A way available to test PHP CGI builds.
in [Windows]
|
From: Venkat Raman Don on 2 Feb 2010 01:37 Hi, Let's don't argue just for the heck of it. Can I know why browser is not a good thing. And what solution you are arriving into. And can you help me share the coding for that solution if we find it better. And BTW my solution doesn't involve comparing screenshot. I think I have given enough reason why not to use GD/COM. Do you still think using COM/GD is a viable option? Or you have another alternative? I already told you we can use Mozilla based browser which can be used across platform. Ideally a solution should be scalable and work on different matrix. Thanks, Don. -----Original Message----- From: Pierre Joye [mailto:pierre.php(a)gmail.com] Sent: Monday, February 01, 2010 1:04 PM To: Venkat Raman Don Cc: php-windows(a)lists.php.net Subject: Re: [PHP-WIN] A way available to test PHP CGI builds. On Mon, Feb 1, 2010 at 9:59 PM, Venkat Raman Don <Don.Raman(a)microsoft.com> wrote: > Hi, > > A good idea about using COM/GD. Let's talk about some pro/cons of it and the solution I gave. > 1. COM never worked with default IIS configuration because of permission issues. This means again you are not simulating a real world scenario. > 2. COM based solution will never have the ability to be ported to UNIX world. > 3. GD can be used for screenshot comparison but any screenshot comparison will have some error of margin. Plus your thing (expected output) will be dependent on screen resolution/display settings etc. If your code has to be portable going this route is not a good idea in my view. > > Regarding dependency on IE, I see no reason why we cannot use a Mozilla based browser and automate it. That way it can work on any platform though it will have its own challenges but not the ones we will face using GD and COM. Sorry, I was not clear. My point was: There is no need of a browser to do http requests and compare the output. For test client side results (visuallly, js, etc.), there are excellent tools out there like selenium, but that's not what we discuss here. Cheers, -- Pierre @pierrejoye | http://blog.thepimp.net | http://www.libgd.org -- PHP Windows Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
From: Pierre Joye on 2 Feb 2010 04:20 hi, Let me start from scratch: To test the output of a http request, no browser is required. PHP alone suffices, or wget/curl cmd line (or other similar tools). PHP supports HTTP (incl. headers), the diff can be generated easily, see run-tests.php. The only reason why a browser would be required is to compare the visual output or how js behaves. That's where GD could help, to compare screenshots of the rendered page. But that's not what we neeed here. That's what I was trying to explain. Cheers, -- Pierre On Tue, Feb 2, 2010 at 7:37 AM, Venkat Raman Don <Don.Raman(a)microsoft.com> wrote: > Hi, > > Let's don't argue just for the heck of it. > > Can I know why browser is not a good thing. And what solution you are arriving into. And can you help me share the coding for that solution if we find it better. And BTW my solution doesn't involve comparing screenshot. > > I think I have given enough reason why not to use GD/COM. Do you still think using COM/GD is a viable option? Or you have another alternative? I already told you we can use Mozilla based browser which can be used across platform. Ideally a solution should be scalable and work on different matrix. > > Thanks, > Don. > > -----Original Message----- > From: Pierre Joye [mailto:pierre.php(a)gmail.com] > Sent: Monday, February 01, 2010 1:04 PM > To: Venkat Raman Don > Cc: php-windows(a)lists.php.net > Subject: Re: [PHP-WIN] A way available to test PHP CGI builds. > > On Mon, Feb 1, 2010 at 9:59 PM, Venkat Raman Don > <Don.Raman(a)microsoft.com> wrote: >> Hi, >> >> A good idea about using COM/GD. Let's talk about some pro/cons of it and the solution I gave. >> 1. COM never worked with default IIS configuration because of permission issues. This means again you are not simulating a real world scenario. >> 2. COM based solution will never have the ability to be ported to UNIX world. >> 3. GD can be used for screenshot comparison but any screenshot comparison will have some error of margin. Plus your thing (expected output) will be dependent on screen resolution/display settings etc. If your code has