Menu
▾
▴
vectorlinux-devel
|
From: M0E L. <m0...@gm...> - 2007-08-09 19:28:54
|
I am not familiar with the code to the existing text-mode installer, and being the n00b that I am in gambas, I'm trying to work on the tools -> Repair Lilo option on the installer. I'd like to know what this needs to do. Will we offer options for both, LILO and GRUB? or just LILO? If we offer both, maybe change the button the read somethng like "Setup boot manager" and then make the user choose between lilo or grub. After the user makes the selection... How is the the configuration going to be done? Are the shell commands like "lilo" and "grubconfig" available on the temporary system ? is there a plan already in place for the algorithm??? Please fill me in here. |
|
From: Jose J. R. <jo...@gm...> - 2007-08-09 20:23:54
|
On 8/9/07, M0E Lnx <m0...@gm...> wrote: > I am not familiar with the code to the existing text-mode installer, > and being the n00b that I am in gambas, I'm trying to work on the > tools -> Repair Lilo option on the installer. > Good starting point... :) > > I'd like to know what this needs to do. > Will we offer options for both, LILO and GRUB? or just LILO? > If we offer both, maybe change the button the read somethng like > "Setup boot manager" and then make the user choose between lilo or > grub. > Agreed, there was no grub support on VL back when we started this... ;) > > After the user makes the selection... > How is the the configuration going to be done? > Are the shell commands like "lilo" and "grubconfig" available on the > temporary system ? > is there a plan already in place for the algorithm??? > > Please fill me in here. > No idea, this was just part of the wishlist... :( Regards, Joe1962 |
|
From: M0E L. <m0...@gm...> - 2007-08-09 20:51:22
|
After speaking to uel on IRC, I think I got enough info to start it... Let's see how this goes... Looks like we'll have 2 options here The current installer can do Xdialog... Now, I'm not sure which would be best, to call the current script and let it do it's xdialog thing, or to re-master the entire thing in gambas.. I'm open for suggestions. On 8/9/07, Jose J. Rodriguez <jo...@gm...> wrote: > On 8/9/07, M0E Lnx <m0...@gm...> wrote: > > I am not familiar with the code to the existing text-mode installer, > > and being the n00b that I am in gambas, I'm trying to work on the > > tools -> Repair Lilo option on the installer. > > > > Good starting point... :) > > > > > I'd like to know what this needs to do. > > Will we offer options for both, LILO and GRUB? or just LILO? > > If we offer both, maybe change the button the read somethng like > > "Setup boot manager" and then make the user choose between lilo or > > grub. > > > > Agreed, there was no grub support on VL back when we started this... ;) > > > > > After the user makes the selection... > > How is the the configuration going to be done? > > Are the shell commands like "lilo" and "grubconfig" available on the > > temporary system ? > > is there a plan already in place for the algorithm??? > > > > Please fill me in here. > > > > No idea, this was just part of the wishlist... :( > > Regards, > Joe1962 > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. > Still grepping through log files to find problems? Stop. > Now Search log events and configuration files using AJAX and a browser. > Download your FREE copy of Splunk now >> http://get.splunk.com/ > _______________________________________________ > Vectorlinux-devel mailing list > Vec...@li... > https://lists.sourceforge.net/lists/listinfo/vectorlinux-devel > |
|
From: M0E L. <m0...@gm...> - 2007-08-10 20:56:25
|
I decided to go back and try a different approach to the boot manager setup process. This time using the existing tools already in the system. We have vliloconf which pretty much already does everything we need it to do.. * Lets the user chose between LILO or GRUB * Sets up the bootsplash, append, Os's and all that other good stuff.. * Uses Xdialog when available. To make this work, we would need to call it straight from frmTools, and rename the "Repair LILO" button to something like "Setup Boot Manager" or something... This also means that we would not need to show the frmConfig window. The only downside I see to this is that it will not be themed like the rest of the installer... Unless there was a way to somehow embed this into the form like the gparted window. * cough* Joe1962 *cough* Other than that, I think this could be a good approach, it works for sure, and it would save us the time it would take to write the whole thing over... But again, I'd like to get input... What does every1 think ?? On 8/9/07, M0E Lnx <m0...@gm...> wrote: > After speaking to uel on IRC, I think I got enough info to start it... > Let's see how this goes... > > Looks like we'll have 2 options here > > The current installer can do Xdialog... > Now, I'm not sure which would be best, to call the current script and > let it do it's xdialog thing, or to re-master the entire thing in > gambas.. > > I'm open for suggestions. > > On 8/9/07, Jose J. Rodriguez <jo...