QTP Vs Selenium. The research and the result.

[BS19]

This is very interesting for me, it took out some myths regarding Selenium and its popularity.

http://www.starbase.co.uk/attachments/article/171/White%20Paper.%20Selenium%20vs.%20HP%20QuickTest%20Professional.pdf

[BS19]

I am still (always) a big fan of Selenium and its power and its beauty.
But this research may be correct!

[Jim Evans]

<sarcasm>Oh, I'm sure the authors of this "study" have no bias toward one technology or another. I'm certainly going to start suggesting people use proprietary tools from now on.</sarcasm>

The following is taken from the "About VIP" section of the "study" (VIP is the author of the "study"):

VIP is an HP Software Elite Partner and has earned a reputation for its expertise related to the HP suite of Business Technology Optimization (BTO) solutions including HP Quality Center, HP QuickTest Professional, HP Performance Center, HP LoadRunner and HP Business Availability Center.

Looking at the "study," I have to ask, "Qui bono?" Who profits from this? I keep putting "study" in quotes, because the paper linked to is nothing more than marketing propaganda from a company with a vested interest in locking people into a proprietary solution, HP's QuickTest Pro in this case. I appreciate good marketing literature as much as the next guy, but lets call it what it is, and not attempt to pass it off as some kind of objective look at the pros and cons of OSS vs. proprietary solutions.

[Marcus Merrell]

This is disgusting.

I couldn't abandoned QTP fast enough when I discovered Selenium, and my ROI was *significantly* different from what they're reporting. Between 2004 and 2007, we had 7 team members working constantly. We had around a thousand scripts, running about a 78% success rate--2-3 FTEs were required just to fix broken and unreliable tests. We took all their "official" training, used every facet of the tool we could, filed bugs judiciously with Mercury, and even spoke at their conferences.

When a colleague of mine coded up a Page Object in Java, using the WebDriver concepts demonstrated by Simon at GTAC, we abandoned the 3-year QTP effort WITHIN 24 HOURS. After one year we had nearly a thousand scripts running at a 96% success rate, and that was after we had lost 4 team members (we were down to 3--we didn't require extra people just to maintain scripts). None of us were Java developers before WebDriver came along.

I understand that my "sample size" is too small to consider this a legitimate refutation of that "study", but it's got me thinking--maybe we're not doing enough to help people Do It Right. This should be a topic of hallway discussion at SeConf 2013 in Boston (only costs $300 to attend, unlike HP Universe's $1795).

I'd like to produce a similar white paper refuting some of these points, but that would mean I'd have to use QTP again for the comparison. I break out in hives just thinking about it...

MM

--
You received this message because you are subscribed to the Google Groups "webdriver" group.
To unsubscribe from this group and stop receiving emails from it, send an email to webdriver+...@googlegroups.com.
To post to this group, send email to webd...@googlegroups.com.
Visit this group at http://groups.google.com/group/webdriver?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.

[Mark Collin]

Having a quick read it’s also wrong:

 

“To complete the Selenium PoC , we found it necessary to install and configure seven different components: Sun Java JDK, Mozilla Firefox, Firebug, Selenium IDE, Selenium Remote Control, the Selenium library, and the Eclipse IDE”

 

Any Selenium professional will not install Selenium IDE, and Selenium RC is deprecated.  So really it’s the Selenium Library and Eclipse (You don’t need to install an IDE, you can write Java in Notepad, but I’ll give them that as everybody uses an IDE these days although I would suggest IntelliJ IDEA.  Installing FireFox is a straw man argument, if you are automating a web page you are going to have to install browser anyway, QTP is no different in that respect.  Firebug is helpful, but FireFox has its own dev tools these days, FireBug is by no means a requirement.

 

So really you need to install Java (download and one click install), an IDE (IntelliJ IDEA is a download and one click install) and Selenium, wow that sounds hard.

 

Going on to the next bit it talks about recording scripts in Selenium IDE and then totally refactoring them, hell just write then direct, Selenium professionals don’t use Selenium IDE.  Start up the Selenium RC server??  What the one that’s been deprecated for over a year, and something you don’t need to start any more??