to be portable going this route is not a good idea in my view. >> >> Regarding dependency on IE, I see no reason why we cannot use a Mozilla based browser and automate it. That way it can work on any platform though it will have its own challenges but not the ones we will face using GD and COM. > > Sorry, I was not clear. My point was: There is no need of a browser to > do http requests and compare the output. > > For test client side results (visuallly, js, etc.), there are > excellent tools out there like selenium, but that's not what we > discuss here. > > Cheers, > -- > Pierre > > @pierrejoye | http://blog.thepimp.net | http://www.libgd.org > > -- > PHP Windows Mailing List (http://www.php.net/) > To unsubscribe, visit: http://www.php.net/unsub.php > > > -- Pierre @pierrejoye | http://blog.thepimp.net | http://www.libgd.org
From: Venkat Raman Don on 2 Feb 2010 11:06 Hi, Pierre, I understand that there are alternate ways. I was just asking why the browser way is bad or not acceptable? Or you are okay with browser way and suggesting an alternative. If you have concerns regarding using browser to test, I would like to know it as it also impacts WINCACHE quality overall. Thanks, Don. -----Original Message----- From: Pierre Joye [mailto:pierre.php(a)gmail.com] Sent: Tuesday, February 02, 2010 1:21 AM To: Venkat Raman Don Cc: php-windows(a)lists.php.net Subject: Re: [PHP-WIN] A way available to test PHP CGI builds. hi, Let me start from scratch: To test the output of a http request, no browser is required. PHP alone suffices, or wget/curl cmd line (or other similar tools). PHP supports HTTP (incl. headers), the diff can be generated easily, see run-tests.php. The only reason why a browser would be required is to compare the visual output or how js behaves. That's where GD could help, to compare screenshots of the rendered page. But that's not what we neeed here. That's what I was trying to explain. Cheers, -- Pierre On Tue, Feb 2, 2010 at 7:37 AM, Venkat Raman Don <Don.Raman(a)microsoft.com> wrote: > Hi, > > Let's don't argue just for the heck of it. > > Can I know why browser is not a good thing. And what solution you are arriving into. And can you help me share the coding for that solution if we find it better. And BTW my solution doesn't involve comparing screenshot. > > I think I have given enough reason why not to use GD/COM. Do you still think using COM/GD is a viable option? Or you have another alternative? I already told you we can use Mozilla based browser which can be used across platform. Ideally a solution should be scalable and work on different matrix. > > Thanks, > Don. > > -----Original Message----- > From: Pierre Joye [mailto:pierre.php(a)gmail.com] > Sent: Monday, February 01, 2010 1:04 PM > To: Venkat Raman Don > Cc: php-windows(a)lists.php.net > Subject: Re: [PHP-WIN] A way available to test PHP CGI builds. > > On Mon, Feb 1, 2010 at 9:59 PM, Venkat Raman Don > <Don.Raman(a)microsoft.com> wrote: >> Hi, >> >> A good idea about using COM/GD. Let's talk about some pro/cons of it and the solution I gave. >> 1. COM never worked with default IIS configuration because of permission issues. This means again you are not simulating a real world scenario. >> 2. COM based solution will never have the ability to be ported to UNIX world. >> 3. GD can be used for screenshot comparison but any screenshot comparison will have some error of margin. Plus your thing (expected output) will be dependent on screen resolution/display settings etc. If your code has to be portable going this route is not a good idea in my view. >> >> Regarding dependency on IE, I see no reason why we cannot use a Mozilla based browser and automate it. That way it can work on any platform though it will have its own challenges but not the ones we will face using GD and COM. > > Sorry, I was not clear. My point was: There is no need of a browser to > do http requests and compare the output. > > For test client side results (visuallly, js, etc.), there are > excellent tools out there like selenium, but that's not what we > discuss here. > > Cheers, > -- > Pierre > > @pierrejoye | http://blog.thepimp.net | http://www.libgd.org > > -- > PHP Windows Mailing List (http://www.php.net/) > To unsubscribe, visit: http://www.php.net/unsub.php > > > -- Pierre @pierrejoye | http://blog.thepimp.net | http://www.libgd.org
