Multibooter

UPDATE 21Apr2025: Workarounds were found for all issues. VT Hash Check v1.67, modified by genieautravail for WinXP, is an excellent program. Highly recommended. Thanks so much for your work of many hours modding VT Hash Check for use under WinXP. ISSUE #1: Your modded package installed fine, but when I made the 1st run I got the following message: "No API key configured. A VirusTotal.com API key is required in order to use this application. Would you like to open the settings window and enter a key now?" -> Yes I signed up for a new user account at virustotal.com, incl. email address. An email with a validation code was then sent to this email address. Being a little cautious about giving up privacy, I had entered a rarely used email address @yandex.com . Eventually the free API key was displayed at virustotal.com and I pasted the API key into the settings of VT Hash Check ISSUE #2: Then an error msg came up: "VT Hash Check - Settings: VTHash.exe - Unable to Locate Component. The application has failed to start because Normaliz.dll was not found. Re-installing the application may fix the problem" Normaliz.dll was not on my computer. Windows XP had been installed on my computers initially as SP2, followed by an update with SP3 later on, which didnot update Internet Explorer. I still have the original Internet Explorer 6.0.2900 under WinXP, I have not installed MS Internet Explorer 8 for WinXP (8Mar2009). The issue was FIXED by extracting normaliz.dll (v6.0.5441.0, not digitally signed) from IE8-WindowsXP-x86-ENU.exe, which is the installer of MS Internet Explorer 8 for WinXP (8Mar2009), and copied normaliz.dll to \Windows\system32\ UPDATE: normaliz.dll can be copied into the install-to folder of VT Hash Check instead of into \Windows\system32\, both locations work OK. IE8-WindowsXP-x86-ENU.exe can be downloaded here: http://web.archive.org/web/20130507062858/http://download.microsoft.com/download/C/C/0/CC0BD555-33DD-411E-936B-73AC6F95AE11/IE8-WindowsXP-x86-ENU.exe I also had to Permit Outgoing Connections for VT Hash Check in my Kerio Firewall. Virus Total is owned by Google, so I opened the gates for Google. GREAT, an easy to use program. works under WinXP, just needs a right-click on the file to be checked. I checked several files with it, for about 10 minutes... until it stopped working. ISSUE #3: About 15 minutes after signing up and getting the API key from virustotal.com the following error message was displayed instead of the results table: "Your virus total account is not allowed to perform that action" When I double-clicked in the Settings window on the green checkmark next to the API key I got a message about an invalid API key. How could a valid, just downloaded key turn invalid after 15 minutes? I rebooted, waited for an hour, but still the same message about the key. I then installed the whole MS Internet Explorer 8 for WinXP (8Mar2009), instead of just copying normaliz.dll, still the same message about the key. Today, one day after installing VT Hash Check, I tried again, same message about the API key.

The issue of preserving the original folder dates seems to be an issue of NTFS partitions, not of FAT32 partitions, For example, when copying FTP sites to an NTFS partition, the folder dates of the folders copied to an NTFS target partition get set to the current date. With Beyond Compare you can "Touch" the folder dates on the target NTFS partition in a 2nd step. But when you add a comment file etc later on, the folder date gets changed to the current date and the original folder date is lost. The original folder date provides sometimes important information, e.g. when software was available. The file and folder dates at FTP sites I have named in my notes as the "FTP upload date". I always try to copy FTP sites to a target FAT32 partition, except if there are files >4GB at the FTP site; in that case I copy the file >4GB separately to an NTFS partition.. As a workaround for maintaining the original folder dates I have tried to stick to FAT32 partitions as much as possible. This has limited me in my archive mainly to 1TB and 2TB HDDs, on which I have always created 4 large logical FAT32 partitions and the remaining space as 1 logical NTFS partition, 4+1 has worked fine for me over time. Having more than 4 FAT32 partitions on GPT HDDs >2TB would eventually result in a drive letter overflow, when also connecting a similarly partitioned backup HDD for copying data from one drive to the other. I also try to maintain some backward compatibility to Win98 for my archive. Win98 has a partition size limit for FAT32 partitions, cannot access NTFS without additional tools and cannot use stuff archived stuff on GPT HDDs >2TB. I therefore consider HDDs >2TB, with huge NTFS partition(s), as useful mainly for movies, music, bittorrent downloads, .isos, stuff from archive.org, .gho and .tbi partition backups, not-yet-processed eMule downloads, infected stuff downloaded, stuff never accessed under Win98, stuff accessed only under Win10, etc, i.e. where the creation date of the holding folder is of secondary importance. All other stuff I prefer to archive on FAT32 partitions, where the folder dates do not get modified when adding stuff into folders. Under FAT32 I do not need to ask myself "is this the original folder date?", especially because I use Beyond Compare for file handling.

