You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@geronimo.apache.org by Jason Dillon <ja...@planet57.com> on 2007/06/08 00:52:33 UTC
Yo yo... Xvfb, selenium blah
So, I hacked up something in the latest 1.0-beta-2-SNAPSHOT of the
selenium-maven-plugin. Its deployed, and some site docs updated.
http://mojo.codehaus.org/selenium-maven-plugin/examples/headless-
with-xvfb.html
Still needs more meat, but I wanna make sure it works before I start
investing in more docs.
But basically, you need to configure an execution *before* start-
server, for xvfb. It will be default background, etc. See the goal
doc for more details:
http://mojo.codehaus.org/selenium-maven-plugin/xvfb-mojo.html
Currently it will default to using DISPLAY=":1", not sure if that is
correct or not. Then it will fork an Xvfb process, write out a
target/selenium/display.properties file, which has the DISPLAY= in
it. And I updated the start-server goal to look for that file, and
if it exists set environment variables for each key/pair in that file.
I think that should be sufficient to get headless muck working...
though I've not tested any of it ;-)
I think you will need to make the Xvfb process suid, as in:
chmod u+s `which Xvfb`
This puppy needs to be run as root, or at least AFAIK it does.
Anyways, give it a shot and lemme know how well it works, or if it
works at all. I will be around tomorrow to fix it up as needed.
Maybe my network will be stable then so I can run some tests on the
gbuild.org systems and better see how well it works out.
Aighty?
Cheers,
--jason
Re: Yo yo... Xvfb, selenium blah
Posted by Jason Dillon <ja...@planet57.com>.
On Jun 8, 2007, at 3:43 AM, Daniel Alheiros wrote:
> Hi
>
> Sorry if I'm saying something stupid, but it's been a long time
> since I read
> the forum discussions.
>
> Wouldn't be enough to use the Java standard headless support?
> It was introduced in JDK 1.4.2 (
> http://java.sun.com/j2se/1.4.2/docs/guide/awt/
> AWTChanges.html#headless)
Selenium launches web browsers, like Firefox... which isn't a Java
application so AWT configuration isn't relative.
> I've used both xvfb and AWT headless and I could say that the
> second option
> was much simpler to set up and more compatible.
I agree that headless AWT is easier, but in this case it doesn't apply.
--jason
> Regards,
> Daniel
>
>
> On 8/6/07 04:50, "Jason Dillon" <ja...@planet57.com> wrote:
>
>> And now I think I've got the auto-detection of free display and
>> rough/
>> good-enough verification of Xvfb working now.
>>
>> Seems to work... so I've updated the docs a wee bit and deployed a
>> new 1.0-beta-2-SNAPSHOT.
>>
>> I think I'm down working on this for now. So Prasad, please have a
>> look and let me know if this will work for you. If so, I'll start a
>> vote for a release later next week.
>>
>> On revision from the notes below, the default display is now :20, and
>> it will try and detect a free display from :20 to :30 (and then give
>> up and barf).
>>
>> * * *
>>
>> BTW, the Xvfb config needs to be put into a profile, and explicitly
>> activated, otherwise it will hose folks who don't have Xvfb installed
>> (and the profile must cause the execution of the xvfb goal before
>> start-server). I suppose if we really want I could probably figure
>> out how to detect when Xvfb is needed and if its available, but...
>> for now lets just make it explicitly activated for folks that need
>> the testsuite to run on a headless system.
>>
>> Cheers,
>>
>> --jason
>>
>>
>> On Jun 7, 2007, at 8:05 PM, Matt Hogstrom wrote:
>>
>>> Cool shtuffes ... your a busy beaver.
>>>
>>> On Jun 7, 2007, at 6:52 PM, Jason Dillon wrote:
>>>
>>>> So, I hacked up something in the latest 1.0-beta-2-SNAPSHOT of the
>>>> selenium-maven-plugin. Its deployed, and some site docs updated.
>>>>
>>>> http://mojo.codehaus.org/selenium-maven-plugin/examples/
>>>> headless-with-xvfb.html
>>>>
>>>> Still needs more meat, but I wanna make sure it works before I
>>>> start investing in more docs.