@gm...> wrote: > > On 8/9/07, M0E Lnx <m0...@gm...> wrote: > > > I am not familiar with the code to the existing text-mode installer, > > > and being the n00b that I am in gambas, I'm trying to work on the > > > tools -> Repair Lilo option on the installer. > > > > > > > Good starting point... :) > > > > > > > > I'd like to know what this needs to do. > > > Will we offer options for both, LILO and GRUB? or just LILO? > > > If we offer both, maybe change the button the read somethng like > > > "Setup boot manager" and then make the user choose between lilo or > > > grub. > > > > > > > Agreed, there was no grub support on VL back when we started this... ;) > > > > > > > > After the user makes the selection... > > > How is the the configuration going to be done? > > > Are the shell commands like "lilo" and "grubconfig" available on the > > > temporary system ? > > > is there a plan already in place for the algorithm??? > > > > > > Please fill me in here. > > > > > > > No idea, this was just part of the wishlist... :( > > > > Regards, > > Joe1962 > > > > ------------------------------------------------------------------------- > > This SF.net email is sponsored by: Splunk Inc. > > Still grepping through log files to find problems? Stop. > > Now Search log events and configuration files using AJAX and a browser. > > Download your FREE copy of Splunk now >> http://get.splunk.com/ > > _______________________________________________ > > Vectorlinux-devel mailing list > > Vec...@li... > > https://lists.sourceforge.net/lists/listinfo/vectorlinux-devel > > > |
|
From: <ea...@gm...> - 2007-08-10 21:05:57
|
That would be indeed a time saver...however, maybe the installer code shoul= d be as much "pure" gambas as possible. But I don't want to trample on your idea either! Maybe vliloconf can be used initially until the installer can actually install something, at which point more attention can be dedicated to gettin= g everything "Gambas'ized" ;) Eug=E9ne |
|
From: Jose J. R. <jo...@gm...> - 2007-08-10 22:34:08
|
On 8/10/07, Eug=E9ne Suter <ea...@gm...> wrote: > That would be indeed a time saver...however, maybe the installer code sho= uld > be as much "pure" gambas as possible. I support the pure Gambas ideal as far as possible. > But I don't want to trample on your idea either! > Maybe vliloconf can be used initially until the installer can actually > install something, at which point more attention can be dedicated to gett= ing > everything "Gambas'ized" ;) This is pretty much what I was going to reply to M0E's post, but you got ahead of me. Let's use what we have for a first "draft", then work from there. Regards, Joe1962 |
|
From: M0E L. <m0...@gm...> - 2007-08-11 00:40:21
|
I agree... And that would also give it a consistant look.. I'll see what I can do to write a gambas based equivalent... all it takes to fire up the existing tool is a line of code anyway ;) On 8/10/07, Jose J. Rodriguez <jo...@gm...> wrote: > On 8/10/07, Eug=E9ne Suter <ea...@gm...> wrote: > > That would be indeed a time saver...however, maybe the installer code s= hould > > be as much "pure" gambas as possible. > > I support the pure Gambas ideal as far as possible. > > > But I don't want to trample on your idea either! > > Maybe vliloconf can be used initially until the installer can actually > > install something, at which point more attention can be dedicated to ge= tting > > everything "Gambas'ized" ;) > > This is pretty much what I was going to reply to M0E's post, but you > got ahead of me. Let's use what we have for a first "draft", then work > from there. > > Regards, > Joe1962 > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. > Still grepping through log files to find problems? Stop. > Now Search log events and configuration files using AJAX and a browser. > Download your FREE copy of Splunk now >> http://get.splunk.com/ > _______________________________________________ > Vectorlinux-devel mailing list > Vec...@li... > https://lists.sourceforge.net/lists/listinfo/vectorlinux-devel > |
|
From: Jose J. R. <jo...@gm...> - 2007-08-11 03:59:27
|
On 8/10/07, M0E Lnx <m0...@gm...> wrote: > Unless there was a way to somehow embed this into the form like the > gparted window. * cough* Joe1962 *cough* > I'll try and check that out next week (after an important new customer demo we'll be doing). Or you can easily figure out how I did it for cfdisk and gparted in the installer code... ;) Regards, Joe1962 |
|
From: <ea...@gm...> - 2007-08-14 20:09:11
|
