Re: [xml-dev] Re: [docbook-apps] Small!! Lightweight!! xslt processorwhi

[Dimitre Novatchev <dnovatchev@gmail.com>]

> If all you want to do is what you did with "4MB Machines" you can still do that today in 4MB ...
> Oh you want Video , and full HTML5 styling, cross referencing against a GB of database and a fancy GUI ... oh well ...
> Blame the programmers.
> Expectations of data size, processing and output change ... but not the processing power to support it ... funny.

Even the Curiosity Rover has 256MB RAM [1]... so running XSLT on a
Mars robot is just a matter of time. What more demanding application
than this could there be?


[1] http://www.theregister.co.uk/2012/08/07/curiosity_software_upgrade/

Cheers,
Dimitre.

On Fri, Aug 17, 2012 at 10:35 AM, David Lee <dlee@calldei.com> wrote:
> Agree completely.
> The statement "IT was done with ..." and "doing the same"  is entirely meaningless out of context.
> Has this alleged person looked at the footprint of modern OS's which "do the same" as say DOS did with 640k?
> Is he asking for exactly the same data and same output and same programmer work effort as before ?  I doubt it.
>
> If you are in a memory and CPU constrained environment some tools and techniques don't work well.
> Welcome to the real world.   As a Mobile developer for a decade or so I can attest that back  in the "Good Old Days" of Palm Programming we had to make do with not only 4MB of memory (storage AND RAM) but also a processor speed in the glacial era.
> So no we didn't do XSLT or HTML for that matter.  But we did well with a customized binary XML encoding scheme and custom display rendering software.
>
> Its all a tradeoff.   Want easy to use and debug languages ? You need some CPU and memory.
> Constrained by CPU and Memory ? write a million lines of C code and limit your expectations to that which can be done without the document being fully read into memory.  And be willing to pay the programmers to work a LOT harder for a LOT longer.  Oh couple that with a backend DB infrastructure that normalizes the data efficiently for the device so you DONT have to read the whole document into memory to process it.
>
> If all you want to do is what you did with "4MB Machines" you can still do that today in 4MB ...
> Oh you want Video , and full HTML5 styling, cross referencing against a GB of database and a fancy GUI ... oh well ...
> Blame the programmers.
> Expectations of data size, processing and output change ... but not the processing power to support it ... funny.
>
>
>
>
>
> ----------------------------------------
> David A. Lee
> dlee@calldei.com
> http://www.xmlsh.org
>
> -----Original Message-----
> From: Jirka Kosek [mailto:jirka@kosek.cz]
> Sent: Friday, August 17, 2012 1:23 PM
> To: Dan Shelton
> Cc: Jeff Chimene; DocBook Apps; xml-dev@lists.xml.org
> Subject: [xml-dev] Re: [docbook-apps] Small!! Lightweight!! xslt processor which is standalone!! and runs Docbook/XSL stylesheets?
>
> On 17.8.2012 16:31, Dan Shelton wrote:
>
>> A colleague with two Dr. (Med/IT) and one Prof. in IT already called
>> XML and XSLT "a failure" because the processing requirements have
>> become insane - IT was once done with 4MB machines, doing the same
>> with today's machines and XML/XSLT goes up to 400MB as minimum. And
>> admittedly, I have no arguments to prove him wrong - XML processing
>> takes a lot of memory (why?) and XSLT processing is... eating memory.
>> Lots of memory. There doesn't seem to be a "small" solution.
>
> I wouldn't say this is failure of XML/XSLT. Most users are not so
> tightly constrained by memory so they are not pushing very hard for
> memory economical implementations. So this is particulary failure of
> implementations and partialy failure of users who are not demanding more
> and who are not prepared to pay for more.
>
> What's quite surprising that very efficient Java implementations like
> Saxon can outperform many C-based implementations both in terms of
> memory usage and performance.
>
> However if your main objective is something small and C based you can
> give a try to Sablotron (https://sourceforge.net/projects/sablotron/)
> and Xalan-C (http://xml.apache.org/xalan-c/). It has been while since
> last time I compiled them from source (more then decade) but they
> definitively should have less then 80 MB of dependencies.
>
>                                 Jirka
>
> --
> ------------------------------------------------------------------
>   Jirka Kosek      e-mail: jirka@kosek.cz      http://xmlguru.cz
> ------------------------------------------------------------------
>        Professional XML consulting and training services
>   DocBook customization, custom XSLT/XSL-FO document processing
> ------------------------------------------------------------------
>  OASIS DocBook TC member, W3C Invited Expert, ISO JTC1/SC34 member
> ------------------------------------------------------------------
>



-- 
Cheers,
Dimitre Novatchev
---------------------------------------
Truly great madness cannot be achieved without significant intelligence.
---------------------------------------
To invent, you need a good imagination and a pile of junk
-------------------------------------
Never fight an inanimate object
-------------------------------------
To avoid situations in which you might make mistakes may be the
biggest mistake of all
------------------------------------
Quality means doing it right when no one is looking.
-------------------------------------
You've achieved success in your field when you don't know whether what
you're doing is work or play
-------------------------------------
Facts do not cease to exist because they are ignored.
-------------------------------------
I finally figured out the only reason to be alive is to enjoy it.