cmake rant

[Timo Kreuzer]

Hi,

Several times now cmake build has been broken. Time for some action!

Last meeting I asked everyone to test/use cmake. It was also mentioned 
that if questions arise, we (Amine and me) would be happy to help out. I 
can't remember anyone has asked how it works, so I assume noone had any 
problems. There's also a pretty good wiki entry describing the whole 
procedure for n00bs.

Now people tell me it's complicated, people are complaining that its 
ridiculous to have 2 build systems, etc.
And probably noone has ever tried it.

We really need to move on.
I don't like having 2 build systems as well.

Current blocker is the debugging which has some issues, Arty is working 
on that. Another problem is a boot problem on real hardware, but no I 
don't know on which configuration it doesn't work, so we need more 
people testing it on their real hardware setup and report any issues.

Here's a list with current issues:
http://www.reactos.org/wiki/CMake_Todo

So please:
If you are missing something, let us know.
If you like to make it better, make suggestions.
But stop ignoring cmake!

If noone cares and everyone just thinks he can give a s^Z damn until we 
officially switch, then we can as well delete all cmake stuff and keep 
rbuild. It has a lot of awesome advantages, like you only have to type 
one command to build everything and you don't need to install cmake and 
you can export whatever you want from kernel32 even if the functions 
don't exist. Also you can enjoy the rbuild-loop again and again.

Or we can do it the hard way and delete rbuild, so people are forced to 
use cmake. I'm sure this approach would be *really* appreaciated.

Thanks,
Timo

[Ged Murphy]

I'm all for deleting rbuild now and dealing with the consequences.
That would be better than the current situation, which seems rather
ridiculous at the moment.

Right now if you change something, you have to :
- edit the cmakefiles.txt file
- edit the rbuild file
- possibly edit a pspec and spec file
- build with rbuild to check for breakages
- build with cmake to check for breakages

Only then can you commit your changes. 
It's not ideal for the people working on this stuff regularly.

Ged.

-----Original Message-----
From: [e-mail] [mailto:[e-mail]] On
Behalf Of Timo Kreuzer
Sent: 21 July 2011 17:39
To: ReactOS Development List
Subject: [ros-dev] cmake rant

Hi,

Several times now cmake build has been broken. Time for some action!

Last meeting I asked everyone to test/use cmake. It was also mentioned that
if questions arise, we (Amine and me) would be happy to help out. I can't
remember anyone has asked how it works, so I assume noone had any problems.
There's also a pretty good wiki entry describing the whole procedure for
n00bs.

Now people tell me it's complicated, people are complaining that its
ridiculous to have 2 build systems, etc.
And probably noone has ever tried it.

We really need to move on.
I don't like having 2 build systems as well.

Current blocker is the debugging which has some issues, Arty is working on
that. Another problem is a boot problem on real hardware, but no I don't
know on which configuration it doesn't work, so we need more people testing
it on their real hardware setup and report any issues.

Here's a list with current issues:
http://www.reactos.org/wiki/CMake_Todo

So please:
If you are missing something, let us know.
If you like to make it better, make suggestions.
But stop ignoring cmake!

If noone cares and everyone just thinks he can give a s^Z damn until we
officially switch, then we can as well delete all cmake stuff and keep
rbuild. It has a lot of awesome advantages, like you only have to type one
command to build everything and you don't need to install cmake and you can
export whatever you want from kernel32 even if the functions don't exist.
Also you can enjoy the rbuild-loop again and again.

Or we can do it the hard way and delete rbuild, so people are forced to use
cmake. I'm sure this approach would be *really* appreaciated.

Thanks,
Timo

_______________________________________________
Ros-dev mailing list
[e-mail]
http://www.reactos.org/mailman/listinfo/ros-dev

[Jérôme Gardou]

+1 for Ged.

Le 21/07/2011 18:51, Ged Murphy a écrit :
> I'm all for deleting rbuild now and dealing with the consequences.
> That would be better than the current situation, which seems rather
> ridiculous at the moment.
>
> Right now if you change something, you have to :
> edit the cmakefiles.txt file
> edit the rbuild file
> possibly edit a pspec and spec file
> build with rbuild to check for breakages
> build with cmake to check for breakages
>
> Only then can you commit your changes.
> It's not ideal for the people working on this stuff regularly.
>
> Ged.
>
> -----Original Message-----
> From: [e-mail] [mailto:[e-mail]] On
> Behalf Of Timo Kreuzer
> Sent: 21 July 2011 17:39
> To: ReactOS Development List
> Subject: [ros-dev] cmake rant
>
> Hi,
>
> Several times now cmake build has been broken. Time for some action!
>
> Last meeting I asked everyone to test/use cmake. It was also mentioned that
> if questions arise, we (Amine and me) would be happy to help out. I can't
> remember anyone has asked how it works, so I assume noone had any problems.
> There's also a pretty good wiki entry describing the whole procedure for
> n00bs.
>
> Now people tell me it's complicated, people are complaining that its
> ridiculous to have 2 build systems, etc.
> And probably noone has ever tried it.
>
> We really need to move on.
> I don't like having 2 build systems as well.
>
> Current blocker is the debugging which has some issues, Arty is working on
> that. Another problem is a boot problem on real hardware, but no I don't
> know on which configuration it doesn't work, so we need more people testing
> it on their real hardware setup and report any issues.
>
> Here's a list with current issues:
> http://www.reactos.org/wiki/CMake_Todo
>
> So please:
> If you are missing something, let us know.
> If you like to make it better, make suggestions.
> But stop ignoring cmake!
>
> If noone cares and everyone just thinks he can give a s^Z damn until we
> officially switch, then we can as well delete all cmake stuff and keep
> rbuild. It has a lot of awesome advantages, like you only have to type one
> command to build everything and you don't need to install cmake and you can
> export whatever you want from kernel32 even if the functions don't exist.
> Also you can enjoy the rbuild-loop again and again.
>
> Or we can do it the hard way and delete rbuild, so people are forced to use
> cmake. I'm sure this approach would be *really* appreaciated.
>
> Thanks,
> Timo
>
> Ros-dev mailing list
> [e-mail]
> http://www.reactos.org/mailman/listinfo/ros-dev
>
> Ros-dev mailing list
> [e-mail]
> http://www.reactos.org/mailman/listinfo/ros-dev

