Is the new geometry backward-compatible? If i recreate the partitions, will they be visible when i regress test revisions before 54511?
Why should we have to erase existing partitions in the first place ? ReactOS has to be able to read valid partitions without destroying them. Kind regards, Sylvain Petreolle
----- Mail original -----
De : "caemyr@myopera.com" caemyr@myopera.com À : ros-dev@reactos.org Cc : Envoyé le : Dimanche 27 Novembre 2011 15h23 Objet : [ros-dev] HDD Geometry switch
Is the new geometry backward-compatible? If i recreate the partitions, will they be visible when i regress test revisions before 54511?
-- With best regards Caemyr
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
Sylvain Petreolle wrote:
Why should we have to erase existing partitions in the first place ? ReactOS has to be able to read valid partitions without destroying them.
Kind regards, Sylvain Petreolle
The partition table must be recreated beause the geometry of the disk changes. In most cases users will not be able to access a partition created by ReactOS using Windows or Linux because of the different geometries. I tested the old geometry using gparted and it always reported a damaged partition table until I switched to the new geometry.
Unfortunately there is no reliable method to derive the disk geometry from the partition table. That's why ReactOS cannot detect the old geometry in a reliabe way. And since Windows and Linux do not care about the reported geometry of the harddisk and always use the 63/255 geometry instead, I don't see any reason not to use the new geometry.
BTW, Windows and Linx switched to the 'new' geometry more than 10 years ago. It's time for us to get rid of this ancient feature (or burden) too.
PS: GPT partitions are geometry agnostic. For them a partition is just a sequence of sectors.
Regards Eric
Hi,
in most of the cases the old CHS-geometry isn't used anymore (it can handle only 8.4 GB btw), so access to the old/new partitions should be possible to both the old and the new versions. There is one exception: some bootloaders do have problems with changing the geometry, because they need e.g. some additional fixed spaces (e.g. the sectors directly behind the mbr), hardcoded values etc.
Regards, Michael Fritscher
[00:18:28] <rafalh> iirc it affects only big drives [00:18:41] <rafalh> I have 9 gb drive and I see it in ros without recreating
Could someone please make a reason and decide if partitions need to be recreated or not? This is all really confusing.
The partition table must be recreated beause the geometry of the disk changes. In most cases users will not be able to access a partition created by ReactOS using Windows or Linux because of the different geometries. I tested the old geometry using gparted and it always reported a damaged partition table until I switched to the new geometry.
Unfortunately there is no reliable method to derive the disk geometry from the partition table. That's why ReactOS cannot detect the old geometry in a reliabe way. And since Windows and Linux do not care about the reported geometry of the harddisk and always use the 63/255 geometry instead, I don't see any reason not to use the new geometry.
Why can't we use the Int 13h drive parameters returned by the BIOS that FreeLoader saves in the CM_DISK_GEOMETRY_DEVICE_DATA struct in the registry for each disk? That's got all the information we need to easily distinguish between 32/64 and 63/255 disks. Class2 would need a bit of modification to parse the hardware resources for the disk in order to find the struct, but I could make a patch if you think it's an idea worth pursuing.
Regards, Cameron
caemyr@myopera.com wrote:
Is the new geometry backward-compatible? If i recreate the partitions, will they be visible when i regress test revisions before 54511?
This is a one-way road. If you want to test older revisions, you must delete and re-create partitions again.
Regards Eric
Am 27.11.2011 16:48, schrieb Eric Kohl:
caemyr@myopera.com wrote:
Is the new geometry backward-compatible? If i recreate the partitions, will they be visible when i regress test revisions before 54511?
This is a one-way road. If you want to test older revisions, you must delete and re-create partitions again.
I don't really understand that. The partitions we currently use are valid. windows can read them. So should reactos. With the new code as with the old code. If the new code cannot deal with the old partitions then it's broken imo.
Timo
Sir, you do realise that you just fucked up regression testing on real hardware, as well as made one on VM way more cumbersome???
Did you consider the impact of this change upon other people at all?
On Sunday, November 27, 2011 4:48 PM, "Eric Kohl" eric.kohl@t-online.de wrote:
caemyr@myopera.com wrote:
Is the new geometry backward-compatible? If i recreate the partitions, will they be visible when i regress test revisions before 54511?
This is a one-way road. If you want to test older revisions, you must delete and re-create partitions again.
Regards Eric
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
Also, knowing the consequences, did you bother to issue any message to our community, to inform them about this?? Not every ReactOS user is subsribed to ros-dev.
Ah wait... you ignored that as well. You will let others handle the problem as usual.
I have 5 sets for testing, 2 real hardware and 3 vm. Each one with two partitions, one for ros and other for all the installers and drivers.
So now, whenever i cross the border of this change, i will have to wipe all, recreate the partitions and copy the stuff back? It will take no less than half an hour for VM. On real hardware, just the copying can take up to 1 hour.
Do you value your time available to spend on ReactOS? Then perhaps you could also think about others as well?
On Sunday, November 27, 2011 4:48 PM, "Eric Kohl" eric.kohl@t-online.de wrote:
caemyr@myopera.com wrote:
Is the new geometry backward-compatible? If i recreate the partitions, will they be visible when i regress test revisions before 54511?
This is a one-way road. If you want to test older revisions, you must delete and re-create partitions again.
Regards Eric
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
caemyr@myopera.com wrote:
Do you value your time available to spend on ReactOS? Then perhaps you could also think about others as well?
Constructively asking, what other way would we have?