Well, a little update on whats happening with the drive partitioning: I'm trying to get the installer to check for available space before installing anything (since this was one of the flaws that the text installe= r has: it will crash and burn if space runs out). This is easy enough to accomplish for the full-auto partitioning mode, sinc= e the installer creates the partitions, it knows how much space there is available. But for the "free-space mode" its a different story: the unused space value= s have to be obtained some other way, and from reading parted's documentation and asking on the #parted channel, libparted does not provide any way to know the amount of unused space on an *unmounted* partition. One possible way would be to mount ALL the partitions found and parse df, but thats a somewhat messy approach. The best would be to get the amount of unused space when the partitions are still unmounted (so these values can be displayed to the user before selecting a partition). I've opened a thread on the GParted forum about this ( http://gparted-forum.surf4.info/viewtopic.php?id=3D862), since they have obviously found a way to do this. Any ideas on solving this problem some other way? Cheers, Eug=E9ne |
|
From: M0E L. <m0...@gm...> - 2007-08-14 20:28:44
|
I would say mount the go the messy way... * Mount the partitions * Get your values * un-mount the partitions * move on... On 8/14/07, Eug=E9ne Suter <ea...@gm...> wrote: > Well, a little update on whats happening with the drive partitioning: > I'm trying to get the installer to check for available space before > installing anything (since this was one of the flaws that the text instal= ler > has: it will crash and burn if space runs out). > > This is easy enough to accomplish for the full-auto partitioning mode, si= nce > the installer creates the partitions, it knows how much space there is > available. > But for the "free-space mode" its a different story: the unused space val= ues > have to be obtained some other way, and from reading parted's documentati= on > and asking on the #parted channel, libparted does not provide any way to > know the amount of unused space on an *unmounted* partition. > > One possible way would be to mount ALL the partitions found and parse df, > but thats a somewhat messy approach. > The best would be to get the amount of unused space when the partitions a= re > still unmounted (so these values can be displayed to the user before > selecting a partition). > > I've opened a thread on the GParted forum about this > (http://gparted-forum.surf4.info/viewtopic.php?id=3D862), > since they have obviously found a way to do this. > Any ideas on solving this problem some other way? > > Cheers, > > Eug=E9ne > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. > Still grepping through log files to find problems? Stop. > Now Search log events and configuration files using AJAX and a browser. > Download your FREE copy of Splunk now >> http://get.splunk.com/ > _______________________________________________ > Vectorlinux-devel mailing list > Vec...@li... > https://lists.sourceforge.net/lists/listinfo/vectorlinux-devel > > |
|
From: uel a. <ue...@gm...> - 2007-08-14 20:29:42
|
you shouldnt need to know the unused space on a partition only the partitio= n size. since we should format the partitions, shouldnt we? Uel On 8/14/07, Eug=E9ne Suter <ea...@gm...> wrote: > > Well, a little update on whats happening with the drive partitioning: > I'm trying to get the installer to check for available space before > installing anything (since this was one of the flaws that the text instal= ler > has: it will crash and burn if space runs out). > > This is easy enough to accomplish for the full-auto partitioning mode, > since the installer creates the partitions, it knows how much space there= is > available. > But for the "free-space mode" its a different story: the unused space > values have to be obtained some other way, and from reading parted's > documentation and asking on the #parted channel, libparted does not provi= de > any way to know the amount of unused space on an *unmounted* partition. > > One possible way would be to mount ALL the partitions found and parse df, > but thats a somewhat messy approach. > The best would be to get the amount of unused space when the partitions > are still unmounted (so these values can be displayed to the user before > selecting a partition). > > I've opened a thread on the GParted forum about this ( > http://gparted-forum.surf4.info/viewtopic.php?id=3D862), since they have > obviously found a way to do this. > Any ideas on solving this problem some other way? > > Cheers, > > Eug=E9ne > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. > Still grepping through log files to find problems? Stop. > Now Search log events and configuration files using AJAX and a browser. > Download your FREE copy of Splunk now >> http://get.splunk.com/ > _______________________________________________ > Vectorlinux-devel mailing list > Vec...@li... > https://lists.sourceforge.net/lists/listinfo/vectorlinux-devel > > |
|
From: M0E L. <m0...@gm...> - 2007-08-14 20:40:28
|
