Friends, time is limited, hurry up and test grub4dos! - China DOS Union Forum

Please fully test, urgent! If this version is okay, it may become the final version!!! Everyone must work together!!!

http://grub4dos.jot.com/

After testing for a few days (maybe only two or three days!), we will release the official version. The reason for being so urgent is to lay the groundwork for the release of WinGRUB. The overall structure of the version will no longer be changed, but if bugs are found, they will still be fixed.

The official version should be released at the latest by December 31st, if there are no abnormalities.

Due to insufficient time, the optimization plan for A20 has been canceled, but other plans have been implemented. Including the search for the default configuration file of grub.exe has been completed.

Now grub.exe first searches for the menu.lst file in the current directory. If it fails, it then searches for the menu.lst file in the root directory of the current disk. If it also fails, it looks for the (hd0,0)/menu.lst file. If it fails again, then the built-in preset_menu will take control.

The search for menu.lst in the disk and directory where grub.exe is located has been canceled. Because of insufficient time, this code cannot be written. It will not be written in the future either, because it is relatively complicated.

Now the support for win98 by the emulated disk has been strengthened, and many faults and obstacles have been eliminated. (It runs satisfactorily on my machine). Everyone is best to test this too. Test the emulated disk with --mem and without --mem to see if it is normal under win98. Please report whether there is an improvement in stability when accessing the emulated disk after entering win98 compared with the previous old version.

Thank you friends for working together.

－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－

Download addresses for grub4dos and related tools added in 2007:

http://download.gna.org/grub4dos/
http://download.gna.org/grubutil/

English and Chinese documents for grub4dos added in 2007:

http://grub4dos.sourceforge.net/wiki

The original GNU GRUB mailing list archives are very valuable for developers:
http://lists.gnu.org/archive/html/bug-grub/

[ Last edited by 不点 on 2007-4-15 at 02:10 AM ]

These days I installed a new computer and used GRUB4DOS again. I found that the latest version of GRUB.EXE has to prob ROM INT every time it starts, causing it not to run or crash in many environments (even in pure DOS with only some simple drivers loaded), while the previous versions could. It is suggested to add an option to disable prob so that GRUB.EXE can still start in some environments where there are actually no problems.

This problem can't be solved now. Let's release the official version first and solve it slowly later.

Friends with this problem can temporarily continue to use the old version.

------------

The previous old version had no detection function, or it was a simple system identification function. The recovery of interrupt vectors was specially handled according to the characteristics of different systems. For different versions of MS-DOS, we spent a lot of space to handle it. This can be said to be very cumbersome.

Now we use a unified method for detection, and any DOS, any driver, and any TSR are equal. The detection method is concise and clear.

The next step, we only need to find corresponding measures for those crash failure situations. This is the next task after 0.4.2.

----------------

Now you can upload the problematic driver program and TSR (along with the commands in config.sys and autoexec.bat). I can study the causes of crashes and failures in the virtual machine.

GRUB for DOS 0.4.2 official version has been released! Thanks to the support of all friends on this site!

Download:

http://grub.linuxeden.com/wakka.php?wakka=UploadFile

http://sourceforge.net/projects/grub4dos/

http://grub4dos.jot.com/

http://sarovar.org/projects/grub4dos/

http://grub4dos.freespaces.com/

It seems that there is no difference at all between the official version and the version of December 28th.

I happen to need it, but I don't know which version is more suitable. Please recommend it, experts.

I don't know how to use it, but I still download it to try, thank you

Agree with the statement of floor 6. The menu also shows the words "2006-12-28".

The date in ChangeLog.txt also seems to be incorrect, such as "2006-10-23 version 0.4.2"

[ Last edited by martin325 on 2007-1-5 at 11:27 PM ]

Test found that the 0.4.2 official version's map --mem method cannot boot a floppy image larger than 2.88MB, but pre11 can.

title Test 10th image: 10.img
map --mem (hd0,0)/img/10.img (fd0)
map --hook
chainloader (fd0)+1
rootnoverify (fd0)

I've been using it all the time, and I haven't found any major problems.

