[Openicc] Fwd: Settings bundles (Robert Krawitz, Edmund Ronald)

[edmund ronald]

Actually, we expect the format to be parsed - and subtrees copied and
pasted - by humans like YOU. But using XML doesn't mean that it has to
be as horrible as what you describe.

I'll ask Robert to paste a sample of today's XML curve representation
here so you can judge for yourself.

Edmund

On Fri, Feb 18, 2011 at 6:36 PM, Chris Murphy <lists at colorremedies.com> wrote:
> For what it's worth, I am increasingly not a huge XML fan due in no small part to CxF. It is a nightmare to sort out visually. It's like reading code that pretends to be plain English readable. It's far worse than a CGATS file to a mortal. I understand the need for extended beyond basic CGATS files, but CxF is user hostile to look at inside.
>
> I'm seeing files that repeat the date of each measurement 10000 times. There are 5000 patches represented in the data set, which have 5000 reference device values, each with date/time stamp, and 5000 measurement values, each with a time stamp. The word cyan is repeated 10000 times instead of once in a single column. The word tag is repeated 25000 times. It such a complete mess to sift through. I guess fully 90% of the content of the file is repetitive, extraneously, useless garbage, visually speaking.
>
> And the measurement data is not co-located with the device data. The top 1/2 of the CxF contains LAB values, the bottom 1/2 the device values. On and on. It's just nasty. It is definitely not easier than a CGATS file if you want something human readable. These particular XML files are only slightly more transparent than a binary ICC file.
>
> So if the idea is for these files to be readable by users of any sort (non-programmers) then I'd reject XML. If the idea is for the files to be more easily parsed by programmers for troubleshooting and engineering purposes, then the format doesn't matter to me much, and XML is fine.
>
>
> Chris Murphy
>
> On Feb 16, 2011, at 7:50 PM, Ann L McCarthy wrote:
>
>> Edmund,
>>
>> XML is my vote.
>>
>> Best regards,
>> Ann McCarthy
>> Imaging Systems R&D
>> Lexmark International, Inc.
>>
>>
>>
>>
>>
>> edmund ronald <edmundronald at gmail.com>
>> Sent by: openicc-bounces+almccart=lexmark.com at lists.freedesktop.org
>> 02/16/2011 02:51 AM
>> Please respond to
>> Open ICC Color Managment <openicc at lists.freedesktop.org>
>>
>> To
>> Open ICC Color Managment <openicc at lists.freedesktop.org>
>> cc
>> Subject
>> Re: [Openicc] Fwd: Settings bundles (Robert Krawitz, Edmund Ronald)
>>
>>
>>
>>
>>
>> Graeme,
>>
>> Yes we will doubtless provide such a facility which interfaces with Argyll.
>> We need to create a data format in which we will store *our¨*
>> (Gutenprint) settings.
>> Our question is XML or JSON? We believe that third party programmers
>> may find it easier to write little programs that run in Javascript in
>> a browser, and maybe have a graphic interface, rather than do C or
>> C++..
>>
>> Edmund
>>
>>
>>
>> On Wed, Feb 16, 2011 at 5:23 AM, Graeme Gill <graeme at argyllcms.com> wrote:
>>
>> > An Argyll linearisation tool (printcal) and the resulting curves has been
>> > available and have been used for some time. If you want to provide a
>> > facility
>> > to import such curves into a printing system, or (better yet) use them
>> > directly, it would not be hard to do so.
>> >
>> > Graeme Gill.
>> > _______________________________________________
>> > openicc mailing list
>> > openicc at lists.freedesktop.org
>> > http://lists.freedesktop.org/mailman/listinfo/openicc
>> >
>> _______________________________________________
>> openicc mailing list
>> openicc at lists.freedesktop.org
>> http://lists.freedesktop.org/mailman/listinfo/openicc
>>
>> _______________________________________________
>> openicc mailing list
>> openicc at lists.freedesktop.org
>> http://lists.freedesktop.org/mailman/listinfo/openicc
>
> _______________________________________________
> openicc mailing list
> openicc at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/openicc
>