 

I could go on but I just gave up reading this, it’s obviously written by somebody who had no clue about Selenium.

[Pankaj Nakhat]

This is the most stupid comparison I have ever seen. I am sure author does not understand selenium well, by reading this its evident. 

Also, most important space where QTP fails, its its integration with other tools and allowing to use different programming interfaces.  Has any one ever tried running QTP tests with CI tools, I can rest assure QTP will die in next few years for sure, at least in the web testing space. 

Regards,
Pankaj Nakhat
www.pankajnakhat.com
t : @pnakhat

[Jason Huggins]

I like page 12. They're saying that if you know Selenium, you can get
a job that pays $20,000 more. Um, awesome!

(hugs)

[BS19]

@hugs
You made me laugh, thanks that gave me a relief between these heated arguments !

If VIP has some kind of relationship with HP, then this is completely unfair and some sort of cheating.

I worked with QTP and Selenium as well.
Though, personally I support and enjoy Java+Selenium ...here is my opinion (objective):

The only huge advantage for QTP is it is very very easy to learn and write scripts for different applications. All plugins come with QTP.
The main disadvantage for QTP is its cost - almost $10000. Where is Selenium is completely free and powerful.

1. In India, there is no big difference between Selenium and QTP testers in terms of salary.
Both get the same salary. Maybe due to the scarcity, Selenium tester can get a little bit more.That's all.

2. Learning and writing Java is difficult when compared with VBS (for a tester with average IQ).
3. Debugging Java libraries will take more time,  when compared with QTP/its VBS.
4. All the 3rd party components and other things should be installed/managed by the tester (for Selenium) - Where as this is done by QTP.


Finally, based on Indeed.com job trend, Selenium jobs are going up and up.
And QTP jobs are going down.

Personally also I am seeing Selenium jobs are increasing.

For

[BS19]

As I thought their comparison is unbiased and fair, I posted this.

But, If this is some sort of manipulation, this thread can be removed or modified as this gives bad impression.
And, if they are manipulating the reality, then they are trying to attack the open source movement / free software movement for some corporation's profits which is not acceptable.

[Mark Collin]

You missed out the main advantage for QTP, it’s not designed to just automate browsers.

 

Selenium is a tool specifically focussed at browser automation (and since everything is becoming a JavaScript/HTML5 app these days it is in a very good position).  A true comparison between QTP and Selenium is like comparing Apples and Oranges.

 

On to your points:

 

1.       In the UK the average rate for jobs mentioning Selenium seems higher than jobs mentioning QTP.  There is a caveat however; there are a lot of development roles that want a developer with Selenium experience mixed in with the tester roles.  On the other side there I could find no developer roles that mention QTP experience.  Based on this I would suggest that the rate for testers is pretty static, the numbers are skewed because people want developers with Selenium experience (and rightly or wrongly developers in the UK command a higher premium than testers).

2.       I disagree.  Learning any language is hard, if you have experience in one already you are going to find it easier.  Testers are no less intelligent than anybody else on average; nothing beats a good book/tutorial/course that teaches the language fundamentals properly.

3.       This comes down to experience.  It would take me a lot longer to debug a problem with QTP than it would a problem with Selenium.  The fact I can see the entire Selenium codebase makes things much easier should I want to go down that route.  Do you really mean that “Paid support where somebody will debug your problems for you” is easier to get with QTP?

4.       Sounds like a bit of a straw man argument to me to be honest.  What plugins are you talking about?  Plugins for IDE are pretty simple to install if that’s what you mean.  Selenium doesn’t really have plugins in the QTP sense.

 

I can give you one good reason why QTP jobs may be declining while Selenium jobs are increasing that hasn’t been mentioned, Linux and OSX.  QTP requires windows, over the last 5 years I haven’t worked with windows as my main development platform at all, it has always been Linux or OSX.  These days when I go into a new office all the developers are invariably working on Mac’s.  IMHO windows days are numbered in development teams, all developers want a *nix based OS unless they are building something specifically for Windows and in the web arena it’s pretty rare to see things being built specifically for windows.

