You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@velocity.apache.org by Justin Wells <jr...@semiotek.com> on 2000/08/26 08:00:23 UTC
Velocity, WebMacro, and the future
I looked at Velocity and I see that you've got it up to about where WM
was at version 0.30, prior to the first release. Not bad for three weeks
of work, you've obviously got some competent people working on it. You
have a long way to go with it though, and all of that work will be
repeating things I've done with WebMacro.
You should all know that I've worked very hard over the last six months
to create an open source license that I believed would be better than
both the MPL and the Apache license, but very much in the same spirit.
It was intended to be a better license for Java programs than the MPL,
and one that was simple and easy for average developers to read.
You can see the product of all that work here:
http://shimari.com/SPL/
The ASF and Jon have had many concerns with this license over time, but
I've worked quite hard to remove everything from it that was a concern.
So far as I know, all of those concerns have been addressed. It was my
intention to fix any concerns that remained.
However, I'm an extremely busy person and I guess I was too slow in
seeing this process pushed through to completion. I still believe in
the SPL, but in the meantime it's obvious I should have provided you
all with an alternative until the SPL is ready for prime time.
I apologize for that.
There's no doubt that my slowness has resulted in this fork in WM. I
want to say that it was honestly because I've been busy. If I'd known
you were so intent on forking, I would have MPL'd it for you prior to
now. I didn't realize how pressing your needs were I guess, and I'm
sorry for that too.
I admit this all pissed me off at first. But I'm not really capable of
holding grudges so I'm not pissed off anymore. Instead I am interested
in doing whatever it takes to mend the fences and bring everyone back
under a common source tree.
There's no way that we could bring Velocity up to where WebMacro is
in short order. It's got too far to go, and I have pressing deadlines
with respect to WM. We're rolling WM out on one of the largest sites
on the net soon, and I just can't throw that all away to go and work
on Velocity--it's a good start, but only a fraction of the way to
where WebMacro is now.
Later this evening I am going to release a snapshot of WebMacro that
has a huge performance gain over the previous snapshot. My feeling
these days is that WM has a theoretical performance advantage over
JSP, and if it isn't already faster than JSP, it will be soon. I
intend to make WebMacro the highest performance page generation
system that there is--so that it can work under great tools like
Turbine to give us all an edge over the non-Java world.
The snapshot will come out under the MPL.
I know there are a lot of egos involved here, and some of you have put
good long hours into Velocity so far. But if we don't mend the fences
and move forward together it's going to cost us all.
You guys have a difficult choice to make now. I'm putting WebMacro
under a license you can work with. It's way ahead of where you are. It
would be bad for us all to fork. We all know that, deep down.
Are we going to bury the hatchet and move the world forward? Or are
we going to split resources between moving WebMacro foward and you
reimplementing what I've already done?
I may be difficult to deal with sometimes, mostly owing to how busy I've
been over the last while. I could really use your help, and I don't
hold grudges.
Let's be friends.
Justin
Re: Velocity, WebMacro, and the future
Posted by Justin Wells <jr...@semiotek.com>.
On Sat, Aug 26, 2000 at 02:19:46AM -0700, Jon Stevens wrote:
> on 8/26/2000 2:13 AM, "Justin Wells" <jr...@semiotek.com> wrote:
>
> > Ok, on this condition: We rename the project Apache-Webmacro, we put in
> > place the WebMacro code base, and we work on merging your better parser
> > into WM's parsing framework (make it implement org.webmacro.parser and
> > support the DirectiveBuilder architecture).
>
> Sorry. No more "conditions" games. You already burned your condition limit
> with the previous one.
No games Jon. It was a serious offer.
My patience is running out. WebMacro will do fine with or without you,
you just have too far to go to catch up. But, you can draw off enough
resources that it'll hurt my ability to compete with JSP.
Right now I'm pretty pissed off with you for violating my trust. And I
can sure tell you're operating on emotion as well. But logically, what
we need to do here is wind up with your parser integrated into WM
under a license we both like.
It would be tough to merge your parser in under WM's tree but we could
do it. You really haven't implemented anything else. But you will soon,
and every line of code you write from here on in will only make it more
difficult for us to merge. If we don't do it now, later we'll regret
it and it'll be too late.
> I think Apache Velocity is a much cooler name than WebMacro. I have always
> had a negative view on products that used the word "Web" in their name.
> Totally overused concept.
That sounds caustic but I think you wouldn't even write that if in the
back of your mind you didn't already know what the right thing to do
is here. My name is associated with "WebMacro" and I don't plan to give
up credit for the hard work I've done, nor do I plan to take a year
long step back from what I'm on the verge of accomplishing.