Announcing this change some days in advance won't make things easier. And sticking to the old geometry till eternity is no solution either. Apart from this, it's a known fact that developers are only available on ros-dev or #reactos-dev. We simply have no other official communication channels for announcements like this.
In any case, I recommend you to choose your words more carefully towards someone who is part of the project for more than 10 years. I guess, just a few of today's developers even knew that remnants like this old disk geometry were still in use.
Cheers,
Colin
I have chosen my words very carefully. I have been only few years within the project but this weekend's change has locked me out from testing, according to Eric's own words.
Others say different things, but i assume they cannot be right, since the commiter said what he said?
As for communication, we have also webpage, news and forum. Any word there? Not a single one.
Regards
On Monday, November 28, 2011 12:48 AM, "Colin Finck" colin@reactos.org wrote:
caemyr@myopera.com wrote:
Do you value your time available to spend on ReactOS? Then perhaps you could also think about others as well?
Constructively asking, what other way would we have?
Announcing this change some days in advance won't make things easier. And sticking to the old geometry till eternity is no solution either. Apart from this, it's a known fact that developers are only available on ros-dev or #reactos-dev. We simply have no other official communication channels for announcements like this.
In any case, I recommend you to choose your words more carefully towards someone who is part of the project for more than 10 years. I guess, just a few of today's developers even knew that remnants like this old disk geometry were still in use.
Cheers,
Colin
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
Hello Olaf,
don't you think you are overreacting a bit? Did you have a look at my patch? After witnessing your reaction, I do not think you had a look at the patch. It changes exactly 4 (four!!!!) lines of code! See the patch below:
/* Start of Patch */ --- trunk/reactos/drivers/storage/class/class2/class2.c 2011/04/23 10:52:01 51437 +++ trunk/reactos/drivers/storage/class/class2/class2.c 2011/11/27 14:18:40 54511 @@ -29,12 +29,12 @@ #define START_UNIT_TIMEOUT 30
/* Disk layout used by Windows NT4 and earlier versions. */ -#define DEFAULT_SECTORS_PER_TRACK 32 -#define DEFAULT_TRACKS_PER_CYLINDER 64 +//#define DEFAULT_SECTORS_PER_TRACK 32 +//#define DEFAULT_TRACKS_PER_CYLINDER 64
/* Disk layout used by Windows 2000 and later versions. */ -//#define DEFAULT_SECTORS_PER_TRACK 63 -//#define DEFAULT_TRACKS_PER_CYLINDER 255 +#define DEFAULT_SECTORS_PER_TRACK 63 +#define DEFAULT_TRACKS_PER_CYLINDER 255
NTSTATUS NTAPI /* End of Patch */
If you really need the old geometry, you can easily change the file drivers/storage/class/class2/class2.c to enable the old geometry again. You only have to enable the lines 32 and 33 and disable the lines 36 and 37. This local modification will not be overwritten by SVN updates.
BTW, after the last WINE synch of rpcrt4.dll I spent 3 weekends to fix the service manager and my local changes. Did I complain? No! Because that's the price we have to pay if we want to improve ReactOS. Sometimes we have to take one step back if we want to get two steps ahead. If we stopped improving ReactOS as soon as the first developer or tester complained about a particular change, we would still boot ReactOS using the DOS loader and plan to release version 0.1 within the next 5 years. Stop crying and get over it.
Regards, Eric
caemyr@myopera.com wrote:
I have chosen my words very carefully. I have been only few years within the project but this weekend's change has locked me out from testing, according to Eric's own words.
Others say different things, but i assume they cannot be right, since the commiter said what he said?
As for communication, we have also webpage, news and forum. Any word there? Not a single one.
Regards
On Monday, November 28, 2011 12:48 AM, "Colin Finck"colin@reactos.org wrote:
caemyr@myopera.com wrote:
Do you value your time available to spend on ReactOS? Then perhaps you could also think about others as well?
Constructively asking, what other way would we have?
Announcing this change some days in advance won't make things easier. And sticking to the old geometry till eternity is no solution either. Apart from this, it's a known fact that developers are only available on ros-dev or #reactos-dev. We simply have no other official communication channels for announcements like this.
In any case, I recommend you to choose your words more carefully towards someone who is part of the project for more than 10 years. I guess, just a few of today's developers even knew that remnants like this old disk geometry were still in use.
Cheers,
Colin
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
Should i mention again, that i`m not a dev and i never posed as one? You could simply set it as optional flag. You could wait with it at least after the release is done and we can concentrate on new changes. You could push this change with, for example, major compiler change that would block rebuilding anyway.
There could be several more things possibly done to minimize the impact of this. You used neither. This is what i`m angry about. Look at CMake, how long the transition is taking place, how are its effects taken into consideration. This change needed at least 10% of it. It got none and i cant stop thinking its simply because most of devs will even know its there. Its a pity that this does not extend to all of us:/
Lets finish the discussion. The odds are all against me. We will simply need to live with its effects. I just hope that my pessimism is not correct. I hate being right:/
End of Drama. Sorry to everyone, especially Eric, who is feeling offended by my outbreak, I cannot promise it wont happen in the future again, so i`m not doing so.
