Leaderboard

== Windows 9x Web Helper == = Overview = Windows 9x Web Helper is a Bash script, file name 9xweb (25 Kb), designed to improve internet accessibility. Scratch coded using Microsoft Notepad, it is free to use, modify and share, using Unlicense (https://unlicense.org). Tested in vanilla Windows 98. Provided an Enhanced Cygwin-Lite install is functional, it should run on enhanced Windows 98 and is confirmed to run in Windows 95. = Links = Overview: Project screenshots, download link and more information. MSFN forum member login required to download. https://msfn.org/board/topic/177106-running-vanilla-windows-98-in-2020/page/17/?tab=comments#comment-1176154 Pre-requisite: The only pre-requisite is a functional Enhanced Cygwin-Lite install, here's more information and download links. https://msfn.org/board/topic/177106-running-vanilla-windows-98-in-2020/page/17/?tab=comments#comment-1175895 FAQ: https://msfn.org/board/topic/181417-windows-9x-web-helper-9xweb/page/3/?tab=comments#comment-1186792

Notice: since upstream merged a big branch, it will take more time than usual to get changes landing in my tree. as a result today's builds may or may not be done. for v27(g3) tree, since AF upstream broke the build, g3 based builds may not be available this week if I can't fix it. for boc, since upstream vanished its repo, I'm now thinking of building icedove-uxp instead.

This is a problem in the system localization files, this bug is on the system, with the POSReady update in all languages except English. The correct message should be that the dll is missing a function. Edit: It started without errors for me. All updates have been installed, including the new Visual C ++ 2015-2019 Redistributable.

I'm working on an Haswell+-specific issue, which is completely unrelated to the Ryzen problems. Luckily someone posted kernel dumps of their Ryzen crashes earlier on (I have no Ryzen myself, and literally no room for it), so I'll try to correct those issues once I start seriously modifying ntoskrnl (I already changed the version to 6.1 and moved the export table to a new section at the end of file, and have no issues). And didn't you say that you ran Vista on Ryzen too? Were there any issues with your config?

actually it does, there are changes that preventing it to work in XP: 1. mozilla removes large chunks of code for XP while adding non-XP compatible APIs usage 2. rust compiler can't produce XP compatible code

I wouldn't go that far. I am banned from the Pale Moon Forum because I made the innocent mistake of screen-capping a Roy Tam build to illustrate an about:config setting I was discussing. That was close to two years ago and I'm still banned. He got hostile with me so I got "defensive" - then I was banned -- permanently. I don't mind if he's allowed back at MSFN. BUT it is my sincere view that he can NOT be allowed to post in threads DEDICATED to XP !!! I run XP for a REASON. I own legit copies of 98, 98SE, 2000, ME, NT, XP, 2003, 7 Pro, 7 Enterprise, 10 Anniversary, and 10 LTSB. But I run XP for a REASON. I consider myself a "tweaker" and an "optimiser". I can do tasks in XP at 45-degrees Celsius that 7 and 10 runs up the CPU to over 80 and at 2 to 3 times the RAM. Again, I run XP for a REASON. And I am SOOO very thankful that we have forums DEDICATED to XP.

