[Bret]
> > > This is complete bull. You are either not actually following the list
> > > or you are completely deluded.
[Marcus]
> > I am not deluded. I don't particularly like your tone one bit! I read all
> > the posts to the list every day. I am offering my opinion of what can
> > happen as a product grows in use and reaches a wider audience. You are
> > entitled to feel that my opinion is BS of course, but do you have to be so
> > rude?
[Jonathan]
> I am completely in agreement with Bret when it comes to recorders for testing,
> but I do agree with Marcus that this statement was a bit over the top. If
> Marcus
> wants to write and maintain a recorder, and there is a mailing list and
> community of practice around it, that is cool with me. I donÈt really want to
> support the recorded scripts either, but this was harsh.
There is an issue with recorders for testing. Different people have
different views, and i think that it is fair for people to challenge
my views in this area. A completely separate issue concerns the
relationship between commercial tools and open-source communities.
It was in response to my suggestion that we consider stop providing
support for recorder users on this list, and that a new list be
created, that he made statements that i claim are bull. Specifically:
"I'm not aware that it [WebRecorder] has in fact directly contributed
to any extra support requests..."
I say this is bull because it meets the definition of bull: it is
false, it is knowingly false, and it is declared with the expectation
that it will be heard as false. (This is the definition from "On
Bullshit" by Harry Frankfort.)
1. It is false. In Marcus's own email, it elsewhere acknowledged that
he'd made a support request and that he knew of one other request. A
simple google search of the archives will uncover more.
2. It is knowingly false. Since he acknowledges these support requests
elsewhere in his email, it is clear that he knows it is false.
3. He knows that we will read it as false. This is harder to prove,
but if you look at the rhetorical sweep of the paragraph it is found
in, you will see that he is both trying to deny that these support
requests exist, but also and at the same time, suggest that they are
an inseparable part of success. Reread the paragraph and judge for
yourself.
> I did not provide this version of WebRecorder in order to create any extra
> burden for you, and I am sorry if you feel that is all it has achieved. I'm
> not aware that it has in fact directly contributed to any extra support
> requests, but I suppose as a product grows in popularity, as Watir clearly
> is doing, and whether or not WebRecorder is in some way responsible for
> that, the support burden will increase as more and more people get to know
> about it, many of whom will be less technical than the earlier adopters.
> That's almost a fact of life - a side effect of success.
Several people in this community have made it clear that we don't want
to support recorder users. It was this sentiment that lead to Marcus's
rhetorical bull.
Is this harsh? Is it harsh to call a liar a liar? I decided to call it
what it was. ("Bullshit" is the current vernacular, but the original
term is "bull." I used the older term specifically to avoid the
scatalogical emphasis, which would make me more vulnerable to the
charge of being over the top. The only other English word for bull is
"humbug," which is truly obscure.)
Marcus's response has been to take WebRecorder off his website. No one
asked him to do this. We asked him to create a separate list.
But it is important for us to realize that as a commercial tool
vendor, he was always free to do this. It is his right. It's his code.
We can support it, help its users, report bugs in it, make
sugggestions and then he can turn around and decide that he wants to
start charge $100 or $1,000 a copy. Or he can take it offline
completely.
We can't do that. Watir is open-source. It belongs to the community.
No one can take it away. And no one can tell us what to do with it.
Not even me. Suppose that I were totally opposed to FireFox support.
There still would be nothing i could do to stop Angrez from creating
FireWatir. The code is public and there is nothing i could do legally
or morally to stop him, except argue against it.
I hate to say it, but i think it is time for us to discuss creating an
open-source recorder. Actually i'd prefer more of a spy tool, but i
think we need to create something that we can support (possibly on a
separate list) and that supports our vision for what a useful recorder
would be.
My personal view is that a really useful recorder would need to be
tuned for specific applications. You have to use different
identification strategies with different applications and different
controls. You all know this, and much of traffic on this list is
really focussed on helping others learn what the best strategies are
for their applications. A truly useful recorder would be tuned to
understand that a tests for a particular application should use names
for text_fields, but identify dropdowns by their prior text. That's my
view.
Watir has resulted from years of interaction between the vision of a
few of us for what a testing tool should be and frequent and detailed
feedback from many users, many of them completely new to the tool,
about what they expect the tool to do.
I suspect that it was also take years to create a good, usable
open-source recorder that actually teaches users about automation as
it is used instead of letting them think, even for a moment, that they
somehow can be successful without understanding the details.
Where should we start?
_______________________________________________
Wtr-general mailing list
[email protected]
http://rubyforge.org/mailman/listinfo/wtr-general