Well, if this is for the "auto-partition" mode, then I think we definitely need to know how much space we got to play with ;) That way, the installer can allocate the necessary space for the necessary partitions. On 8/14/07, uel archuletta <ue...@gm...> wrote: > you shouldnt need to know the unused space on a partition only the partit= ion > size. since we should format the partitions, shouldnt we? > > Uel > > On 8/14/07, Eug=E9ne Suter <ea...@gm...> wrote: > > > > Well, a little update on whats happening with the drive partitioning: > > I'm trying to get the installer to check for available space before > installing anything (since this was one of the flaws that the text instal= ler > has: it will crash and burn if space runs out). > > > > This is easy enough to accomplish for the full-auto partitioning mode, > since the installer creates the partitions, it knows how much space there= is > available. > > But for the "free-space mode" its a different story: the unused space > values have to be obtained some other way, and from reading parted's > documentation and asking on the #parted channel, libparted does not provi= de > any way to know the amount of unused space on an *unmounted* partition. > > > > One possible way would be to mount ALL the partitions found and parse d= f, > but thats a somewhat messy approach. > > The best would be to get the amount of unused space when the partitions > are still unmounted (so these values can be displayed to the user before > selecting a partition). > > > > I've opened a thread on the GParted forum about this > (http://gparted-forum.surf4.info/viewtopic.php?id=3D862 ), > since they have obviously found a way to do this. > > Any ideas on solving this problem some other way? > > > > Cheers, > > > > Eug=E9ne > > > > > ------------------------------------------------------------------------- > > This SF.net email is sponsored by: Splunk Inc. > > Still grepping through log files to find problems? Stop. > > Now Search log events and configuration files using AJAX and a browser. > > Download your FREE copy of Splunk now >> http://get.splunk.com/ > > _______________________________________________ > > Vectorlinux-devel mailing list > > Vec...@li... > > > https://lists.sourceforge.net/lists/listinfo/vectorlinux-devel > > > > > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. > Still grepping through log files to find problems? Stop. > Now Search log events and configuration files using AJAX and a browser. > Download your FREE copy of Splunk now >> http://get.splunk.com/ > _______________________________________________ > Vectorlinux-devel mailing list > Vec...@li... > https://lists.sourceforge.net/lists/listinfo/vectorlinux-devel > > |
|
From: <ea...@gm...> - 2007-08-14 20:58:43
|
Well, for the automatic partitioning mode, the entire drive is erased and new partition sizes are calculated and compared to the amount of space that the packages will occupy. But for the free-space mode (in which only free space from an existing partition is used), it is not going to make any modifications to the drive at all until it is sure that the packages will fit in that unused space...thats why its important to know how much unused space that selected partition has. Using the df method is my last resort...its *really* messy (especially if the user has many partitions). On 14/08/07, M0E Lnx <m0...@gm...> wrote: > > Well, if this is for the "auto-partition" mode, then I think we > definitely need to know how much space we got to play with ;) > > That way, the installer can allocate the necessary space for the > necessary partitions. > > On 8/14/07, uel archuletta <ue...@gm...> wrote: > > you shouldnt need to know the unused space on a partition only the > partition > > size. since we should format the partitions, shouldnt we? > > > > Uel > > |
|
From: M0E L. <m0...@gm...> - 2007-08-14 21:06:56
|
Does this assume the available "free space" is on a linux friendly-formatted partition? or does this require resizing a hughe vfat or ntfs and creating the necessary ext* partitions? On 8/14/07, Eug=E9ne Suter <ea...@gm...> wrote: > Well, for the automatic partitioning mode, the entire drive is erased and > new partition sizes are calculated and compared to the amount of space th= at > the packages will occupy. > > But for the free-space mode (in which only free space from an existing > partition is used), it is not going to make any modifications to the driv= e > at all until it is sure that the packages will fit in that unused > space...thats why its important to know how much unused space that select= ed > partition has. > > Using the df method is my last resort...its *really* messy (especially if > the user has many partitions). > > > > On 14/08/07, M0E Lnx < m0...@gm...> wrote: > > Well, if this is for the "auto-partition" mode, then I think we > > definitely need to know how much space we got to play with ;) > > > > That way, the installer can allocate the necessary space for the > > necessary partitions. > > > > On 8/14/07, uel archuletta < ue...@gm...> wrote: > > > you shouldnt need to know the unused space on a partition only the > partition > > > size. since we should format the partitions, shouldnt we? > > > > > > Uel > > > > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. > Still grepping through log files to find problems? Stop. > Now Search log events and configuration files using AJAX and a browser. > Download your FREE copy of Splunk now >> http://get.splunk.com/ > _______________________________________________ > Vectorlinux-devel mailing list > Vec...@li... > https://lists.sourceforge.net/lists/listinfo/vectorlinux-devel > > |