Changelog and milestones: 20200112 Beta - 1st Beta 20200203 Beta - 2nd Beta - View YT video description - New sleepSecs variable for file system race condition 20200405 Beta - 3rd Beta - Improved main screen readability - Modified video Wget command - Improved video_fetch and getURL logic - Remove non-ascii text from video description - New delete video option after video and information view - More strategic sleepSecs for file system race condition - Decreased sleepSecs variable from 3 to 2 seconds - Additional Video Fetch help - Video title cleanup 20200411 - Confirmed to work in Windows 95. 20200522 Beta - 4th Beta - Removed duplicate video_fetch clipboard retrieval - Modified showClip clipboard cat command (query reduce stackdumps) - Disabled Video Fetch Wget user agent string (uncomment as desired) - Modified getHTML and getMedia Wget commands - Reset several variables after Video Fetch - Adjusted $sleepSecs instances (file system race condition) - New $preview variable (off, 5, +5 toggle) - New $description variable (yes, no, ask) - Updated Video Fetch [h]elp 20200712 Beta - 5th Beta - Broke proxy https://proxybrowser.xyz, new http://dontfilter.us - Quote wrapped all /dev/clipboard instances - Added '&disable_polymer=1' just in case note - New $AskKeepvid variable for 'ERROR: 403 Forbidden' YouTube - Simplified Wget command - Silenced Wget kill after YouTube preview for clean output - YouTube title cleanup: excess underscores, period tail - New YouTube search, specify search or Trending, uses Googlebot UA - YouTube search results include: video durations, single query cache - Updated Video Fetch [h]elp 20200805 Beta - 6th Beta - YouTube 'polymer' compatible: custom search, trending search, video fetch - Initiate new search directly from search results screen - More efficient getFormat HTML file chunking - Modified $sleepSec instances - Minor UI tweaks 20201124 Beta - 7th Beta - USER CONFIG $browser2: old seamonkey.exe, new k-meleonw9x.exe - USER CONFIG $sEngine1: old https://google.com, new http://google.com - USER CONFIG $sEngine2: old https://search.yahoo.com, new https://lite.duckduckgo.com/lite - Main screen 'archive.org cache' renamed 'Wayback Machine' - Function loadUrl improved $url grep - Function loadUrl now accepts FTP URLs - Modified newURL function prompt 'enter new URL >OR CONTENT< to clipboard' (ie. bookmark tag) - Clipboard content now displayed regardless of '--Invalid URL--' - Clipboard output asthetically trimmed to one line, does not alter actual system clipboard - Clipboard content >1 line (page break) denoted 'clipboard_content ...' - Clipboard content with >68 characters denoted 'clipboard_content ..' - Tag user selected titles from search_YT and cache_YT functions, denoted => - Fixed YouTube title extractions containing embedded double quotes - More efficient $ytURL extraction in get_search_URL function - More robust and efficient $vidTitle processing in getFormat function - More efficient $vidText processing in extract_info function - More robust title grep in search_YT function - Improved titles cleanup in search_YT function - Modified Wget commands - Modified $sleepSecs and sleep instances - Hopefully fixed pathways and prompts - Revised Main screen and Video Fetch [h]elp - Miscellaneous fixes and UI tweaks

For anyone using YouTube via 9xweb in vanilla Windows 98, both SMPlayer/MPlayer v0.6.7 and VLC v0.8.6d work, discussed in and around this post: https://msfn.org/board/topic/177106-running-vanilla-windows-98-in-2020/page/5/?tab=comments#comment-1173021 TinyUpload is working again (yeah), even *.exe files up to 50 MB per file..so dangerous :) In case other links go down here are the goods. The player files should be readily available elsewhere, codecs might be harder to obtain. The files are untouched from original download other than renaming. No JavaScript required. SMPlayer v0.6.7 (SMPlayer_0.6.7.exe, 13.0 MB): http://s000.tinyupload.com/index.php?file_id=32421385938599921768 Codecs for SMPlayer's mplayer (SMPlayer_MPlayer_codecs.zip, 9.57 MB). In a default install, extract files into C:\Program Files\SMPlayer\mplayer\codecs: http://s000.tinyupload.com/index.php?file_id=33821390908445725051 VLC v0.8.6d (VLC_0.8.6d.exe, 9.34 MB): http://s000.tinyupload.com/index.php?file_id=34990583317191199660 Directing the 9xweb script to launch SMPlayer's 'mplayer', the default USER CONFIG entry, instead of 'smplayer' improves video launch time considerably. For SMPlayer's mplayer: f key toggles full screen, number 3/4 keys adjust brightness, arrow keys skip back and forth, spacebar pauses, q key exits. An 800 MHz system, 32 MB AGP graphics, 1152 x 864 resolution and 16-bit colour can play Format 18 videos full screen. Lesser hardware may need to use a smaller window, enable skipped frames, etc.