I have checked the source code. openvpn-2.5.4.zip does contain the string "dev_interface_list_size", while openvpn-2.3.18.zip does not contain this string. The source code files of OpenVPN are located at https://build.openvpn.net/downloads/releases/ So this bug of official OpenVPN v2.5.4 under WinXP SP3, or something similar, will very likely not occur with v2.3.18 under WinXP SP2. An actual test, however, would confirm whether eMule using OpenVPN v2.3.18 does work OK under WinXP SP2 and also under WinXP SP3.

Thanks , so there is no major issue using OpenVPN v2.5.4 on my SSE2 desktop computer. I wasn't sure whether "OpenVPN v2.5.4 Windows XP SP3 bug" in schtrom's article refers to the official v2.5.4 or to his modded v2.5.4. So the severe bug of the official v2.5.4 was resolved by schtrom's modded v2.5.4. Does OpenVPN v2.3.18 under WinXP SP2 (i.e. on an SSE-only computer) have the same severe bug "the returned list size is 1 on Windows XP SP3 for an empty list" as the official OpenVPN v2.5.4 under WinXP SP3 (i.e. on an SSE2 computer)?

1) OpenVPN v2.5.4 (schtrom mod) Thanks , OpenVPN 2.5.4 (schtrom mod) is working fine under WinXP on my SSE2 desktop computer. Your link to the free VPN server at VPNBook was very helpful for learning about OpenVPN and for testing various parameters in the .ovpn OpenVPN config file. The 2 settings with which I was able to get OpenVPN v2.5.4 to work with the .ovpn config file from VPNBook were: - in Device Manager -> TAP-Windows Adapter V9 -> Advanced tab: -> set Media Status: Always connected - added the parameter "ip-win32 ipapi" to the .ovpn config file from VPNBook (or later to the .ovpn file from ExpressVPN) OpenVPN 2.5.4 (schtrom mod), however, does NOT work on my Inspiron 7500 laptop (SSE-only, 650MHz Pentium 3) under WinXP, I have tried many tricks to no avail. 2) OpenVPN v2.3.18 OpenVPN v2.3.18 (last official version for WinXP), however, DOES work OK under SSE-only. Download-link of the Installer (32-bit), Windows XP: http://swupdate.openvpn.org/community/releases/openvpn-install-2.3.18-I001-i686.exe Download links of other builds of v2.3.18 for Windows (not tested): - Installer (64-bit), Windows XP http://swupdate.openvpn.org/community/releases/openvpn-install-2.3.18-I001-x86_64.exe - Installer (32-bit), Windows Vista and later https://swupdate.openvpn.org/community/releases/openvpn-install-2.3.18-I601-i686.exe - Installer (64-bit), Windows Vista and later https://swupdate.openvpn.org/community/releases/openvpn-install-2.3.18-I601-x86_64.exe The .ovpn file from VPNBook doesNOT work with OpenVPN v2.3.18 under SSE-only. The .ovpn file from ExpressVPN, however, DOES work with SSE-only if the parameter "cipher AES-256-GCM" is changed to "cipher AES-256-CBC" in the .ovpn config file VPNBook apparently doesnot work with the parameter "cipher AES-256-CBC". ExpressVPN has a free 7-day trial. 