Re: [htmltmpl] Free Htmltemplate: Java implementation of HTML::Template

[<hniksic@xe...>]

Philip S Tellis <philip.tellis@...> writes:

> Sometime on Feb 18, Hrvoje Niksic assembled some asciibets to say:
>
>> Allow me to announce the availability of a Java implementation of
>> HTML::Template.  The home page is at:
>> 
>>     http://htmltemplate.inet.hr/
>
> Uh, there already was a java version of html template at 
> http://html-tmpl-java.sourceforge.net/
>
> why start a new project?

It's not new and, more importantly, it didn't start out as a fork.
When I started writing Free Htmltemplate in 2002, I wasn't originally
aware of your version.  I even announced a release back in 2002:

    http://www.geocrawler.com/archives/3/23294/2002/10/0/9958609/

Later I did come across your HTML.Template, but I found that I
preferred some of the design and implementation choices I've made.
I'll list some examples that convinced me to stick with a separate
implementation:

* HTML.Template didn't seem to offer a caching facility.  One of the
  strong points of Free Htmltemplate is the ability to request
  transparent caching of compiled templates (including optimizations
  like blind cache), but also to simply retrieve the CompiledTemplate
  object and use it later.

* Free Htmltemplate is written with multithreaded environment in mind
  from the start.  For example, the transparent caching code is
  careful to synchronize accesses to its data structures, while making
  sure that synchronization doesn't degrade into serialization through
  long-standing locks.

* I had different ideas of what the translation of HTML::Template's
  API to Java should look like.  For example, I don't have methods
  that accept Object[] arguments that closely follow Perl's subroutine
  call model.  The Free Htmltemplate approach seems more "Java-like"
  to me, while retaining the spirit of the original API.  This is
  surely deep in the realms of personal taste, but I invite you to
  take a look and judge for yourself.

* I used the new API's such as Collections, Iterators, etc. and did
  not want to revert to Hashtables, Vectors, and Enumerations.  This
  is a matter of taste as well.

* My simple-minded benchmarks showed that my implementation was
  faster.  This might simple be a consequence of the caching.

None of this is meant as an argument that my program is in any way
"better" than yours, just as an explanation of why I didn't just pick
up your version when I became aware of it.

Re: [htmltmpl] Free Htmltemplate: Java implementation of HTML::Template

[<hniksic@xe...>]

Sam Tregar <sam@...> writes:

> If I wanted to run some benchmarks with your code what JVM should I
> use?

Simply use the latest revision of Sun's 1.4.x.  That's what I use.

Re: [htmltmpl] Free Htmltemplate: Java implementation of HTML::Template

[Sam Tregar <sam@tr...>]

On Wed, 18 Feb 2004, [iso-8859-2] Hrvoje Nik?i? wrote:

> Good point, I really should back up such a claim with numbers.  I
> originally didn't go out of my way to make Htmltemplate very fast, I
> simply coded it the "obvious" way.  But then I did some benchmarks
> and, to my surprise, found it to be a good deal faster than I
> expected.  The benchmark was admittedly *very* simple-minded: take a
> template, compile it once, and run the template transformation over
> and over to see how many transformations per second I can squeeze.

What kind of complexity did you put into your test template?  I've
found that deep loops with plenty of IF/ELSE logic are a good way to
stress HTML::Template.

> The code generated by HTML::Template::JIT still contains calls to
> Perl's library and object system, doesn't it?  To me that's quite
> different from typical hand-written C code, which I'd expect to always
> beat Java by a wide margin.

It makes a few, mostly to get param values from SVs, but it makes a
lot fewer than you might imagine.  The template text itself and all
the branching and looping logic is compiled into pure C code.

> How do you benchmark HTML::Template::JIT?  I can try to repeat the
> same benchmark with Free Htmltemplate to see if I still get the
> speedup I got when I last compared them.

I use a script called benchmark.pl in the scripts/ directory of the
HTML::Template::JIT distribution.  It uses Benchmark.pm in a fairly
standard fashion.

If I wanted to run some benchmarks with your code what JVM should I
use?  I think I have some version of Sun's 1.4 JRE installed but I
seem to remember something about it not being very fast.

-sam

Re: [htmltmpl] Free Htmltemplate: Java implementation of HTML::Template

[Philip S Tellis <philip.tellis@gm...>]

Sometime on Feb 18, Hrvoje Niksic assembled some asciibets to say:

> Allow me to announce the availability of a Java implementation of
> HTML::Template.  The home page is at:
> 
>     http://htmltemplate.inet.hr/

Uh, there already was a java version of html template at 
http://html-tmpl-java.sourceforge.net/

why start a new project?