WM's architecture is far more flexible than you seem to realize,
and the performance improvements I've made apparently haven't even
dawned on you yet. That sounds patronizing, but really you'll
understand after you've worked on this problem as long as I have.
You view WM as a plug-in to Turbine but that's actually not how most
people use it. It's an important use, and one I want to support, but
there are lots of people with a different requirement. You're seeing
WM through your own eyes only, and missing most of the way it
advantages other users.
Unless you support those users all you'll have is a turbine-plugin.
Justin
Re: Velocity, WebMacro, and the future
Posted by Jon Stevens <jo...@latchkey.com>.
on 8/26/2000 2:13 AM, "Justin Wells" <jr...@semiotek.com> wrote:
> Ok, on this condition: We rename the project Apache-Webmacro, we put in
> place the WebMacro code base, and we work on merging your better parser
> into WM's parsing framework (make it implement org.webmacro.parser and
> support the DirectiveBuilder architecture).
Sorry. No more "conditions" games. You already burned your condition limit
with the previous one.
I think Apache Velocity is a much cooler name than WebMacro. I have always
had a negative view on products that used the word "Web" in their name.
Totally overused concept.
Once you MPL/APL your code, we will consider implementing the parts of your
arch that we like.
> I would like to lead the project, and I would like people commit to
> performance as a primary goal. There are a lot of design decisions in
> WM's core which as primary maintainer I'm not sure other people really
> understand and I'm a bit worried that people will break things without
> realizing it.
Nope. That is not how the Jakarta Project works. For the guidelines, please
see the jakarta.apache.org website. This is all well documented.
-jon
--
Scarab -
Java Servlet Based - Open Source
Bug/Issue Tracking System
<http://scarab.tigris.org/>
Re: Velocity, WebMacro, and the future
Posted by Justin Wells <jr...@semiotek.com>.
On Sat, Aug 26, 2000 at 01:51:52AM -0700, Jon Stevens wrote:
> Ok. Come merge with the Velocity project.
Ok, on this condition: We rename the project Apache-Webmacro, we put in
place the WebMacro code base, and we work on merging your better parser
into WM's parsing framework (make it implement org.webmacro.parser and
support the DirectiveBuilder architecture).
I would like to lead the project, and I would like people commit to
performance as a primary goal. There are a lot of design decisions in
WM's core which as primary maintainer I'm not sure other people really
understand and I'm a bit worried that people will break things without
realizing it.
Justin
Re: Velocity, WebMacro, and the future
Posted by Jon Stevens <jo...@latchkey.com>.
on 8/26/2000 1:49 AM, "Justin Wells" <jr...@semiotek.com> wrote:
> Velocity has a long way to go to catch up to where WebMacro is now, and for
> what I'm doing today, I need something that is already there. Not something
> that will take another year to materialize.
>
> Note the parser you keep talking about is only 800 lines of code in WM.
> It's a small, small piece of the puzzle and you haven't even begun to
> work on the rest. Caching is also a very small part of it.
Oh master, please tell me more about this puzzle you speak of!
Dude. It is a parser, an introspection engine, a template serializer and a
caching mechanism. We are not talking rocket science here and we have proven
that by re-implementing most of that in 2 weeks time. Another 2 weeks and we
will be 2x as far, if not more because we have the developer resources of
some of the best Java engineers out there.
> Yeah I don't, and I still don't, but I'm not as stuck on the license
> as you are.
Then why did you waste 5 months of my time???????? If you weren't as stuck,
then you should have simply gone with the APL and not wasted everyone's
time.
> I'm sorry you had to put resources into it,
> but right now the right thing to do is merge and move forward together.
Ok. Come merge with the Velocity project.
-jon
--
Scarab -
Java Servlet Based - Open Source
Bug/Issue Tracking System
<http://scarab.tigris.org/>
Re: Velocity, WebMacro, and the future
Posted by Justin Wells <jr...@semiotek.com>.
> > I admit this all pissed me off at first. But I'm not really capable of
> > holding grudges so I'm not pissed off anymore. Instead I am interested
> > in doing whatever it takes to mend the fences and bring everyone back
> > under a common source tree.
>
> Whatever it takes? Ok. I will vote +1 to include you as a developer for
> Velocity so that you can kill WebMacro. Sounds perfect to me? How about the
> rest of you?
Velocity has a long way to go to catch up to where WebMacro is now, and for
what I'm doing today, I need something that is already there. Not something
that will take another year to materialize.
Note the parser you keep talking about is only 800 lines of code in WM.
