Via Simon Phipps comes news there is a patent non-assertion clause associated with Adobe’s decision to take the fully open, de jure route to PDF standardisation. This is a maraschino cherry on top of the original news. Duane Nickull apparently turned this update around in double quick time so take a bow, sir (but why are you using that proprietary blogging platform with the bad service levels, I can’t read the clarification?). Open source developers everywhere now largely know where they stand.
Even Sam Ruby with his simple line in the sand definition of open standards.
A standard is one that has multiple, inter-operable, independent implementations. An open standard, at least in the software world, is one where at least one of those implementations is open source.
So in fact according to this country-dividing (its a line in the sand thing, TE Lawrence and all that) definition Adobe PDF isn’t open. Yet. But then that isn’t Adobe’s problem.
If you want to build an open source implementation have at it, open source developers. There appear to be no reason why you and PDF should not be wed… On that note however Andrew Shebanow from Adobe’s blog is quite interesting. While I tend to agree that “history will view this announcement as an historic one for Adobe and even for the software industry as a whole,” its in the comments after a question from Rosyna that the shoe drops.
Who is going to build a test suite for PDF conformance, and what are the trademark and certification issues? There are still many many governance issues to consider.
Oh yeah – while I am handing out the sweetness and light the news that developers are going to be able to use Flex Data Services directly from Javascript without needing Adobe Flex application clients makes a hell of a lot of business sense, and helps bring us towards The Synchronised Web, where data is data, and access methods dont matter.
disclaimer: Duane is a good friend of mine and unfairly gave me some credit for Adobe’s standardisation decision. So I am well disposed these folks… Adobe is not a client but it soon will be. Any analyst likes to feel listened to. It certainly helps that its almost like my impassioned plea for more business openness was being listened to… 😉
James Ward says:
February 1, 2007 at 8:41 pm
Hi James,
There are numerous Open Source libraries that read and write PDF. I use a number of them on Linux – Evince most frequently to view PDFs.
I don’t totally buy that definition of Open Standards though. While I agree that Open Standards work best when there is an Open Source RI (see Geir’s recently blog on this), there are/were plenty of “Open Standards” without Open Source implementations. With the exception of the Lynx browser (which no one really used at the time), before Netscape went Open Source, were there any Open Source Browsers? Would this have made the W3C’s HTML standard not an Open Standard?
Geir’s Blog:
http://blogs.codehaus.org/people/geir/archives/001507_another_example_of_why_open_source_ris_for_specs_is_a_good_thing.html
-James
Luis Villa says:
February 2, 2007 at 4:48 am
A full open source community pdf implementation seems unlikely; lots of open source apps implement partial pdf support (abiword, evince/kpdf, OOo all come to mind), but I’d guess that most people just don’t find pdf interesting enough to justify implementing the entire spec. If it does happen, it’ll be from someone like Novell or IBM who has a corporate itch to scratch.
Andrew Shebanow says:
February 3, 2007 at 4:49 pm
I think that at this point there are enough implementations of PDF to qualify under Sam Ruby’s definition. He doesn’t say that all those implementations must be open source, just one of them. I personally think iText qualifies – I’ve used it myself before I came to Adobe.
As for conformance suites, I agree this is an area where more work is needed. That is why I brought up the issue in response to Rosyna’s question. I personally look forward to seeing a good conformance solution come out of the ISO effort.