Philip

-- 
All men know the utility of useful things;
but they do not know the utility of futility.
		-- Chuang-tzu

Re: [htmltmpl] Free Htmltemplate: Java implementation of HTML::Template

[<hniksic@xe...>]

Sam Tregar <sam@...> writes:

> On Wed, 18 Feb 2004, Hrvoje Niksic wrote:
>
>> Allow me to announce the availability of a Java implementation of
>> HTML::Template.  The home page is at:
>>
>>     http://htmltemplate.inet.hr/
>
> Wow, great work!  I love the website and your excellent
> documentation.

Thanks!

> I do, however, have a bone to pick.  You claim your system is much
> faster than HTML::Template and HTML::Template::JIT:
>
>    Free htmltemplate is significantly faster than HTML::Template in
>    rendering pre-compiled templates. Even with the JIT in place,
>    which according to Sam speeds up template rendering by a factor
>    of 4 to 8, htmltemplate might still be as fast or faster.
>
> Can you back this up with numbers?

Good point, I really should back up such a claim with numbers.  I
originally didn't go out of my way to make Htmltemplate very fast, I
simply coded it the "obvious" way.  But then I did some benchmarks
and, to my surprise, found it to be a good deal faster than I
expected.  The benchmark was admittedly *very* simple-minded: take a
template, compile it once, and run the template transformation over
and over to see how many transformations per second I can squeeze.
The result was astounding 3000 renderings per second, which was
totally beyond my expectations.  I first thought that I did something
wrong, that the code wasn't even running, but adding the output showed
that I was indeed producing the intended output 3000 times in a row in
about 1 second.

If my memory serves me, the same kind of benchmark showed
HTML::Template performing about 300 renderings per second.  The
reasoning that follows is summed up by the above paragraph: if JIT
speeds up HTML::Template 8 times, Htmltemplate might still be faster.
But I really haven't checked it.

I guess I should reword that paragraph to make it clear that it's
based on speculation and not on actual measurements of
HTML::Template::JIT's performance.

> Java has improved a great deal, granted, but I'd be pretty shocked
> if it could beat the highly-optimized C code generated by
> HTML::Template::JIT.

The code generated by HTML::Template::JIT still contains calls to
Perl's library and object system, doesn't it?  To me that's quite
different from typical hand-written C code, which I'd expect to always
beat Java by a wide margin.

How do you benchmark HTML::Template::JIT?  I can try to repeat the
same benchmark with Free Htmltemplate to see if I still get the
speedup I got when I last compared them.

Re: [htmltmpl] Free Htmltemplate: Java implementation of HTML::Template

[Sam Tregar <sam@tr...>]

On Wed, 18 Feb 2004, Hrvoje Niksic wrote:

> Allow me to announce the availability of a Java implementation of
> HTML::Template.  The home page is at:
>
>     http://htmltemplate.inet.hr/

Wow, great work!  I love the website and your excellent documentation.
Feel free to make announcements of new versions on the list.

I do, however, have a bone to pick.  You claim your system is much
faster than HTML::Template and HTML::Template::JIT:

   Free htmltemplate is significantly faster than HTML::Template in
   rendering pre-compiled templates. Even with the JIT in place, which
   according to Sam speeds up template rendering by a factor of 4 to
   8, htmltemplate might still be as fast or faster.

Can you back this up with numbers?  Java has improved a great deal,
granted, but I'd be pretty shocked if it could beat the
highly-optimized C code generated by HTML::Template::JIT.

-sam

[htmltmpl] Free Htmltemplate: Java implementation of HTML::Template

[Hrvoje Niksic <hniksic@xe...>]

Allow me to announce the availability of a Java implementation of
HTML::Template.  The home page is at:

    http://htmltemplate.inet.hr/

I apologize in advance for the off-topicness of this announcement.
Questions and discussion should be directed to me, unless they also
concern HTML::Template.  New versions will not be announced here, the
only purpose of this announcement is to alert people to the existence
of another HTML::Template implementation.

If you are forced or prefer to work in Java and would like to use
HTML::Template templates from your Servlets, this might be the thing
for you:

* Free Htmltemplate's API is designed to be familiar to users of
  HTML::Template, although the naming is more suited to Java.

* The template syntax is intended to be 100% compatible with
  HTML::Template.  Where differences exist, they are documented -- see
  http://htmltemplate.inet.hr/deviations.html .

* Free Htmltemplate is currently licensed under the GNU GPL.  If you
  like it and would like to use it in commercial projects, let me know
  and I'll consider changing the license to LGPL or to something
  BSD-like.

Please note that Free Htmltemplate has nothing to do with the
HTML.Template written by Philip Tellis, also in Java.