|
From: <ea...@gm...> - 2007-08-14 21:34:22
|
Parted allows resizing of ntfs and fat partitions as well, so the idea is to allow the user to chose *any* partition he/she wishes to "steal" space from, even windows partitions. On 14/08/07, M0E Lnx <m0...@gm...> wrote: > > Does this assume the available "free space" is on a linux > friendly-formatted partition? or does this require resizing a hughe > vfat or ntfs and creating the necessary ext* partitions? > |
|
From: <ea...@gm...> - 2007-08-14 23:01:38
|
Seems that parted can be used for this after all (according to the post in the gparted forum): Lets say the partition you want to check is a fat32 partition on /dev/hda1. > > >parted /dev/hda1 > > bla bla bla > > >print 1 > > Minor: 1 > Flags: > File System: fat32 > Size: 10.0GB (100%) > Minimum size: 8369MB (83.7%) <------- RIGHT HERE > Maximum size: 10.0GB (100%) > > 100% - 83.7% = 16.3% free > > That's how GParted does it. Hope this helps. > > Patrick Verner > www.partedmagic.com Problem is, this only works for me on non ext* filesystems. Running parted against one of my ext3 backup partitions gives this error: root:# parted /dev/hda print 5 > Error: File system has an incompatible feature enabled. Anyone else get this problem on ext3 partitions as well? |
|
From: uel a. <ue...@gm...> - 2007-08-14 23:18:50
|
Sorry I didnt think about wanting to resize. i dont have any ext3 partitions but I do get this error with ext2 root:# parted /dev/hdb print 5 Error: File system was not cleanly unmounted! You should run e2fsck. Modifying an unclean file system could cause severe corruption. Ignore/Cancel? i Minor: 5 Flags: File System: ext2 Size: 10.5GB (4.20%) Minimum size: 9001MB (3.60%) Maximum size: 250GB (100%) and it seems the version of parted i have doesnt support ntfs root:# parted /dev/hda print 1 No Implementation: Support for opening ntfs file systems is not implemented yet. but works well with fat32 root:# parted /dev/hdb print 1 Minor: 1 Flags: boot, lba File System: fat32 Size: 105GB (41.9%) Minimum size: 97.5GB (39.0%) Maximum size: 250GB (100%) uel On 8/14/07, Eug=E9ne Suter <ea...@gm...> wrote: > > Seems that parted can be used for this after all (according to the post i= n > the gparted forum): > > > Lets say the partition you want to check is a fat32 partition on > > /dev/hda1. > > > > >parted /dev/hda1 > > > > bla bla bla > > > > >print 1 > > > > Minor: 1 > > Flags: > > File System: fat32 > > Size: 10.0GB (100%) > > Minimum size: 8369MB (83.7%) <------- RIGHT HERE > > Maximum size: 10.0GB (100%) > > > > 100% - 83.7% =3D 16.3% free > > > > That's how GParted does it. Hope this helps. > > > > Patrick Verner > > www.partedmagic.com > > > > Problem is, this only works for me on non ext* filesystems. > Running parted against one of my ext3 backup partitions gives this error: > > > root:# parted /dev/hda print 5 > > Error: File system has an incompatible feature enabled. > > > Anyone else get this problem on ext3 partitions as well? > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. > Still grepping through log files to find problems? Stop. > Now Search log events and configuration files using AJAX and a browser. > Download your FREE copy of Splunk now >> http://get.splunk.com/ > _______________________________________________ > Vectorlinux-devel mailing list > Vec...@li... > https://lists.sourceforge.net/lists/listinfo/vectorlinux-devel > > |
|
From: M0E L. <m0...@gm...> - 2007-08-15 00:34:52
|
