История версий «Grab All Files – Downloader & Pages to HTML» – 25 версий
Grab All Files – Downloader & Pages to HTML от Grab All Files – Bulk PDF & File Downloader
История версий «Grab All Files – Downloader & Pages to HTML» – 25 версий
Будьте осторожны со старыми версиями! Эти версии отображаются только для сведения и в целях тестирования.Вы всегда должны использовать последнюю версию дополнения.
Последняя версия
Версия 5.8.9
Выпущено 27 сент. 2026 г. — 5,41 МБРаботает с firefox 128.0 и вышеPage collection: a page that shows only a CAPTCHA is now skipped instead of stopping the whole run, and a public form with a CAPTCHA no longer stops it; pages refused with HTTP 429/503 are retried once more at the end, and Reset now also stops the background count of candidate pages. NCPSSD: PDF saving survives restarts and resumes without resending, and when the site's sign-in has expired the extension says so and offers a button to open the site and sign in, then retry. Saved marks: a file that failed in this run is no longer shown as saved; on Firefox, files saved in an earlier session are offered again, because Firefox cannot report files deleted outside the browser. Active Capture no longer saves a file under another row's name.Исходный код выпущен на условиях: Все права защищены
Старые версии
Версия 5.8.8
Выпущено 26 сент. 2026 г. — 5,38 МБРаботает с firefox 128.0 и вышеNCPSSD (National Center for Philosophy and Social Sciences Documentation): PDFs are now found on its English-language journal pages, for the Chinese journals it lists, and in its ancient-books viewer, including articles whose PDF address is only issued when clicked; PDF saving is more reliable, and a failed article can be retried after opening its page. Saved marks are more accurate: the extension re-checks whether a saved file still exists, clears the mark when the file was deleted, keeps it when it cannot tell, and old browser download history no longer brings a mark back after a reset; resuming an interrupted save no longer re-selects files already saved. Page collection: pages answered with HTTP 429/503 without Retry-After are retried again, long rate-limit waits no longer open real tabs one after another, and CAPTCHA pages with long text are no longer saved as normal pages. Active Capture no longer saves the same file twice when a download starts late.Исходный код выпущен на условиях: Все права защищены
Версия 5.8.7
Выпущено 24 сент. 2026 г. — 5,35 МБРаботает с firefox 128.0 и вышеFixed repeated image-site permission prompts after access was already granted. If the browser cannot confirm a granted image site, collection makes a bounded retry and shows a clear result instead of waiting for the same permission again. The add-page result also makes the next page selection clearer when no new page was saved.Исходный код выпущен на условиях: Все права защищены
Версия 5.8.6
Выпущено 23 сент. 2026 г. — 5,35 МБРаботает с firefox 128.0 и вышеFirefox can add pages to the same HTML without a Save As dialog when the previous file is verified in the current Downloads/GrabAllFiles folder. Otherwise, choose the previous file in the dialog. Linked files are saved individually rather than automatically bundled into a ZIP. Back up an HTML you have edited before adding pages.Исходный код выпущен на условиях: Все права защищены
Версия 5.8.5
Выпущено 22 сент. 2026 г. — 5,34 МБРаботает с firefox 128.0 и вышеImproved collection recovery, permission guidance and candidate selection. Firefox opens a dedicated popup and supports adding pages to a previous HTML download without automatically deleting the old file. Rejected filenames use a safe alternative. Public contact forms are less likely to be mistaken for login checks. Improved handling of server-requested waiting periods.Исходный код выпущен на условиях: Все права защищены
Версия 5.8.4
Выпущено 21 сент. 2026 г. — 5,3 МБРаботает с firefox 128.0 и вышеLinked files: the total size you keep is now yours to set -- unlimited by default, instead of a hidden 2 GB ceiling that silently downloaded and discarded everything past it. Before a run you can see how many linked files there are and roughly how much they weigh; during one, how much has landed. RSS/Atom feeds are no longer collected as files, a site that stops answering is dropped instead of retried file by file, and permission for a redirect target is asked for once, up front. The result screen no longer offers "choose the next page" when every page left would fail the same way, and says why a candidate could not be counted. Crawls no longer stop at a site's other host, and saved files can be named by title or by their original name.Исходный код выпущен на условиях: Все права защищены
Версия 5.8.3