Support, hope to increase ISO support!

Back then, the teacher said to learn assembly well, but I didn't listen!
Now it's all good! Assembly is still completely incomprehensible, and I can't mix 16bit 32bit code. It's natural that I don't understand boot loader code
Using the previous guess_xxx was not bad! Anyway, everyone's environment is similar
Strongly support Wengier

The original guess dos versions has been deleted. That method just works, but it's not perfect.

The old method can only restore a few key interrupt vectors, while the new automatic detection fully restores all interrupt vectors from 00 to 7F. For interrupt numbers above 80h, basically no need to care (no need to restore). The new method is perfect, and it will bring a more real BIOS environment than the old method, thus bringing greater stability to the boot of the operating system (that is, the operation of the operating system after chainloader + boot).

If you encounter a detection failure, please upload the drivers and TSRs that caused the failure, which may help solve the problem as soon as possible.

In addition, if there is a problem with map --mem, you can also report it in detail. Don't just say it briefly, that's not useful. I can use map --mem on my real machine to boot a win98 virtual hard disk image, which is 37M in size, far exceeding 2.88M.

If you don't report, this problem may continue to exist in the new version, and then you may have to use the old version all the time.

Floor1　不点	Posted 2006-12-24 13:32
银牌会员　Posts 1,115　Credits 2,491
Please fully test, urgent! If this version is okay, it may become the final version!!! Everyone must work together!!! http://grub4dos.jot.com/ After testing for a few days (maybe only two or three days!), we will release the official version. The reason for being so urgent is to lay the groundwork for the release of WinGRUB. The overall structure of the version will no longer be changed, but if bugs are found, they will still be fixed. The official version should be released at the latest by December 31st, if there are no abnormalities. Due to insufficient time, the optimization plan for A20 has been canceled, but other plans have been implemented. Including the search for the default configuration file of grub.exe has been completed. Now grub.exe first searches for the menu.lst file in the current directory. If it fails, it then searches for the menu.lst file in the root directory of the current disk. If it also fails, it looks for the (hd0,0)/menu.lst file. If it fails again, then the built-in preset_menu will take control. The search for menu.lst in the disk and directory where grub.exe is located has been canceled. Because of insufficient time, this code cannot be written. It will not be written in the future either, because it is relatively complicated. Now the support for win98 by the emulated disk has been strengthened, and many faults and obstacles have been eliminated. (It runs satisfactorily on my machine). Everyone is best to test this too. Test the emulated disk with --mem and without --mem to see if it is normal under win98. Please report whether there is an improvement in stability when accessing the emulated disk after entering win98 compared with the previous old version. Thank you friends for working together. －－－－－－－－－－－－－－－－－－－－－－－－－－－－－－－ Download addresses for grub4dos and related tools added in 2007: http://download.gna.org/grub4dos/ http://download.gna.org/grubutil/ English and Chinese documents for grub4dos added in 2007: http://grub4dos.sourceforge.net/wiki The original GNU GRUB mailing list archives are very valuable for developers: http://lists.gnu.org/archive/html/bug-grub/ [ Last edited by 不点 on 2007-4-15 at 02:10 AM ]

Floor2　Wengier	Posted 2006-12-29 22:23
系统支持　Posts 10,521　Credits 27,736
These days I installed a new computer and used GRUB4DOS again. I found that the latest version of GRUB.EXE has to prob ROM INT every time it starts, causing it not to run or crash in many environments (even in pure DOS with only some simple drivers loaded), while the previous versions could. It is suggested to add an option to disable prob so that GRUB.EXE can still start in some environments where there are actually no problems.

Floor3　不点	Posted 2006-12-30 01:15
银牌会员　Posts 1,115　Credits 2,491
This problem can't be solved now. Let's release the official version first and solve it slowly later. Friends with this problem can temporarily continue to use the old version. ------------ The previous old version had no detection function, or it was a simple system identification function. The recovery of interrupt vectors was specially handled according to the characteristics of different systems. For different versions of MS-DOS, we spent a lot of space to handle it. This can be said to be very cumbersome. Now we use a unified method for detection, and any DOS, any driver, and any TSR are equal. The detection method is concise and clear. The next step, we only need to find corresponding measures for those crash failure situations. This is the next task after 0.4.2. ---------------- Now you can upload the problematic driver program and TSR (along with the commands in config.sys and autoexec.bat). I can study the causes of crashes and failures in the virtual machine.