Regards
On Monday, November 28, 2011 11:00 PM, "Eric Kohl" eric.kohl@t-online.de wrote:
Hello Olaf,
don't you think you are overreacting a bit? Did you have a look at my patch? After witnessing your reaction, I do not think you had a look at the patch. It changes exactly 4 (four!!!!) lines of code! See the patch below:
/* Start of Patch */ --- trunk/reactos/drivers/storage/class/class2/class2.c 2011/04/23 10:52:01 51437 +++ trunk/reactos/drivers/storage/class/class2/class2.c 2011/11/27 14:18:40 54511 @@ -29,12 +29,12 @@ #define START_UNIT_TIMEOUT 30
/* Disk layout used by Windows NT4 and earlier versions. */ -#define DEFAULT_SECTORS_PER_TRACK 32 -#define DEFAULT_TRACKS_PER_CYLINDER 64 +//#define DEFAULT_SECTORS_PER_TRACK 32 +//#define DEFAULT_TRACKS_PER_CYLINDER 64
/* Disk layout used by Windows 2000 and later versions. */ -//#define DEFAULT_SECTORS_PER_TRACK 63 -//#define DEFAULT_TRACKS_PER_CYLINDER 255 +#define DEFAULT_SECTORS_PER_TRACK 63 +#define DEFAULT_TRACKS_PER_CYLINDER 255
NTSTATUS NTAPI /* End of Patch */
If you really need the old geometry, you can easily change the file drivers/storage/class/class2/class2.c to enable the old geometry again. You only have to enable the lines 32 and 33 and disable the lines 36 and 37. This local modification will not be overwritten by SVN updates.
BTW, after the last WINE synch of rpcrt4.dll I spent 3 weekends to fix the service manager and my local changes. Did I complain? No! Because that's the price we have to pay if we want to improve ReactOS. Sometimes we have to take one step back if we want to get two steps ahead. If we stopped improving ReactOS as soon as the first developer or tester complained about a particular change, we would still boot ReactOS using the DOS loader and plan to release version 0.1 within the next 5 years. Stop crying and get over it.
Regards, Eric
caemyr@myopera.com wrote:
I have chosen my words very carefully. I have been only few years within the project but this weekend's change has locked me out from testing, according to Eric's own words.
Others say different things, but i assume they cannot be right, since the commiter said what he said?
As for communication, we have also webpage, news and forum. Any word there? Not a single one.
Regards
On Monday, November 28, 2011 12:48 AM, "Colin Finck"colin@reactos.org wrote:
caemyr@myopera.com wrote:
Do you value your time available to spend on ReactOS? Then perhaps you could also think about others as well?
Constructively asking, what other way would we have?
Announcing this change some days in advance won't make things easier. And sticking to the old geometry till eternity is no solution either. Apart from this, it's a known fact that developers are only available on ros-dev or #reactos-dev. We simply have no other official communication channels for announcements like this.
In any case, I recommend you to choose your words more carefully towards someone who is part of the project for more than 10 years. I guess, just a few of today's developers even knew that remnants like this old disk geometry were still in use.
Cheers,
Colin
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
caemyr@myopera.com wrote:
Should i mention again, that i`m not a dev and i never posed as one? You could simply set it as optional flag. You could wait with it at least after the release is done and we can concentrate on new changes. You could push this change with, for example, major compiler change that would block rebuilding anyway.
Olaf, I announced the change! You did not respond! Please tell me how long do I have to wait for a response? A week? Two weeks? Six months?
Since you insist on not being a developer, I attached a special class2.c file which is configured to use the old geometry. Replace the drivers\storage\class\class2\class2.c file with the attached one and build ReactOS and you are using the old geometry again.
There could be several more things possibly done to minimize the impact of this. You used neither. This is what i`m angry about. Look at CMake, how long the transition is taking place, how are its effects taken into consideration. This change needed at least 10% of it. It got none and i cant stop thinking its simply because most of devs will even know its there. Its a pity that this does not extend to all of us:/
Please do not compare my tiny patch with the CMake changes.
Lets finish the discussion. The odds are all against me. We will simply need to live with its effects. I just hope that my pessimism is not correct. I hate being right:/
End of Drama. Sorry to everyone, especially Eric, who is feeling offended by my outbreak, I cannot promise it wont happen in the future again, so i`m not doing so.
I am not feeling offended at all.
Regards Eric
On 28.11.2011 3:48, Colin Finck wrote:
caemyr@myopera.com wrote:
Do you value your time available to spend on ReactOS? Then perhaps you could also think about others as well?
Constructively asking, what other way would we have?
Announcing this change some days in advance won't make things easier. And sticking to the old geometry till eternity is no solution either. Apart from this, it's a known fact that developers are only available on ros-dev or #reactos-dev. We simply have no other official communication channels for announcements like this.
In any case, I recommend you to choose your words more carefully towards someone who is part of the project for more than 10 years. I guess, just a few of today's developers even knew that remnants like this old disk geometry were still in use.
Cheers,
Colin
No drama please!
Indeed let's think constructively, and also let's try to find a good solution. But before this, do we really have the problem?
WBR, Aleksey.
No drama please!
Indeed let's think constructively, and also let's try to find a good solution. But before this, do we really have the problem?
WBR, Aleksey.
Hello Aleksey,
perhaps my warning about the potential problems with the new disk geometry was a bit too loud. But I thought that I should warn everyone about potential issues instead of ignoring them.
The only situations where the new disk geometry could cause real problems require a harddisk in CHS mode. Calculating an LBA address from a CHS address or vice versa require the correct disk geometry.
If the harddisk uses the LBA mode, disk checkers can cause trouble when they try to fix the disk. This can happen because partitions might not be cylinder-aligned.
Right now, I don't see any other issues. But I might be wrong.
Regards Eric
Do you have to work at being this rude and disrespectful or does it just come naturally to you? This is the a great example of why the reactos forums are such a nasty place to visit. You must be quite proud.