3) Questions Question 1: schtrom indicates that v2.5.4 contains a "severe bug" "OpenVPN v2.5.4 contains a severe bug on Windows XP SP3" https://openvpn-for-windows-xp.sourceforge.io/#topic15 Does this mean that his mod v2.5.4 works OK under Windows Server 2003 R2 SP2 (as shown in his screen shots), but not under Windows XP SP3? How serious is this bug? Does this bug not occur with v2.3.18 under WinXP SP2? i.e. will this bug not show up, for example, if eMule is run under VPN on a Pentium3/SSE-only computer with OpenVPN v2.3.18 under WinXP SP2? i.e. would the non-occurrence of this bug be an indication of a special use of Windows XP SP2. Is Windows XP SP2 the only version of Windows in which OpenVPN v2.3 works, without this bug, with an old Pentium3/SSE-only? Kai Schtrom indicates as a major benefit of v2.5.4: "new TAP-Driver with a network speed of 1 GBit/s instead of 10 MBit/s". 800 kilobytes/s, however, is a useful speed download limitation for eMule. Faster download speeds may reduce the uptime until eMule crashes, especially after 14+ days of continuous uptime on an old and weak computer. eMule is a constant trickle from a huge pool of old stuff (e.g. 2000+ items simultaneously on the transfer list, even on a weak computer), while bittorrent is a very fast downloader if the file is still around. So the download speed of the new TAP driver for OpenVPN is definitely relevant for bittorrent, but less so for eMule. Question 2: Any experiences with VPN providers which have servers in many countries (for working around country-specific requirements) AND where port forwarding can be set, so that eMule works OK under VPN? vypr.com? airvpn.org?

jason88fr (altogether 5 posts, all at this link) at forum.kaspersky.com posted in your link https://forum.kaspersky.com/topic/prevent-blocking-of-vpn-21565/ on November 17, 2021 a screenshot of ExpressVPN being flagged by Kaspersky (no idea which version). My ancient version of Kaspersky, in contrast, did NOT flag anything when I updated Kaspersky via the same ExpressVPN. I had not selected any Protection components during the installation of my ancient version of Kaspersky and I was using OpenVPN, not the proprietary VPN software by ExpressVPN. "Since September 2021, ExpressVPN has been a subsidiary of Kape Technologies, a company wholly owned by Israeli billionaire Teddy Sagi" from: https://en.wikipedia.org/wiki/ExpressVPN

I have updated the virus definitions of my ancient version of Kaspersky Ok under WinXP via VPN. 16,832,335 signatures, database release date 4/12/2025 11:14 AM. I have used OpenVPN v2.5.4 (schtrom mod) under WinXP, with a .ovpn config file by ExpressVPN, VPN server location in Ireland. TAP-Windows Adapter V9 was set in Device Manager, Advanced tab, Media Status to Always connected. I have used for this signature update a 7-day free trial by ExpressVPN. ExpressVPN worked fine for me with Kaspersky and also with uTorrent under WinXP, but not with eMule, no port forwarding. The screenshot below shows the Update running via VPN and the System Tray with the green [=connected] OpenVPN icon. The blue shield of the Kerio v2.1.5 Firewall icon shows a green arrow indicating the Permitted Outgoing Connection during the signature update. The ability to obtain signature updates of my ancient version of Kaspersky via VPN may perhaps be of use in some countries

The corresponding bootable CD (i.e. not the installer version) of Acronis Disk Director v12.5.163 (26Dec2018) can be downloaded from: http://dl.acronis.com/s/AcronisDiskDirector12.5Workstation_163_en-US.iso It works OK with my old Pentium 3 laptop (700MHz, SSE-only) and detects OK a HDD in an external docking station connected via eSATA to a SATA PC Card in the laptop. Is an excellent CD.

