Re: Is Cactus successful (was RE: [PROPOSAL] Tapestry joins Jakarta)

[Andrew C. Oliver]

That's awesome information.  Thank you for this!  

I got no reply from Javaworld or Dr. Dobbs, but I had not tried JDJ.  I 
wrote Tony Stintes and got no reply,
but I think this: 
http://www.javaworld.com/javaworld/javaqa/2002-05/01-qa-0503-excel3.html 
was sufficient
reply.  

Somehow we managed to get coverage in some German professional journals, 
etc.  But I'm not entirely sure
why "accidental marketing" works so much better in countries where 
English is not the native language maybe they try harder
to find things?

Once again, thank you so much for this.  It was very helpful to me!

-Andy

Steve Downey wrote:

I don't think I've ever NOT gotten a reply. It has been a few years since I've 
written, but I doubt things have changed that much.

Getting enough decent technical material for a magazine is always a problem. 
Writing can be fixed by an editor, although it's better if it doesn't have to 
be fixed too much. Technical content can't be fixed by an editor. 

You do need to check that you're hitting the right person. The thing to do is 
to check the author's guidelines from the publication. It's usually a link on 
the main page of the website. I think JDJ has a web form for proposals, for 
example, so that they don't get lost in someones inbox. It can be a little 
bit of a problem, as a lot of the editors work virtually, and hand-off can be 
botched. 

As long as you're clear who you are, being a principal is not an issue. Think 
about it from the other side. Would you rather read an article about JUnit by 
Erich Gamma or John Doe? You should know more about POI than anyone, and are 
in a better position to write about how to use it than anyone. 


On Tuesday 22 October 2002 11:19 am, Andrew C. Oliver wrote:
 

Cool!  Suggestions?  I have never actually gotten a reply from an editor
himself.

Of course for project validty, it may be better to have a non-prinicipal
contribute (if I write about POI it will not likely be viewed as
objective but someone who has written about other APIs will probably get
more credibility there), but that is another story.

Steve Downey wrote:
   

On Sunday 20 October 2002 02:03 pm, Andrew C. Oliver wrote:
 

Release more often, announce the releases.  While you may have had
articles published, I've never actually seen one.  (I've seen them on
Maven, Tomcat, Velocity, Struts to no end, Cocoon, Struts).  I found the
best approach to this is to spam magazines and complain that they
haven't covered it.  Find and article on say JUnit and write the author
"but you haven't written on web app testing with cactus!"
   

No, it's not the best approach.

The best way is call up an editor and say, "I'd like to write an article
for you about Cactus. Cactus is , will interest you readers because
of , and the article will be about N words long, not counting
source code. Can I email you an outline?"

Editors have the incredibly difficult job of keeping the covers of the
magazine apart. They desparately need articles. They can't pay as much as
would seem fair, but enough to buy decent toys.



--
To unsubscribe, e-mail:   
For additional commands, e-mail: 
 



--
To unsubscribe, e-mail:   
For additional commands, e-mail: 


 




--
To unsubscribe, e-mail:   
For additional commands, e-mail: 

Re: Is Cactus successful (was RE: [PROPOSAL] Tapestry joins Jakarta)

[Steve Downey]

I don't think I've ever NOT gotten a reply. It has been a few years since I've 
written, but I doubt things have changed that much.

Getting enough decent technical material for a magazine is always a problem. 
Writing can be fixed by an editor, although it's better if it doesn't have to 
be fixed too much. Technical content can't be fixed by an editor. 

You do need to check that you're hitting the right person. The thing to do is 
to check the author's guidelines from the publication. It's usually a link on 
the main page of the website. I think JDJ has a web form for proposals, for 
example, so that they don't get lost in someones inbox. It can be a little 
bit of a problem, as a lot of the editors work virtually, and hand-off can be 
botched. 

As long as you're clear who you are, being a principal is not an issue. Think 
about it from the other side. Would you rather read an article about JUnit by 
Erich Gamma or John Doe? You should know more about POI than anyone, and are 
in a better position to write about how to use it than anyone. 