== Windows 9x Web Helper (Discontinued) == = Overview = Windows 9x Web Helper is a Bash script, file name 9xweb (25 Kb), designed to improve internet accessibility. Scratch coded using Microsoft Notepad, it is free to use, modify and share, using Unlicense (https://unlicense.org). Tested in vanilla Windows 98. Provided an Enhanced Cygwin-Lite install is functional, it should run on enhanced Windows 98 and is confirmed to run in Windows 95. The script runs from a Bash terminal in a COMMAND.COM window. It is configurable and web browser/media player agnostic. The screenshots are fairly self explanatory. Entering [h]elp ('h' key) provides additional information for each of the three screens: main screen, bookmarks, video fetch. = Screenshots = Main screen. https://i.postimg.cc/hjjXp8WD/9xweb-01-main-screen.png Bookmarks screen. https://i.postimg.cc/gkq6c2DP/9xweb-02-bookmarks-screen.png Video Fetch screen. https://i.postimg.cc/C1XYGKvy/9xweb-03-video-fetch-screen.png = Pre-requisites = This 9xweb script. Enhanced Cygwin-Lite install (see detailed January 7th post three up). Stable network connection. Favourite web browser(s). Favourite MP4 capable media player(s). Web browsing with JavaScript disabled will save system resources, especially if using MP4 autoplay. All default USER CONFIG section URLs work without JavaScript. = Preparation = Install pre-requisites. Move the 9xweb script to C:\cygwin\bin. = Configuration = Modify the 'USER CONFIG' section of the 9xweb script as desired. = Usage = Launch Cygwin-Lite via desktop or system tray shortcut. [1] Enter command '9xweb' at the Bash terminal prompt (or 9, tab, Enter keys). [1] Although [q]uit is preferred to cleanly exit, pressing Ctrl-c can be used to terminate downloads and force exit. The script is smart enough to properly re-initialize the next time it is run. [1] Wait for the busy hourglass cursor to settle before proceeding. = Web Proxies = Two web proxies can be configured, which can be toggled and accessed from the main screen. As web proxies act as a man-in-the-middle, they are a potential security weakness. It is recommended they only be used for non-sensitive browsing when no available web browser can natively access a site. At the time of this post, RetroZilla v2.2 with the about:config tweaks mentioned earlier can access almost all sites without issue. = Internet Archive = The main screen archive.org functions use the Wayback Machine. This is useful to view a snapshot of a website in time or access sites that are no longer active. Selecting most recent will load the most recent snapshot available. Unfortunately not all websites have been archived. Specifying a date will load a snapshot of the site on the specified date, if available. Otherwise the most recent date available after the specified date gets loaded. For example, specifying 2011 may display a snapshot from February 2011, the earliest available in 2011. Review the 'archive.org' URL in the browser to confirm the actual snapshot date. To get something later in the year enter a more specific query, such as 201106. = Bookmarks = The bookmarks screen displays only the URL, no description. URLs are usually descriptive enough and this conveniently prevents most longer bookmarks from screen wrapping. A descriptive entry can be added as a 'bookmark', to be used as a 'sticky note' or provide a description for the URL above. For example, bookmark a non-descriptive youtube[dot]com URL from clipboard. Then select clipboard [n]ew and enter a description to the clipboard, such as '^^^ CONFIG.SYS info ^^^'. Although screen output will display '---Invalid URL---', select [ b ]ookmark clipboard. This description has now been bookmarked as a note below the non-descriptive URL (see screenshot). URLs and notes can be removed using [d]elete bookmark. = Video = AFAIK there is no longer a working mechanism for vanilla Windows 98 to access youtube[dot]com or any other multimedia site content, as the OS is not capable of meeting the recommended modern browser requirements. The primary purpose of video fetch is to provide some multimedia access, especially Windows 98, DOS and retro computing related educational media. Note video fetch provides partial URL support only, some downloads will fail with 403 Forbidden errors. To me this is acceptable, respect the content provider's wishes, just choose another less restrictive video. Both regular (short) and channel (long containing 'list=' and/or 'index=') URLs work. To keep the code simple only