 

Maybe HP will pull their finger out and start supporting Linux or OSX, I doubt it though.

 

From: webd...@googlegroups.com [mailto:webd...@googlegroups.com] On Behalf Of BS19
Sent: 03 March 2013 06:11
To: webd...@googlegroups.com
Subject: Re: [webdriver] Re: QTP Vs Selenium. The research and the result.

 

@hugs

--

[BS19]

Thanks Mark, for the details!
I am working on Ubuntu (Linux flavor), but would like to know why developers prefer Unix or Linux or Mac. Any specific advantage for them?

[Andreas Tolf Tolfsen]

On Sat, Mar 02, 2013 at 06:15:15AM -0800, BS19 wrote:
> This is very interesting for me, it took out some myths regarding
> Selenium and its popularity.
>

> http://www.starbase.co.uk/attachments/article/171/White%20Paper%20Selenium%20vs.%20HP%20QuickTest%20Professional.pdf

Oh I had a good laugh reading this. This is a prime example of a
company spreading FUD about freely available automation tools such as
Selenium. However, the fact that Selenium is seen as a real contender
by VIP is worth noting.

I'm surprised they get away with the quasi-scientific tone in this
document. They claim to be performing a comparison, but there isn't a
single link to data they claim to compare. I feel like holding up a
[citation needed] sign.

To their defense, they are clear about where their sympathies lie, but
when you present a blatantly useless illustrations such as the one on
page 6, it just makes you look pathetic.

I've never used QTP, but as someone pointed out they appear to be
comparing apples and oranges, without knowing anything about fruit. If
you try to eat an orange the same way you eat an apple, you're going to
end up very miscontent.

[Mark Collin]

A couple of things off the top of my head:

• The terminal is actually useful.
• It's very easy to setup and configure things (apt-get install maven3 tomcat7 is much easier than the faff you have to go through on windows, swap apt-get to brew if you are a Mac user).
• Stability.
  

If I sat down and thought about it I could probably come up with a list as long as your arm.� This being said, you should already know.� Why did you pick Ubuntu instead of Windows?� Don't you find life easier on Ubuntu?

On 03/03/2013 16:13, BS19 wrote:

Thanks Mark, for the details!
I am working on Ubuntu (Linux flavor), but would like to know why developers prefer Unix or Linux or Mac. Any specific advantage for them?

�

On Sunday, March 3, 2013 1:58:34 AM UTC-8, Mark Collin wrote:

You missed out the main advantage for QTP, it�s not designed to just automate browsers.

�

Selenium is a tool specifically focussed at browser automation (and since everything is becoming a JavaScript/HTML5 app these days it is in a very good position).� A true comparison between QTP and Selenium is like comparing Apples and Oranges.

�

On to your points:

�

1.������ In the UK the average rate for jobs mentioning Selenium seems higher than jobs mentioning QTP.� There is a caveat however; there are a lot of development roles that want a developer with Selenium experience mixed in with the tester roles.� On the other side there I could find no developer roles that mention QTP experience.� Based on this I would suggest that the rate for testers is pretty static, the numbers are skewed because people want developers with Selenium experience (and rightly or wrongly developers in the UK command a higher premium than testers).

2.������ I disagree.� Learning any language is hard, if you have experience in one already you are going to find it easier.� Testers are no less intelligent than anybody else on average; nothing beats a good book/tutorial/course that teaches the language fundamentals properly.

3.������ This comes down to experience.� It would take me a lot longer to debug a problem with QTP than it would a problem with Selenium.� The fact I can see the entire Selenium codebase makes things much easier should I want to go down that route.� Do you really mean that �Paid support where somebody will debug your problems for you� is easier to get with QTP?

4.������ Sounds like a bit of a straw man argument to me to be honest.� What plugins are you talking about?� Plugins for IDE are pretty simple to install if that�s what you mean.� Selenium doesn�t really have plugins in the QTP sense.