So then, you do need to know the ammount of free space left on the device so you can then resize which ever partition is the largest and steal space to make your Linux friendly partition. This involves * an analysis of the drive to determine what partitions are present. * If in Wind OS, most likely one hughe partition for the system and mayne a backup / restore partition * Maybe analyze the size of each partition to determine which one is the largest and also to get a correct label for your new partitons (/dev/hdxx) On 8/14/07, uel archuletta <ue...@gm...> wrote: > Sorry I didnt think about wanting to resize. > i dont have any ext3 partitions but I do get this error with ext2 > root:# parted /dev/hdb print 5 > Error: File system was not cleanly unmounted! You should run e2fsck. > Modifying > an unclean file system could cause severe corruption. > Ignore/Cancel? i > > Minor: 5 > Flags: > File System: ext2 > Size: 10.5GB (4.20%) > Minimum size: 9001MB (3.60%) > Maximum size: 250GB (100%) > > and it seems the version of parted i have doesnt support ntfs > root:# parted /dev/hda print 1 > No Implementation: Support for opening ntfs file systems is not implement= ed > yet. > but works well with fat32 > root:# parted /dev/hdb print 1 > > Minor: 1 > Flags: boot, lba > File System: fat32 > Size: 105GB (41.9%) > Minimum size: 97.5GB (39.0%) > Maximum size: 250GB (100%) > > uel > > On 8/14/07, Eug=E9ne Suter <ea...@gm...> wrote: > > > > Seems that parted can be used for this after all (according to the post= in > the gparted forum): > > > > > > > > > Lets say the partition you want to check is a fat32 partition on > /dev/hda1. > > > > > > >parted /dev/hda1 > > > > > > bla bla bla > > > > > > >print 1 > > > > > > Minor: 1 > > > Flags: > > > File System: fat32 > > > Size: 10.0GB (100%) > > > Minimum size: 8369MB (83.7%) <------- RIGHT HERE > > > Maximum size: 10.0GB (100%) > > > > > > 100% - 83.7% =3D 16.3% free > > > > > > That's how GParted does it. Hope this helps. > > > > > > Patrick Verner > > > www.partedmagic.com > > > > > > > > Problem is, this only works for me on non ext* filesystems. > > Running parted against one of my ext3 backup partitions gives this erro= r: > > > > > > > > > root:# parted /dev/hda print 5 > > > Error: File system has an incompatible feature enabled. > > > > Anyone else get this problem on ext3 partitions as well? > > > > > > > ------------------------------------------------------------------------- > > This SF.net email is sponsored by: Splunk Inc. > > Still grepping through log files to find problems? Stop. > > Now Search log events and configuration files using AJAX and a browser. > > Download your FREE copy of Splunk now >> http://get.splunk.com/ > > _______________________________________________ > > Vectorlinux-devel mailing list > > Vec...@li... > > > https://lists.sourceforge.net/lists/listinfo/vectorlinux-devel > > > > > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. > Still grepping through log files to find problems? Stop. > Now Search log events and configuration files using AJAX and a browser. > Download your FREE copy of Splunk now >> http://get.splunk.com/ > _______________________________________________ > Vectorlinux-devel mailing list > Vec...@li... > https://lists.sourceforge.net/lists/listinfo/vectorlinux-devel > > |
|
From: Sriram D. <sri...@gm...> - 2007-08-15 17:39:22
|
Is it part of your plan to allow multiple partitions to be marked for using free space from? if so then going with your mount approach may not be that bad. show a lis tof all partitions + free space on partition. let the user make a decision on which partitions should be resized. the installer can provide a basic visual check that the required free space has been achieved. and then un mount every thing and go on with your usual algorithm of resizing the selected partitions and using the resulting free space. This approach makes good use of mounting all partitions and hopefully will handle ext partitions also. cheers ram On 8/14/07, M0E Lnx <m0...@gm...> wrote: > > So then, you do need to know the ammount of free space left on the > device so you can then resize which ever partition is the largest and > steal space to make your Linux friendly partition. > This involves > * an analysis of the drive to determine what partitions are present. > * If in Wind OS, most likely one hughe partition for the system and > mayne a backup / restore partition > * Maybe analyze the size of each partition to determine which one is > the largest and also to get a correct label for your new partitons > (/dev/hdxx) > > > > On 8/14/07, uel archuletta <ue...@gm...> wrote: > > Sorry I didnt think about wanting to resize. > > i dont have any ext3 partitions but I do get this error with ext2 > > root:# parted /dev/hdb print 5 > > Error: File system was not cleanly unmounted! You should run e2fsck. > > Modifying > > an unclean file system could cause severe corruption. > > Ignore/Cancel? i > > > > Minor: 5 > > Flags: > > File System: ext2 > > Size: 10.5GB (4.20%) > > Minimum size: 9001MB (3.60%) > > Maximum size: 250GB (100%) > > > > and it seems the version of parted i have doesnt support ntfs > > root:# parted /dev/hda print 1 > > No Implementation: Support for opening ntfs file systems is not > implemented > > yet. > > but works well with fat32 > > root:# parted /dev/hdb print 1 > > > > Minor: 1 > > Flags: boot, lba > > File System: fat32 > > Size: 105GB (41.9%) > > Minimum size: 97.5GB (39.0%) > > Maximum size: 250GB (100%) > > > > uel > > > > On 8/14/07, Eug=E9ne Suter <ea...@gm...