On Tuesday 22 October 2002 11:19 am, Andrew C. Oliver wrote:
> Cool!  Suggestions?  I have never actually gotten a reply from an editor
> himself.
>
> Of course for project validty, it may be better to have a non-prinicipal
> contribute (if I write about POI it will not likely be viewed as
> objective but someone who has written about other APIs will probably get
> more credibility there), but that is another story.
>
> Steve Downey wrote:
> >On Sunday 20 October 2002 02:03 pm, Andrew C. Oliver wrote:
> >>Release more often, announce the releases.  While you may have had
> >>articles published, I've never actually seen one.  (I've seen them on
> >>Maven, Tomcat, Velocity, Struts to no end, Cocoon, Struts).  I found the
> >>best approach to this is to spam magazines and complain that they
> >>haven't covered it.  Find and article on say JUnit and write the author
> >>"but you haven't written on web app testing with cactus!"
> >
> >No, it's not the best approach.
> >
> >The best way is call up an editor and say, "I'd like to write an article
> > for you about Cactus. Cactus is , will interest you readers because
> > of , and the article will be about N words long, not counting
> > source code. Can I email you an outline?"
> >
> >Editors have the incredibly difficult job of keeping the covers of the
> >magazine apart. They desparately need articles. They can't pay as much as
> >would seem fair, but enough to buy decent toys.
> >
> >
> >
> >--
> >To unsubscribe, e-mail:   
> >For additional commands, e-mail: 


--
To unsubscribe, e-mail:   
For additional commands, e-mail: 

Re: Is Cactus successful (was RE: [PROPOSAL] Tapestry joins Jakarta)

[Andrew C. Oliver]

Cool!  Suggestions?  I have never actually gotten a reply from an editor 
himself.  

Of course for project validty, it may be better to have a non-prinicipal 
contribute (if I write about POI it will not likely be viewed as 
objective but someone who has written about other APIs will probably get 
more credibility there), but that is another story.

Steve Downey wrote:

On Sunday 20 October 2002 02:03 pm, Andrew C. Oliver wrote:
 

Release more often, announce the releases.  While you may have had
articles published, I've never actually seen one.  (I've seen them on
Maven, Tomcat, Velocity, Struts to no end, Cocoon, Struts).  I found the
best approach to this is to spam magazines and complain that they
haven't covered it.  Find and article on say JUnit and write the author
"but you haven't written on web app testing with cactus!"

   

No, it's not the best approach. 

The best way is call up an editor and say, "I'd like to write an article for 
you about Cactus. Cactus is , will interest you readers because of 
, and the article will be about N words long, not counting source code. 
Can I email you an outline?"

Editors have the incredibly difficult job of keeping the covers of the 
magazine apart. They desparately need articles. They can't pay as much as 
would seem fair, but enough to buy decent toys. 



--
To unsubscribe, e-mail:   
For additional commands, e-mail: 


 




--
To unsubscribe, e-mail:   
For additional commands, e-mail: 

Re: Is Cactus successful (was RE: [PROPOSAL] Tapestry joins Jakarta)

[Steve Downey]

On Sunday 20 October 2002 02:03 pm, Andrew C. Oliver wrote:
> Release more often, announce the releases.  While you may have had
> articles published, I've never actually seen one.  (I've seen them on
> Maven, Tomcat, Velocity, Struts to no end, Cocoon, Struts).  I found the
> best approach to this is to spam magazines and complain that they
> haven't covered it.  Find and article on say JUnit and write the author
> "but you haven't written on web app testing with cactus!"
>
No, it's not the best approach. 

The best way is call up an editor and say, "I'd like to write an article for 
you about Cactus. Cactus is , will interest you readers because of 
, and the article will be about N words long, not counting source code. 
Can I email you an outline?"

Editors have the incredibly difficult job of keeping the covers of the 
magazine apart. They desparately need articles. They can't pay as much as 
would seem fair, but enough to buy decent toys. 



--
To unsubscribe, e-mail:   
For additional commands, e-mail: 

Re: Is Cactus successful (was RE: [PROPOSAL] Tapestry joins Jakarta)

[Andrew C. Oliver]

> > 
> > Just FYI, you have failed to provide sufficient outside-of-jakarta
> > marketing.  
> 
> Thanks for the information! You know what I like about you? It is the
> faith that you have in yourself and in the fact that you know it all...
> 
I'm sorry to have insulted you.  I was only trying to help.  Based on
what you said, the observations I'd had from where I work and what I'd
observed.  If my observations were incorrect I apologize.