[Gabriel ilardi]

-1 imho, not at least debugging is possible, we keep accumulating regressions and not being able to debug is not an option to me...
> Date: Thu, 21 Jul 2011 18:57:28 +0200
> From: [e-mail]
> To: [e-mail]
> Subject: Re: [ros-dev] cmake rant
>
> +1 for Ged.
>
> Le 21/07/2011 18:51, Ged Murphy a écrit :
> > I'm all for deleting rbuild now and dealing with the consequences.
> > That would be better than the current situation, which seems rather
> > ridiculous at the moment.
> >
> > Right now if you change something, you have to :
> > edit the cmakefiles.txt file
> > edit the rbuild file
> > possibly edit a pspec and spec file
> > build with rbuild to check for breakages
> > build with cmake to check for breakages
> >
> > Only then can you commit your changes.
> > It's not ideal for the people working on this stuff regularly.
> >
> > Ged.
> >
> > -----Original Message-----
> > From: [e-mail] [mailto:[e-mail]] On
> > Behalf Of Timo Kreuzer
> > Sent: 21 July 2011 17:39
> > To: ReactOS Development List
> > Subject: [ros-dev] cmake rant
> >
> > Hi,
> >
> > Several times now cmake build has been broken. Time for some action!
> >
> > Last meeting I asked everyone to test/use cmake. It was also mentioned that
> > if questions arise, we (Amine and me) would be happy to help out. I can't
> > remember anyone has asked how it works, so I assume noone had any problems.
> > There's also a pretty good wiki entry describing the whole procedure for
> > n00bs.
> >
> > Now people tell me it's complicated, people are complaining that its
> > ridiculous to have 2 build systems, etc.
> > And probably noone has ever tried it.
> >
> > We really need to move on.
> > I don't like having 2 build systems as well.
> >
> > Current blocker is the debugging which has some issues, Arty is working on
> > that. Another problem is a boot problem on real hardware, but no I don't
> > know on which configuration it doesn't work, so we need more people testing
> > it on their real hardware setup and report any issues.
> >
> > Here's a list with current issues:
> > http://www.reactos.org/wiki/CMake_Todo
> >
> > So please:
> > If you are missing something, let us know.
> > If you like to make it better, make suggestions.
> > But stop ignoring cmake!
> >
> > If noone cares and everyone just thinks he can give a s^Z damn until we
> > officially switch, then we can as well delete all cmake stuff and keep
> > rbuild. It has a lot of awesome advantages, like you only have to type one
> > command to build everything and you don't need to install cmake and you can
> > export whatever you want from kernel32 even if the functions don't exist.
> > Also you can enjoy the rbuild-loop again and again.
> >
> > Or we can do it the hard way and delete rbuild, so people are forced to use
> > cmake. I'm sure this approach would be *really* appreaciated.
> >
> > Thanks,
> > Timo
> >
> > Ros-dev mailing list
> > [e-mail]
> > http://www.reactos.org/mailman/listinfo/ros-dev
> >
> > Ros-dev mailing list
> > [e-mail]
> > http://www.reactos.org/mailman/listinfo/ros-dev
>
> Ros-dev mailing list
> [e-mail]
> http://www.reactos.org/mailman/listinfo/ros-dev

[Javier Agustìn Fernàndez Arroyo]

we do need a working cmake RosBE version before purging rbuild

On Thu, Jul 21, 2011 at 7:16 PM, Gabriel ilardi [e-mail]:
> -1 imho, not at least debugging is possible, we keep accumulating
> regressions and not being able to debug is not an option to me...
> > Date: Thu, 21 Jul 2011 18:57:28 +0200
> > From: [e-mail]
> > To: [e-mail]
> > Subject: Re: [ros-dev] cmake rant
> > +1 for Ged.
> >
> > Le 21/07/2011 18:51, Ged Murphy a écrit :
> > > I'm all for deleting rbuild now and dealing with the consequences.
> > > That would be better than the current situation, which seems rather
> > > ridiculous at the moment.
> > >
> > > Right now if you change something, you have to :
> > > edit the cmakefiles.txt file
> > > edit the rbuild file
> > > possibly edit a pspec and spec file
> > > build with rbuild to check for breakages
> > > build with cmake to check for breakages
> > >
> > > Only then can you commit your changes.
> > > It's not ideal for the people working on this stuff regularly.
> > >
> > > Ged.
> > >
> > > -----Original Message-----
> > > From: [e-mail] [mailto:[e-mail]]
>
> On
> > > Behalf Of Timo Kreuzer
> > > Sent: 21 July 2011 17:39
> > > To: ReactOS Development List
> > > Subject: [ros-dev] cmake rant
> > >
> > > Hi,
> > >
> > > Several times now cmake build has been broken. Time for some action!
> > >
> > > Last meeting I asked everyone to test/use cmake. It was also mentioned
>
> that
> > > if questions arise, we (Amine and me) would be happy to help out. I
>
> can't
> > > remember anyone has asked how it works, so I assume noone had any
>
> problems.
> > > There's also a pretty good wiki entry describing the whole procedure
>
> for
> > > n00bs.
> > >
> > > Now people tell me it's complicated, people are complaining that its
> > > ridiculous to have 2 build systems, etc.
> > > And probably noone has ever tried it.
> > >
> > > We really need to move on.
> > > I don't like having 2 build systems as well.
> > >
> > > Current blocker is the debugging which has some issues, Arty is working
>
> on
> > > that. Another problem is a boot problem on real hardware, but no I
>
> don't
> > > know on which configuration it doesn't work, so we need more people
>
> testing
> > > it on their real hardware setup and report any issues.
> > >
> > > Here's a list with current issues:
> > > http://www.reactos.org/wiki/CMake_Todo
> > >
> > > So please:
> > > If you are missing something, let us know.
> > > If you like to make it better, make suggestions.
> > > But stop ignoring cmake!
> > >
> > > If noone cares and everyone just thinks he can give a s^Z damn until we
> > > officially switch, then we can as well delete all cmake stuff and keep
> > > rbuild. It has a lot of awesome advantages, like you only have to type
>
> one
> > > command to build everything and you don't need to install cmake and you
>
> can
> > > export whatever you want from kernel32 even if the functions don't
>
> exist.
> > > Also you can enjoy the rbuild-loop again and again.
> > >
> > > Or we can do it the hard way and delete rbuild, so people are forced to
>
> use
> > > cmake. I'm sure this approach would be *really* appreaciated.
> > >
> > > Thanks,
> > > Timo
> > >
> > > Ros-dev mailing list
> > > [e-mail]
> > > http://www.reactos.org/mailman/listinfo/ros-dev
> > >
> > > Ros-dev mailing list
> > > [e-mail]
> > > http://www.reactos.org/mailman/listinfo/ros-dev
> >
> > Ros-dev mailing list
> > [e-mail]
> > http://www.reactos.org/mailman/listinfo/ros-dev
>
> Ros-dev mailing list
> [e-mail]
> http://www.reactos.org/mailman/listinfo/ros-dev