1) Maybe it works in an external Firewire enclosure, connected to a Firewire card in the PC? 2) Maybe the issue is related to different workings on NTFS under WinXP and Win10? "The inability to process version 2.0 of the $LogFile by versions of Windows older than 8.0 results in an unnecessary invocation of the CHKDSK disk repair utility. This is particularly a concern in a multi-boot scenario involving pre- and post-8.0 versions of Windows, or when frequently moving a storage device between older and newer versions. A Windows Registry setting exists to prevent the automatic upgrade of the $LogFile to the newer version. The problem can also be dealt with by disabling Hybrid Boot" from: https://en.wikipedia.org/wiki/NTFS#Versions The Wikipedia article lists in the footnote https://learn.microsoft.com/en-us/archive/technet-wiki/15645.windows-8-volume-compatibility-considerations-with-prior-versions-of-windows

Docu of Paragon Hard Disk Manager 15 and GPT Loader - http://download.paragon-software.com/doc/Paragon-Hard-Disk-Manager-15-Premium-User-Manual.pdf text on p.258: "GPT Loader enables to use all disk space of modern 2.2TB+ drives under Windows XP, but only for storing data, not for the Windows XP accommodation. Only internally connected single 2.2TB+ drives are supported, not external storages, or those combined into RAID (Redundant Array of Independent Disks)." This text has remained unchanged in the manual from the initial release of the manual thru the manual for SP5 Therefore it may be quite a task to get the Paragon GPT Loader to work with external HDDs >2TB - http://download.paragon-software.com/doc/Paragon-Hard-Disk-Manager-15-Premium-SP2-User-Manual.pdf - http://download.paragon-software.com/doc/Paragon-Hard-Disk-Manager-15-Premium-SP3-Manual.pdf pp.14-15 indicates: "Supported Media - Support of both MBR and GPT hard disks (2.2TB+ disks included) - IDE, SCSI and SATA hard disks - SSD (Solid State Drive) - AFD (Advanced Format Drive) - Non-512B sector size drives - FireWire (i.e. IEEE1394), USB 1.0, USB 2.0, USB 3.0 hard disks " HDDs with 512 byte sector size and eSATA connections are NOT listed. - http://download.paragon-software.com/doc/HDM15Premium_SP5_en270517pdf.pdf - The Suite edition apparently does not contain the GPT Loader, GPT Loader is not mentioned in the manual: http://download.paragon-software.com:80/doc/HDM15Suite_SP3_en.pdf Good search key in the .pdf manuals above: "Loader" http://download.paragon-software.com/doc/GPTLoader_RG_081111.pdf

My old Inspiron 7500 laptop (SSE-only, Pentium 3, 700MHz) can handle three internal HDDs. It has an internal 120GB IDE HDD for booting various opsys, a 320GB IDE HDD as data drive in the right-bay and a left-bay for various left-bay modules, e.g. a combo of floppy drive+HDD in a single module. In have inserted a 1TB Advanced Format laptop GPT HDD into a SATA ODD caddy in a left-bay module, with the Paragon GPT Loader installed under WinXP, i.e. I was using the 1TB SATA GPT HDD as a 3rd internal HDD, not as an external HDD connected via an external docking station to the eSATA PC Card. The 1TB GPT HDD seems to work great as 3rd HDD in the SATA ODD caddy under WinXP. System Commander v9.04 booted without any issue into WinXP. The laptop also booted without any issue into Win98, no drive letter shift etc. The 1TB SATA GPT HDD was listed in Win98 Device Manager as "Generic IDE Type 02", but no drive letter was assigned, Win98 was working OK. The old Inspiron 7500 laptop (SSE-only, Pentium 3, 700MHz) has a Phoenix BIOS 4.0 Release 6.0 (1985-1998) and displayed the 1TB SATA GPT HDD as 65,535MB. Since the 1TB GPT HDD seems to work OK inside this old Pentium 3 laptop, I may buy eventually a laptop HDD > 2TB for the left-bay module with the SATA ODD caddy, for further experimenting. How about a 25-year-old 700MHz Pentium 3 laptop, with a 3TB or 4TB data HDD inside? Or maybe an SSD, because the max.height of the HDD/SSD which fits into the SATA ODD caddy is 13mm, not 15mm. The major issue, however, is the slow data transfer speed between the internal GPT HDD and the other components inside the laptop. Data transfer via external docking station + eSATA PC Card would be maybe 2-5 times faster (my guess) than via internal HDD. In the old Inspiron 7500 an internal SSD is only marginally faster than a 5400rpm HDD, I have reverted from a 120GB IDE SSD to a 120GB IDE HDD for the boot drive.