�

I can give you one good reason why QTP jobs may be declining while Selenium jobs are increasing that hasn�t been mentioned, Linux and OSX.� QTP requires windows, over the last 5 years I haven�t worked with windows as my main development platform at all, it has always been Linux or OSX.� These days when I go into a new office all the developers are invariably working on Mac�s.� IMHO windows days are numbered in development teams, all developers want a *nix based OS unless they are building something specifically for Windows and in the web arena it�s pretty rare to see things being built specifically for windows.

�

Maybe HP will pull their finger out and start supporting Linux or OSX, I doubt it though.

�

From: webd...@googlegroups.com [mailto:webd...@googlegroups.com] On Behalf Of BS19
Sent: 03 March 2013 06:11
To: webd...@googlegroups.com
Subject: Re: [webdriver] Re: QTP Vs Selenium. The research and the result.

�

@hugs
You made me laugh, thanks that gave me a relief between these heated arguments !

If VIP has some kind of relationship with HP, then this is completely unfair and some sort of cheating.

I worked with QTP and Selenium as well.
Though, personally I support and enjoy Java+Selenium ...here is my opinion (objective):

The only huge advantage for QTP is it is very very easy to learn and write scripts for different applications. All plugins come with QTP.
The main disadvantage for QTP is its cost - almost $10000. Where is Selenium is completely free and powerful.

1. In India, there is no big difference between Selenium and QTP testers in terms of salary.
Both get the same salary. Maybe due to the scarcity, Selenium tester can get a little bit more.That's all.

2. Learning and writing Java is difficult when compared with VBS (for a tester with average IQ).

3. Debugging Java libraries will take more time,� when compared with QTP/its VBS.

4. All the 3rd party components and other things should be installed/managed by the tester (for Selenium) - Where as this is done by QTP.


Finally, based on Indeed.com job trend, Selenium jobs are going up and up.
And QTP jobs are going down.

Personally also I am seeing Selenium jobs are increasing.

For

--
You received this message because you are subscribed to the Google Groups "webdriver" group.
To unsubscribe from this group and stop receiving emails from it, send an email to webdriver+...@googlegroups.com.
To post to this group, send email to webd...@googlegroups.com.
Visit this group at http://groups.google.com/group/webdriver?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.

�
�

--
You received this message because you are subscribed to the Google Groups "webdriver" group.
To unsubscribe from this group and stop receiving emails from it, send an email to webdriver+...@googlegroups.com.
To post to this group, send email to webd...@googlegroups.com.
Visit this group at http://groups.google.com/group/webdriver?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.

�
�

[Marcus Merrell]

This is interesting--the blog post from which Starbase originally posted this "white paper" has vanished. I posted a lengthy comment there (much less inflammatory, much more fact-based than my "disgusting" email from Saturday morning), which later said "awaiting moderation". This morning I looked, couldn't find the original post... then looked in my browser history, and all links to the blog post were gone. I even checked the author's posting history and searched for "Hidden Costs of Open Source".

I posted another comment under a post called "Regular Expressions", and that one got moderated and made it through... it's still there:

http://blog.starbase.co.uk/regular-expressions/

...as is the Google cache:

http://webcache.googleusercontent.com/search?q=cache:kwFhs5CkjNEJ:blog.starbase.co.uk/the-hidden-costs-of-open-source/+&cd=1&hl=en&ct=clnk&gl=us

I don't care to speculate about why it's gone, but I hope they're taking my advice. There is definitely a case to be made for using QTP in certain circumstances, and I'd even be glad help them craft it... but they need to be truthful about the case they're making, and not just put out a bunch of hogwash and fear-mongering, hoping nobody will call them on it.

In any event, the (silly little) shark fin is below the water...

MM

On Mon, Mar 4, 2013 at 3:55 AM, Mark Collin <mark....@lazeryattack.com> wrote:

A couple of things off the top of my head:

• The terminal is actually useful.
• It's very easy to setup and configure things (apt-get install maven3 tomcat7 is much easier than the faff you have to go through on windows, swap apt-get to brew if you are a Mac user).
• Stability.
  