[Igor Paliychuk]

You don't need RosBE for using cmake. For building ROS with nmake you need
only VS command prompt and installed cmake. The same for gcc - you need gcc
and cmake installed. That's all (if i'm right).

2011/7/21 Javier Agustìn Fernàndez Arroyo [e-mail]
> we do need a working cmake RosBE version before purging rbuild
>
> On Thu, Jul 21, 2011 at 7:16 PM, Gabriel ilardi [e-mail]:
> > -1 imho, not at least debugging is possible, we keep accumulating
> > regressions and not being able to debug is not an option to me...
> > > Date: Thu, 21 Jul 2011 18:57:28 +0200
> > > From: [e-mail]
> > > To: [e-mail]
> > > Subject: Re: [ros-dev] cmake rant
> > > +1 for Ged.
> > >
> > > Le 21/07/2011 18:51, Ged Murphy a écrit :
> > > > I'm all for deleting rbuild now and dealing with the consequences.
> > > > That would be better than the current situation, which seems rather
> > > > ridiculous at the moment.
> > > >
> > > > Right now if you change something, you have to :
> > > > edit the cmakefiles.txt file
> > > > edit the rbuild file
> > > > possibly edit a pspec and spec file
> > > > build with rbuild to check for breakages
> > > > build with cmake to check for breakages
> > > >
> > > > Only then can you commit your changes.
> > > > It's not ideal for the people working on this stuff regularly.
> > > >
> > > > Ged.
> > > >
> > > > -----Original Message-----
> > > > From: [e-mail] [mailto:[e-mail]]
> >
> > On
> > > > Behalf Of Timo Kreuzer
> > > > Sent: 21 July 2011 17:39
> > > > To: ReactOS Development List
> > > > Subject: [ros-dev] cmake rant
> > > >
> > > > Hi,
> > > >
> > > > Several times now cmake build has been broken. Time for some action!
> > > >
> > > > Last meeting I asked everyone to test/use cmake. It was also mentioned
> >
> > that
> > > > if questions arise, we (Amine and me) would be happy to help out. I
> >
> > can't
> > > > remember anyone has asked how it works, so I assume noone had any
> >
> > problems.
> > > > There's also a pretty good wiki entry describing the whole procedure
> >
> > for
> > > > n00bs.
> > > >
> > > > Now people tell me it's complicated, people are complaining that its
> > > > ridiculous to have 2 build systems, etc.
> > > > And probably noone has ever tried it.
> > > >
> > > > We really need to move on.
> > > > I don't like having 2 build systems as well.
> > > >
> > > > Current blocker is the debugging which has some issues, Arty is
> >
> > working on
> > > > that. Another problem is a boot problem on real hardware, but no I
> >
> > don't
> > > > know on which configuration it doesn't work, so we need more people
> >
> > testing
> > > > it on their real hardware setup and report any issues.
> > > >
> > > > Here's a list with current issues:
> > > > http://www.reactos.org/wiki/CMake_Todo
> > > >
> > > > So please:
> > > > If you are missing something, let us know.
> > > > If you like to make it better, make suggestions.
> > > > But stop ignoring cmake!
> > > >
> > > > If noone cares and everyone just thinks he can give a s^Z damn until
> >
> > we
> > > > officially switch, then we can as well delete all cmake stuff and keep
> > > > rbuild. It has a lot of awesome advantages, like you only have to type
> >
> > one
> > > > command to build everything and you don't need to install cmake and
> >
> > you can
> > > > export whatever you want from kernel32 even if the functions don't
> >
> > exist.
> > > > Also you can enjoy the rbuild-loop again and again.
> > > >
> > > > Or we can do it the hard way and delete rbuild, so people are forced
> >
> > to use
> > > > cmake. I'm sure this approach would be *really* appreaciated.
> > > >
> > > > Thanks,
> > > > Timo
> > > >
> > > > Ros-dev mailing list
> > > > [e-mail]
> > > > http://www.reactos.org/mailman/listinfo/ros-dev
> > > >
> > > > Ros-dev mailing list
> > > > [e-mail]
> > > > http://www.reactos.org/mailman/listinfo/ros-dev
> > >
> > > Ros-dev mailing list
> > > [e-mail]
> > > http://www.reactos.org/mailman/listinfo/ros-dev
> >
> > Ros-dev mailing list
> > [e-mail]
> > http://www.reactos.org/mailman/listinfo/ros-dev
>
> Ros-dev mailing list
> [e-mail]
> http://www.reactos.org/mailman/listinfo/ros-dev

[Aleksey Bragin]

I didn’t hear many positive opinions when I forced everyone to switch over to the new ntdll loader. I minusone this proposal too. CMake should get wider usage first, and get the regressions sorted out. After all, it’s a build system change, not a migration to BSD kernel, it’s a doable task.