-----Original Message----- From: ros-dev-bounces@reactos.org [mailto:ros-dev-bounces@reactos.org] On Behalf Of caemyr@myopera.com Sent: 27 November 2011 23:07 To: ReactOS Development List Subject: Re: [ros-dev] HDD Geometry switch
Also, knowing the consequences, did you bother to issue any message to our community, to inform them about this?? Not every ReactOS user is subsribed to ros-dev.
Ah wait... you ignored that as well. You will let others handle the problem as usual.
I have 5 sets for testing, 2 real hardware and 3 vm. Each one with two partitions, one for ros and other for all the installers and drivers.
So now, whenever i cross the border of this change, i will have to wipe all, recreate the partitions and copy the stuff back? It will take no less than half an hour for VM. On real hardware, just the copying can take up to 1 hour.
Do you value your time available to spend on ReactOS? Then perhaps you could also think about others as well?
On Sunday, November 27, 2011 4:48 PM, "Eric Kohl" eric.kohl@t-online.de wrote:
caemyr@myopera.com wrote:
Is the new geometry backward-compatible? If i recreate the partitions,
will they be visible when i regress test revisions before 54511?
This is a one-way road. If you want to test older revisions, you must delete and re-create partitions again.
Regards Eric
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
"No drama please!" - Aleksey Bragin
2011/11/28 Ged Murphy gedmurphy.maillists@gmail.com:
Do you have to work at being this rude and disrespectful or does it just come naturally to you? This is the a great example of why the reactos forums are such a nasty place to visit. You must be quite proud.
-----Original Message----- From: ros-dev-bounces@reactos.org [mailto:ros-dev-bounces@reactos.org] On Behalf Of caemyr@myopera.com Sent: 27 November 2011 23:07 To: ReactOS Development List Subject: Re: [ros-dev] HDD Geometry switch
Also, knowing the consequences, did you bother to issue any message to our community, to inform them about this?? Not every ReactOS user is subsribed to ros-dev.
Ah wait... you ignored that as well. You will let others handle the problem as usual.
I have 5 sets for testing, 2 real hardware and 3 vm. Each one with two partitions, one for ros and other for all the installers and drivers.
So now, whenever i cross the border of this change, i will have to wipe all, recreate the partitions and copy the stuff back? It will take no less than half an hour for VM. On real hardware, just the copying can take up to 1 hour.
Do you value your time available to spend on ReactOS? Then perhaps you could also think about others as well?
On Sunday, November 27, 2011 4:48 PM, "Eric Kohl" eric.kohl@t-online.de wrote:
caemyr@myopera.com wrote:
Is the new geometry backward-compatible? If i recreate the partitions,
will they be visible when i regress test revisions before 54511?
This is a one-way road. If you want to test older revisions, you must delete and re-create partitions again.
Regards Eric
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
-- With best regards Caemyr
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
"But before this, do we really have the problem?" - Aleksey Bragin
On Mon, Nov 28, 2011 at 11:49 AM, Magnus Johnsson magnusjjj@gmail.comwrote:
"No drama please!" - Aleksey Bragin
2011/11/28 Ged Murphy gedmurphy.maillists@gmail.com:
Do you have to work at being this rude and disrespectful or does it just come naturally to you? This is the a great example of why the reactos forums are such a nasty
place
to visit. You must be quite proud.
-----Original Message----- From: ros-dev-bounces@reactos.org [mailto:ros-dev-bounces@reactos.org]
On
Behalf Of caemyr@myopera.com Sent: 27 November 2011 23:07 To: ReactOS Development List Subject: Re: [ros-dev] HDD Geometry switch
Also, knowing the consequences, did you bother to issue any message to
our
community, to inform them about this?? Not every ReactOS user is
subsribed
to ros-dev.
Ah wait... you ignored that as well. You will let others handle the
problem
as usual.
I have 5 sets for testing, 2 real hardware and 3 vm. Each one with two partitions, one for ros and other for all the installers and drivers.
So now, whenever i cross the border of this change, i will have to wipe
all,
recreate the partitions and copy the stuff back? It will take no less
than
half an hour for VM. On real hardware, just the copying can take up to 1 hour.
Do you value your time available to spend on ReactOS? Then perhaps you
could
also think about others as well?
On Sunday, November 27, 2011 4:48 PM, "Eric Kohl" <eric.kohl@t-online.de
wrote:
caemyr@myopera.com wrote:
Is the new geometry backward-compatible? If i recreate the partitions,
will they be visible when i regress test revisions before 54511?
This is a one-way road. If you want to test older revisions, you must delete and re-create partitions again.
Regards Eric
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
-- With best regards Caemyr
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
Hi,
I support this demand. Instead of whining, could you check if your partitions still work with ReactOS? As far as I know, this should not affect small disks. So, try and complain after.
Furthermore, I would like to precise that we cannot stay 15 years back just because the switch is somehow a pain. Everyone may have to recreate its partitions. You have to understand that we cannot pretend to provide a modern OS with a deprecated architecture that can hardly leave closely to other OSes. This is definitely not selfish.
Thanks Eric for helping bringing ReactOS into a more suitable shape.
Finally, I would like to point that starting a drama is pointless. We are currently in a critical step in the OS development, we have to work all together and not fighting one against the others.
So, please test first, and then, depending on your ability to use (or not) your old partitions with the new geometry, we will see how we can help you.
Regards, Pierre
PS: Regarding communication step, nightly builds tester should read what changed before testing and then be aware of the problem. For release, a note can be put into the release announcement. PPS: If ReactOS is still considered as alpha, this is typically for such cases where a single change can have a real impact on data. We cannot wait for beta or even stable (not to speak about gold) to apply such changes.