-Andy
 
-- 
http://www.superlinksoftware.com - software solutions for business
http://jakarta.apache.org/poi - Excel/Word/OLE 2 Compound Document in
Java
http://krysalis.sourceforge.net/centipede - the best build/project
structure
a guy/gal could have! - Make Ant simple on complex Projects!
The avalanche has already started. It is too late for the pebbles to
vote.
-Ambassador Kosh


--
To unsubscribe, e-mail:   
For additional commands, e-mail: 

Re: Is Cactus successful (was RE: [PROPOSAL] Tapestry joins Jakarta)

[Ellis Teer]

If you check the cactus stats
(http://jakarta.apache.org/cactus/stats/index.html), you will find it is
receiving quite a lot of attention (1500-2500 visits per day). It gets
between 500-1500 downloads per day which is quite honorable for such a
"niche" project (not only it is unit testing but only J2EE-related unit
testing).
Note: I don't like too much the stats from webalizer and I have started
using awstats
(http://jakarta.apache.org/~vmassol/awstats/awstats.jakarta.apache.org.h
tml) as an experiment. Much better stats I think (except the history).


Thank you for mentioning that stats package as a sidenote.  Thinking of 
using it for a project now instead of analog/report magic.

With Cactus, my impression is that a sizable number of developers know 
about it, are interested in it, think it is a good idea, but are too 
'busy' in their day jobs to use unit testing, or more complex 
server-side unit testing.  And for those who have pet projects, those 
projects are usually small enough that the benefits of setting up test 
cases in the mind are outweighed by adding another feature.

You mentioned later in this thread what I believe is a key point.  You 
are doing something new and the public/users/possible committers need to 
learn to adopt unit testing and then 'IUT'.  And while people are 
thinking about it, they will browse the page, and even download a copy. 
 But until they use it for a project and see the advantages its 
adoption/attraction to developers may be slow.  I think the 
documentation and polish helps find developers more than hurts, IUT is 
just ahead of their needs though it doesn't need to be.

It reminds me of how digital video records have been slow to take off. 
But their time is coming one way or another.

--
Ellis Teer
www.sitepen.com


--
To unsubscribe, e-mail:   
For additional commands, e-mail: 

RE: Is Cactus successful (was RE: [PROPOSAL] Tapestry joins Jakarta)

[Vincent Massol]

> -Original Message-
> From: Andrew C. Oliver [mailto:acoliver@;apache.org]
> Sent: 20 October 2002 19:03
> To: Vincent Massol
> Cc: 'Jakarta General List'
> Subject: RE: Is Cactus successful (was RE: [PROPOSAL] Tapestry joins
> Jakarta)
> 
> 
> > > I'm sorry to have insulted you.  I was only trying to help.  Based
on
> > > what you said, the observations I'd had from where I work and what
I'd
> > > observed.  If my observations were incorrect I apologize.
> >
> > Hey, your observations may be right and I appreciate your help! ;-)
> >
> > My strong reaction was probably due more to the tone of your email.
> > Saying to someone: "you have failed to provide ..." is not a very
nice
> > thing to say. Saying, "maybe you should try to work more on the
> > marketing side" is softer! :-). My turn to apologize if my reaction
was
> > too strong ;-)
> >
> 
> Ahh... word connotation.  I need a Jon page.. . I suck at finding the
> right way of saying things in email.  I'm told I'm abrasive in email,
> but not in person.  The funny thing is I say the same things... They
> just come off differently.  Maybe its my facial expressions or tone of
> voice...  (though my wife always misinterprets my facial expressions)
> 
> > Ok I want to be positive and try to see if there's anything I can do
to
> > improve the overall Cactus community. You say Cactus might be
missing
> > some marketing muscles. I would like to believe that. From the
> > information in my email do you still think there's more the Cactus
team
> > could do?
> >
> 
> Release more often, announce the releases.  While you may have had
> articles published, I've never actually seen one.  (I've seen them on
> Maven, Tomcat, Velocity, Struts to no end, Cocoon, Struts).  I found
the
> best approach to this is to spam magazines and complain that they
> haven't covered it.  Find and article on say JUnit and write the
author
> "but you haven't written on web app testing with cactus!"
> 
> Some people do the JUG tour.
> 

The discussion I had started was not about how to grow Cactus user base.
We are very happy with that (well beyond my expectations actually and
progressing steadily) but more on how to grow the committer's base. Of
course growing the number of people who knows about Cactus may have an
effect on the number of committers but I am not so sure about that
(that's not my observations so far).