I’m all to help with hunting down CMake problems btw.

WBR,
Aleksey Bragin.

From: Gabriel ilardi 
Sent: Thursday, July 21, 2011 9:16 PM
To: ReactOS Development List 
Subject: Re: [ros-dev] cmake rant

-1 imho, not at least debugging is possible, we keep accumulating regressions and not being able to debug is not an option to me...
> Date: Thu, 21 Jul 2011 18:57:28 +0200
> From: [e-mail]
> To: [e-mail]
> Subject: Re: [ros-dev] cmake rant
>
> +1 for Ged.
>
> Le 21/07/2011 18:51, Ged Murphy a écrit :
> > I'm all for deleting rbuild now and dealing with the consequences.
> > That would be better than the current situation, which seems rather
> > ridiculous at the moment.
> >
> > Right now if you change something, you have to :
> > edit the cmakefiles.txt file
> > edit the rbuild file
> > possibly edit a pspec and spec file
> > build with rbuild to check for breakages
> > build with cmake to check for breakages
> >
> > Only then can you commit your changes.
> > It's not ideal for the people working on this stuff regularly.
> >
> > Ged.
> >
> > -----Original Message-----
> > From: [e-mail] [mailto:[e-mail]] On
> > Behalf Of Timo Kreuzer
> > Sent: 21 July 2011 17:39
> > To: ReactOS Development List
> > Subject: [ros-dev] cmake rant
> >
> > Hi,
> >
> > Several times now cmake build has been broken. Time for some action!
> >
> > Last meeting I asked everyone to test/use cmake. It was also mentioned that
> > if questions arise, we (Amine and me) would be happy to help out. I can't
> > remember anyone has asked how it works, so I assume noone had any problems.
> > There's also a pretty good wiki entry describing the whole procedure for
> > n00bs.
> >
> > Now people tell me it's complicated, people are complaining that its
> > ridiculous to have 2 build systems, etc.
> > And probably noone has ever tried it.
> >
> > We really need to move on.
> > I don't like having 2 build systems as well.
> >
> > Current blocker is the debugging which has some issues, Arty is working on
> > that. Another problem is a boot problem on real hardware, but no I don't
> > know on which configuration it doesn't work, so we need more people testing
> > it on their real hardware setup and report any issues.
> >
> > Here's a list with current issues:
> > http://www.reactos.org/wiki/CMake_Todo
> >
> > So please:
> > If you are missing something, let us know.
> > If you like to make it better, make suggestions.
> > But stop ignoring cmake!
> >
> > If noone cares and everyone just thinks he can give a s^Z damn until we
> > officially switch, then we can as well delete all cmake stuff and keep
> > rbuild. It has a lot of awesome advantages, like you only have to type one
> > command to build everything and you don't need to install cmake and you can
> > export whatever you want from kernel32 even if the functions don't exist.
> > Also you can enjoy the rbuild-loop again and again.
> >
> > Or we can do it the hard way and delete rbuild, so people are forced to use
> > cmake. I'm sure this approach would be *really* appreaciated.
> >
> > Thanks,
> > Timo
> >
> > Ros-dev mailing list
> > [e-mail]
> > http://www.reactos.org/mailman/listinfo/ros-dev
> >
> > Ros-dev mailing list
> > [e-mail]
> > http://www.reactos.org/mailman/listinfo/ros-dev
>
> Ros-dev mailing list
> [e-mail]
> http://www.reactos.org/mailman/listinfo/ros-dev

--------------------------------------------------------------------------------
_______________________________________________
Ros-dev mailing list
[e-mail]
http://www.reactos.org/mailman/listinfo/ros-dev

CMake wont get ANY wider usage, if its going to be broken all the time. To prevent that, we need either to slow down development in any other areas, or stop it completely, as we are basically doubling the work.

I vote +1, but only if debugging is fixed or reverted back to working state. We can deal with other issues, but this one is basic.

Arty?

On Fri, 22 Jul 2011 00:30 +0400, "Aleksey Bragin" [e-mail] wrote:
> I didn’t hear many positive opinions when I forced everyone to switch
> over to the new ntdll loader. I minusone this proposal too. CMake should
> get wider usage first, and get the regressions sorted out. After all,
> it’s a build system change, not a migration to BSD kernel, it’s a doable
> task.

With best regards
Caemyr

[Timo Kreuzer]

Am 22.07.2011 00:38, schrieb [e-mail]:
> CMake wont get ANY wider usage, if its going to be broken all the time. To prevent that, we need either to slow down development in any other areas, or stop it completely, as we are basically doubling the work.