It's a small, small piece of the puzzle and you haven't even begun to
work on the rest. Caching is also a very small part of it.
> ---------------------------------------------------------------------
> Let me quote you once again:
> > I don't like the MPL's handling of some things.
> ---------------------------------------------------------------------
Yeah I don't, and I still don't, but I'm not as stuck on the license
as you are. I was trying to create something better. Now I think it's
more important to avoid a code fork. I don't see that as hard to
understand: I'd prefer one code base under an inferior license to
two codebases with one under a better license.
Right now you're really angry. So was I when I first read your post, but
I calmed myself down. We both need to do what is best for the user base
here, and that is *not* two separate implementations.
There's no way, even with my help, that Velocity is going to be where
WebMacro is anytime soon. I'm sorry you had to put resources into it,
but right now the right thing to do is merge and move forward together.
Justin
Re: Velocity, WebMacro, and the future
Posted by Jon Stevens <jo...@latchkey.com>.
Wow Justin, I just keep repeating the same things over and over and over to
you and you continue not to listen. Here is yet another go around of the
repeat for one last time...
on 8/25/2000 11:00 PM, "Justin Wells" <jr...@semiotek.com> wrote:
> I looked at Velocity and I see that you've got it up to about where WM
> was at version 0.30, prior to the first release. Not bad for three weeks
> of work, you've obviously got some competent people working on it. You
> have a long way to go with it though, and all of that work will be
> repeating things I've done with WebMacro.
Yes. All that hard work is repeating things you have done with WebMacro. The
point is to come up with a clone that is available under a non-restrictive
license and isn't encumbered with the baggage that WebMacro carries (which
includes you at this point).
I would also like to go on to say that while you may be ahead in performance
for a short time (mainly because you have caching), you are not ahead in the
parser. There are several things that the Velocity parser is light years
ahead of WebMacro's parser on.
> You should all know that I've worked very hard over the last six months
> to create an open source license that I believed would be better than
> both the MPL and the Apache license, but very much in the same spirit.
> It was intended to be a better license for Java programs than the MPL,
> and one that was simple and easy for average developers to read.
I'm sorry, there isn't anything better than the Apache License as far as I'm
concerned other than maybe the Artistic one (or not one at all). Note that
I'm not counting the warranty disclaimer stuff which you pointed out has
issues in Canada.
> The ASF and Jon have had many concerns with this license over time, but
> I've worked quite hard to remove everything from it that was a concern.
> So far as I know, all of those concerns have been addressed. It was my
> intention to fix any concerns that remained.
The fact that you don't give credit to the rest of the people who also
worked hard to improve your license really bothers me. Also, some of the
stuff that we had to remove from your license was seriously misconstrued and
clearly states your intentions of trying to control the source code. That in
itself scares me because I really don't trust that your license is trying to
do something nasty that might be hidden within it. A lot of what your
license does is very confusing from the point of view that it locks things
down and then opens things up later. That creates a lot of questionable
situations and is exactly one of the problems that the GPL is filled with.
Lastly, I will continue to repeat to you that the ASF is not interested in
the SPL at all and will not dedicate the resources necessary to approve your
license. This is not my decision, this came from the head of the ASF, Roy.
There really is no benefit in approving the SPL given that your software was
so easy to clone.
> However, I'm an extremely busy person and I guess I was too slow in
> seeing this process pushed through to completion.
WRONG! Again, the problem was the fact that you gave me a completely absurd
requirement. Let me quote your private email since you don't seem to be
listening to me at all.
---------------------------------------------------------------------------
Justin said:
> OK, here's a challenge for you. I will put it under a Canadian-ized version
> of one of those licenses if you (hopefully with the help of other apache and
> collab people) convince Altavista.com to use WebMacro. That's not
> as far fetched as it sounds, they've been evaluating it. Having
> Altavista use WM would be worth a lot to me.
I responded:
> I would love to spend a few hours either emailing or talking with the people
> at Altavista.com and do my best to try to convince them that WebMacro is a
> good solution for them. Feel free to introduce them to me.
>
> However, I will NOT agree towards making the license MPL a condition that they
> choose or do not choose to use WM especially since their final decision is out
> of both of our control. That just is not a fair condition.
>
> You either change the license it on your own or you don't change it at all.
> Period.
Justin responded:
> I would compromise and MPL it if doing so got WM onto some site
> like AltaVista, otherwise I want to continue to act according to
> my beliefs.
---------------------------------------------------------------------------
That is the most damaging thing of our entire 5 months of conversations and
I'm going to call you on it.
> I still believe in
> the SPL, but in the meantime it's obvious I should have provided you
> all with an alternative until the SPL is ready for prime time.
