated from the average length of its first frames and the size of the file, since the whole file wasn't read.
:aac/no-audio	warning	No ADTS frame was found after the tags, so the file has no duration.
:file/unknown-format	warning	The file isn't in a format that its reader knows, so nothing was read from it. Media files can be MP3, MP4, Ogg with Opus or Vorbis, or AAC in ADTS frames, and images PNG, JPEG, GIF or WebP, while AVIF, HEIC, HEIF, SVG, BMP, TIFF and ICO images are only recognised by their type.
:file/truncated	warning	The file ends before its structures say it does.
:file/malformed	warning	The file doesn't read as its format says.
:file/duration	info	The duration of the media file or stream differs from the one that the feed declares by more than 10 seconds and 2 percent. A host that inserts ads can explain it, but then a position in this file may not be the same moment in another file of the episode, such as its video.
:file/incomplete	warning	The bytes that the reader needed didn't arrive, or a limit stopped the reading, so nothing was read.
:file/unreachable	warning	The file couldn't be fetched: the request failed, or the server answered with a status that isn't a success.
:file/unverifiable	info	This podcast:integrity can't be checked. A PGP signature needs the key of the publisher, and an sri value must name a sha256, sha384 or sha512 hash.
:check/unreachable	error	The file couldn't be fetched: the name of its host wasn't found, the host couldn't be reached, the TLS handshake failed, or the request took too long. The :type of wary-fetch says which, when the platform tells.
:check/status	error	The server answered with a status that isn't a success, so a listener doesn't get the file.
:check/downgrade	error	A redirect leads from https to http, which clients refuse to follow.
:check/redirects	warning	The file is more than five redirects away, more than the check and some clients follow.
:check/moved	warning	The file moved for good, with a 301 or 308 redirect, so the feed can name its new URL.
:check/head	error	The server doesn't answer a HEAD request for the file. RFC 9110 requires HEAD of every general-purpose server, and directories use it to check files.
:check/type	warning	The server gives the file a different media type than the feed declares.
:check/media-type	error	The chosen requirements don't take a media file of the :type that the server gives it, e.g. YouTube, which makes its videos from audio files.
:check/no-reader	info	A reading of the file was asked for, but the media reader knows no format of the :type that the feed declares, e.g. audio/x-wav, so its duration and codecs weren't checked.
:check/no-type	warning	The feed declares no media type for the file, which RSS 2.0 and the podcast namespace require. The :served type is the one that the server gives it.
:check/length	info	The size of the file differs from the length that the feed declares. A host that inserts ads sends a file of a different size to each request, so the length can't be exact.
:check/ranges	warning	The server doesn't answer a request for a byte range with that range. RFC 9110 makes ranges optional but recommends them, players need them to seek and resume, and Apple Podcasts requires them.
:check/format	warning	The media file is in a different format than the type that the feed declares.
:check/duration	info	The duration of the media file differs from the one that the feed declares by more than 10 seconds and 2 percent, which a host that inserts ads can explain.
:check/slow-start	warning	The movie of the MP4 file comes after its media data, so a player must fetch the end of the file before it can play the start. Writing the movie first fixes it.
:check/codecs	warning	The codecs that the alternate enclosure declares are those of neither its file nor any variant of its stream, so an app that asks the platform whether it can play them gets the wrong answer.
:check/height	warning	The height that the alternate enclosure declares is that of neither its video nor any variant of its stream.
:check/unread	warning	The file arrived, but its content wasn't read, because a request for more of its bytes failed or its reader failed.
:check/integrity	error	The media file doesn't match the hash of its podcast:integrity.
:check/unverified	info	The media file wasn't compared with its podcast:integrity, because a PGP signature needs the key of the publisher, the integrity names no hash that Subresource Integrity knows, or the whole file didn't arrive.
:check/image-type	info	The server gives the image a different media type than its bytes have, such as image/jpg for a JPEG. Browsers go by the bytes of an image.
:check/image-size	warning	The size of the image differs from the width, height or aspect ratio that its podcast:image declares.
:check/artwork-size	error	The image of an itunes:image isn't a square of the size that the chosen requirements ask for, e.g. 1400 to 3000 pixels for Apple.
:check/artwork-type	error	The image of an itunes:image isn't of a type that the chosen requirements ask for, e.g. JPEG or PNG for Apple, or it's a JPEG that apps and browsers can't decode: lossless, hierarchical, arithmetic-coded, JPEG-LS, or with more than 8 bits per sample.
:check/artwork-color	error	The image of an itunes:image isn't in RGB colours, as the chosen requirements ask.
:check/artwork-alpha	error	The image of an itunes:image has an alpha channel or transparency, which the chosen requirements don't allow.
:check/artwork-extension	warning	The file extension in the URL of an itunes:image doesn't match the type of its image, as the chosen requirements ask.
:check/last-modified	warning	The server answers a HEAD request for the artwork without Last-Modified, which the chosen requirements ask for.
:check/cors	warning	The server doesn't let pages on other sites read the file, so web players can't show it. The Access-Control-Allow-Origin header of the answer, or of a redirect on the way, decides this, and the podcast namespace requires CORS for transcripts.
:check/empty	error	The transcript or chapters file holds no segments or chapters in the format that the feed declares.
:check/feed-cors	info	The server doesn't let pages on other sites read the feed, so a web app can't read it without a server of its own.
:check/validators	info	The server sends the feed without ETag or Last-Modified, so a client can't ask whether it changed, and has to fetch it whole every time.
:check/feed-type	warning	The server sends the feed as a media type that isn't one of XML, the :served, e.g. text/html, so a browser can show it as a page and an app or a directory can refuse it. Serve it as application/rss+xml or application/xml.
:check/feed-size	warning	The feed is larger than the requirements take: its :bytes against their :max-bytes.
:check/unchecked	warning	The check reached a limit of files or time, and skipped the files after it, which the diagnostic counts.
:validate/rss-version	warning	The version of the rss element isn't 2.0. The :value is the one it has. A feed built from data has none, which emit writes.
:validate/missing-element	error	An element that RSS 2.0 requires is missing.
:validate/missing-enclosure	warning	An item has no enclosure, so there's nothing to play.
:validate/duplicate-guid	error	An item has the guid of an earlier item, so apps, directories and sync servers take the two for one episode. The :first-path is that of the earlier item.
:validate/duplicate-enclosure	warning	An item has the enclosure URL of an earlier item, e.g. a trailer and the episode it's cut from. Apple keeps only the first, so its requirements make it an error. The :first-path is that of the earlier item.
:validate/unknown-media-type	warning	The type of an enclosure is neither audio, video nor a playlist of either, e.g. application/octet-stream, so apps may not play it. The :value is the type.
:validate/unknown-length	info	The length of an enclosure is 0, which says that it isn't known, e.g. under dynamic ad insertion, so an app can't tell the size of the file before it downloads it.
:validate/unknown-category	warning	Apple Podcasts doesn't list the category of a podcast, so the show isn't listed under it. The :value is its name, and the :replacement is the path of the one that Apple lists in another case, if any. The requirements of Apple and PSP-1 make it an error.
:validate/renamed-category	warning	The category is one that Apple Podcasts renamed or dropped in 2019, which it still reads. The :replacement is the path of the category and perhaps the subcategory that stands for it now.
:validate/escaped-category	warning	The ampersand of a category's name was escaped twice, e.g. Society & Culture, so it names no category. The :replacement is the name meant. The requirements of Apple and PSP-1 make it an error.
:validate/wrong-subcategory	warning	The subcategory belongs under another category, or stands alone where a category belongs. The :replacement is the path where Apple Podcasts lists it. The requirements of Apple and PSP-1 make it an error.
:validate/category-depth	warning	A subcategory holds a third level, which Apple Podcasts doesn't take. The requirements of Apple and PSP-1 make it an error.
:validate/missing-attribute	error	A required attribute is missing.
:validate/wrong-shape	error	A value doesn't have the shape that its element takes: a bare value where a map of attributes belongs, or a map or sequence where a single value does.
:validate/missing-child	error	A container lacks a child that the spec requires.
:validate/invalid-structure	error	The children of a container contradict the spec, such as a time split with both recipients and a remote item.
:validate/invalid-value	error	A value isn't of its type, such as a string where a date belongs or :no where a boolean does, or it's a negative number, which no quantity in a feed can be. The :value is the value.
:validate/invalid-enum	warning	A value isn't one of those that the spec lists. The :value is the value.
:validate/invalid-url	warning	A URL isn't absolute. The :value is the URL.
:validate/http-url	info	A URL uses http where the podcast namespace requires https.
:validate/invalid-language	warning	A language isn't a language tag such as en or en-US, e.g. en_US, en-UK or english. The :value is the value, and the :reason says why: a reason from langtag-clj with its :subtag, e.g. :unregistered-subtag for a subtag that the IANA registry doesn't have, or :not-iso-639 for a language that isn't two or three letters.
:validate/invalid-email	info	An email address isn't one, alone or followed by a name in parentheses, e.g. an author given by name alone. The :value is the value.
:validate/invalid-media-type	warning	A media type isn't a type and a subtype such as audio/mpeg. The :value is the value.
:validate/invalid-guid	warning	A podcast:guid isn't a UUID, so the library, like the Podcast Index, identifies the feed by the guid derived from its URL. The :value is the guid.
:validate/unknown-role	info	The role of a person isn't in the Podcast Taxonomy. The :value is the role.
:validate/unknown-element	info	The registry doesn't know the element, so it's kept under :xml/unknown.
:validate/unknown-key	warning	The registry doesn't define the :key, e.g. the typo :rss/titel, or a key on the channel that isn't an element, since a parse keeps no attributes there.
:validate/deprecated	info	The spec of the element has deprecated it.
:validate/required	error	A required element of the chosen requirements is missing. The :key is its key in the model.
:validate/recommended	warning	A recommended element of the chosen requirements is missing. The :key is its key in the model. When every item lacks it, it's one diagnostic of all the items, whose :count is the number of items.
:gpodder/unknown-episodes	info	Episode actions from a gpodder server name episodes of feeds that aren't fetched yet. They wait in the cursor, and the next pull tries them again.
:gpodder/future-actions	warning	Episode actions from a gpodder server name a time more than a minute after the pull, as a device with a fast clock writes them, so they changed nothing.
:gpodder/malformed-actions	warning	Episode actions from a gpodder server name no feed, no media file or no readable time, and were left out.
:gpodder/unnamed-episodes	info	Episodes changed, but the server wasn't told, because the gpodder API names an episode by the URL of its media, and that URL isn't known.
:gpodder/rejected-url	warning	A gpodder server refused a URL that was sent to it, so it keeps no subscription or episode action for it.
:library/save-failed	warning	The library or its search index couldn't be written to the store in full. The :reason says why.
:library/unreadable-document	warning	A document of the saved library didn't read: library.edn, so none of its parts were restored, or the document of a feed, which is fetched again. The :name says which, and the :reason why.
:library/not-loaded	warning	The feeds that :feed-guids names are due, but the library holds them in their documents alone, and no :store was given to load them from, so they weren't fetched.
:library/newer-format	error	The saved library is of a newer version than this one reads, so none of it was restored, and a save won't write over it. The :version says which.
:library/malformed-profile	warning	The saved profile breaks the rules of a profile, so the library kept the profile it had, and the next sync brings back what the other devices hold. The :reason says what's wrong.
:library/index-rebuilt	info	The search index was built again from the feeds, since some of its documents were missing or didn't read. The :reason says why.
:library/push-not-applied	info	A WebSub hub pushed a feed that couldn't be applied, so the feed is due at once and the next refresh fetches it. The :reason says why, e.g. that the body isn't a feed, or that the library holds the feed only in its document.
:podping/unavailable	warning	No node of the Hive chain answered a call of the poll. The result has the pings of the blocks that were read before that call.
:podping/not-notified	warning	The service didn't take the word that a feed changed: the call failed, as the :reason says, or the service answered with the :status of a failure, e.g. 401 for a token that it doesn't know.
:comments/unsupported	info	Comments over the protocol of the thread can't be read.
:comments/no-appview	warning	An AT Protocol thread wasn't read, because the options named no AppView.
:comments/no-relay	warning	A Nostr thread wasn't read, because neither the options nor the root named a relay.
:comments/malformed	warning	The URI of the root post, or a post, a collection of replies, an author or a thread that a server sent, isn't in a form that can be read.
:comments/unreachable	warning	A post, a collection of replies or an author couldn't be fetched: the request failed, or the server answered with an error. With :retry?, the next fetch of the thread asks again.
:comments/refused	warning	A server refused an unsigned request, as a server in secure mode does.
:comments/untrusted	warning	A server sent a post or an author from another server, which wasn't used, since a server only speaks for its own posts and authors.
:comments/through-api	info	ActivityPub didn't give the root post, so the thread was read through the Mastodon API of the server of the root post, which can give fewer replies.
:comments/withheld	info	Some replies were left out, because they aren't public, their authors ask apps to show them only to signed-in users, or a moderation service hid them. An AppView also holds back a reply when its author and the author of the post above it block each other.
:comments/unverified	warning	Events from a Nostr relay were left out, and the diagnostic counts them. Such an event has a wrong id or signature, or isn't one that the request asked for.
:comments/truncated	warning	A size limit cut some replies, which the diagnostic counts. Its :limit names the limit: :page for the 1,000 items of a page of replies, :id for the 2,048 characters of an id, :thread for the 10,000 comments of a thread, or :depth for the 64 levels of replies of an AT Protocol thread.
:comments/incomplete	info	The fetch reached a limit before it read the whole thread, or a server didn't answer for a while. Fetch the thread again to read more.
:value/not-an-object	warning	The record isn't a JSON object, so it describes no payment.
:value/text-for-number	info	Numbers came as text, as some apps write them, and were read as numbers.
:value/rounded	info	Amounts in millisatoshis had a fraction, and were rounded.
:value/field-alias	info	Fields came under other names, as some apps write them, and were read under their own names.
:value/unreadable	warning	Some fields don't hold a value of their type, and were left out.
:value/guid-as-item-id	info	The itemID holds the guid of the episode, as Breez writes it, and was read as that guid.
:value/action-alias	info	The action came as streaming, as Podverse writes it, and was read as stream.
:value/unknown-action	warning	The action isn't boost, stream or auto, so the payment has no action.
:value/unknown-keys	info	The record has fields that its format doesn't define, which keep their names.
:value/no-podcast	warning	The record names no podcast: no guid, feed id, feed URL or title.
:value/claimed-amount	warning	The sender claims a different amount than the node received, so the payment keeps the amount of the node.
:value/claimed-recipient	warning	The sender claims that the payment went to a different recipient than the one that received it.
:value/no-metadata	warning	The comment names no URL of metadata, or the URL answered without the x-rss-payment header.
:value/custom-record	error	The records of the payment leave out the custom key and value of the recipient, because the key isn't a number, it's a type that the Lightning protocol reserves, or the value is missing. The payment would reach the node but not the wallet of the recipient, so don't send it.
:value/remote-unresolved	warning	The value block of the remote item of a time split couldn't be fetched, so only the fees of that split are paid until a later fetch finds it. The Podcast Index or the remote feed didn't answer, refused the request or answered with an error, or there were neither credentials for the index nor a feed URL to ask.
:nostr/signer-refused	warning	The signer didn't sign, e.g. because the listener said no, and :reason is what it said.
:nostr/signer-approval	info	The remote signer asks the listener to approve the request on the web page at :url first. Open the page, and ask again once the listener approved.
:nostr/signer-silent	warning	The remote signer didn't answer in time. The request went out, so the signer can still sign it.
:nostr/signer-unreachable	warning	No relay of the connection took the request, so it reached no signer, and :relays says why.
:nostr/signer-unreadable	error	The answer of the signer can't be read, or the signer signed another event than the one it was asked to, and :reason says which.
:nostr/request-too-large	warning	The request to the remote signer has more bytes than an encrypted payload of NIP-44 holds, so it wasn't sent.
:nostr/not-published	warning	No relay took the event, and :relays says why.
:value/no-recipients	error	The value block names no recipient, so outside its time splits a payment goes to nobody.
:value/fees-over	error	The fees of the recipients add up to more than 100 percent, so they take the whole of each payment, and the last of them less than it asks.
:value/no-shares	error	No recipient but those of the fees has a split, so after the fees a payment goes nowhere.
:value/same-recipient	warning	Two recipients have the same address and custom record, so one wallet gets both shares, which is likely a mistake.
:value/recipient-address	error	The address of the recipient doesn't have the form of its :type: the public key of a node, hexadecimal text of 33 bytes that starts with 02 or 03, or a Lightning address of the form user@domain on the public internet. No payment can reach it.
:value/empty-time-split	error	The time split names neither a remote item nor a recipient, so during it a payment goes to the fees alone.
:value/mixed-time-split	warning	The time split names a remote item and recipients too, which the podcast namespace doesn't allow, and an app pays the remote item.
:value/time-splits-overlap	warning	The time split starts before the one before it ends, so an app pays the recipients of the earlier one until it ends.
:value/no-method	warning	The app gave no way to pay a recipient of this :type, so it wasn't paid.
:value/not-an-address	warning	The text isn't a Lightning address of the form user@domain.
:value/unreachable	warning	The request failed, because the server didn't answer.
:value/lnurl-error	warning	The service of the Lightning address answered with an error, and said why.
:value/malformed	warning	The service of the Lightning address answered with something that LNURL-pay doesn't define.
:value/amount-out-of-range	warning	The Lightning address doesn't accept this amount.
:value/wrong-amount	error	The invoice asks for a different amount than the one requested, so don't pay it.
:value/unknown-success-action	error	The answer has a success action that LUD-09 doesn't define, or a URL on another host, so don't pay the invoice.
:value/unverifiable	info	The service of the Lightning address gave no URL to ask whether the invoice is paid, which LUD-21 defines.
:value/unproven	error	The service of the Lightning address says that the invoice is paid, but gives no preimage that proves it.
:value/no-keysend	info	The domain of the Lightning address names no node for keysend, so pay it with LNURL-pay.
:value/insecure	error	The answer came over plain http, where anyone on the way can change it, so don't pay with it.
:value/no-wallet-info	info	No relay of the connection has the info event of the wallet, so it isn't known which commands the wallet accepts. The wallet says so when it gets a command that it doesn't accept.
:value/wallet-unsupported	warning	The wallet doesn't accept the command: its :methods don't name the :method of the command, or its :encryption lacks NIP-44 version 2, which is what podcast-clj writes. No request was sent.
:value/command-too-large	warning	The command is larger than the 65,535 bytes that a request to a wallet can hold. The record of a payment includes the titles of the show and the episode, and the message of the listener. No request was sent.
:value/wallet-unreachable	warning	No relay of the connection took the request: each gave no connection, refused the subscription or refused the event, and :relays says why. The request reached no wallet, so it's safe to try again.
:value/wallet-silent	error	The wallet didn't answer in time. The request went out, so the wallet can still carry out the command until :expires-at. Don't send the payment again before then, and afterwards ask the wallet whether it paid.
:value/wallet-refused	warning	The wallet answered with an error and didn't carry out the command. The :code is the code of NIP-47, such as INSUFFICIENT_BALANCE, QUOTA_EXCEEDED or PAYMENT_FAILED, and :reason is the text of the wallet. A WebLN provider gives a text and no code, also when the user said no.
