struts-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jonathan Revusky <revu...@wanadoo.es>
Subject Re: Why did Struts development stagnate?
Date Wed, 29 Mar 2006 19:09:38 GMT
Niall Pemberton wrote:
> ----- Original Message ----- 
> From: "Jonathan Revusky" <revusky@wanadoo.es>
> Sent: Saturday, March 25, 2006 11:27 PM
> 
> 
>>It still seems broadly on-topic to me. It's certainly a legitimate,
>>well-formulated question.
>>
>>Seriously, the only other possibility I see is struts-dev. If it's
>>off-topic on both struts-user and struts-dev, then the question really
>>is (as I am starting to suppose) basically taboo.
> 
> 
> 
> The question isn't taboo - I posed the same kind of thing (and offered one
> perspective) in an earlier thread:
> 
> http://article.gmane.org/gmane.comp.jakarta.struts.user/122903
> 
> However I don't think what I said in that thread was the whole story -
> clearly frameworks such as WebWork succeeded and I assume they were a
> volunteer effort as well.

Yes, the bulk of your explanation there seemed to be that Struts was an 
all-volunteer effort and so on.

This could not possibly be why it fell behind Webwork.

> 
> We currently have 22 committers on Struts - 

Out of curiosity, what is your rough guess as to how many of these 22 
people committed any code in the last... year, let's say.

> but levels of activity vary
> widely and I would say that the type of talented people it takes to drive a
> project forward (and I don't include myself in that group) no longer have an
> interest in doing so on the Action 1 side - for various reasons. People such
> as Craig put their effort into developing the JSF standard and see that as
> the future for web development and that is where they now concentrate their
> effort. Don was doing alot of work inovating with Struts Ti 

Well, I was not aware of this. However, you mean that Struts TI was a 
complete rewrite of the framework? I mean, was there a tacit assumption 
there that Struts 1.x could not be evolved forward and required a 
complete rewrite?

> and had the
> offer to merge not come along from WebWork - we would probably be seeing the
> fruits of his efforts as Action2 and not even discussing "stagnation" at
> this point. Ted was AWOL doing C# for a while (hes been "back" for a while
> which is good :-), Martin seems focused on javascript etc. etc. So I guess
> this leads to the next question "Well why didn't we attract new talented
> people into the project that would drive Struts forward?" This I don't
> know - seems that lots of people decided to go invent their own web
> framework (YAWF) rather than get involved with Struts. Some of that is
> certainly their own egos being the "founder of a framework" and some of it I
> believe is the compatibility issue - its far easier to write a brand new
> shiny web framework when not hampered by backwards compatibility. Whether we
> as a community "put them off" I have no knowledge - but I've never seem that
> proferred anywhere as a reason. It was always something like "Struts sucks
> because of x, y and z and my brand new shiny framework does it better".
> Course its far easier to invent a new framework by looking at existing ones
> and seeing how you can improve them. Back to the "new people" question
> though - its not my perspective that we have lots of people knocking at the
> door trying to give us contributions and we're turning them away. I believe
> its easy to become a Struts committer - you offer reasonable code, are
> helpful in the community (e.g. answering questions on the user list), been
> around a while and don't start flame wars or attack people personally - then
> you get asked. Theres probably 2/3 people who probably think they should
> have been asked, but haven't - they may or may no have a point - but besides
> them I don't see it as a case of Struts excluding people and I don't have an
> explanation for why there are not hoards of people wanting to join.

Well, first of all, on the question of people going off and doing their 
own framework, you have to basically figure that some of these people 
just didn't think that they could apply their ideas in this setting. If 
somebody with a fire in their belly and some innovative ideas had showed 
up here and wanted to work on that, would they have been able to do so?

After all, the fact remains that everybody knows that any work they do 
under the ASF umbrella will get much more attention and usage than it 
would otherwise. This is the main (probably the only) reason that the 
Webwork people have come here. So, a priori, your saying that you aren't 
attracting collaborators is really quite odd, isn't it?

The thing is, Niall, that pretty much all the times you get a new 
collaborator, that person was first a user. Typically that someone is a 
"power user", and is pushing the limits of what the tool can do, and 
starts donating code to make the tool more powerful, and next thing you 
know, the guy is a collaborator.

Now, you've got a lot of users, so that this basic mechanism doesn't 
operate is rather odd.