I am using driver 1.3.68.2 of 13Apr2007 under WinXP for the Vantec eSATA PC Card (has a Silicon Image Sil 3512 SATALink Controller) in my SSE-only Pentium 3 laptop, has worked great with HDDs <2TB. It came on the original driver CD of the eSATA PC Card. Under Win98 I use v1.3.65.2. Where did you find v1.3.71.1?

I had first attempted to install Paragon Hard Disk Manager 15 build 813 without having installed .NET FrameWork 4, but the installation stopped with the msg: ".NET 4.0 not found. Please install .NET 4.0 for further installation of the product" After having installed .NET Framework 4 the installation completed OK. The installer also modified I:\WINXPSP2\Microsoft.NET\Framework\v4.0.30319\ngen_service.log I can also confirm that build 813 requires MS Visual C ++ 2010 x86 Redistributable v10.0.40219, which was already installed on my SSE-only Pentium 3 laptop. (this specific build of C++ Redist is required). install_xp.reg and install.reg of Hard Disk Manager 15 build 813 have the same MD5 values as those of your build 779. The MD5 checksum of vid_custom.dll, also in the install-to, is CFC582BB605FFF80C2BACD5D43C0E485 Total Uninstall indicates a quite sophisticated installation, when just the GPT Loader is selected for installation with build 813. For example, 7 .exe files are added to \Windows\Installer\{619A89DE-5F01-11E2-85E8-000C2982512D}\ and a lot of registry entries. \Windows\Installer\d80b3.msi was also added by the installer and is digitally signed by Paragon 11Nov2015 but the signature is not valid. d80b3.msi is 6,334,976 bytes and has the MD5 B253CAE901F0D54D962DEF76E219135E and contains, among others, the 7 .exe files.

Paragon 16 v11 does not install/work under WinXP, requires a more recent opsys, if I remember right. No idea about the GPT loader v11 from MediaFire, it is digitally signed OK. Maybe v11.0.0.175 is for Toshiba HDDs? Toshiba is mentioned in "Installer v1.0.2.cmd"

I have restored the original WinXP SP3 disk.sys, partmgr.sys and sp3.cab on my SSE-only Pentium 3. I installed .NET Framework 4 because this is required by Paragon Hard Disk Manager 15. I don't like software requiring .NET Framework. I then installed Paragon Hard Disk Manager 15 v10.1.25.813 (Premium, 299-PME, 11Nov2015), entered the registration details, and selected only GPT Loader in the installation menu (I did not select Paragon Hard Disk Manager and did not select HotCore Driver). The installation was under Total Uninstall, for identifying the registry entries etc made during the installation, and for a clean uninstall. I do not have any HDD >2TB. The GPT Loader seems to read and write OK with a 1TB GPT HDD in a docking station connected via eSATA to an eSATA PC Card. I don't have \gpt_loader\ from other builds of Paragon Hard Disk Manager. Here is the MD5 of the 3 files in the folder of build 813: MD5 checksums of Paragon Hard Disk Manager 15 build 813: gpt_loader.sys: B0AFB269EBA85F9EC264035640319138 gpt_loader.inf: B04C4E8E712B6CB283C4AB1C53F06226 gpt_loader.cat: 9FC7222187B2A809C21E543E928686BE BTW, Total Uninstall lists 150.30MB space used by the installation when only the GPT Loader is selected. The main culprit is a 143MB .msi in \Documents and Settings\

