The Paloose project is currently run by me, . I recently retired after a (too) long career designing (amongst other things) computer hardware and software; I am now a professional music engraver, something I did as a hobby during my working life. I started Paloose to keep myself amused and to support my other interest, .

I have always been interested in literate programming ever since Donald Knuth created the first weave/tangle WEB system for producing TeX and Metafont. AntWeave is an implementation of the NoWeb literate-programming tool originally designed by Norman Ramsey. I have used NoWeb on many projects in the past to great success and recently wanted to revive my use of the system. Rather than use a standalone program (as the original one NoWeb does) it is based on a Java and XML/XSL approach and runs as a Task within the Ant build system. By using this approach it is completely universal on any system that supports Ant (and by implication Java and Java-XML). It removes the necessity to build system specific version of the original and processors; which can be a tricky process when starting from scratch without the required binaries.

The main AntWeave site is where you can .

It can be also downloaded from . Feel free to contribute — please contact me if you want to.

The original name I chose was "Papoose"; in the UK this is defined as "". Unfortunately it does not seem have this meaning in the US, indeed, with some people, it is a derogatory term. However, it occurred to me that a simple change of letter would give "Paloose", which is another term for the Appaloosa horse.

This is something dear to my heart as, until recently, I was privileged to have one of my own ('Zitty') who I rode until she reached the grand old age of 29 and then became a "house pet" until she died aged 33. I now have a fine American Quarter Horse named "Fred" (born 2005, his dad is from Pawnee, Oklahoma) who is coming along nicely as a promising Western horse.

Paloose is released under the GPL-3. If anyone wishes to get involved then please me. I intend to use it for all my small sites and will continue to do so until Cocoon is supported by the smaller ISPs that I use. It is also interesting (and slightly depressing) that the latest Cocoon does not seem to be as easy to use as Cocoon Version 2.1.x. I use this latter version on my local servers and have no plans (yet) to upgrade to Cocoon Version 2.2 or later. While I am prepared to admit that this is my problem, I know others who have had the same experience. As a result perhaps Paloose will fulfil the requirements of those who want a quick Cocoon-like engine for their small Web sites.

You should have received a copy of the GNU General Public License along with Paloose when you downloaded it. If not, see

This is quite a difficult question as I have limited time and am not trying to do a complete Cocoon implementation in PHP5 — a pointless exercise anyway. Now that I have added a simple (limited) data caching scheme to the pipeline all of what I intended to add is complete. I am gradually refining the code and improving the documentation and comments.

Yes please. I suggest the following or the . Scale it as you feel fit.

Why don't you use SAX, like Cocoon, instead of DOM?

I thought about this quite hard initially. I concluded that using a DOM approach was far simpler. It has the penalty of being slower but Paloose is only intended for small volume/traffic sites and ultimate performance is not such an issue. I have seen no reason to revise this decision in the light of typical site usage. Changing to a SAX approach would mean a major code change and I really do not have the time at present as my Music Publishing venture () and my horses have more call on my time.

Assuming that you are already using the caching system and want more speed, it is possible, but at the expense of a little effort (and possible obscurity), to tailor the Paloose code and your own sitemaps and XSL. The implications and method are described on a in more detail.

I have done considerable work on various ways of caching within Paloose and versions after 1.3.0 use pipeline data caching on certain generators and transformers. I confess that I am in a quandry over this as the gains from using a cache seem to be minimal and not really worth the effort except in some key cases — see for more on this. It is not the end of the story, but more performance increases will probably come from elsewhere.

It is useful to provide some background to some of the decisions that I made for flows and forms. This is where Paloose diverges from Cocoon most. The latter is based on Java and Javascript which was not available to me (conceptually, not practically, as Paloose is based solely on PHP5). Thus I had to base what I did on a pure PHP approach while including the concepts of Cocoon flows and forms. I made several false starts which arose from several design problems:

All of these problems seemed intractable at first with decisions made about one influencing, detrimentally, a decision made elsewhere. Mirroring the Cocoon forms template, was initially desirable for commonality with Cocoon. But it soon became apparent that this was going to make the whole approach far too complicated for what I wanted. Much code was wasted in exploring this but I believe it was the right decision. In the end I based the Paloose forms () on the JXForms that older version of Cocoon used. (I had produced some sites with this in the past so I was reasonably familiar with it). is not radically different from JXForms but it does not slavishly follow the latter. I believe that is suitable for the restrictions I had set on Paloose and, as I have said elsewhere, if you have a site that requires all the facilities of then you should probably be using Cocoon anyway.

Once I had settled on there was the problem of how to implement the flow script, which in Cocoon uses Javascript. The Cocoon approach is to take a "snapshot" of the Javascript engine (a simplistic way of describing it) in order to maintain continuity between client requests. I was unable to find a sensible way of doing this with PHP5 (which does not mean to say that one does not exist). So I had to find a means of providing the user with a simple method of continuations based solely on a PHP5 approach. I believe I have while keeping to the spirit of Cocoon.

All these solutions, however successful, have made me diverge from the strict Cocoon approach so please read the documentation very carefully.

These have been done by my son who is the GUI art director for and I am very grateful to him for the help he has given me in their design. However I have to take the blame for the Paloose site: any good points are due to him, the bad ones are all mine.

The addition of a simple transformer is all that is necessary; it is explained in more detail on .

How do I port Paloose Sites to Cocoon?

How do I port Cocoon Sites to Paloose?

This is really just the reverse of the above. The only issue might be one of performance so it is worth bearing in mind that Paloose is obviously slower than Cocoon.

I have written an example component that is a "template" for writing generator, transformer and serializer components; also included is an example exceptions class used by the template. Current releases of Paloose include it in the folder . Also have a look at the various existing components that make up Paloose.

Yes. The can detect the iPhone to allow separate pipelines to be run.

The usual problem here that you have not set up the file up correctly. Remember that Apache will serve all your requests, including ones that are destined for Paloose unless you tell it otherwise. For example if you want requests for, say, , to be processed by Paloose then you would have to add the following line into your file:

in the file will cause an infinite loop.

This means you are still running PHP4. You need to run PHP5. Speak to your ISP to provide PHP5. The other reason may be that your file has not been set properly to use PHP5.

See , this is the usual problem. Also make sure that you do not overwrite a system that is required. That is why it is best to have your site in a separate folder.

Yes the basic schema is written in . They are not an essential to running Paloose or understanding, they are merely given out of interest; although I would urge the use of schemas to reduce errors in your XML files.

They are pretty versions of the raw RELAX NG which are easier to read. If anyone would like a copy of the transforms necessary to produce the pretty-printed version please and I will consider putting them up as a download.

This is a fairly contentious subject. Sitemaps are built in such expandable form that producing a schema in something like or is very difficult — they simply do not have a rich enough structure to do what I think is necessary without having to use Schematron additions. The closest that I got to producing one used . I wrote an extensible set of schemas which worked fine, but are now sadly out of date. I also produced schemas for Dublin Core, Jakarta Ant and RDF (although RDF is a major problem because it is impossible to write a universal validating XML schema for it — it ain't the right sort of grammar and the current 2004 specification for RDF is flawed), all of which can be found on the home page. DSD-2 is the best schema language we never had, sad ....