On 11/28/2011 12:22 PM, Javier Agustìn Fernàndez Arroyo wrote:
"But before this, do we really have the problem?" - Aleksey Bragin
On Mon, Nov 28, 2011 at 11:49 AM, Magnus Johnsson <magnusjjj@gmail.com mailto:magnusjjj@gmail.com> wrote:
"No drama please!" - Aleksey Bragin 2011/11/28 Ged Murphy <gedmurphy.maillists@gmail.com <mailto:gedmurphy.maillists@gmail.com>>: > Do you have to work at being this rude and disrespectful or does it just > come naturally to you? > This is the a great example of why the reactos forums are such a nasty place > to visit. > You must be quite proud. > > > -----Original Message----- > From: ros-dev-bounces@reactos.org <mailto:ros-dev-bounces@reactos.org> [mailto:ros-dev-bounces@reactos.org <mailto:ros-dev-bounces@reactos.org>] On > Behalf Of caemyr@myopera.com <mailto:caemyr@myopera.com> > Sent: 27 November 2011 23:07 > To: ReactOS Development List > Subject: Re: [ros-dev] HDD Geometry switch > > Also, knowing the consequences, did you bother to issue any message to our > community, to inform them about this?? Not every ReactOS user is subsribed > to ros-dev. > > Ah wait... you ignored that as well. You will let others handle the problem > as usual. > > I have 5 sets for testing, 2 real hardware and 3 vm. Each one with two > partitions, one for ros and other for all the installers and drivers. > > So now, whenever i cross the border of this change, i will have to wipe all, > recreate the partitions and copy the stuff back? It will take no less than > half an hour for VM. On real hardware, just the copying can take up to 1 > hour. > > Do you value your time available to spend on ReactOS? Then perhaps you could > also think about others as well? > > On Sunday, November 27, 2011 4:48 PM, "Eric Kohl" <eric.kohl@t-online.de <mailto:eric.kohl@t-online.de>> > wrote: >> caemyr@myopera.com <mailto:caemyr@myopera.com> wrote: >> > Is the new geometry backward-compatible? If i recreate the partitions, > will they be visible when i regress test revisions before 54511? >> > >> >> This is a one-way road. If you want to test older revisions, you must >> delete and re-create partitions again. >> >> >> Regards >> Eric >> >> _______________________________________________ >> Ros-dev mailing list >> Ros-dev@reactos.org <mailto:Ros-dev@reactos.org> >> http://www.reactos.org/mailman/listinfo/ros-dev >> > > > -- > With best regards > Caemyr > > _______________________________________________ > Ros-dev mailing list > Ros-dev@reactos.org <mailto:Ros-dev@reactos.org> > http://www.reactos.org/mailman/listinfo/ros-dev > > > _______________________________________________ > Ros-dev mailing list > Ros-dev@reactos.org <mailto:Ros-dev@reactos.org> > http://www.reactos.org/mailman/listinfo/ros-dev > _______________________________________________ Ros-dev mailing list Ros-dev@reactos.org <mailto:Ros-dev@reactos.org> http://www.reactos.org/mailman/listinfo/ros-dev
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
Eric, Olaf - I need your knowledge and experience.
Please tell me exact scenarios when things break. Then we could look into those and fix the issues without going back to the old scheme and without ruining testing setups.
It would greatly help resolving this situation if you could give answers to those questions.
WBR, Aleksey.
On 28.11.2011 15:22, Javier Agustìn Fernàndez Arroyo wrote:
"But before this, do we really have the problem?" - Aleksey Bragin
On Mon, Nov 28, 2011 at 11:49 AM, Magnus Johnsson <magnusjjj@gmail.com mailto:magnusjjj@gmail.com> wrote:
"No drama please!" - Aleksey Bragin
Was I and my work treated any better? I`m fed up with such selfish approach in this project. Had I tried to impose such serious change as this one, in the way it was did here, i would be kicked out the next day.
I spent many months and money collecting hardware to test ReactOS with, and now i can trash all of it. Same as usual, you expect us testers to be waiting on call to to the bloody dirty job, but on the other hand have no problems making their work miserable. Yet you dare to bash me for lack of respect??
You might not value my time i spend for this project, but i do.
Good day to you all
On Monday, November 28, 2011 10:44 AM, "Ged Murphy" gedmurphy.maillists@gmail.com wrote:
Do you have to work at being this rude and disrespectful or does it just come naturally to you? This is the a great example of why the reactos forums are such a nasty place to visit. You must be quite proud.
Instead of rushing in to insult people in the way you did, why not try to discuss things.
Firstly, is there actually a problem? You might find that your partitions can still be read and this was just a miscommunication. If they can't, explain the problem and request that the change be reverted for now while the problem is further discussed.
Do you have any suggestions on how this change could be better integrated? You might find that this is one of those all or nothing changes which is unavoidable. It's not really a valid argument to stick with deprecated technology or configurations just because it disrupts testing rigs. Sometimes there's no other way than to make the leap and deal with the outcome as best as possible.
I think Eric did everything he needed to notify people that this change was coming. ReactOS' primary communication medium is the ros-dev mailing list, and despite what anyone else may think, it's definitely not the #reactos-dev IRC channel. He notified that the change was coming, there were no objections and then the change came some days later. If only more high impact changes were done like this.
Lastly, I find it a little petty and completely untrue to say you would be 'kicked out the next day' or that no one values the time you spend.. You know your work on reactos is just as respected and important as anyone else's.
Ged.