Its really not that bad. Its only few cases (like 1 out of 30 commits) 
that changes rbuild/cmake files, so only in those cases it would be 
neccessary to do additional work and check if it builds. If you only 
change source files, its very unlikely to break build.
So for now until kdbg is fixed and hopefully everyone knows how to use 
cmake, I'd appreciate if you would take the extra efford in these cases 
(it won't be that way forever!) and also fix the cmake files. This way 
you will also learn how to work with cmake files. Thanks to Alex for at 
least trying to not break cmake builds, Ged, your "the cmake people will 
manage that" approach is disapproved :)

Thanks,
Timo

[Zachary Gorden]

While Alex was trying to avoid build breakages, that recent incident with
his deletion of functions does highlight quite nicely why rbuild needs to be
put down.

On Thu, Jul 21, 2011 at 7:38 PM, Timo Kreuzer [e-mail] wrote:
> Am 22.07.2011 00:38, schrieb [e-mail]:
>
> CMake wont get ANY wider usage, if its going to be broken all the time. To
> > prevent that, we need either to slow down development in any other areas, or
> > stop it completely, as we are basically doubling the work.
>
> Its really not that bad. Its only few cases (like 1 out of 30 commits) that
> changes rbuild/cmake files, so only in those cases it would be neccessary to
> do additional work and check if it builds. If you only change source files,
> its very unlikely to break build.
> So for now until kdbg is fixed and hopefully everyone knows how to use
> cmake, I'd appreciate if you would take the extra efford in these cases (it
> won't be that way forever!) and also fix the cmake files. This way you will
> also learn how to work with cmake files. Thanks to Alex for at least trying
> to not break cmake builds, Ged, your "the cmake people will manage that"
> approach is disapproved :)
>
> Thanks,
> Timo
>
> ______________________________**_________________
> Ros-dev mailing list
> [e-mail]
> http://www.reactos.org/**mailman/listinfo/ros-devhttp://www.reactos.org/mailman/listinfo/ros-dev

[Sylvain Petreolle]

+1, linking on thin air should never be possible.
> De : Zachary Gorden [e-mail]
> À : ReactOS Development List [e-mail]
> Envoyé le : Vendredi 22 Juillet 2011 2h43
> Objet : Re: [ros-dev] cmake rant
>
> While Alex was trying to avoid build breakages, that recent incident with his deletion of functions does highlight quite nicely why rbuild needs to be put down.
>
> On Thu, Jul 21, 2011 at 7:38 PM, Timo Kreuzer [e-mail] wrote:
>
> Am 22.07.2011 00:38, schrieb [e-mail]:
> > CMake wont get ANY wider usage, if its going to be broken all the time. To prevent that, we need either to slow down development in any other areas, or stop it completely, as we are basically doubling the work.
> > >

Its really not that bad. Its only few cases (like 1 out of 30 commits) that changes rbuild/cmake files, so only in those cases it would be neccessary to do additional work and check if it builds. If you only change source files, its very unlikely to break build.
> > So for now until kdbg is fixed and hopefully everyone knows how to use cmake, I'd appreciate if you would take the extra efford in these cases (it won't be that way forever!) and also fix the cmake files. This way you will also learn how to work with cmake files. Thanks to Alex for at least trying to not break cmake builds, Ged, your "the cmake people will manage that" approach is disapproved :)
> >
> > Thanks,
> > Timo
> >
> > Ros-dev mailing list
> > [e-mail]
> > http://www.reactos.org/mailman/listinfo/ros-dev
>
> Ros-dev mailing list
> [e-mail]
> http://www.reactos.org/mailman/listinfo/ros-dev

[Ged Murphy]

Timo Kreuzer wrote:
> Ged, your "the cmake people will manage that" approach is disapproved :)

I wasn't looking for an approval, I was stating a fact
The fact is, if I check anything into reactos, I won't bother to check
whether it builds with cmake until it's the default build.
When it is the default build, I won't be bothering to check whether it
builds with rbuild.
> From my point of view, it's either cmake or rbuild. We can't have both.

Ged.

[Timo Kreuzer]

Am 25.07.2011 10:27, schrieb Ged Murphy:
> I wasn't looking for an approval, I was stating a fact The fact is, if 
> I check anything into reactos, I won't bother to check whether it 
> builds with cmake until it's the default build. When it is the default 
> build, I won't be bothering to check whether it builds with rbuild.

I hope you recognize that this is extremely egoistic.
I'm not going to explain WHY we have 2 build systems atm, because it 
should should know by now.
If you have any interest in moving forward, you should consider 
supporting changes, instead of blocking them.

Thanks,
Timo

[Ged Murphy]

Timo Kreuzer wrote:
> I hope you recognize that this is extremely egoistic.

I wouldn't say egotistical, I'd say it's an 'unwillingness to conform to a
broken model' or 'understanding when a method of practice is incorrect and
choosing not to follow it'
As it happens I probably won't be checking any code in, I was merely making
a point, so you have nothing to worry about :)
> If you have any interest in moving forward, you should consider supporting

changes, instead of blocking them.

I'm heavily supporting it, I'm calling for rbuild to be removed and cmake to
be default.
Other than attending its Christmas nativity play, I'm not sure what more I
can do.

Ged.

[James Tabor]

I'm for the change! Just update the dumb dumb "wiki" pages so us from
remote parts of the world can do it from rbuild tree w/o  any hang
ups....

Thanks,
James

On Thu, Jul 21, 2011 at 3:30 PM, Aleksey Bragin [e-mail] wrote:
> I didn’t hear many positive opinions when I forced everyone to switch over
> to the new ntdll loader. I minusone this proposal too. CMake should get
> wider usage first, and get the regressions sorted out. After all, it’s a
> build system change, not a migration to BSD kernel, it’s a doable task.
>
> I’m all to help with hunting down CMake problems btw.
>
> WBR,
> Aleksey Bragin.

[Timo Kreuzer]

Am 23.07.2011 00:38, schrieb James Tabor:
> I'm for the change! Just update the dumb dumb "wiki" pages so us from
> remote parts of the world can do it from rbuild tree w/o  any hang
> ups....