>>>>
>>>> But basically, you need to configure an execution *before* start-
>>>> server, for xvfb. It will be default background, etc. See the
>>>> goal doc for more details:
>>>>
>>>> http://mojo.codehaus.org/selenium-maven-plugin/xvfb-mojo.html
>>>>
>>>> Currently it will default to using DISPLAY=":1", not sure if that
>>>> is correct or not. Then it will fork an Xvfb process, write out a
>>>> target/selenium/display.properties file, which has the DISPLAY= in
>>>> it. And I updated the start-server goal to look for that file,
>>>> and if it exists set environment variables for each key/pair in
>>>> that file.
>>>>
>>>> I think that should be sufficient to get headless muck working...
>>>> though I've not tested any of it ;-)
>>>>
>>>> I think you will need to make the Xvfb process suid, as in:
>>>>
>>>> chmod u+s `which Xvfb`
>>>>
>>>> This puppy needs to be run as root, or at least AFAIK it does.
>>>>
>>>> Anyways, give it a shot and lemme know how well it works, or if it
>>>> works at all. I will be around tomorrow to fix it up as needed.
>>>> Maybe my network will be stable then so I can run some tests on
>>>> the gbuild.org systems and better see how well it works out.
>>>>
>>>> Aighty?
>>>>
>>>> Cheers,
>>>>
>>>> --jason
>>>>
>>>
>>
>
>
> http://www.bbc.co.uk/
> This e-mail (and any attachments) is confidential and may contain
> personal views which are not the views of the BBC unless
> specifically stated.
> If you have received it in error, please delete it from your system.
> Do not use, copy or disclose the information in any way nor act in
> reliance on it and notify the sender immediately.
> Please note that the BBC monitors e-mails sent or received.
> Further communication will signify your consent to this.
>
Re: Yo yo... Xvfb, selenium blah
Posted by Daniel Alheiros <Da...@bbc.co.uk>.
Hi
Sorry if I'm saying something stupid, but it's been a long time since I read
the forum discussions.
Wouldn't be enough to use the Java standard headless support?
It was introduced in JDK 1.4.2 (
http://java.sun.com/j2se/1.4.2/docs/guide/awt/AWTChanges.html#headless)
I've used both xvfb and AWT headless and I could say that the second option
was much simpler to set up and more compatible.
Regards,
Daniel
On 8/6/07 04:50, "Jason Dillon" <ja...@planet57.com> wrote:
> And now I think I've got the auto-detection of free display and rough/
> good-enough verification of Xvfb working now.
>
> Seems to work... so I've updated the docs a wee bit and deployed a
> new 1.0-beta-2-SNAPSHOT.
>
> I think I'm down working on this for now. So Prasad, please have a
> look and let me know if this will work for you. If so, I'll start a
> vote for a release later next week.
>
> On revision from the notes below, the default display is now :20, and
> it will try and detect a free display from :20 to :30 (and then give
> up and barf).
>
> * * *
>
> BTW, the Xvfb config needs to be put into a profile, and explicitly
> activated, otherwise it will hose folks who don't have Xvfb installed
> (and the profile must cause the execution of the xvfb goal before
> start-server). I suppose if we really want I could probably figure
> out how to detect when Xvfb is needed and if its available, but...
> for now lets just make it explicitly activated for folks that need
> the testsuite to run on a headless system.
>
> Cheers,
>
> --jason
>
>
> On Jun 7, 2007, at 8:05 PM, Matt Hogstrom wrote:
>
>> Cool shtuffes ... your a busy beaver.
>>
>> On Jun 7, 2007, at 6:52 PM, Jason Dillon wrote:
>>
>>> So, I hacked up something in the latest 1.0-beta-2-SNAPSHOT of the
>>> selenium-maven-plugin. Its deployed, and some site docs updated.
>>>
>>> http://mojo.codehaus.org/selenium-maven-plugin/examples/
>>> headless-with-xvfb.html
>>>
>>> Still needs more meat, but I wanna make sure it works before I
>>> start investing in more docs.