Выпущено 21 сент. 2026 г. — 5,25 МБРаботает с firefox 128.0 и вышеPage collector: add pages to an HTML you already saved (Free plan too), save the HTML before waiting on a document permission, list duplicate pages apart from failures, show image/file counts per candidate, and mark “Saved” only when the file is really there. Deep scans no longer report success when the session expired or site access was revoked mid-crawl; they stop and tell you how to resume. Page hooks left by Active Capture now expire automatically. Fixed folder-save and right-click download filenames, ZIP name collisions, stalled-download hints, table merged cells and math in AI packages, and a clock-skew bug that deleted paid license keys. 78 fixes from a full code audit. Lazy-loaded images are now captured, and the image embedding limits (100 files / 25 MiB by default) can be raised in the detailed settings.Исходный код выпущен на условиях: Все права защищены
Версия 5.8.2
Выпущено 21 сент. 2026 г. — 5,24 МБРаботает с firefox 128.0 и вышеPage collector: add pages to an HTML you already saved (Free plan too), save the HTML before waiting on a document permission, list duplicate pages apart from failures, show image/file counts per candidate, and mark “Saved” only when the file is really there. Deep scans no longer report success when the session expired or site access was revoked mid-crawl; they stop and tell you how to resume. Page hooks left by Active Capture now expire automatically. Fixed folder-save and right-click download filenames, ZIP name collisions, stalled-download hints, table merged cells and math in AI packages, and a clock-skew bug that deleted paid license keys. 78 fixes from a full code audit.Исходный код выпущен на условиях: Все права защищены
Версия 5.7.7
Выпущено 9 сент. 2026 г. — 3,52 МБРаботает с firefox 128.0 и вышеImproved protection of existing files and cleanup after interrupted ZIP extraction. HTML reports now distinguish saved files from pending text extraction. Fixed stale collection progress updates and save-location persistence errors.Исходный код выпущен на условиях: Все права защищены
Версия 5.7.6
Выпущено 8 сент. 2026 г. — 3,52 МБРаботает с firefox 128.0 и вышеImproved save-folder selection before collection and placed the HTML filename above the destination. Clarified Firefox download settings. Improved scanning, file handling and collection reliability.Исходный код выпущен на условиях: Все права защищены
Версия 5.7.5
Выпущено 6 сент. 2026 г. — 3,45 МБРаботает с firefox 128.0 и вышеFirefox now opens in a standard window so you can adjust its width. Added space below the Download button. Improved scan cancellation, saved-file status, download filenames and resume handling. Fixed sensitive-data masking in AI exports and Japanese text decoding. Updated PDF and Excel readers.Исходный код выпущен на условиях: Все права защищены
Версия 5.7.4
Выпущено 5 сент. 2026 г. — 2,8 МБРаботает с firefox 128.0 и вышеA one-off backend hiccup can no longer wipe a paid licence key: the extension now treats only an explicit server rejection as a rejection and keeps working on its offline grace otherwise. 'Hand to AI' now recovers files the direct fetch cannot reach (public CDNs, strict-cookie portals) the same way Download and ZIP do. Page Collector no longer mistakes version-numbered pages (/v20.11.0) for files, no longer pre-selects sign-out links, upgrades same-site http links to https, and keeps server-declared Japanese filenames. Rescans keep icon-style download buttons, forms with a hidden field named 'action' post to the right address, and cancelling a save dialog on Firefox no longer shows an error. 27 fixes from a full-code review; see the changelog.Исходный код выпущен на условиях: Все права защищены
Версия 5.7.3
Выпущено 4 сент. 2026 г. — 2,78 МБРаботает с firefox 128.0 и вышеDocument downloads are safer and their results are clearer. Page Collector now refuses sign-in screens and error pages returned under Word, Excel or PDF filenames, and validates PDF and OpenXML signatures before saving or extracting text. Bulk download now detects HTML even when its opening tag is split across streaming chunks. If a website is saved instead of the requested file, the completion card stays open with a warning instead of showing a clean success or celebration. Warnings are scoped to the current run, and public-CDN retries now include the exact job, document and progress position in diagnostics.Исходный код выпущен на условиях: Все права защищены
Версия 5.7.1