Floor4　不点	Posted 2007-01-01 09:53
银牌会员　Posts 1,115　Credits 2,491
GRUB for DOS 0.4.2 official version has been released! Thanks to the support of all friends on this site! Download: http://grub.linuxeden.com/wakka.php?wakka=UploadFile http://sourceforge.net/projects/grub4dos/ http://grub4dos.jot.com/ http://sarovar.org/projects/grub4dos/ http://grub4dos.freespaces.com/

Floor5　tangtai	Posted 2007-01-03 01:12
等待验证用户　Posts 314　Credits 640
Bump

Floor6　taojiang	Posted 2007-01-04 01:05
初级用户　Posts 66　Credits 145
It seems that there is no difference at all between the official version and the version of December 28th.

Floor7　scriptor	Posted 2007-01-04 02:24
银牌会员　Posts 555　Credits 1,187
I happen to need it, but I don't know which version is more suitable. Please recommend it, experts.

Floor8　dl04118	Posted 2007-01-05 06:18
初级用户　Posts 32　Credits 199
I don't know how to use it, but I still download it to try, thank you

Floor9　martin325	Posted 2007-01-05 23:25
银牌会员　Posts 603　Credits 1,582
Agree with the statement of floor 6. The menu also shows the words "2006-12-28". The date in ChangeLog.txt also seems to be incorrect, such as "2006-10-23 version 0.4.2" [ Last edited by martin325 on 2007-1-5 at 11:27 PM ]

Floor10　fastslz	Posted 2007-01-11 03:09
铂金会员　Posts 2,315　Credits 5,493　From 上海
Test found that the 0.4.2 official version's map --mem method cannot boot a floppy image larger than 2.88MB, but pre11 can. title Test 10th image: 10.img map --mem (hd0,0)/img/10.img (fd0) map --hook chainloader (fd0)+1 rootnoverify (fd0)

Floor11　GOTOmsdos	Posted 2007-01-11 03:18
铂金会员　Posts 1,827　Credits 5,154
I've been using it all the time, and I haven't found any major problems.

Floor12　store88	Posted 2007-01-11 03:52
中级用户　Posts 125　Credits 416

Floor13　xieliuyun	Posted 2007-01-11 06:15
初级用户　Posts 7　Credits 120
Support, hope to increase ISO support!

Floor14　AlwaysInherit	Posted 2007-01-12 05:18
初级用户　Posts 46　Credits 112
Back then, the teacher said to learn assembly well, but I didn't listen! Now it's all good! Assembly is still completely incomprehensible, and I can't mix 16bit 32bit code. It's natural that I don't understand boot loader code Using the previous guess_xxx was not bad! Anyway, everyone's environment is similar Strongly support Wengier

Floor15　不点	Posted 2007-01-12 07:12
银牌会员　Posts 1,115　Credits 2,491
The original guess dos versions has been deleted. That method just works, but it's not perfect. The old method can only restore a few key interrupt vectors, while the new automatic detection fully restores all interrupt vectors from 00 to 7F. For interrupt numbers above 80h, basically no need to care (no need to restore). The new method is perfect, and it will bring a more real BIOS environment than the old method, thus bringing greater stability to the boot of the operating system (that is, the operation of the operating system after chainloader + boot). If you encounter a detection failure, please upload the drivers and TSRs that caused the failure, which may help solve the problem as soon as possible. In addition, if there is a problem with map --mem, you can also report it in detail. Don't just say it briefly, that's not useful. I can use map --mem on my real machine to boot a win98 virtual hard disk image, which is 37M in size, far exceeding 2.88M. If you don't report, this problem may continue to exist in the new version, and then you may have to use the old version all the time.