>
> I apologize for that.
No, what you should apologize for is your absurd requirement that I help you
get Altavista to use your software in return for an MPL license.
I will repeat to you again that no matter how much effort you put into your
SPL, the ASF is not going to accept it. Roy already told you this and for
some reason you seem to continue to completely ignore that fact.
> There's no doubt that my slowness has resulted in this fork in WM. I
> want to say that it was honestly because I've been busy. If I'd known
> you were so intent on forking, I would have MPL'd it for you prior to
> now. I didn't realize how pressing your needs were I guess, and I'm
> sorry for that too.
Bullshit. The reason for the fork had nothing to do with the SPL, it had to
do with your stupid requirement for altavista.com. Up until you started
playing stupid games with me, I was willing to work with you towards making
things happen with your license. As you can see above, I even gave you the
chance to really define what you meant by getting av.com to use your
software. You had your chance and you decided to screw it up by playing
games. Not cool.
> I admit this all pissed me off at first. But I'm not really capable of
> holding grudges so I'm not pissed off anymore. Instead I am interested
> in doing whatever it takes to mend the fences and bring everyone back
> under a common source tree.
Whatever it takes? Ok. I will vote +1 to include you as a developer for
Velocity so that you can kill WebMacro. Sounds perfect to me? How about the
rest of you?
> There's no way that we could bring Velocity up to where WebMacro is
> in short order. It's got too far to go, and I have pressing deadlines
> with respect to WM. We're rolling WM out on one of the largest sites
> on the net soon, and I just can't throw that all away to go and work
> on Velocity--it's a good start, but only a fraction of the way to
> where WebMacro is now.
I don't agree with that at all. Velocity is working just fine and is quite
functional as is.
> Later this evening I am going to release a snapshot of WebMacro that
> has a huge performance gain over the previous snapshot. My feeling
> these days is that WM has a theoretical performance advantage over
> JSP, and if it isn't already faster than JSP, it will be soon. I
> intend to make WebMacro the highest performance page generation
> system that there is--so that it can work under great tools like
> Turbine to give us all an edge over the non-Java world.
>
> The snapshot will come out under the MPL.
You are encouraging the fork of the code base by doing so because you are
now forcing your users to choose based on different criteria other than
licenses. You refused to do this before there was competition and now you
are suddenly changing your mind.
That is the most absurd reason I have heard for changing a license. I
thought that you had all these beliefs about software and now you are just
throwing that away suddenly? Give me a break!
---------------------------------------------------------------------
Let me quote you once again:
> I don't like the MPL's handling of some things.
---------------------------------------------------------------------
I find that totally silly that you have now decided to MPL WebMacro given
that you don't even like the license. Make up your mind!
Also, the fact that you are putting your code under the MPL vs. the APL
still shows to me that you just don't get it. While the MPL is an acceptable
license, there is an even less restrictive license, the APL which is what I
will continue to prefer (as do plenty of other people). So, putting things
under the MPL still doesn't magically solve the license issues.
Call me a license bigot, but at least I'm open, consistent and clear with my
beliefs.
> I know there are a lot of egos involved here, and some of you have put
> good long hours into Velocity so far. But if we don't mend the fences
> and move forward together it's going to cost us all.
>
> You guys have a difficult choice to make now. I'm putting WebMacro
> under a license you can work with. It's way ahead of where you are. It
> would be bad for us all to fork. We all know that, deep down.
>
> Are we going to bury the hatchet and move the world forward? Or are
> we going to split resources between moving WebMacro foward and you
> reimplementing what I've already done?
It isn't a difficult choice at all. Your MPL of WM doesn't solve any of the
technical issues in WM's parser that Velocity has already solved.
If you decide to MPL (or even APL) WM, that simply means that we can now
pillage your code as needed to increase the functionality of Velocity more
quickly. It isn't nearly as much work to simply use your code to continue to
create a better product. Of course you could do the same with Velocity code.
But I don't care because that isn't an issue for me.
I vote to split resources (you can come work with us if you want though) and
I am continue to work on Velocity with whomever wants to.
I'm totally tired of playing your games.
> I may be difficult to deal with sometimes, mostly owing to how busy I've
> been over the last while. I could really use your help, and I don't
> hold grudges.
I don't hold grudges at all. The issue here is that you had your chance and
you gave it up. End of story. You pushed things this far. The damage is
already done, regardless of grudges. You can't take that back.
thanks,
-jon
--
Scarab -
Java Servlet Based - Open Source
Bug/Issue Tracking System
<http://scarab.tigris.org/>