My belief is that Cactus is too much viewed as a finished product:
- it works
- it has nice documentation
- it has a nice build process
- it has a quite responsive mailing list (a bit less lately as I have
slightly less time with Maven/other stuff and I'm trying to see if
others jump in when I don't answer 30 seconds after the question has
been posted ;-). It seems to be working! I have the feeling Cactus is
getting more ML participation ...).
- steady releases (9 versions/releases in 16 months, 1 stable releases/4
months).

It is indeed a finished product but there are still so many interesting
things to do to make it even way better ... ;-) (documented on the
todo/goal page).

> > You said the persons in your team were pondering about using Cactus.
> > That means they already know about it. So marketing is good! Maybe
> > documentation need to be improved?
> >
> 
> No not on my team.. . Just people I know at work ;-)
> 
> If I had a need for cactus at the moment the documentation doesn't
look
> too bad.

It isn't but needs reorganization to prevent misreadings like the one
you mention below ... :-)

> 
> > FYI, we have actually started working on the Cactus front-end (Maven
> > plugin, Eclipse plugin, standalone testing application, Ant tasks,
etc)
> > which should make it easier to use (this was one weak-point noted by
the
> > users - the entry barrier).
> >
> 
> Ahh.  i actually use Centipede and haven't figured out what eclipse
> wants from me (I want CVS + compilation...it gives me either or).

Let's not go there... Ok just a little then... Personally I don't want
compilation, which Eclipse gives me ;-) (I'm talking about dynamic
compilation). I want to know right away the result of a change I make to
a class across all my project, I want to know right away if I've broken
any coding convention as I type (using checkstyle for example), I want
to know the code impacted by my Aspects (AOP/AspectJ), if I make a
change deep in the directory structure, I want it to be surfaced so that
I know what is not committed, etc.

If you can resist, try not to answer this as it will lead to another OT
thread ;-) (you've started!).

> 
> The b

RE: Is Cactus successful (was RE: [PROPOSAL] Tapestry joins Jakarta)

[Vincent Massol]

> -Original Message-
> From: Andrew C. Oliver [mailto:acoliver@;apache.org]
> Sent: 20 October 2002 18:32
> To: 'Jakarta General List'
> Subject: Re: Is Cactus successful (was RE: [PROPOSAL] Tapestry joins
> Jakarta)
> 
> 
> > >
> > > Just FYI, you have failed to provide sufficient outside-of-jakarta
> > > marketing.
> >
> > Thanks for the information! You know what I like about you? It is
the
> > faith that you have in yourself and in the fact that you know it
all...
> >
> I'm sorry to have insulted you.  I was only trying to help.  Based on
> what you said, the observations I'd had from where I work and what I'd
> observed.  If my observations were incorrect I apologize.

Hey, your observations may be right and I appreciate your help! ;-)

My strong reaction was probably due more to the tone of your email.
Saying to someone: "you have failed to provide ..." is not a very nice
thing to say. Saying, "maybe you should try to work more on the
marketing side" is softer! :-). My turn to apologize if my reaction was
too strong ;-)

Ok I want to be positive and try to see if there's anything I can do to
improve the overall Cactus community. You say Cactus might be missing
some marketing muscles. I would like to believe that. From the
information in my email do you still think there's more the Cactus team
could do?

You said the persons in your team were pondering about using Cactus.
That means they already know about it. So marketing is good! Maybe
documentation need to be improved?

FYI, we have actually started working on the Cactus front-end (Maven
plugin, Eclipse plugin, standalone testing application, Ant tasks, etc)
which should make it easier to use (this was one weak-point noted by the
users - the entry barrier).