What I have noticed is that the communication with your user community 
is rather poor. Basically, for all of it, the bulk of your users seem 
completely clued out as to what is going on with the Webwork merger.

For example, you get people flaming me because I am saying that Webwork 
is better than Struts. They say "stop bashing Struts". But I am saying 
exactly what the Struts developers are saying! They have accepted that 
Webwork is better than Struts! So am I supposed to be more catholic than 
the pope?

Also these people assume that I must be a Webwork developer. Somebody 
wrote a spoof of me in which I was praising Webwork to the skies! I have 
nothing to do with Webwork. I have never even used it. When I say 
Webwork is better, I am simply echoing what the Struts PMC are already 
saying.

So, I mean, some of this is just going on because people don't know 
what's going on. I see a real communications failure.

If people really knew that the current Struts 1.x codebase is being 
abandoned, you would think that there would be a lot more threads on 
this list about migration issues. "I've got this Struts 1.x App and I 
just was having a look at Webwork, which is going to be Struts Action 2 
and have various questions about how my app can be migrated...." I don't 
see threads like that, which means to me that you have not communicated 
to  your rank and file users what is really going on here.

Now, if there really is a problem in terms of user<->developer 
communication here, it would explain why the process whereby certain 
power users become collaborators is not happening as often as it should. 
And this would be a factor in the stagnation.

Certainly, given the size of the user community, even if 1 in 100 people 
eventually became committers via that process, you would have a lot of 
active committers.

That a community like webwork with far fewer users nonetheless has a 
more active, real developer team, is really something to look at.

Certainly, in earlier discussions, most people just seemed to think that 
it was really hard to become a commmitters. So if that is a 
misconception, it is a widely held one. There's something odd going on here.

> 
> Another answer to the question is "it hasn't stagnated - 


Stop, Niall, stop. That's not an answer. :-) Let's not go around 
completely in circles.

> we've moved on to
> Shale" and that is the future for existing Struts users. 

Well, if that is the case, you haven't communicated it to your users.

I grant that if you are going to communicate something to your users, 
you should probably have a consistent message. The Action/Shale 
cohabitation seems to almost preclude having a consistent message.

Anyway, JSF/Shale is just something completely different 
paradigmatically and the idea of that as "Struts 2" is really quite odd.

> Clearly there are
> quite a few people that will disagree with this - but also alot that will
> say "great I buy JSF as the future and I'm glad the Struts project has an
> offering that supports this".

Well, unless you are offering migration tools or a compatibility layer 
or something, how does it benefit your users that Shale is under the 
"Struts umbrella" any more than if it was a separate project? I mean, 
it's a paradigmatic shift that you have to get head around either way 
and existing apps would need to be refactored.

> 
> At the end of the day though this does seem academic,  - since we now have two
> offering for whatever camp you fall into (component orientated or action
> orientated) and from my point of view the really good thing about the
> WebWork merger is not only the great software were getting - but also the
> talented new blood thats coming into the project.

Well, if you accept that the Webwork people just ran the better project, 
you guys failed to keep Struts 1.x going at least in terms of innovation 
and development, then by that logic, the current Struts PMC should just 
step down probably and let the Webwork people run the show.

If the same PMC that presided over technical stagnation before is going 
to remain the managers of the project, then I think it isn't an academic 
question. You have to examine the mistakes you made before.


> 
> So I've given my answer to the question - now can we let this list get back
> to helping and answering user questions - which is its main purpose?

Niall, I don't know what you're talking about here. I see no sign that 
the list stopped helping people and answering their questions due to the 
presence of this thread.

You were giving some signs that you now were willing to talk about this. 
You've had a certain say about this now. You've stepped forward and said 
the topic is not taboo. Well, now you're saying, let's not talk about it 
any more, i.e. I broke the taboo temporarily to get this guy off my 
back, but nudge nudge, wink, wink, the topic really is taboo.

Okay, maybe that wasn't your intent, but if not, and the topic isn't 
taboo, how do you know other people don't have opinions to express now?

Again, the idea that this is an either-or proposition and the list has 
to choose between talking about this and helping people by answering 
technical questions is actually absurd, isn't it?

Jonathan Revusky
--
lead developer, FreeMarker project, http://freemarker.org/

> 
> Niall


---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@struts.apache.org
For additional commands, e-mail: user-help@struts.apache.org


Mime
View raw message