> wrote: > > > > > > Seems that parted can be used for this after all (according to the > post in > > the gparted forum): > > > > > > > > > > > > > Lets say the partition you want to check is a fat32 partition on > > /dev/hda1. > > > > > > > > >parted /dev/hda1 > > > > > > > > bla bla bla > > > > > > > > >print 1 > > > > > > > > Minor: 1 > > > > Flags: > > > > File System: fat32 > > > > Size: 10.0GB (100%) > > > > Minimum size: 8369MB (83.7%) <------- RIGHT HERE > > > > Maximum size: 10.0GB (100%) > > > > > > > > 100% - 83.7% =3D 16.3% free > > > > > > > > That's how GParted does it. Hope this helps. > > > > > > > > Patrick Verner > > > > www.partedmagic.com > > > > > > > > > > > > Problem is, this only works for me on non ext* filesystems. > > > Running parted against one of my ext3 backup partitions gives this > error: > > > > > > > > > > > > > root:# parted /dev/hda print 5 > > > > Error: File system has an incompatible feature enabled. > > > > > > Anyone else get this problem on ext3 partitions as well? > > > > > > > > > > > > ------------------------------------------------------------------------- > > > This SF.net email is sponsored by: Splunk Inc. > > > Still grepping through log files to find problems? Stop. > > > Now Search log events and configuration files using AJAX and a > browser. > > > Download your FREE copy of Splunk now >> http://get.splunk.com/ > > > _______________________________________________ > > > Vectorlinux-devel mailing list > > > Vec...@li... > > > > > https://lists.sourceforge.net/lists/listinfo/vectorlinux-devel > > > > > > > > > > > > > ------------------------------------------------------------------------- > > This SF.net email is sponsored by: Splunk Inc. > > Still grepping through log files to find problems? Stop. > > Now Search log events and configuration files using AJAX and a browser. > > Download your FREE copy of Splunk now >> http://get.splunk.com/ > > _______________________________________________ > > Vectorlinux-devel mailing list > > Vec...@li... > > https://lists.sourceforge.net/lists/listinfo/vectorlinux-devel > > > > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. > Still grepping through log files to find problems? Stop. > Now Search log events and configuration files using AJAX and a browser. > Download your FREE copy of Splunk now >> http://get.splunk.com/ > _______________________________________________ > Vectorlinux-devel mailing list > Vec...@li... > https://lists.sourceforge.net/lists/listinfo/vectorlinux-devel > |
|
From: Jose J. R. <jo...@gm...> - 2007-08-15 17:55:42
|
On 8/15/07, Sriram Durbha <sri...@gm...> wrote: > Is it part of your plan to allow multiple partitions to be marked for using > free space from? There is UI and code in place that will allow selecting multiple HDDs and partitions. Now, regarding the current discussion and commits, I am sorry I don't have a working VL since last week and cannot yet review the code (so going by my bad memory here), because I have been migrating to another laptop as the previous one was giving loads of weird issues. However, I have been reading the commits on this list and it seems to me (though I might be wrong for the reasons stated) that there is some duplication going on. There is already infrastructure in place to hold all the HDDs and partitions in global array structures. IIRC, this also includes the HDD and partition sizes. The thing that was missing, I think, was a similar way to hold the free space. Regards, Joe1962 |
|
From: M0E L. <m0...@gm...> - 2007-08-15 18:21:47
|
I hope we're not going around in circles.... Quick... get your laptop..... On 8/15/07, Jose J. Rodriguez <jo...@gm...> wrote: > On 8/15/07, Sriram Durbha <sri...@gm...> wrote: > > Is it part of your plan to allow multiple partitions to be marked for using > > free space from? > > There is UI and code in place that will allow selecting multiple HDDs > and partitions. > > Now, regarding the current discussion and commits, I am sorry I don't > have a working VL since last week and cannot yet review the code (so > going by my bad memory here), because I have been migrating to another > laptop as the previous one was giving loads of weird issues. However, > I have been reading the commits on this list and it seems to me > (though I might be wrong for the reasons stated) that there is some > duplication going on. There is already infrastructure in place to hold > all the HDDs and partitions in global array structures. IIRC, this > also includes the HDD and partition sizes. The thing that was missing, > I think, was a similar way to hold the free space. > > Regards, > Joe1962 > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. > Still grepping through log files to find problems? Stop. > Now Search log events and configuration files using AJAX and a browser. > Download your FREE copy of Splunk now >> http://get.splunk.com/ > _______________________________________________ > Vectorlinux-devel mailing list > Vec...@li... > https://lists.sourceforge.net/lists/listinfo/vectorlinux-devel > |