Thanks
-Vincent

> 
> -Andy
> 
> --
> http://www.superlinksoftware.com - software solutions for business
> http://jakarta.apache.org/poi - Excel/Word/OLE 2 Compound Document in
> Java
> http://krysalis.sourceforge.net/centipede - the best build/project
> structure
>   a guy/gal could have! - Make Ant simple on complex
> Projects!
> The avalanche has already started. It is too late for the pebbles to
> vote.
> -Ambassador Kosh
> 
> 
> --
> To unsubscribe, e-mail:
<mailto:general-unsubscribe@;jakarta.apache.org>
> For additional commands, e-mail:
<mailto:general-help@;jakarta.apache.org>



--
To unsubscribe, e-mail:   <mailto:general-unsubscribe@;jakarta.apache.org>
For additional commands, e-mail: <mailto:general-help@;jakarta.apache.org>

RE: Is Cactus successful (was RE: [PROPOSAL] Tapestry joins Jakarta)

[Henri Yandell]

On Sun, 20 Oct 2002, Vincent Massol wrote:

>
>
> Ok I want to be positive and try to see if there's anything I can do to
> improve the overall Cactus community. You say Cactus might be missing
> some marketing muscles. I would like to believe that. From the
> information in my email do you still think there's more the Cactus team
> could do?

Even ignoring my Jakarta-Apache involvement, I had become aware of Cactus
through articles [i think] and mention in the media. However, there's no
real meme [I think is the phrase] for Cactus. I know that Cactus is Java,
I know it tests web pages somehow (least it's hooked up in there) but I
don't have a good idea as to why I would choose it, how it differs from
httpunit etc. [A quick look at the website shows that my initial
assumption was off a touch, it tests Java server-side components].

So for me personally, and I suspect other people who are aware of Cactus,
there's not a real understanding first off of where it fits in, the level
of effort to use etc.

[adding it to article list to write at some point, though I'm sure there
are many out there. ]

Probably not much use, but a report from a prospective customer can
sometimes be of interest. I'd say Cactus is marketing the brand okay, but
not the meme. Not that I really understand brands or memes :)

Hen


--
To unsubscribe, e-mail:   
For additional commands, e-mail: 

RE: Is Cactus successful (was RE: [PROPOSAL] Tapestry joins Jakarta)

[Vincent Massol]

> -Original Message-
> From: Henri Yandell [mailto:bayard@;generationjava.com]
> Sent: 20 October 2002 20:36
> To: Jakarta General List
> Cc: [EMAIL PROTECTED]
> Subject: RE: Is Cactus successful (was RE: [PROPOSAL] Tapestry joins
> Jakarta)
> 
> 
> 
> On Sun, 20 Oct 2002, Vincent Massol wrote:
> 
> >
> >
> > Ok I want to be positive and try to see if there's anything I can do
to
> > improve the overall Cactus community. You say Cactus might be
missing
> > some marketing muscles. I would like to believe that. From the
> > information in my email do you still think there's more the Cactus
team
> > could do?
> 
> Even ignoring my Jakarta-Apache involvement, I had become aware of
Cactus
> through articles [i think] and mention in the media. However, there's
no
> real meme [I think is the phrase] for Cactus. I know that Cactus is
Java,
> I know it tests web pages somehow (least it's hooked up in there) but
I
> don't have a good idea as to why I would choose it, how it differs
from
> httpunit etc. [A quick look at the website shows that my initial
> assumption was off a touch, it tests Java server-side components].

Hehe ... I think you may have nailed the exact problem! Even your last
sentence is not completely correct which proves your point (although it
is written on the web site!)... ;-) It's about *unit* testing java
server-side components (although at the moment it is more restricted to
unit testing J2EE components but that's not the only goal). And it's
about doing it in-container (inside the container).

Would "In-container Unit Testing" or "Integration Unit Testing" be a
nice meme?

I think the problem also comes from the fact that unit tests are still
relatively new. And what Cactus is doing is even newer. Thus there is
not yet any global knowledge of what IUT is about ... Cactus was built
to explore this road and is indeed a precursor ;-)

> 
> So for me personally, and I suspect other people who are aware of
Cactus,
> there's not a real understanding first off of where it fits in, the
level
> of effort to use etc.
> 

Yes, you are right. Everything is described on the web site (included
comparisons with other strategies), etc. BUT the problem is that you
have to read it first ... 

In that sense, Andrew was right. There is a need to do a lot of
evangelization on the concept of IUT so that it enters our global mind.