If I sat down and thought about it I could probably come up with a list as long as your arm.  This being said, you should already know.  Why did you pick Ubuntu instead of Windows?  Don't you find life easier on Ubuntu?

On 03/03/2013 16:13, BS19 wrote:

Thanks Mark, for the details!
I am working on Ubuntu (Linux flavor), but would like to know why developers prefer Unix or Linux or Mac. Any specific advantage for them?
 

On Sunday, March 3, 2013 1:58:34 AM UTC-8, Mark Collin wrote:

You missed out the main advantage for QTP, it’s not designed to just automate browsers.

 

Selenium is a tool specifically focussed at browser automation (and since everything is becoming a JavaScript/HTML5 app these days it is in a very good position).  A true comparison between QTP and Selenium is like comparing Apples and Oranges.

 

On to your points:

 

1.       In the UK the average rate for jobs mentioning Selenium seems higher than jobs mentioning QTP.  There is a caveat however; there are a lot of development roles that want a developer with Selenium experience mixed in with the tester roles.  On the other side there I could find no developer roles that mention QTP experience.  Based on this I would suggest that the rate for testers is pretty static, the numbers are skewed because people want developers with Selenium experience (and rightly or wrongly developers in the UK command a higher premium than testers).

2.       I disagree.  Learning any language is hard, if you have experience in one already you are going to find it easier.  Testers are no less intelligent than anybody else on average; nothing beats a good book/tutorial/course that teaches the language fundamentals properly.

3.       This comes down to experience.  It would take me a lot longer to debug a problem with QTP than it would a problem with Selenium.  The fact I can see the entire Selenium codebase makes things much easier should I want to go down that route.  Do you really mean that “Paid support where somebody will debug your problems for you” is easier to get with QTP?

4.       Sounds like a bit of a straw man argument to me to be honest.  What plugins are you talking about?  Plugins for IDE are pretty simple to install if that’s what you mean.  Selenium doesn’t really have plugins in the QTP sense.

 

I can give you one good reason why QTP jobs may be declining while Selenium jobs are increasing that hasn’t been mentioned, Linux and OSX.  QTP requires windows, over the last 5 years I haven’t worked with windows as my main development platform at all, it has always been Linux or OSX.  These days when I go into a new office all the developers are invariably working on Mac’s.  IMHO windows days are numbered in development teams, all developers want a *nix based OS unless they are building something specifically for Windows and in the web arena it’s pretty rare to see things being built specifically for windows.

 

Maybe HP will pull their finger out and start supporting Linux or OSX, I doubt it though.

 

From: webd...@googlegroups.com [mailto:webd...@googlegroups.com] On Behalf Of BS19

Sent: 03 March 2013 06:11
To: webd...@googlegroups.com
Subject: Re: [webdriver] Re: QTP Vs Selenium. The research and the result.

 

@hugs

You made me laugh, thanks that gave me a relief between these heated arguments !

If VIP has some kind of relationship with HP, then this is completely unfair and some sort of cheating.

I worked with QTP and Selenium as well.
Though, personally I support and enjoy Java+Selenium ...here is my opinion (objective):

The only huge advantage for QTP is it is very very easy to learn and write scripts for different applications. All plugins come with QTP.
The main disadvantage for QTP is its cost - almost $10000. Where is Selenium is completely free and powerful.

1. In India, there is no big difference between Selenium and QTP testers in terms of salary.
Both get the same salary. Maybe due to the scarcity, Selenium tester can get a little bit more.That's all.

2. Learning and writing Java is difficult when compared with VBS (for a tester with average IQ).

3. Debugging Java libraries will take more time,  when compared with QTP/its VBS.

4. All the 3rd party components and other things should be installed/managed by the tester (for Selenium) - Where as this is done by QTP.


Finally, based on Indeed.com job trend, Selenium jobs are going up and up.
And QTP jobs are going down.

Personally also I am seeing Selenium jobs are increasing.