From: Ferenc Kovacs on 2 Feb 2010 11:29 On Tue, Feb 2, 2010 at 5:06 PM, Venkat Raman Don <Don.Raman(a)microsoft.com> wrote: > Hi, > > Pierre, I understand that there are alternate ways. I was just asking why the browser way is bad or not acceptable? Or you are okay with browser way and suggesting an alternative. If you have concerns regarding using browser to test, I would like to know it as it also impacts WINCACHE quality overall. > Why are you using something what isn't needed? If you dont need the rendered layout, executed client side js, then what are the advantages using a browser over a php module? Tyrael > Thanks, > Don. > > -----Original Message----- > From: Pierre Joye [mailto:pierre.php(a)gmail.com] > Sent: Tuesday, February 02, 2010 1:21 AM > To: Venkat Raman Don > Cc: php-windows(a)lists.php.net > Subject: Re: [PHP-WIN] A way available to test PHP CGI builds. > > hi, > > Let me start from scratch: > > To test the output of a http request, no browser is required. PHP > alone suffices, or wget/curl cmd line (or other similar tools). PHP > supports HTTP (incl. headers), the diff can be generated easily, see > run-tests.php. > > The only reason why a browser would be required is to compare the > visual output or how js behaves. That's where GD could help, to > compare screenshots of the rendered page. But that's not what we neeed > here. > > That's what I was trying to explain. > > Cheers, > -- > Pierre > > On Tue, Feb 2, 2010 at 7:37 AM, Venkat Raman Don > <Don.Raman(a)microsoft.com> wrote: >> Hi, >> >> Let's don't argue just for the heck of it. >> >> Can I know why browser is not a good thing. And what solution you are arriving into. And can you help me share the coding for that solution if we find it better. And BTW my solution doesn't involve comparing screenshot. >> >> I think I have given enough reason why not to use GD/COM. Do you still think using COM/GD is a viable option? Or you have another alternative? I already told you we can use Mozilla based browser which can be used across platform. Ideally a solution should be scalable and work on different matrix. >> >> Thanks, >> Don. >> >> -----Original Message----- >> From: Pierre Joye [mailto:pierre.php(a)gmail.com] >> Sent: Monday, February 01, 2010 1:04 PM >> To: Venkat Raman Don >> Cc: php-windows(a)lists.php.net >> Subject: Re: [PHP-WIN] A way available to test PHP CGI builds. >> >> On Mon, Feb 1, 2010 at 9:59 PM, Venkat Raman Don >> <Don.Raman(a)microsoft.com> wrote: >>> Hi, >>> >>> A good idea about using COM/GD. Let's talk about some pro/cons of it and the solution I gave. >>> 1. COM never worked with default IIS configuration because of permission issues. This means again you are not simulating a real world scenario. >>> 2. COM based solution will never have the ability to be ported to UNIX world. >>> 3. GD can be used for screenshot comparison but any screenshot comparison will have some error of margin. Plus your thing (expected output) will be dependent on screen resolution/display settings etc. If your code has to be portable going this route is not a good idea in my view. >>> >>> Regarding dependency on IE, I see no reason why we cannot use a Mozilla based browser and automate it. That way it can work on any platform though it will have its own challenges but not the ones we will face using GD and COM. >> >> Sorry, I was not clear. My point was: There is no need of a browser to >> do http requests and compare the output. >> >> For test client side results (visuallly, js, etc.), there are >> excellent tools out there like selenium, but that's not what we >> discuss here. >> >> Cheers, >> -- >> Pierre >> >> @pierrejoye | http://blog.thepimp.net | http://www.libgd.org >> >> -- >> PHP Windows Mailing List (http://www.php.net/) >> To unsubscribe, visit: http://www.php.net/unsub.php >> >> >> > > > > -- > Pierre > > @pierrejoye | http://blog.thepimp.net | http://www.libgd.org > > > -- > PHP Windows Mailing List (http://www.php.net/) > To unsubscribe, visit: http://www.php.net/unsub.php > >