-----Original Message----- From: ros-dev-bounces@reactos.org [mailto:ros-dev-bounces@reactos.org] On Behalf Of caemyr@myopera.com Sent: 28 November 2011 11:41 To: ReactOS Development List Subject: Re: [ros-dev] HDD Geometry switch
Was I and my work treated any better? I`m fed up with such selfish approach in this project. Had I tried to impose such serious change as this one, in the way it was did here, i would be kicked out the next day.
I spent many months and money collecting hardware to test ReactOS with, and now i can trash all of it. Same as usual, you expect us testers to be waiting on call to to the bloody dirty job, but on the other hand have no problems making their work miserable. Yet you dare to bash me for lack of respect??
You might not value my time i spend for this project, but i do.
Good day to you all
On Monday, November 28, 2011 10:44 AM, "Ged Murphy" gedmurphy.maillists@gmail.com wrote:
Do you have to work at being this rude and disrespectful or does it just come naturally to you? This is the a great example of why the reactos forums are such a nasty place to visit. You must be quite proud.
Yes there is a problem
Does anyone care for any suggestions?? Despite ROS being in feature-freeze, pre-release period, such change is simply commited without any consultation.
Since when you should discuss changes AFTER applying them??
This discussion is pointless. I indeed whined too much. I should simply follow other testers we had previously and stop wasting my time. No wonder why we always had few times less testers than devs, if the primer are simply ignored.
Let me stop being rude and disrespectful, by finishing this discussion.
Next time when you be needing regression testing, please email Eric insted.
Regards
On Monday, November 28, 2011 12:29 PM, "Ged Murphy" gedmurphy.maillists@gmail.com wrote:
Instead of rushing in to insult people in the way you did, why not try to discuss things.
Firstly, is there actually a problem? You might find that your partitions can still be read and this was just a miscommunication. If they can't, explain the problem and request that the change be reverted for now while the problem is further discussed.
Do you have any suggestions on how this change could be better integrated? You might find that this is one of those all or nothing changes which is unavoidable. It's not really a valid argument to stick with deprecated technology or configurations just because it disrupts testing rigs. Sometimes there's no other way than to make the leap and deal with the outcome as best as possible.
I think Eric did everything he needed to notify people that this change was coming. ReactOS' primary communication medium is the ros-dev mailing list, and despite what anyone else may think, it's definitely not the #reactos-dev IRC channel. He notified that the change was coming, there were no objections and then the change came some days later. If only more high impact changes were done like this.
Lastly, I find it a little petty and completely untrue to say you would be 'kicked out the next day' or that no one values the time you spend.. You know your work on reactos is just as respected and important as anyone else's.
Ged.
-----Original Message----- From: ros-dev-bounces@reactos.org [mailto:ros-dev-bounces@reactos.org] On Behalf Of caemyr@myopera.com Sent: 28 November 2011 11:41 To: ReactOS Development List Subject: Re: [ros-dev] HDD Geometry switch
Was I and my work treated any better? I`m fed up with such selfish approach in this project. Had I tried to impose such serious change as this one, in the way it was did here, i would be kicked out the next day.
I spent many months and money collecting hardware to test ReactOS with, and now i can trash all of it. Same as usual, you expect us testers to be waiting on call to to the bloody dirty job, but on the other hand have no problems making their work miserable. Yet you dare to bash me for lack of respect??
You might not value my time i spend for this project, but i do.
Good day to you all
On Monday, November 28, 2011 10:44 AM, "Ged Murphy" gedmurphy.maillists@gmail.com wrote:
Do you have to work at being this rude and disrespectful or does it just come naturally to you? This is the a great example of why the reactos forums are such a nasty place to visit. You must be quite proud.
-- With best regards Caemyr
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
Let's please stop this drama. It won't bring anything good in the project. I wouldn't like to see people leaving the project or questioning their involvement in it. Especially for two constants change.
Eric motivated its changes. Olaf, you motivated your (and perhaps others) grievances. The only thing we ask you now is to test whether this change effectively affects your disks and your capacity to regress-test ReactOS. If you answer, we'll probably be able to help you. Otherwise, we won't.
Let's discuss this as gentlemen. If it starts being a real problem for several testers, then I propose we push this as a topic for the next meeting (in 3 days).
On a side note, more than a year ago, Eric & I already talked about a geometry change in ReactOS. Perhaps should have we warned earlier this would have to be done. Would have it changed something? I'm not sure.
On 11/28/2011 01:44 PM, caemyr@myopera.com wrote:
Yes there is a problem
Does anyone care for any suggestions?? Despite ROS being in feature-freeze, pre-release period, such change is simply commited without any consultation.
Since when you should discuss changes AFTER applying them??
This discussion is pointless. I indeed whined too much. I should simply follow other testers we had previously and stop wasting my time. No wonder why we always had few times less testers than devs, if the primer are simply ignored.
Let me stop being rude and disrespectful, by finishing this discussion.
Next time when you be needing regression testing, please email Eric insted.
Regards
On Monday, November 28, 2011 12:29 PM, "Ged Murphy"gedmurphy.maillists@gmail.com wrote:
Instead of rushing in to insult people in the way you did, why not try to discuss things.
Firstly, is there actually a problem? You might find that your partitions can still be read and this was just a miscommunication. If they can't, explain the problem and request that the change be reverted for now while the problem is further discussed.
Do you have any suggestions on how this change could be better integrated? You might find that this is one of those all or nothing changes which is unavoidable. It's not really a valid argument to stick with deprecated technology or configurations just because it disrupts testing rigs. Sometimes there's no other way than to make the leap and deal with the outcome as best as possible.