Formats 18 (640 x 360 MP4) and 22 (1280 x 720 MP4) are supported. = Limitations = Enhanced Cygwin-Lite and the Bash shell are limited and not fully stable. The Unix environment may create new browser or media player configurations in c:/cygwin/home/system_name/ when first launched (eg. SMPlayer's mplayer). The browser or media player may dump data to the terminal and clutter 9xweb's output. Some software is configurable. For example, first run using SMPlayer's mplayer will create a new configuration directory c:/cygwin/$HOME/mplayer. Adding this to c:/cygwin/$HOME/mplayer/config will stop undesirable output: really-quiet=TRUE The default 9xweb_bookmarks.txt file location is c:/cygwin/$HOME/, which can be changed in USER CONFIG, and will be lost if Cygwin-Lite is fully un-installed. = Bugs = Report bugs via this thread or preferably private message. If fixable new releases will be posted here. Please do not send bug reports for Wget 403 Forbidden errors. = Disclaimer = This software is freely provided in the hope it will be useful. Use at own risk. End-user must accept full responsibility, including potential data loss or system issues. Review UNLICENSE information at the top of the 9xweb script. Edit: 20200112 Beta - 1st Beta 20200203 Beta - 2nd Beta - View YT video description - New sleepSecs variable for file system race condition 20200405 Beta - 3rd Beta - Improved main screen readability - Modified video Wget command - Improved video_fetch and getURL logic - Remove non-ascii text from video description - New delete video option after video and information view - More strategic sleepSecs for file system race condition - Decreased sleepSecs variable from 3 to 2 seconds - Additional Video Fetch help - Video title cleanup 20200411 - Confirmed to work in Windows 95. 20200522 - 4th Beta - Removed duplicate video_fetch clipboard retrieval - Modified showClip clipboard cat command (query reduce stackdumps) - Disabled Video Fetch Wget user agent string (uncomment as desired) - Modified getHTML and getMedia Wget commands - Reset several variables after Video Fetch - Adjusted $sleepSecs instances (file system race condition) - New $preview variable (off, 5, +5 toggle) - New $description variable (yes, no, ask) - Updated Video Fetch [h]elp 20200712 Beta - 5th Beta - Broke proxy https://proxybrowser.xyz, new http://dontfilter.us - Quote wrapped all /dev/clipboard instances - Added '&disable_polymer=1' just in case note - New $AskKeepvid variable for 'ERROR: 403 Forbidden' YouTube - Simplified Wget command - Silenced Wget kill after YouTube preview for clean output - YouTube title cleanup: excess underscores, period tail - New YouTube search, specify search or Trending, uses Googlebot UA - YouTube search results include: video durations, single query cache - Updated Video Fetch [h]elp 20200805 Beta - 6th Beta - YouTube 'polymer' compatible: custom search, trending search, video fetch - Initiate new search directly from search results screen - More efficient getFormat HTML file chunking - Modified $sleepSec instances - Minor UI tweaks 20201124 Beta - 7th Beta - USER CONFIG $browser2: old seamonkey.exe, new k-meleonw9x.exe - USER CONFIG $sEngine1: old https://google.com, new http://google.com - USER CONFIG $sEngine2: old https://search.yahoo.com, new https://lite.duckduckgo.com/lite - Main screen 'archive.org cache' renamed 'Wayback Machine' - Function loadUrl improved $url grep - Function loadUrl now accepts FTP URLs - Modified newURL function prompt 'enter new URL >OR CONTENT< to clipboard' (ie. bookmark tag) - Clipboard content now displayed regardless of '--Invalid URL--' - Clipboard output asthetically trimmed to one line, does not alter actual system clipboard - Clipboard content >1 line (page break) denoted 'clipboard_content ...' - Clipboard content with >68 characters denoted 'clipboard_content ..' - Tag user selected titles from search_YT and cache_YT functions, denoted => - Fixed YouTube title extractions containing embedded double quotes - More efficient $ytURL extraction in get_search_URL function - More robust and efficient $vidTitle processing in getFormat function - More efficient $vidText processing in extract_info function - More robust title grep in search_YT function - Improved titles cleanup in search_YT function - Modified Wget commands - Modified $sleepSecs and sleep instances - Hopefully fixed pathways and prompts - Revised Main screen and Video Fetch [h]elp - Miscellaneous fixes and UI tweaks Note: Project discontinued.