For

--
You received this message because you are subscribed to the Google Groups "webdriver" group.
To unsubscribe from this group and stop receiving emails from it, send an email to webdriver+...@googlegroups.com.
To post to this group, send email to webd...@googlegroups.com.
Visit this group at http://groups.google.com/group/webdriver?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

--
You received this message because you are subscribed to the Google Groups "webdriver" group.
To unsubscribe from this group and stop receiving emails from it, send an email to webdriver+...@googlegroups.com.
To post to this group, send email to webd...@googlegroups.com.
Visit this group at http://groups.google.com/group/webdriver?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[David]

Thought I'd add some comments to this thread:

* for why people might be using Macs, it's a variant of Unix so has advantage Mark mentioned, or closer to it than Windows. And in today's popularity of Apple, you can't ignore the hyper factor plus the rising Safari (and iOS/mobile Safari) browser user base for your website/web app. Years ago, you wouldn't need to care about them.

* if Java is a problem for switching from QTP to Selenium, there is an bridge path option - http://htejera.users.sourceforge.net/vbswebdriver/index.html - that fits the best of both worlds, VBScript w/ Selenium.

* Where Selenium is lacking outside of web automation, one can base work off projects like Appium or just Selenium's JSONWireProtocol to create mobile/desktop UI automation tool/framework that uses the same Selenium APIs (or most of it anyways) and same supported languages. 

[Simon Stewart]

Finally had a chance to read this. My major takeaways:

* Our docs need to be clarified.

* The role of IDE as a bootstrapping tool isn't made clear.

* It appears as if being a QA person with Selenium skills is worth $20k to your salary.

I'd much rather we looked through the FUD to the substantive claims they're making.

Simon

[BS19]

"* It appears as if being a QA person with Selenium skills is worth $20k to your salary.""""""" 

:)

I am glad I am a Selenium developer, but not with a good salary (in India). But I like java + Selenium.

The research above reminded of one or two movies - Resident Evil and Predators.
In those movies, one corporation (just for the sake of their profits) almost creates a world in which they rule and control everything and does its business to increase their profits.

My observation is  - When I reply to some topic on the forum, all the previous conversion also being kept intact.
This may increase the page size thus requires more server space to host this site.
Maybe the all previous conversation will work as a reference, but no need to be in the same reply thread.

[BS19]

I am sorry it is not Predators it is

Aliens.

[Bill Ross]

If the context isn't in the thread, it is harder to help.

Bill

-------- Original message --------
From: BS19 <vemuriba...@gmail.com>
Date:
To: webd...@googlegroups.com
Subject: Re: [webdriver] Re: QTP Vs Selenium. The research and the result.

[anita valentinova]

I just would like to give another reference (fresh one), an unbiased comparison between Telerik Test Studio and Selenium (in case someone is looking for a comparison of the two tools).

[darrell]

Unbiased? Did you read Jim Evans' first post on this thread? This is not an unbiased comparison of Telerik Test Studio (TTS) and Selenium. This is marketing material. Note the paragraph about the author, "Jim Homes is the Developer Advocate for Telerik Test Studio".

Jim Homes is essentially comparing TTS to Selenium IDE and WebDriver from the perspective of a record and playback tool. What the article points out is that someone with limited knowledge of software development and test automation tools can use TTS out of the box much faster than WebDriver. What the article does not point out is that TTS is limited by what the tool can do. WebDriver is part of a solution and is limited by the person using it.

Essentially, if I want to 'go from zero to 60' then TTS will do it up to twice as fast as WebDriver. However what Jim Homes fails to point out is that if I want to 'go from zero to 100' then TTS will take ten times longer than WebDriver with the appropriate libraries. Different people will appreciate the tools for different reasons.

Case in point, I recently worked with two testers. One had years of experience as a manual tester (Tester1). The other had many years as a software developer (Tester2). Tester1 could craft a test plan which found 80% of the most critical defects within a day or two. It took them a week to automate the tests with HP QC. Tester2 could craft a test plan which found 40% of the most critical defects within two days. They automated the tests with WebDriver as them created the tests. We tried to get Tester1 using WebDriver. They couldn't not get anything working after a week and Tester2 had to take over their automation. I would imagine that Tester2 might have been able to automate everything with HP QC but it would have taken him a longer time, e.g. 40% of the defects in 7 business days. So total time spent was:

• Tester1: HP QC, 80% of automation, 7 business days.
• Tester1: WebDriver, 0% of automation, 5 business days.
• Tester2: HP QC, 40% of automation, 7 business days.
• Tester2: WebDriver, 40% of automation, 2 business days.
  