I think Eric did everything he needed to notify people that this change was coming. ReactOS' primary communication medium is the ros-dev mailing list, and despite what anyone else may think, it's definitely not the #reactos-dev IRC channel. He notified that the change was coming, there were no objections and then the change came some days later. If only more high impact changes were done like this.
Lastly, I find it a little petty and completely untrue to say you would be 'kicked out the next day' or that no one values the time you spend.. You know your work on reactos is just as respected and important as anyone else's.
Ged.
-----Original Message----- From: ros-dev-bounces@reactos.org [mailto:ros-dev-bounces@reactos.org] On Behalf Of caemyr@myopera.com Sent: 28 November 2011 11:41 To: ReactOS Development List Subject: Re: [ros-dev] HDD Geometry switch
Was I and my work treated any better? I`m fed up with such selfish approach in this project. Had I tried to impose such serious change as this one, in the way it was did here, i would be kicked out the next day.
I spent many months and money collecting hardware to test ReactOS with, and now i can trash all of it. Same as usual, you expect us testers to be waiting on call to to the bloody dirty job, but on the other hand have no problems making their work miserable. Yet you dare to bash me for lack of respect??
You might not value my time i spend for this project, but i do.
Good day to you all
On Monday, November 28, 2011 10:44 AM, "Ged Murphy" gedmurphy.maillists@gmail.com wrote:
Do you have to work at being this rude and disrespectful or does it just come naturally to you? This is the a great example of why the reactos forums are such a nasty place to visit. You must be quite proud.
-- With best regards Caemyr
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
Ummm... Eric wrote the e-mail a WEEK before applying them.
*NOBODY* had any objections then.
Also you keep avoiding the question that Aleksey and 15 other asked: HAVE YOU TESTED THIS ACTUALLY BREAKS ANYTHING????
Finally, everyone needs to realize there is a chance of data loss with a pre-alpha OS. I suspect there will be many other changes that will require a destructive re-format. When I was a tester for Windows Whistler betas, every new release required a complete reformat of the disk. Windows Me preview formatted c: when you formatted a:, and a newer beta physically damaged the disk by sending wrong commands to the disk heads.
If you, when testing ReactOS, or any other OS, are expecting anything else than a fatal destruction of your testing rig, and that every release which does not do so is a miracle, you're not an OS tester.
Best regards, Alex Ionescu
On Mon, Nov 28, 2011 at 2:44 AM, caemyr@myopera.com wrote:
Yes there is a problem
Does anyone care for any suggestions?? Despite ROS being in feature-freeze, pre-release period, such change is simply commited without any consultation.
Since when you should discuss changes AFTER applying them??
This discussion is pointless. I indeed whined too much. I should simply follow other testers we had previously and stop wasting my time. No wonder why we always had few times less testers than devs, if the primer are simply ignored.
Let me stop being rude and disrespectful, by finishing this discussion.
Next time when you be needing regression testing, please email Eric insted.
Regards
On Monday, November 28, 2011 12:29 PM, "Ged Murphy" < gedmurphy.maillists@gmail.com> wrote:
Instead of rushing in to insult people in the way you did, why not try to discuss things.
Firstly, is there actually a problem? You might find that your partitions can still be read and this was just a miscommunication. If they can't, explain the problem and request that the change be reverted for now while the problem is further discussed.
Do you have any suggestions on how this change could be better integrated? You might find that this is one of those all or nothing changes which is unavoidable. It's not really a valid argument to stick with deprecated technology or configurations just because it disrupts testing rigs. Sometimes there's no other way than to make the leap and deal with the outcome as best as possible.
I think Eric did everything he needed to notify people that this change was coming. ReactOS' primary communication medium is the ros-dev mailing list, and despite what anyone else may think, it's definitely not the #reactos-dev IRC channel. He notified that the change was coming, there were no objections and then the change came some days later. If only more high impact changes were done like this.
Lastly, I find it a little petty and completely untrue to say you would be 'kicked out the next day' or that no one values the time you spend.. You know your work on reactos is just as respected and important as anyone else's.
Ged.
-----Original Message----- From: ros-dev-bounces@reactos.org [mailto:ros-dev-bounces@reactos.org]
On
Behalf Of caemyr@myopera.com Sent: 28 November 2011 11:41 To: ReactOS Development List Subject: Re: [ros-dev] HDD Geometry switch
Was I and my work treated any better? I`m fed up with such selfish approach in this project. Had I tried to impose such serious change as this one, in the way it was did here, i would be kicked out the next day.
I spent many months and money collecting hardware to test ReactOS with, and now i can trash all of it. Same as usual, you expect us testers to be waiting on call to to the bloody dirty job, but on the other hand have no problems making their work miserable. Yet you dare to bash me for lack of respect??
You might not value my time i spend for this project, but i do.
Good day to you all
On Monday, November 28, 2011 10:44 AM, "Ged Murphy" gedmurphy.maillists@gmail.com wrote:
Do you have to work at being this rude and disrespectful or does it
just
come naturally to you? This is the a great example of why the reactos forums are such a nasty place to visit. You must be quite proud.
-- With best regards Caemyr
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
-- With best regards Caemyr
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
Please re-read my emails before commenting them.
On Monday, November 28, 2011 8:40 AM, "Alex Ionescu" ionucu@videotron.ca wrote:
Ummm... Eric wrote the e-mail a WEEK before applying them.
*NOBODY* had any objections then.
Also you keep avoiding the question that Aleksey and 15 other asked: HAVE YOU TESTED THIS ACTUALLY BREAKS ANYTHING????