From: Venkat Raman Don on 2 Feb 2010 12:13 Hi, The reason I am using browser is because I wanted to simulate the way server works. Thanks for all the input. I will look into other modules and see if a similar thing can be achieved using PHP itself. Don. -----Original Message----- From: Ferenc Kovacs [mailto:tyra3l(a)gmail.com] Sent: Tuesday, February 02, 2010 8:29 AM To: Venkat Raman Don Cc: Pierre Joye; php-windows(a)lists.php.net Subject: Re: [PHP-WIN] A way available to test PHP CGI builds. On Tue, Feb 2, 2010 at 5:06 PM, Venkat Raman Don <Don.Raman(a)microsoft.com> wrote: > Hi, > > Pierre, I understand that there are alternate ways. I was just asking why the browser way is bad or not acceptable? Or you are okay with browser way and suggesting an alternative. If you have concerns regarding using browser to test, I would like to know it as it also impacts WINCACHE quality overall. > Why are you using something what isn't needed? If you dont need the rendered layout, executed client side js, then what are the advantages using a browser over a php module? Tyrael > Thanks, > Don. > > -----Original Message----- > From: Pierre Joye [mailto:pierre.php(a)gmail.com] > Sent: Tuesday, February 02, 2010 1:21 AM > To: Venkat Raman Don > Cc: php-windows(a)lists.php.net > Subject: Re: [PHP-WIN] A way available to test PHP CGI builds. > > hi, > > Let me start from scratch: > > To test the output of a http request, no browser is required. PHP > alone suffices, or wget/curl cmd line (or other similar tools). PHP > supports HTTP (incl. headers), the diff can be generated easily, see > run-tests.php. > > The only reason why a browser would be required is to compare the > visual output or how js behaves. That's where GD could help, to > compare screenshots of the rendered page. But that's not what we neeed > here. > > That's what I was trying to explain. > > Cheers, > -- > Pierre > > On Tue, Feb 2, 2010 at 7:37 AM, Venkat Raman Don > <Don.Raman(a)microsoft.com> wrote: >> Hi, >> >> Let's don't argue just for the heck of it. >> >> Can I know why browser is not a good thing. And what solution you are arriving into. And can you help me share the coding for that solution if we find it better. And BTW my solution doesn't involve comparing screenshot. >> >> I think I have given enough reason why not to use GD/COM. Do you still think using COM/GD is a viable option? Or you have another alternative? I already told you we can use Mozilla based browser which can be used across platform. Ideally a solution should be scalable and work on different matrix. >> >> Thanks, >> Don. >> >> -----Original Message----- >> From: Pierre Joye [mailto:pierre.php(a)gmail.com] >> Sent: Monday, February 01, 2010 1:04 PM >> To: Venkat Raman Don >> Cc: php-windows(a)lists.php.net >> Subject: Re: [PHP-WIN] A way available to test PHP CGI builds. >> >> On Mon, Feb 1, 2010 at 9:59 PM, Venkat Raman Don >> <Don.Raman(a)microsoft.com> wrote: >>> Hi, >>> >>> A good idea about using COM/GD. Let's talk about some pro/cons of it and the solution I gave. >>> 1. COM never worked with default IIS configuration because of permission issues. This means again you are not simulating a real world scenario. >>> 2. COM based solution will never have the ability to be ported to UNIX world. >>> 3. GD can be used for screenshot comparison but any screenshot comparison will have some error of margin. Plus your thing (expected output) will be dependent on screen resolution/display settings etc. If your code has to be portable going this route is not a good idea in my view. >>> >>> Regarding dependency on IE, I see no reason why we cannot use a Mozilla based browser and automate it. That way it can work on any platform though it will have its own challenges but not the ones we will face using GD and COM. >> >> Sorry, I was not clear. My point was: There is no need of a browser >> to do http requests and compare the output. >> >> For test client side results (visuallly, js, etc.), there are >> excellent tools out there like selenium, but that's not what we >> discuss here. >> >> Cheers, >> -- >> Pierre >> >> @pierrejoye | http://blog.thepimp.net | http://www.libgd.org >> >> -- >> PHP Windows Mailing List (http://www.php.net/) To unsubscribe, visit: >> http://www.php.net/unsub.php >> >> >> > > > > -- > Pierre > > @pierrejoye | http://blog.thepimp.net | http://www.libgd.org > > > -- > PHP Windows Mailing List (http://www.php.net/) To unsubscribe, visit: > http://www.php.net/unsub.php > > -- PHP Windows Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
First
|
Prev
|
Next
|
Last
Pages: 1 2 3 Prev: [PHP-WIN] A way available to test PHP CGI builds. Next: php_dba.dll with DB4 |