Выпущено 3 сент. 2026 г. — 2,77 МБРаботает с firefox 128.0 и вышеFiles served through a public CDN (a 302 to a host that answers “anyone may read”) now land in the folder you chose instead of failing. The download is retried once without credentials when the credentialed request gets no reply — the CDN was open all along; the extension was insisting on showing a badge it did not want. When most of a run cannot be written to the chosen folder, the files go to Downloads/GrabAllFiles automatically with a notice instead of a confirm toast that was easy to miss (19 of 20 files were being left unsaved on a university course page). The action rows under Download and after a download now follow one button system — one filled pill for the action that matters, outlined rectangles for the rest — and the “Retry failed” button is readable again (it had been white text on light grey). Page Collector can read Word and Excel documents, records which save route was used, and shows the Hand to AI button where it belongs.Исходный код выпущен на условиях: Все права защищены
Версия 5.7.0
Выпущено 3 сент. 2026 г. — 2,76 МБРаботает с firefox 128.0 и вышеSelected files can now be handed to an AI in one step. The new Hand to AI button extracts the text of the PDFs, documents and pages you picked — Word and Excel included — into a single ZIP with an approximate token count, so you know before uploading whether it fits. A download that ended early can be resumed without rescanning the site, and a paused page collection now tells you it is waiting. This release also fixes a serious Chrome and Edge problem where saved files lost their name and folder: the GrabAllFiles folder was never created, Name by title did nothing, and ZIP, merged-PDF and AI package names became random text. Saved counts are now checked against what actually reached disk.Исходный код выпущен на условиях: Все права защищены
Версия 5.6.14
Выпущено 2 сент. 2026 г. — 2,51 МБРаботает с firefox 128.0 и вышеFixes for problems found in a review of the last five releases. "Skip sign-in and resume" no longer fails with an error and now closes the sign-in tab it opened. With auto-scroll on, the page you have open is collected fully scrolled, so lazy-loaded text and images are no longer missing. Image hover previews come back for pictures hosted on a CDN. Extracting a ZIP into a folder can retry files that need sign-in. The licence screen no longer tells an activated Pro user they are on the Free plan, and a paused older collection again promises the number of pages it will really resume with.Исходный код выпущен на условиях: Все права защищены
Версия 5.6.13
Выпущено 1 сент. 2026 г. — 2,51 МБРаботает с firefox 128.0 и вышеUnsaved page results now appear directly in the matching candidate row in both Free and Pro: each row shows a Not saved badge, a clear localized reason, and a safe link for opening the page. New Unsaved only and Copy unsaved URLs controls make large result sets easier to review. The separate lower list now appears only for older or dynamically added pages that have no matching candidate, avoiding duplicate explanations. Candidate selection, virtual scrolling, accessibility labels, and completed-result refreshes stay synchronized.Исходный код выпущен на условиях: Все права защищены
Версия 5.6.12
Выпущено 30 авг. 2026 г. — 2,5 МБРаботает с firefox 128.0 и вышеFree Page Collector now shows related-page candidates and keeps exactly one selected: the first candidate is selected by default, and choosing another switches the selection. Candidate rows again show image, directly linked file, and page-link counts; items beyond the first 20 are clearly marked as not yet counted. Exact size and SHA-256 reuse avoids duplicate linked files across runs, recovered local images improve offline HTML, and progress/results are clearer.Исходный код выпущен на условиях: Все права защищены
Версия 5.6.11
Выпущено 30 авг. 2026 г. — 2,47 МБРаботает с firefox 128.0 и вышеFree still includes every feature and the same output quality. File downloads are handled 20 at a time, and page collection now pauses every 10 completed pages for confirmation while keeping one resumable job and one final HTML; Pro skips those plan pauses. Shift range selection is now Pro-only. AI-ready exports default to 25 MB parts, with a configurable size or no splitting. If a file cannot be saved directly to the chosen folder, the extension now asks before using the browser-managed Downloads folder and reports both destinations together.Исходный код выпущен на условиях: Все права защищены
Версия 5.6.9
Выпущено 29 авг. 2026 г. — 2,45 МБРаботает с firefox 128.0 и вышеИсходный код выпущен на условиях: Все права защищены
Версия 5.6.8
Выпущено 28 авг. 2026 г. — 2,42 МБРаботает с firefox 128.0 и вышеИсходный код выпущен на условиях: Все права защищены
Версия 5.6.7
Выпущено 27 авг. 2026 г. — 2,4 МБРаботает с firefox 128.0 и вышеИсходный код выпущен на условиях: Все права защищены
Версия 5.6.5
Выпущено 25 авг. 2026 г. — 2,22 МБРаботает с firefox 128.0 и вышеИсходный код выпущен на условиях: Все права защищены
Версия 5.6.4
Выпущено 22 авг. 2026 г. — 2,17 МБРаботает с firefox 128.0 и вышеИсходный код выпущен на условиях: Все права защищены
Версия 5.6.3
Выпущено 22 авг. 2026 г. — 2,17 МБРаботает с firefox 128.0 и вышеИсходный код выпущен на условиях: Все права защищены