Finally, everyone needs to realize there is a chance of data loss with a pre-alpha OS. I suspect there will be many other changes that will require a destructive re-format. When I was a tester for Windows Whistler betas, every new release required a complete reformat of the disk. Windows Me preview formatted c: when you formatted a:, and a newer beta physically damaged the disk by sending wrong commands to the disk heads.
If you, when testing ReactOS, or any other OS, are expecting anything else than a fatal destruction of your testing rig, and that every release which does not do so is a miracle, you're not an OS tester.
Best regards, Alex Ionescu
On Mon, Nov 28, 2011 at 2:44 AM, caemyr@myopera.com wrote:
Yes there is a problem
Does anyone care for any suggestions?? Despite ROS being in feature-freeze, pre-release period, such change is simply commited without any consultation.
Since when you should discuss changes AFTER applying them??
This discussion is pointless. I indeed whined too much. I should simply follow other testers we had previously and stop wasting my time. No wonder why we always had few times less testers than devs, if the primer are simply ignored.
Let me stop being rude and disrespectful, by finishing this discussion.
Next time when you be needing regression testing, please email Eric insted.
Regards
On Monday, November 28, 2011 12:29 PM, "Ged Murphy" < gedmurphy.maillists@gmail.com> wrote:
Instead of rushing in to insult people in the way you did, why not try to discuss things.
Firstly, is there actually a problem? You might find that your partitions can still be read and this was just a miscommunication. If they can't, explain the problem and request that the change be reverted for now while the problem is further discussed.
Do you have any suggestions on how this change could be better integrated? You might find that this is one of those all or nothing changes which is unavoidable. It's not really a valid argument to stick with deprecated technology or configurations just because it disrupts testing rigs. Sometimes there's no other way than to make the leap and deal with the outcome as best as possible.
I think Eric did everything he needed to notify people that this change was coming. ReactOS' primary communication medium is the ros-dev mailing list, and despite what anyone else may think, it's definitely not the #reactos-dev IRC channel. He notified that the change was coming, there were no objections and then the change came some days later. If only more high impact changes were done like this.
Lastly, I find it a little petty and completely untrue to say you would be 'kicked out the next day' or that no one values the time you spend.. You know your work on reactos is just as respected and important as anyone else's.
Ged.
-----Original Message----- From: ros-dev-bounces@reactos.org [mailto:ros-dev-bounces@reactos.org]
On
Behalf Of caemyr@myopera.com Sent: 28 November 2011 11:41 To: ReactOS Development List Subject: Re: [ros-dev] HDD Geometry switch
Was I and my work treated any better? I`m fed up with such selfish approach in this project. Had I tried to impose such serious change as this one, in the way it was did here, i would be kicked out the next day.
I spent many months and money collecting hardware to test ReactOS with, and now i can trash all of it. Same as usual, you expect us testers to be waiting on call to to the bloody dirty job, but on the other hand have no problems making their work miserable. Yet you dare to bash me for lack of respect??
You might not value my time i spend for this project, but i do.
Good day to you all
On Monday, November 28, 2011 10:44 AM, "Ged Murphy" gedmurphy.maillists@gmail.com wrote:
Do you have to work at being this rude and disrespectful or does it
just
come naturally to you? This is the a great example of why the reactos forums are such a nasty place to visit. You must be quite proud.
-- With best regards Caemyr
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
-- With best regards Caemyr
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
Ros-dev mailing list Ros-dev@reactos.org http://www.reactos.org/mailman/listinfo/ros-dev
@Alex Since you apparently TL:DR'd my previous emails - a short summary. I`m a tester not a dev, i dont have to and i dont really know much about hdd geometry and possible influence of changing it. I incorrectly assumed that any change done will be considered with other peoples in mind, not only the commiter's biddings. I failed. The problem is that now, with every crossing of the drive geometry changing rev i need to recreate all the partitions used and copy back the needed data. In case of real hardware, it may take up to one hour, which is almost a third of my whole free time in my standard working day, a majority of which is taken by ROS.
From what i tested on real hw rig with 54528. Two partition setup on 40GB hdd, 2GB for ROS and rest for data. When installing head on rev with old geometry, it works fine.
Reinstalling rev with old geometry - the bigger partition is unreadable, can only be wiped clean and unreadable.
I couldn't reproduce it on VBox, but also i was not able to create an identical virtual hdd as the one used in real hw rig.
So no problem at all. The rigs are to be dismounted and stored for a year or two, in the meantime the geometry change should be distant enough not to bother anymore. Sorry for the fuss. I should follow Ged's advice and keep posting nice, gentle emails to ros-dev, to have them ignored like most of the previous one.
It was a great timing by the way, with the long-awaited release, we have a serious change in, consequences of which might reveal in near future, days, weeks - or perhaps right after the release.
Again, sorry for raising this as a serious issue. I obviously shouldn't be bothered.
Regards
Hi, I tested the new install on my test hardware and the existing partition worked. Same partition since 2000. James
On Sun, Nov 27, 2011 at 8:23 AM, caemyr@myopera.com wrote:
Is the new geometry backward-compatible? If i recreate the partitions, will they be visible when i regress test revisions before 54511?
-- With best regards Caemyr
-
Aleksey Bragin -
Alex Ionescu -
caemyr＠myopera.com -
Cameron Gutman -
Colin Finck -
Eric Kohl -
Ged Murphy -
James Tabor -
Javier Agustìn Fernàndez Arroyo -
Magnus Johnsson -
Michael Fritscher -
Pierre Schweitzer -
Sylvain Petreolle -
Timo Kreuzer