|
From: <ea...@gm...> - 2007-08-15 19:29:35
|
Indeed, there is some duplication going on...but that can be cleaned up ;) I guess it wouldn't be a problem to add the free space info to the arrays you originally created for the hard-drive information...just trying to find out why exactly ext2/ext3 partitions give those errors in parted first... On 15/08/07, Jose J. Rodriguez <jo...@gm...> wrote: > There is UI and code in place that will allow selecting multiple HDDs > and partitions. > > Now, regarding the current discussion and commits, I am sorry I don't > have a working VL since last week and cannot yet review the code (so > going by my bad memory here), because I have been migrating to another > laptop as the previous one was giving loads of weird issues. However, > I have been reading the commits on this list and it seems to me > (though I might be wrong for the reasons stated) that there is some > duplication going on. There is already infrastructure in place to hold > all the HDDs and partitions in global array structures. IIRC, this > also includes the HDD and partition sizes. The thing that was missing, > I think, was a similar way to hold the free space. > > Regards, > Joe1962 > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. > Still grepping through log files to find problems? Stop. > Now Search log events and configuration files using AJAX and a browser. > Download your FREE copy of Splunk now >> http://get.splunk.com/ > _______________________________________________ > Vectorlinux-devel mailing list > Vec...@li... > https://lists.sourceforge.net/lists/listinfo/vectorlinux-devel > |
|
From: M0E L. <m0...@gm...> - 2007-08-15 20:27:40
|
I wonder if I doubled anything up ? On 8/15/07, Eug=E9ne Suter <ea...@gm...> wrote: > Indeed, there is some duplication going on...but that can be cleaned up ;= ) > I guess it wouldn't be a problem to add the free space info to the arrays > you originally created for the hard-drive information...just trying to fi= nd > out why exactly ext2/ext3 partitions give those errors in parted first... > > > On 15/08/07, Jose J. Rodriguez <jo...@gm...> wrote: > > There is UI and code in place that will allow selecting multiple HDDs > > and partitions. > > > > Now, regarding the current discussion and commits, I am sorry I don't > > have a working VL since last week and cannot yet review the code (so > > going by my bad memory here), because I have been migrating to another > > laptop as the previous one was giving loads of weird issues. However, > > I have been reading the commits on this list and it seems to me > > (though I might be wrong for the reasons stated) that there is some > > duplication going on. There is already infrastructure in place to hold > > all the HDDs and partitions in global array structures. IIRC, this > > also includes the HDD and partition sizes. The thing that was missing, > > I think, was a similar way to hold the free space. > > > > Regards, > > Joe1962 > > > > > ------------------------------------------------------------------------- > > This SF.net email is sponsored by: Splunk Inc. > > Still grepping through log files to find problems? Stop. > > Now Search log events and configuration files using AJAX and a browser. > > Download your FREE copy of Splunk now >> http://get.splunk.com/ > > _______________________________________________ > > Vectorlinux-devel mailing list > > Vec...@li... > > > https://lists.sourceforge.net/lists/listinfo/vectorlinux-devel > > > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. > Still grepping through log files to find problems? Stop. > Now Search log events and configuration files using AJAX and a browser. > Download your FREE copy of Splunk now >> http://get.splunk.com/ > _______________________________________________ > Vectorlinux-devel mailing list > Vec...@li... > https://lists.sourceforge.net/lists/listinfo/vectorlinux-devel > > |
|
From: <ea...@gm...> - 2007-08-15 21:11:22
|
On 15/08/07, M0E Lnx <m0...@gm...> wrote: > > I wonder if I doubled anything up ? No, its me: in the functions for acquiring partition size with fdisk, this Joe has already done and saved the data in an array...all I have to do is use it... :\ |
|
From: <ea...@gm...> - 2007-08-17 21:16:39
|
Looks like there is a solution for the ext2/ext3 partition free space without using parted :) Parsing the output of dumpe2fs (with the -h flag) will give the number of free blocks and the block size. Got this by looking through the GParted source code, so its probably a good idea to give a mention in the installer's comments... Now to add this to the PartInfo array! :) |
×
Want the latest updates on software, tech news, and AI?
Get latest updates about software, tech news, and AI from SourceForge directly in your inbox once a month.
Thanks for helping keep SourceForge clean.
X