>>>
>>> But basically, you need to configure an execution *before* start-
>>> server, for xvfb. It will be default background, etc. See the
>>> goal doc for more details:
>>>
>>> http://mojo.codehaus.org/selenium-maven-plugin/xvfb-mojo.html
>>>
>>> Currently it will default to using DISPLAY=":1", not sure if that
>>> is correct or not. Then it will fork an Xvfb process, write out a
>>> target/selenium/display.properties file, which has the DISPLAY= in
>>> it. And I updated the start-server goal to look for that file,
>>> and if it exists set environment variables for each key/pair in
>>> that file.
>>>
>>> I think that should be sufficient to get headless muck working...
>>> though I've not tested any of it ;-)
>>>
>>> I think you will need to make the Xvfb process suid, as in:
>>>
>>> chmod u+s `which Xvfb`
>>>
>>> This puppy needs to be run as root, or at least AFAIK it does.
>>>
>>> Anyways, give it a shot and lemme know how well it works, or if it
>>> works at all. I will be around tomorrow to fix it up as needed.
>>> Maybe my network will be stable then so I can run some tests on
>>> the gbuild.org systems and better see how well it works out.
>>>
>>> Aighty?
>>>
>>> Cheers,
>>>
>>> --jason
>>>
>>
>
http://www.bbc.co.uk/
This e-mail (and any attachments) is confidential and may contain personal views which are not the views of the BBC unless specifically stated.
If you have received it in error, please delete it from your system.
Do not use, copy or disclose the information in any way nor act in reliance on it and notify the sender immediately.
Please note that the BBC monitors e-mails sent or received.
Further communication will signify your consent to this.
Re: Yo yo... Xvfb, selenium blah
Posted by Jason Dillon <ja...@planet57.com>.
And now I think I've got the auto-detection of free display and rough/
good-enough verification of Xvfb working now.
Seems to work... so I've updated the docs a wee bit and deployed a
new 1.0-beta-2-SNAPSHOT.
I think I'm down working on this for now. So Prasad, please have a
look and let me know if this will work for you. If so, I'll start a
vote for a release later next week.
On revision from the notes below, the default display is now :20, and
it will try and detect a free display from :20 to :30 (and then give
up and barf).
* * *
BTW, the Xvfb config needs to be put into a profile, and explicitly
activated, otherwise it will hose folks who don't have Xvfb installed
(and the profile must cause the execution of the xvfb goal before
start-server). I suppose if we really want I could probably figure
out how to detect when Xvfb is needed and if its available, but...
for now lets just make it explicitly activated for folks that need
the testsuite to run on a headless system.
Cheers,
--jason
On Jun 7, 2007, at 8:05 PM, Matt Hogstrom wrote:
> Cool shtuffes ... your a busy beaver.
>
> On Jun 7, 2007, at 6:52 PM, Jason Dillon wrote:
>
>> So, I hacked up something in the latest 1.0-beta-2-SNAPSHOT of the
>> selenium-maven-plugin. Its deployed, and some site docs updated.
>>
>> http://mojo.codehaus.org/selenium-maven-plugin/examples/
>> headless-with-xvfb.html
>>
>> Still needs more meat, but I wanna make sure it works before I
>> start investing in more docs.
>>
>> But basically, you need to configure an execution *before* start-
>> server, for xvfb. It will be default background, etc. See the
>> goal doc for more details:
>>
>> http://mojo.codehaus.org/selenium-maven-plugin/xvfb-mojo.html
>>
>> Currently it will default to using DISPLAY=":1", not sure if that
>> is correct or not. Then it will fork an Xvfb process, write out a
>> target/selenium/display.properties file, which has the DISPLAY= in
>> it. And I updated the start-server goal to look for that file,
>> and if it exists set environment variables for each key/pair in
>> that file.
>>
>> I think that should be sufficient to get headless muck working...
>> though I've not tested any of it ;-)
>>
>> I think you will need to make the Xvfb process suid, as in:
>>
>> chmod u+s `which Xvfb`
>>
>> This puppy needs to be run as root, or at least AFAIK it does.
>>
>> Anyways, give it a shot and lemme know how well it works, or if it
>> works at all. I will be around tomorrow to fix it up as needed.
>> Maybe my network will be stable then so I can run some tests on
>> the gbuild.org systems and better see how well it works out.
>>
>> Aighty?
>>
>> Cheers,
>>
>> --jason
>>
>
Re: Yo yo... Xvfb, selenium blah
Posted by Matt Hogstrom <ma...@hogstrom.org>.
Cool shtuffes ... your a busy beaver.
On Jun 7, 2007, at 6:52 PM, Jason Dillon wrote:
> So, I hacked up something in the latest 1.0-beta-2-SNAPSHOT of the
> selenium-maven-plugin. Its deployed, and some site docs updated.
>
> http://mojo.codehaus.org/selenium-maven-plugin/examples/
> headless-with-xvfb.html
>
> Still needs more meat, but I wanna make sure it works before I
> start investing in more docs.
>
> But basically, you need to configure an execution *before* start-
> server, for xvfb. It will be default background, etc. See the
> goal doc for more details:
>
> http://mojo.codehaus.org/selenium-maven-plugin/xvfb-mojo.html
>
> Currently it will default to using DISPLAY=":1", not sure if that
> is correct or not. Then it will fork an Xvfb process, write out a
> target/selenium/display.properties file, which has the DISPLAY= in
> it. And I updated the start-server goal to look for that file, and
> if it exists set environment variables for each key/pair in that file.
>
> I think that should be sufficient to get headless muck working...
> though I've not tested any of it ;-)
>
> I think you will need to make the Xvfb process suid, as in:
>
> chmod u+s `which Xvfb`
>
> This puppy needs to be run as root, or at least AFAIK it does.
>
> Anyways, give it a shot and lemme know how well it works, or if it
> works at all. I will be around tomorrow to fix it up as needed.
> Maybe my network will be stable then so I can run some tests on the
> gbuild.org systems and better see how well it works out.
>
> Aighty?
>
> Cheers,
>
> --jason
>
Re: Yo yo... Xvfb, selenium blah
Posted by Jason Dillon <ja...@planet57.com>.
FYI, I was able to verify that the basic functionality is working...
yay. I'm working on some X11 display auto detection and verification
muck right now. Should have something working in an hour or so,
maybe less, and then I'll publish new snaps for your testing tomorrow.
--jason
On Jun 7, 2007, at 3:52 PM, Jason Dillon wrote:
> So, I hacked up something in the latest 1.0-beta-2-SNAPSHOT of the
> selenium-maven-plugin. Its deployed, and some site docs updated.
>
> http://mojo.codehaus.org/selenium-maven-plugin/examples/
> headless-with-xvfb.html
>
> Still needs more meat, but I wanna make sure it works before I
> start investing in more docs.
>
> But basically, you need to configure an execution *before* start-
> server, for xvfb. It will be default background, etc. See the
> goal doc for more details:
>
> http://mojo.codehaus.org/selenium-maven-plugin/xvfb-mojo.html
>
> Currently it will default to using DISPLAY=":1", not sure if that
> is correct or not. Then it will fork an Xvfb process, write out a
> target/selenium/display.properties file, which has the DISPLAY= in
> it. And I updated the start-server goal to look for that file, and
> if it exists set environment variables for each key/pair in that file.
>
> I think that should be sufficient to get headless muck working...
> though I've not tested any of it ;-)
>
> I think you will need to make the Xvfb process suid, as in:
>
> chmod u+s `which Xvfb`
>
> This puppy needs to be run as root, or at least AFAIK it does.
>
> Anyways, give it a shot and lemme know how well it works, or if it
> works at all. I will be around tomorrow to fix it up as needed.
> Maybe my network will be stable then so I can run some tests on the
> gbuild.org systems and better see how well it works out.
>
> Aighty?
>
> Cheers,
>
> --jason