From this perspective you might see that HP QC is the obvious choice. The truth of the matter however is that Tester1 is not a programmer. Test automation is programming. When I do a demo of the basic tests I might find that HP QC (or TTS) is better. I get QUICK results. With WebDriver the learning curve is much steeper at first for a non-programmer. So I end up getting nothing automated.

Now, you give me HP QC and it will take me 10 business days to automate 95% BUT I it will take me 5 business days to automate 95%. Additionally, within a year of creating test automation there will be some things I just cannot automate using HP QC without the maintenance costs making it a bad ROI. I will have no problem automating the same scenario using WebDriver and 3rd party libraries.

By the way, in the end I got Tester1 and Tester2 pairing. Tester1 would determine WHAT to test and Tester2 would automate it using WebDriver. End result was:

• Tester1/Tester2: WebDriver, 90% of automation, 3 business days.

That is 6 person-days (two people at 3 days each) for 90% of the automation. Better than 7 person-days for 80% of the automation (which is the best result from HP QC).

So:

• Good tester, bad programmer will do better with HP QC or TTS.
• Good tester, bad programmer fail with WebDriver.
• Bad tester, good programmer will do okay with HP QC or TTS.
• Bad tester, good programmer will do better with WebDriver.

For a one to two week pilot, HP QC or TTS will seem better. WebDriver might end up in complete failure. So statistically, you have better change of some success with record and playback tools. However:

• Good tester paired with good programmer will do okay with HP QC or TTS.
• Good tester paired with good programmer will do better with WebDriver.

Most companies will change the tool rather than change the automation team. But there is no reason not to change the people when you are changing how you test (manual versus automated).

[Jim Evans]

I'd like to chime in here with a word about Jim Holmes before he's pilloried here on this list. Jim has worked very hard to open and maintain bridges with the open-source community. I know he respects the WebDriver project immensely, and I personally consider him a friend. I reached out to him shortly after this marketing campaign started, and though I'm not willing to share the private communication between us, I was satisfied with our conversation. As a side note, Jim is no longer with Telerik (as of Friday), so it's probably best to consider this marketing campaign as coming from the company, not the man.

[Aniket Gadre]

I hope you meant HP QTP, HP QC is a Test Management tool 

Best Regards,

Aniket Gadre

 

“Programming can be fun, so can cryptography; however they should not be combined.”

Visit this group at http://groups.google.com/group/webdriver.
For more options, visit https://groups.google.com/d/optout.

[darrell]

Thanks for the correction. I did mean QTP. 

[WebTestGadfly]

Very good discussion overall.  Of course some of the material in the original article may be biased.  But it is a very healthy discussion notwithstanding.   A cautious reader will always "consider the source" in any one's presentations.

Meanwhile, for someone who has been in the field for a some time I feel it is important to point out that are many products that can test web applications and many of them can effectively remove the GUI complexity of QTP and the need for programming skill in Webdriver/Selenium solutions, respectively.   

But the direction of Webdriver/Selenium IS exciting overall, in particular if it becomes a W3C standard as has been recently proposed.

-WTG

[SuperKevy]

Do you have a reference to this "But the direction of Webdriver/Selenium IS exciting overall, in particular 

if it becomes a W3C standard as has been recently proposed."   I'm currently working with outside contractors who ignore w3c validation to doctype standards - Its like going back to IE3 and IE4 mentalities. 

[Jim Evans]

The proposed spec being worked on by a W3C working group can be found at https://dvcs.w3.org/hg/webdriver/raw-file/tip/webdriver-spec.html. The nice thing about this particular standard is that web site developers don't need to do anything to "follow" the spec; it's the browser vendors who are implementing it.