Paragon Partition Manager 15 Professional x86 v10.1.25.779 is the most recent build of v15 I have come across. Paragon Hard Disk Manager 15 also has a "GPT Loader" selection in its installation menu. I have various builds of Paragon Hard Disk Manager 15 in my archive, among them: v10.1.25.779 (Suite, 32bit, 9Sep2015) v10.1.25.813 (Professional, 299-PRE, 19Nov2015) and v10.1.25.813 (Premium, 299-PME, 11Nov2015) v10.1.25.1125 (Professional, 799-PRE, 12Apr2017) v10.1.25.1137 (Professional, 799-PRE, 24Apr2017) Maybe the more recent build 813 contains an improved GPT Loader? A new release 813, only 2 months after build 779, indicates serious issues with build 779. Build 813 is the last build for SSE-only. I had installed build 813 on my SSE-only Pentium 3 laptop, but removed it again because the Linux ext3 partitions created by Paragon Partition Manager 15 were displayed as BAD by Symantec PartitionMagic 8.05, i.e. the ext3 partitions created by Paragon HDM 15 are not compatible with PartitionMagic, i.e. not backward-compatible with Win98 software.

I still have the (rejected) file replacement with the Windows 2003 Server files disk.sys and partmgrs.sys running on my old Pentium 3 laptop (SSE-only). When I inserted a 2TB MBR HDD into a Sharkoon Combo docking station, connected via eSATA to an eSATA PCCard, a Found New Hardware Found bubble come up and then I got the message: "Hardware Installation. The software you are installing for this hardware Disk drive has not passed Windows Logo testing to verify its compatibility with Windows XP." The HDD had probably been detected already earlier in this docking station, before I used the replacement disk.sys and partmgr.sys. This msg "has not passed Windows Logo testing" may be related to the file replacement, it has been displayed when inserting 2 other MBR HDDs since I replaced disk.sys and partmgr.sys. No such msg has been displayed when I inserted maybe 5 other MBR HDDs. Any idea what may be going on?

Your topic may help to prepare for the inevitable end of Windows XP, nothing lives forever. Kurt_Aust prepared in 2014 a now-pinned topic "VirtualBox Windows 98se step by step" https://msfn.org/board/topic/170785-virtualbox-windows-98se-step-by-step/#comments foreboding the end of Windows 98. His outstanding how-to instructions for Win98 are now at https://forums.virtualbox.org/viewtopic.php?f=28&t=59559 Which other VM software did you consider, when you decided on VirtualBox? Why did you decide on v7? https://forums.virtualbox.org/viewtopic.php?t=104726&start=45

Very interesting comment, but don't give up, your topic is definitely a great topic! I had test-installed MS Virtual PC 2007 v6.0.192 (32bit, SP1, 8May2008) in June 2024 on my ancient Inspiron 7500 laptop (700MHz, Pentium 3, SSE-only). I had wanted to compare the usefulness of Sandboxie to a Virtual PC, but then I gave up because creating an OK .vhd file with a good/clean opsys was a major time-consuming new project,