What wiki page do you mean? The "CMake" page? Is there anything missing, 
I think it was pretty complete.
Or do you mean the "Compiling reactos" page? That one would need to be 
updated, yes.
If there are questions, ask :)

Timo

[victor martinez]

My grandma(neither me) wasn't able to follow the "Cmake" setup wikipage. So more work is needed. The day I'm able to create a build with Cmake method, then I'll update the Wiki accordingly.
For cmake developers is pretty obvious but I failed twice and rejected to perform a third attempt.
:)
> Date: Sat, 23 Jul 2011 17:19:50 +0200
> From: [e-mail]
> To: [e-mail]
> Subject: Re: [ros-dev] cmake rant
>
> Am 23.07.2011 00:38, schrieb James Tabor:
> > I'm for the change! Just update the dumb dumb "wiki" pages so us from
> > remote parts of the world can do it from rbuild tree w/o  any hang
> > ups....
>
> What wiki page do you mean? The "CMake" page? Is there anything missing, 
> I think it was pretty complete.
> Or do you mean the "Compiling reactos" page? That one would need to be 
> updated, yes.
> If there are questions, ask :)
>
> Timo
>
> Ros-dev mailing list
> [e-mail]
> http://www.reactos.org/mailman/listinfo/ros-dev

[Timo Kreuzer]

Am 23.07.2011 18:23, schrieb victor martinez:
> My grandma(neither me) wasn't able to follow the "Cmake" setup 
> wikipage. So more work is needed. The day I'm able to create a build 
> with Cmake method, then I'll update the Wiki accordingly.
> For cmake developers is pretty obvious but I failed twice and rejected 
> to perform a third attempt.
> :)

I talked to your grandma, she said it was easy and you are a n00b, so 
she gave up trying to explain it to you :-p

j/k

[victor martinez]

Baztarz!
j/k

Date: Sat, 23 Jul 2011 19:04:31 +0200
From: [e-mail]
To: [e-mail]
Subject: Re: [ros-dev] cmake rant

Am 23.07.2011 18:23, schrieb victor martinez:

My grandma(neither me) wasn't able to follow the "Cmake" setup
        wikipage. So more work is needed. The day I'm able to create a
        build with Cmake method, then I'll update the Wiki accordingly.

For cmake developers is pretty obvious but I failed twice and
        rejected to perform a third attempt.

:)

I talked to your grandma, she said it was easy and you are a n00b,
    so she gave up trying to explain it to you :-p

j/k

_______________________________________________
Ros-dev mailing list
[e-mail]
http://www.reactos.org/mailman/listinfo/ros-dev

[Elvis]

Your grandma is trying to build ReactOS using the cmake system? Now that 
is one l33t grandma. :)
On 7/24/11 2:23, victor martinez wrote:
> My grandma(neither me) wasn't able to follow the "Cmake" setup wikipage.
> So more work is needed. The day I'm able to create a build with Cmake
> method, then I'll update the Wiki accordingly.
> For cmake developers is pretty obvious but I failed twice and rejected
> to perform a third attempt.
> :)
> > Date: Sat, 23 Jul 2011 17:19:50 +0200
> > From: [e-mail]
> > To: [e-mail]
> > Subject: Re: [ros-dev] cmake rant
> >
> > Am 23.07.2011 00:38, schrieb James Tabor:
> > > I'm for the change! Just update the dumb dumb "wiki" pages so us from
> > > remote parts of the world can do it from rbuild tree w/o any hang
> > > ups....
> >
> > What wiki page do you mean? The "CMake" page? Is there anything missing,
> > I think it was pretty complete.
> > Or do you mean the "Compiling reactos" page? That one would need to be
> > updated, yes.
> > If there are questions, ask :)
> >
> > Timo
> >
> > Ros-dev mailing list
> > [e-mail]
> > http://www.reactos.org/mailman/listinfo/ros-dev
>
> Ros-dev mailing list
> [e-mail]
> http://www.reactos.org/mailman/listinfo/ros-dev

[Amine Khaldi]

Hi,

Being one of the "CMake people", I'm obviously biased toward the adoption ;)

Seriously, it's the logical/rational thing to do.

That said, I just committed a temporary revert so that backtraces are 
reliable again and CMake adoption can really be possible now.

I also want to make sure everyone understands one thing: CMake is 
awesome for us. CMake allows us to have our build system maintained by 
3rd party. It allows us to have simple, readable build system 
macros/functions which means anyone can hack them. It allows us to 
support several ways to compile ros (CMake generators), from using xcode 
to eclipse to VS to nmake and (mingw/unix) make files. It allows us to 
have the MSVC builds we always have been wanting.

There is one tiny catch though: because of the excellent dependency 
tracking, CMake build can be significantly slower compared to rbuild 
build. You won't notice this if you check dependencies yourself and use 
make module/fast, but if you compile full bootcd for example you'll 
notice that it takes a moment to do dependency checking. This very 
slowness is nothing compared to rbuild looping around itself n times 
though ;)

As I'm sure you realize, it's not just about speed. The features we gain 
are well worth it.

As final note: rest assured that we will do our best to make the CMake 
build as fast as possible.

Regards,
Amine.

[Conan Kudo (ニール・ゴンパ)]

Does this mean that we can purge rbuild and update the wiki docs to have
information on how to build ReactOS from source using CMake? Or are there
any other showstopping issues that would prevent this?

On Sat, Jul 23, 2011 at 6:33 PM, Amine Khaldi [e-mail]:
> Hi,
>
> Being one of the "CMake people", I'm obviously biased toward the adoption
> ;)
>
> Seriously, it's the logical/rational thing to do.
>
> That said, I just committed a temporary revert so that backtraces are
> reliable again and CMake adoption can really be possible now.
>
> I also want to make sure everyone understands one thing: CMake is awesome
> for us. CMake allows us to have our build system maintained by 3rd party. It
> allows us to have simple, readable build system macros/functions which means
> anyone can hack them. It allows us to support several ways to compile ros
> (CMake generators), from using xcode to eclipse to VS to nmake and
> (mingw/unix) make files. It allows us to have the MSVC builds we always have
> been wanting.
>
> There is one tiny catch though: because of the excellent dependency
> tracking, CMake build can be significantly slower compared to rbuild build.
> You won't notice this if you check dependencies yourself and use make
> module/fast, but if you compile full bootcd for example you'll notice that
> it takes a moment to do dependency checking. This very slowness is nothing
> compared to rbuild looping around itself n times though ;)
>
> As I'm sure you realize, it's not just about speed. The features we gain
> are well worth it.
>
> As final note: rest assured that we will do our best to make the CMake
> build as fast as possible.
>
> Regards,
> Amine.
>
> ______________________________**_________________
> Ros-dev mailing list
> [e-mail]
> http://www.reactos.org/**mailman/listinfo/ros-devhttp://www.reactos.org/mailman/listinfo/ros-dev

After some checks, it seems that a single header file change takes around 1m to do a quick rebuild. If cmake files are touched, CMake does recheck and it takes around 3-3.5m. The full build  takes around 5m. Please mind that these values are affected by CCache.

On Sat, 23 Jul 2011 18:38 -0500, "Conan Kudo (ニール・ゴンパ)" [e-mail] wrote:
> Does this mean that we can purge rbuild and update the wiki docs to have
> information on how to build ReactOS from source using CMake? Or are there
> any other showstopping issues that would prevent this?
>
> On Sat, Jul 23, 2011 at 6:33 PM, Amine Khaldi
> [e-mail]:
> > Hi,
> >
> > Being one of the "CMake people", I'm obviously biased toward the adoption
> > ;)
> >
> > Seriously, it's the logical/rational thing to do.
> >
> > That said, I just committed a temporary revert so that backtraces are
> > reliable again and CMake adoption can really be possible now.
> >
> > I also want to make sure everyone understands one thing: CMake is awesome
> > for us. CMake allows us to have our build system maintained by 3rd party. It
> > allows us to have simple, readable build system macros/functions which means
> > anyone can hack them. It allows us to support several ways to compile ros
> > (CMake generators), from using xcode to eclipse to VS to nmake and
> > (mingw/unix) make files. It allows us to have the MSVC builds we always have
> > been wanting.
> >
> > There is one tiny catch though: because of the excellent dependency
> > tracking, CMake build can be significantly slower compared to rbuild build.
> > You won't notice this if you check dependencies yourself and use make
> > module/fast, but if you compile full bootcd for example you'll notice that
> > it takes a moment to do dependency checking. This very slowness is nothing
> > compared to rbuild looping around itself n times though ;)
> >
> > As I'm sure you realize, it's not just about speed. The features we gain
> > are well worth it.
> >
> > As final note: rest assured that we will do our best to make the CMake
> > build as fast as possible.
> >
> > Regards,
> > Amine.
> >
> > ______________________________**_________________
> > Ros-dev mailing list
> > [e-mail]
> > http://www.reactos.org/**mailman/listinfo/ros-devhttp://www.reactos.org/mailman/listinfo/ros-dev
>
> Ros-dev mailing list
> [e-mail]
> http://www.reactos.org/mailman/listinfo/ros-dev

With best regards
Caemyr

The test runtime is working again. Seems that debugging is back to working state, even without umode autotracing.

We should now push the cmake transition. Waiting wont change much, will only make it more difficut.

On Sun, 24 Jul 2011 15:48 +0200, [e-mail] wrote:
> After some checks, it seems that a single header file change takes around
> 1m to do a quick rebuild. If cmake files are touched, CMake does recheck
> and it takes around 3-3.5m. The full build  takes around 5m. Please mind
> that these values are affected by CCache.

With best regards
Caemyr

[Ged Murphy]

That’s totally different. The loader is a core component and was a long way off being ready, as Alex has shown with the huge amount of work he’s done on it since it was committed.

It got to the point where most devs had reverted it due to the problems it was causing.
> From what I understand, cmake is pretty much ready to go and quite a few people seem to already be using it as their default build system.

There’s now a usage split in the community between the cmake guys and the rbuild guys.

Both currently have their merits and their issues, but both are largely working.

Ged.

From: [e-mail] [mailto:[e-mail]] On Behalf Of Aleksey Bragin
Sent: 21 July 2011 21:31
To: ReactOS Development List
Subject: Re: [ros-dev] cmake rant

I didn’t hear many positive opinions when I forced everyone to switch over to the new ntdll loader. I minusone this proposal too. CMake should get wider usage first, and get the regressions sorted out. After all, it’s a build system change, not a migration to BSD kernel, it’s a doable task.

I’m all to help with hunting down CMake problems btw.

WBR,

Aleksey Bragin.

From: Gabriel ilardi mailto:[e-mail]

Sent: Thursday, July 21, 2011 9:16 PM

To: ReactOS Development List mailto:[e-mail]

Subject: Re: [ros-dev] cmake rant

-1 imho, not at least debugging is possible, we keep accumulating regressions and not being able to debug is not an option to me...
> Date: Thu, 21 Jul 2011 18:57:28 +0200
> From: [e-mail]
> To: [e-mail]
> Subject: Re: [ros-dev] cmake rant
>
> +1 for Ged.
>
> Le 21/07/2011 18:51, Ged Murphy a écrit :
> > I'm all for deleting rbuild now and dealing with the consequences.
> > That would be better than the current situation, which seems rather
> > ridiculous at the moment.
> >
> > Right now if you change something, you have to :
> > edit the cmakefiles.txt file
> > edit the rbuild file
> > possibly edit a pspec and spec file
> > build with rbuild to check for breakages
> > build with cmake to check for breakages
> >
> > Only then can you commit your changes.
> > It's not ideal for the people working on this stuff regularly.
> >
> > Ged.
> >
> > -----Original Message-----
> > From: [e-mail] [mailto:[e-mail]] On
> > Behalf Of Timo Kreuzer
> > Sent: 21 July 2011 17:39
> > To: ReactOS Development List
> > Subject: [ros-dev] cmake rant
> >
> > Hi,
> >
> > Several times now cmake build has been broken. Time for some action!
> >
> > Last meeting I asked everyone to test/use cmake. It was also mentioned that
> > if questions arise, we (Amine and me) would be happy to help out. I can't
> > remember anyone has asked how it works, so I assume noone had any problems.
> > There's also a pretty good wiki entry describing the whole procedure for
> > n00bs.
> >
> > Now people tell me it's complicated, people are complaining that its
> > ridiculous to have 2 build systems, etc.
> > And probably noone has ever tried it.
> >
> > We really need to move on.
> > I don't like having 2 build systems as well.
> >
> > Current blocker is the debugging which has some issues, Arty is working on
> > that. Another problem is a boot problem on real hardware, but no I don't
> > know on which configuration it doesn't work, so we need more people testing
> > it on their real hardware setup and report any issues.
> >
> > Here's a list with current issues:
> > http://www.reactos.org/wiki/CMake_Todo
> >
> > So please:
> > If you are missing something, let us know.
> > If you like to make it better, make suggestions.
> > But stop ignoring cmake!
> >
> > If noone cares and everyone just thinks he can give a s^Z damn until we
> > officially switch, then we can as well delete all cmake stuff and keep
> > rbuild. It has a lot of awesome advantages, like you only have to type one
> > command to build everything and you don't need to install cmake and you can
> > export whatever you want from kernel32 even if the functions don't exist.
> > Also you can enjoy the rbuild-loop again and again.
> >
> > Or we can do it the hard way and delete rbuild, so people are forced to use
> > cmake. I'm sure this approach would be *really* appreaciated.
> >
> > Thanks,
> > Timo
> >
> > Ros-dev mailing list
> > [e-mail]
> > http://www.reactos.org/mailman/listinfo/ros-dev
> >
> > Ros-dev mailing list
> > [e-mail]
> > http://www.reactos.org/mailman/listinfo/ros-dev
>
> Ros-dev mailing list
> [e-mail]
> http://www.reactos.org/mailman/listinfo/ros-dev

_____

_______________________________________________
Ros-dev mailing list
[e-mail]
http://www.reactos.org/mailman/listinfo/ros-dev

[Conan Kudo (ニール・ゴンパ)]

I agree with Ged on this too. Purging rbuild and forcing everyone to use
CMake is really the only way to get people to use it. So what if we can't
build for a little bit? At least we'll get CMake into a fully workable state
much more quickly with more attention on it. Admittedly, I haven't been
trying out recent revisions of the cmake branch, but that is because I have
a hard time actually downloading from the ReactOS SVN anyway.

On Thu, Jul 21, 2011 at 11:51 AM, Ged Murphy
[e-mail]:
> I'm all for deleting rbuild now and dealing with the consequences.
> That would be better than the current situation, which seems rather
> ridiculous at the moment.
>
> Right now if you change something, you have to :
> edit the cmakefiles.txt file
> edit the rbuild file
> possibly edit a pspec and spec file
> build with rbuild to check for breakages
> build with cmake to check for breakages
>
> Only then can you commit your changes.
> It's not ideal for the people working on this stuff regularly.
>
> Ged.
>
> -----Original Message-----
> From: [e-mail] [mailto:[e-mail]] On
> Behalf Of Timo Kreuzer
> Sent: 21 July 2011 17:39
> To: ReactOS Development List
> Subject: [ros-dev] cmake rant
>
> Hi,
>
> Several times now cmake build has been broken. Time for some action!
>
> Last meeting I asked everyone to test/use cmake. It was also mentioned that
> if questions arise, we (Amine and me) would be happy to help out. I can't
> remember anyone has asked how it works, so I assume noone had any problems.
> There's also a pretty good wiki entry describing the whole procedure for
> n00bs.
>
> Now people tell me it's complicated, people are complaining that its
> ridiculous to have 2 build systems, etc.
> And probably noone has ever tried it.
>
> We really need to move on.
> I don't like having 2 build systems as well.
>
> Current blocker is the debugging which has some issues, Arty is working on
> that. Another problem is a boot problem on real hardware, but no I don't
> know on which configuration it doesn't work, so we need more people testing
> it on their real hardware setup and report any issues.
>
> Here's a list with current issues:
> http://www.reactos.org/wiki/CMake_Todo
>
> So please:
> If you are missing something, let us know.
> If you like to make it better, make suggestions.
> But stop ignoring cmake!
>
> If noone cares and everyone just thinks he can give a s^Z damn until we
> officially switch, then we can as well delete all cmake stuff and keep
> rbuild. It has a lot of awesome advantages, like you only have to type one
> command to build everything and you don't need to install cmake and you can
> export whatever you want from kernel32 even if the functions don't exist.
> Also you can enjoy the rbuild-loop again and again.
>
> Or we can do it the hard way and delete rbuild, so people are forced to use
> cmake. I'm sure this approach would be *really* appreaciated.
>
> Thanks,
> Timo
>
> Ros-dev mailing list
> [e-mail]
> http://www.reactos.org/mailman/listinfo/ros-dev
>
> Ros-dev mailing list
> [e-mail]
> http://www.reactos.org/mailman/listinfo/ros-dev