Thanks for the advice Robin.  The main scalability constraints I am
finding are:

1) Performance - from benchmarks and real world testing approx 100
requests per second give or take is the mean.  From what I've read Axkit
has much greater throughput.

2) Extensibility - Mason is a nice app server package but some parts of it
(i.e. error handling) have become intimately tied to the Mason structure
in the latest branch.  Also most of Mason's functions are tied to a
request, so if I want to manipulate a template from within perl-space
(i.e. on a cron basis have a script evaluate a template's contents) I have
to generate a request object and component stack.  I'm hoping that will be
simpler with axkit.

3) XML compatibility - yes I could use mason as a filter like you
suggested but I'm still limited by the performance constraints of the
component structure.  I'm hoping to use server-side includes or something
similar with Axkit to emulate this kind of template structure - I'm sure
it will be more work but I'm at the point where I need to process
thousands of requests per second as opposed to hundreds and integrate with
other vendors using xml standards.  Axkit seems to be the 'big kids' Mason
equivalent.

Thanks for your help!  I'm sure I'll have more questions as time goes on.

Regards,

Fred Moyer



> Fred Moyer wrote:
>> I have a large mod_perl application which uses perl modules for the
>> application and object processing and relies on HTML::Mason for
>> content delivery and templating.  Needless to say I am starting to hit
>> Mason's scalability limits and am considering replacing Mason with
>> Axkit and porting all the templates.  From what I have learned so far
>> I will need to use either XSLT or XPath for the templating and
>> delivery needs.
>
> AxKit isn't really complicated, it's just that there's a lot of stuff
> (same goes  for XML). I'd say step one is to get a feel for every option
> from 30.000 feet  and then to figure out what you need to get started
> and which options sound  saner to you (knowing that you can look into
> other options later). To be  perfectly honest, it is very difficult if
> not impossible to tell someone else  how they should be using AxKit, it
> really is TIMTOWDI :)
>
> Reading the Wiki (http://axkit.org/wiki/view/AxKit/DefaultPage) if you
> haven't  already is absolutely vital.
>
> There are several ways to generate content (XSP, Providers) and several
> ways of  transforming content for display (XSLT, XPathScript) with
> varying degrees of  dynamicity.
>
> Coming from a Mason background, you may feel more at home with
> XPathScript.  However, it may be that the similarities it bears with
> Mason's templating system  are precisely what you want to get away from.
>
>> Can anyone here who has undergone such a migration offer some advice?
>
> I migrated in a brute-force manner (ie rewrote a lot), quite a long time
> ago  now. I can basically split my former Mason content in two parts:
> the bits where  the code was mostly separated into another module with
> the page only calling the  API, and the bits where the pages had way too
> much code in them.
>
> For the former the switch is easier. You can wrap the same modules into
> an XSP  taglib (using TaglibHelper or SimpleTaglib) in order to generate
> an XML  structure corresponding to your data, and style that with XSLT.
>
> For the latter it is less fun, you may have to refactor quite a bit.
> What I did  was to keep Mason running and port segments of the sites one
> by one, time  allowing. Note that Mason can be used as an input to AxKit
> (using the filter  provider) so that you could modify the HTML in your
> Mason pages to be XML, and  filter that through AxKit to style it with
> the XSLT that styles your pure-AxKit  stuff.
>
>> xml schema definitions
>
> *bzzzzzt* It's very likely that you will never need those :) Doing
> without them  will make your life a *lot* easier.
>
>
> PS: If you don't mind telling us about the "scalability limits" you've
> found in  Mason that are inciting you to switch to AxKit, I'm sure
> people on the list  would be glad to hear about that.
>
>
> --
> Robin Berjon <[EMAIL PROTECTED]>
> Research Engineer, Expway        http://expway.fr/
> 7FC0 6F5F D864 EFB8 08CE  8E74 58E6 D5DB 4889 2488
>
>
> --------------------------------------------------------------------- To
> unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]





---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to