I am continuing here the anti-virus discussion of AstroSkipper's flagged WinXP files uploaded to MediaFire https://msfn.org/board/topic/183352-proxhttpsproxy-and-httpsproxy-in-windows-xp-for-future-use/page/62/#findComment-1276757 so that his topic "ProxHTTPSProxy and HTTPSProxy in Windows XP for future use" will stay focussed on ProxHTTPSProxy and HTTPSProxy. @VistaLover I have tried to load https://opentip.kaspersky.com/ (the page to upload a file to Kaspersky for checking a potentially falsely-flagged file) with Tor Browser v7.5.6 (WinXP, 23Jun2018), but apparently this web page does not work with Tor Browser v7.5.6 under WinXP. The reason that I used Tor Browser is that I do not want Kaspersky to know my IP, for not endangering the ability to update my ancient version of Kaspersky from this IP. Unfortunately I do not have VPN. @AstroSkipper, the co-author of cacert_Updater.exe, appears not to be inclined to clear up the false-flag reporting by Kaspersky, MediaFire and virustotal of his program. @VistaLover could you submit the falsely flagged file "cacert_Updater.exe" to Kaspersky? @VistaLover I have been able to fetch Kaspersky definition updates after the Sep2024 US ban. The last update of my ancient version of Kaspersky was on 8Nov2024. I plan to update again in Feb.2025, or earlier if Kaspersky corrects the false flag and indicates that AstroSkipper's file will not be flagged anymore by their current signature. BTW, I also have in my software archive the server version corresponding to my ancient version of Kaspersky (same build, same release date). Most likely the server version can still be updated, just like my workstation version. The minimum system requirements of the server version are a Pentium 400MHz, 512MB RAM and Windows 2000 Server with Advanced SP4 Rollup1. Windows Server 2003 SP2 and Windows Server 2008 are also indicated as OK. I had tried to install the server version under WinXP SP3, but I got the installer msg: "Current operating system is not supported". I did not continue any further with the server version because I don't have Windows 2000 Server installed on my computers. Nevertheless: an excellent virus checker with (most likely) a current virus signature ... for Windows 2000 !!!

Your old version of April 2022 gets flagged by my ancient version of Kaspersky, signature of 8Nov2024, as follows: detected: Trojan program Trojan.Win32.Gamaredon.gj file: E:\Downloads_5\cacert_Updater_Fixed_Recreated.7z//cacert_Updater.exe No riskware is detected. virustotal: https://www.virustotal.com/gui/file/245eb2147210581ce8965cd1337d1137b48da190e5e5e2dd6627d8eb803dd563 Software which gets flagged has a much shorter life expectancy. A programming job may be considered complete after having ensured that it will not get flagged.

I have a 2TB HDD, labeled "FLAGGED", nearly full with flagged stuff. The flagged stuff is in the subfolders: adware, corrupt, incomplete, infected [=Trojans+viruses], malware, riskware and password-protected. Password-protected installer files require full trust. Your files may enjoy full trust because the source/referring page is msfn.org. The only installer containing pw-protected files which I use is for Beyond Compare, which has my full trust.

It is not unusual that trojans etc only get detected/added to virus signature updates several years after they came out. Again, why not just create a different SFX?

I just saw a description of the Gamaredon trojan flagged by Kaspersky in cacert_Updater.exe: "Gamaredon ... is a Russian, state-sponsored cyber-espionage hacking group with cybersecurity researchers linking them to the FSB (Russian Federal Security Service)" https://www.bleepingcomputer.com/news/security/gamaredon-hackers-start-stealing-data-30-minutes-after-a-breach/ Russian Kaspersky is very unlikely to flag real Russian state-sponsored trojans, cacert_Updater.exe for ProxHTTPSProxy must therefore be clean, and this must be a false flag Why not just create a different SFX?

I used Cert_Updater_v1.6.exe just as an example of another certificate updater, to see whether it also gets flagged when virus-checking. MiTeC EXE Explorer v1.2 indicates for cacert_Updater.exe an incorrect timestamp of 10May2012, even if the most recent file in the .exe, HTTPDL.exe, has the file modification date 13Feb2019. An incorrect timestamp is a little unusual, but not suspicious, many OK .exes have an incorrect timestamp. Maybe, if the content of cacert_Updater.exe (CERTUPD.bat, HTTPDL.exe and URLLINK) is put into a different SFX, it will not be flagged by virustotal/MediaFire?