:value/wallet-unreadable	error	The answer can't be read, and :reason says why: it isn't the signed event of the wallet, it isn't encrypted for this connection, or it answers another request. The wallet may have carried out the command, so ask it before paying again.
:value/no-provider	info	The page has no WebLN provider, which a browser extension or a wallet app gives it.
:value/below-one-sat	warning	WebLN counts a keysend payment in whole sats, and this payment is below 1 sat, so nothing was sent.
:value/not-a-node	warning	The recipient isn't a node with a public key, or the payment has no amount, so keysend can't pay it and nothing was sent. Pay a Lightning address with an invoice.
:analytics/unreadable-pattern	warning	A pattern in the list of user agents doesn't compile on this platform, so its entry was left out.
:analytics/unreadable-line	warning	A line of a server log doesn't read as a request in the format of the log, so it counts as no download.
:analytics/unreachable	warning	OP3 gave no page of downloads: the request failed, or OP3 answered with an error. The result has the downloads of the pages before it, and its continuation token names that page.
:id3/truncated	warning	The bytes end before the tag does, so only the frames that fit were read.
:id3/frame-size	info	A frame gave its size in the other version's form, a plain number in ID3v2.4 or a syncsafe one in ID3v2.3, so it was read that way.
:id3/text-encoding	info	A frame names a text encoding that ID3 doesn't define, so its text was read as ISO-8859-1.
:id3/utf8-latin1	info	A frame declared its text as ISO-8859-1, but the bytes are UTF-8, so the text was read as UTF-8.
:id3/utf16-bom	info	A frame declared its text as UTF-16 without a byte order mark, so the byte order was worked out from where the zero bytes are.
:id3/image-type	info	A picture gave its MIME type loosely, such as JPG or image/jpg, or gave none, so the type was taken from the bytes of the image.
:id3/chapter-end	info	A chapter ends at or before its start, so its end was left out.
:id3/second-tag	info	A second tag follows the first, as when a writer adds a tag in front of an existing one, and both tags were read.
:id3/skipped-frame	info	A frame is compressed or encrypted, and wasn't read.
:id3/malformed	warning	A frame couldn't be parsed and was left out. When its header couldn't be parsed, the frames after it were left out too.
:id3/padding-junk	info	The padding after the frames holds bytes that aren't zeros, because a writer didn't clear it or a frame gave a wrong size.
:id3/too-many-frames	warning	The tag holds more frames than the :max-frames option allows, counting those inside chapters, so the frames after the limit were left out.
:id3/too-many-strings	warning	The text frames of the tag hold more strings than the :max-strings option allows, so from the limit on, each text frame kept only its first string.
:id3/too-many-tags	warning	More tags follow each other than the :max-tags option allows, so the tags after the limit were left out.
:sync/backend-failed	warning	A backend failed in a pull or a push, so the step went on without it, and it keeps its cursor. The :backend names it, and the :reason says why.
:sync/unreadable-document	warning	A document in the store didn't read as a sync document, and was left out until its device writes one that reads.
:sync/other-kind	warning	A document holds another kind of state than the one this device syncs, e.g. that of another app pointed at the same folder, and was left out.
:sync/newer-format	warning	A document was written in a newer version of the state's format than this device reads, and was left out.
:sync/malformed-state	error	The state of this device holds a value that the other devices can't merge, e.g. from an assoc, so the step wrote nothing. The :error says where.
:sync/clock-skew	warning	A device wrote its document more than a minute ahead of the clock of this device. When every other device is named, the clock of this device is probably behind.
:sync/shared-device	warning	The document of this device changed without this device writing it, so two installations, or two tabs of one page, use the same device id.
:sync/retired-device	warning	This device was retired from another device, so nothing was written for it. Its changes no longer reach the other devices, so give the state a new device id to sync again.
:sync/conflict	warning	The shared document changed during every attempt to write it, so the step ended without writing.
:sync/overlapping-step	info	A step began while another step of this device was still running, in this process or another tab of the page, so it did nothing, since two steps at once can write an older state after a newer one.