> [adding it to article list to write at some point, though I'm sure
there
> are many out there. ]

That would be nice ;-)

> 
> Probably not much use, but a report from a prospective customer can
> sometimes be of interest. I'd say Cactus is marketing the brand okay,
but
> not the meme. Not that I really understand brands or memes :)

I think you are completely right :-)

In addition the concept may need to be expanded a bit as it may be too
restrictive. Here are some ideas for the future:
- runtime unit testing (possibly using AOP)
- stress unit testing
(Thus more like tools like Introscope but at a much more agile level.)

It would still be In-Container Unit Testing (ICUT or IUT), though...
 
> 
> Hen
> 

-Vincent



--
To unsubscribe, e-mail:   <mailto:general-unsubscribe@;jakarta.apache.org>
For additional commands, e-mail: <mailto:general-help@;jakarta.apache.org>

RE: Is Cactus successful (was RE: [PROPOSAL] Tapestry joins Jakarta)

[Andrew C. Oliver]

> > I'm sorry to have insulted you.  I was only trying to help.  Based on
> > what you said, the observations I'd had from where I work and what I'd
> > observed.  If my observations were incorrect I apologize.
> 
> Hey, your observations may be right and I appreciate your help! ;-)
> 
> My strong reaction was probably due more to the tone of your email.
> Saying to someone: "you have failed to provide ..." is not a very nice
> thing to say. Saying, "maybe you should try to work more on the
> marketing side" is softer! :-). My turn to apologize if my reaction was
> too strong ;-)
> 

Ahh... word connotation.  I need a Jon page.. . I suck at finding the
right way of saying things in email.  I'm told I'm abrasive in email,
but not in person.  The funny thing is I say the same things... They
just come off differently.  Maybe its my facial expressions or tone of
voice...  (though my wife always misinterprets my facial expressions)

> Ok I want to be positive and try to see if there's anything I can do to
> improve the overall Cactus community. You say Cactus might be missing
> some marketing muscles. I would like to believe that. From the
> information in my email do you still think there's more the Cactus team
> could do?
> 

Release more often, announce the releases.  While you may have had
articles published, I've never actually seen one.  (I've seen them on
Maven, Tomcat, Velocity, Struts to no end, Cocoon, Struts).  I found the
best approach to this is to spam magazines and complain that they
haven't covered it.  Find and article on say JUnit and write the author
"but you haven't written on web app testing with cactus!"  

Some people do the JUG tour.  

> You said the persons in your team were pondering about using Cactus.
> That means they already know about it. So marketing is good! Maybe
> documentation need to be improved?
> 

No not on my team.. . Just people I know at work ;-)

If I had a need for cactus at the moment the documentation doesn't look
too bad.

> FYI, we have actually started working on the Cactus front-end (Maven
> plugin, Eclipse plugin, standalone testing application, Ant tasks, etc)
> which should make it easier to use (this was one weak-point noted by the
> users - the entry barrier).
> 

Ahh.  i actually use Centipede and haven't figured out what eclipse
wants from me (I want CVS + compilation...it gives me either or).

The biggest turnoff for me was this:

"
 This will work with Internet Explorer as the XSL transformation is
performed on the client side (i.e by the browser). I'm not sure about
other browsers.
" 

IE doesn't work well under Linux.  Plus I like mozilla better and even
use it when I have to use Winblows.  

-Andy

> Thanks
> -Vincent
> 
> > 
> > -Andy
> > 
> > --
> > http://www.superlinksoftware.com - software solutions for business
> > http://jakarta.apache.org/poi - Excel/Word/OLE 2 Compound Document in
> > Java
> > http://krysalis.sourceforge.net/centipede - the best build/project
> > structure
> > a guy/gal could have! - Make Ant simple on complex
> > Projects!
> > The avalanche has already started. It is too late for the pebbles to
> > vote.
> > -Ambassador Kosh
> > 
> > 
> > --
> > To unsubscribe, e-mail:
> 
> > For additional commands, e-mail:
> 
> 
> 
-- 
http://www.superlinksoftware.com - software solutions for business
http://jakarta.apache.org/poi - Excel/Word/OLE 2 Compound Document in
Java
http://krysalis.sourceforge.net/centipede - the best build/project
structure
a guy/gal could have! - Make Ant simple on complex Projects!
The avalanche has already started. It is too late for the pebbles to
vote.
-Ambassador Kosh


--
To unsubscribe, e-mail:   
For additional commands, e-mail: