bitbake parsing of IMAGE_INSTALL += # tslib mtd-utils" extremely user unfriendly.

bitbake parsing of IMAGE_INSTALL += # tslib mtd-utils" extremely user unfriendly.

[Leon Woestenberg]

[-- Attachment #1: Type: text/plain, Size: 282 bytes --]

Hello all,

bitbake can really braindump on us when we insert typo's.  The result of
bitbake 1.14 parsing this one wasn't pleasant:

IMAGE_INSTALL += # tslib mtd-utils"

(Yes, it's a typo. No, I wouldn't expect bitbake to give me that much
output :) )

Regards,

Leon.

[-- Attachment #2: Type: text/html, Size: 327 bytes --]

Re: bitbake parsing of IMAGE_INSTALL += # tslib mtd-utils" extremely user unfriendly.

[Richard Purdie]

On Mon, 2012-02-20 at 23:00 +0100, Leon Woestenberg wrote:
> Hello all,
> 
> bitbake can really braindump on us when we insert typo's.  The result
> of bitbake 1.14 parsing this one wasn't pleasant:
> 
> IMAGE_INSTALL += # tslib mtd-utils"
> 
> (Yes, it's a typo. No, I wouldn't expect bitbake to give me that much
> output :) )

What was the output? (I realise I can find out)

With usability issues like this, opening a bug in
bugzilla.yoctoproject.org is a good idea, particularly as we've been
focusing on cleaning up this kind of problem. Reporting them with a
known reproducible test case helps a lot as when you're looking to try
and fix these things, you need a good set of reproducers. 

As a quick plug for the work that's been done already, we've now worked
through many of the issues that were reported so it is worth taking the
time to report them! :)

Cheers,

Richard

Re: bitbake parsing of IMAGE_INSTALL += # tslib mtd-utils" extremely user unfriendly.

[Eric Bénard]

Hi Richard,

Le Tue, 21 Feb 2012 12:56:14 +0000,
Richard Purdie <richard.purdie@linuxfoundation.org> a écrit :
> 
> What was the output? (I realise I can find out)
> 
> With usability issues like this, opening a bug in
> bugzilla.yoctoproject.org is a good idea, particularly as we've been
> focusing on cleaning up this kind of problem. Reporting them with a
> known reproducible test case helps a lot as when you're looking to try
> and fix these things, you need a good set of reproducers. 
> 
> As a quick plug for the work that's been done already, we've now worked
> through many of the issues that were reported so it is worth taking the
> time to report them! :)
> 
I had also a bitbake failure after an error in an image and bitbake
exited without any information concerning the error (even when running
with -DDD I didn't managed to see a log - but I must admit I didn't
search in the whole history).

Basically I had :
XSERVER += "    <==== missing \
	xf86-input-tslib \
	"
and bitbake exit with :
ERROR: Command execution failed: Exited with 1
et nothing more.

Eric



-- 
http://eukrea.com/en/news/104-2012

Re: bitbake parsing of IMAGE_INSTALL += # tslib mtd-utils" extremely user unfriendly.

[Paul Eggleton]

On Thursday 23 February 2012 16:30:32 Eric Bénard wrote:
> Richard Purdie <richard.purdie@linuxfoundation.org> a écrit :
> > What was the output? (I realise I can find out)
> > 
> > With usability issues like this, opening a bug in
> > bugzilla.yoctoproject.org is a good idea, particularly as we've been
> > focusing on cleaning up this kind of problem. Reporting them with a
> > known reproducible test case helps a lot as when you're looking to try
> > and fix these things, you need a good set of reproducers.
> > 
> > As a quick plug for the work that's been done already, we've now worked
> > through many of the issues that were reported so it is worth taking the
> > time to report them! :)
> 
> I had also a bitbake failure after an error in an image and bitbake
> exited without any information concerning the error (even when running
> with -DDD I didn't managed to see a log - but I must admit I didn't
> search in the whole history).
> 
> Basically I had :
> XSERVER += "    <==== missing \
> 	xf86-input-tslib \
> 	"
> and bitbake exit with :
> ERROR: Command execution failed: Exited with 1
> et nothing more.

This got broken last night and fixed last night, so if you update you should 
have the fix.

Cheers,
Paul

-- 

Paul Eggleton
Intel Open Source Technology Centre

Re: bitbake parsing of IMAGE_INSTALL += # tslib mtd-utils" extremely user unfriendly.

[Paul Eggleton]

On Thursday 23 February 2012 15:32:47 Paul Eggleton wrote:
> > Basically I had :
> > XSERVER += "    <==== missing \
> > 
> > 	xf86-input-tslib \
> > 	"
> > 
> > and bitbake exit with :
> > ERROR: Command execution failed: Exited with 1
> > et nothing more.
> 
> This got broken last night and fixed last night, so if you update you should
> have the fix.

Er, actually that's completely false. However I do have a bitbake patch ready 
to submit to fix this.

Cheers,
Paul

-- 

Paul Eggleton
Intel Open Source Technology Centre

Re: bitbake parsing of IMAGE_INSTALL += # tslib mtd-utils" extremely user unfriendly.

[Eric Bénard]

Le Thu, 23 Feb 2012 15:38:33 +0000,
Paul Eggleton <paul.eggleton@linux.intel.com> a écrit :

> On Thursday 23 February 2012 15:32:47 Paul Eggleton wrote:
> > > Basically I had :
> > > XSERVER += "    <==== missing \
> > > 
> > > 	xf86-input-tslib \
> > > 	"
> > > 
> > > and bitbake exit with :
> > > ERROR: Command execution failed: Exited with 1
> > > et nothing more.
> > 
> > This got broken last night and fixed last night, so if you update you should
> > have the fix.
> 
> Er, actually that's completely false. However I do have a bitbake patch ready 
> to submit to fix this.
> 
I confirm this is false I as updated bitbake before sending the mail ;-)

Thanks,
Eric
-- 
http://eukrea.com/en/news/104-2012

Re: bitbake parsing of IMAGE_INSTALL += # tslib mtd-utils" extremely user unfriendly.

[Richard Purdie]

On Thu, 2012-02-23 at 16:30 +0100, Eric Bénard wrote:
> Hi Richard,
> 
> Le Tue, 21 Feb 2012 12:56:14 +0000,
> Richard Purdie <richard.purdie@linuxfoundation.org> a écrit :
> > 
> > What was the output? (I realise I can find out)
> > 
> > With usability issues like this, opening a bug in
> > bugzilla.yoctoproject.org is a good idea, particularly as we've been
> > focusing on cleaning up this kind of problem. Reporting them with a
> > known reproducible test case helps a lot as when you're looking to try
> > and fix these things, you need a good set of reproducers. 
> > 
> > As a quick plug for the work that's been done already, we've now worked
> > through many of the issues that were reported so it is worth taking the
> > time to report them! :)
> > 
> I had also a bitbake failure after an error in an image and bitbake
> exited without any information concerning the error (even when running
> with -DDD I didn't managed to see a log - but I must admit I didn't
> search in the whole history).
> 
> Basically I had :
> XSERVER += "    <==== missing \
> 	xf86-input-tslib \
> 	"
> and bitbake exit with :
> ERROR: Command execution failed: Exited with 1
> et nothing more.

This think this one has been fixed now. Leon's original error is more
problematic though...

Cheers,

Richard

Re: bitbake parsing of IMAGE_INSTALL += # tslib mtd-utils" extremely user unfriendly.

[Richard Purdie]

On Mon, 2012-02-20 at 23:00 +0100, Leon Woestenberg wrote:
> bitbake can really braindump on us when we insert typo's.  The result
> of bitbake 1.14 parsing this one wasn't pleasant:
>
> IMAGE_INSTALL += # tslib mtd-utils"
>
> (Yes, it's a typo. No, I wouldn't expect bitbake to give me that much
> output :) )

I agree we need to fix that. I've proposed a patch to bitbake which
would allow detection and a better error message for something like this
by enforcing quoting of variables. It is a major change in bitbake's
behaviour though so I'm taking feedback on whether we should make the
change.

Cheers,

Richard

Re: bitbake parsing of IMAGE_INSTALL += # tslib mtd-utils" extremely user unfriendly.

[Leon Woestenberg]

[-- Attachment #1: Type: text/plain, Size: 964 bytes --]

Hello Richard,

On Sun, Feb 26, 2012 at 2:06 PM, Richard Purdie <
richard.purdie@linuxfoundation.org> wrote:

> On Mon, 2012-02-20 at 23:00 +0100, Leon Woestenberg wrote:
> > bitbake can really braindump on us when we insert typo's.  The result
> > of bitbake 1.14 parsing this one wasn't pleasant:
> >
> > IMAGE_INSTALL += # tslib mtd-utils"
> >
> > (Yes, it's a typo. No, I wouldn't expect bitbake to give me that much
> > output :) )
>
> I agree we need to fix that. I've proposed a patch to bitbake which
> would allow detection and a better error message for something like this
> by enforcing quoting of variables. It is a major change in bitbake's
> behaviour though so I'm taking feedback on whether we should make the
> change.
>

If we focus on our user interface view, that would mean this will no longer
work:

SOME_BINARY_VARIABLE = 1

but that might be a better compromise than what we currently have.

Regards,

Leon.

[-- Attachment #2: Type: text/html, Size: 1374 bytes --]

Re: bitbake parsing of IMAGE_INSTALL += # tslib mtd-utils" extremely user unfriendly.

[Richard Purdie]

On Sun, 2012-02-26 at 20:35 +0100, Leon Woestenberg wrote:
> On Sun, Feb 26, 2012 at 2:06 PM, Richard Purdie
> <richard.purdie@linuxfoundation.org> wrote:
>         On Mon, 2012-02-20 at 23:00 +0100, Leon Woestenberg wrote:
>         > bitbake can really braindump on us when we insert typo's.
>          The result
>         > of bitbake 1.14 parsing this one wasn't pleasant:
>         >
>         > IMAGE_INSTALL += # tslib mtd-utils"
>         >
>         > (Yes, it's a typo. No, I wouldn't expect bitbake to give me
>         that much
>         > output :) )
>
>         I agree we need to fix that. I've proposed a patch to bitbake
>         which
>         would allow detection and a better error message for something
>         like this
>         by enforcing quoting of variables. It is a major change in
>         bitbake's
>         behaviour though so I'm taking feedback on whether we should
>         make the
>         change.
>
> If we focus on our user interface view, that would mean this will no
> longer work:
>
> SOME_BINARY_VARIABLE = 1
>
> but that might be a better compromise than what we currently have.

FWIW, internally, that is still a string even if it looks a bit like a
binary value so there is no effective change.

Cheers,

Richard