On woensdag 3 augustus 2016 11:01:47 CEST Kael Shipman wrote: > Haha, sorry to have so deeply offended you with my optimism! I'd love to > leave all your dark storm clouds hovering over the scene, but I think > you're wrong. > > Answers below. > > On 08/03/2016 09:29 AM, Adam Tauno Williams wrote: > >> As for marketing, I'm still mashing this around a bit. Most > >> interesting to me is that in a future where everything is open, > >> "marketing" would serve to unify products, rather than divide them. > >> In other words, you wouldn't have GnuCash vs Quickbooks -- you would > >> GnuCash AND Quickbooks, each providing a unique interface over a > >> common, standardized data storage mechanism for extended business > >> data, of which financial data is a part. > > This will never ever ever happen; it is not a viable option, it will > > never ever happen. We have such abundant examples of this as a NON > > -OPTION I am baffled how it still gets floated as a thing. > > > > Quickbooks, Google, et al have no-zero-nada-zilch motivation to > > participate in such a model, and to bear the extra tedium, cost, and > > infringement on product development or time-line. > > > > This is fantasy land thinking - full stop. It imagines that somehow > > the underlying data model is irrelevant, which is FALSE, and is > > recognized by any software development with any real life experience. > The dynamics of the Android platform are actually a wonderful example of > most of what I'm talking about. Think about the environment that the > Android platform has enabled: programs are encouraged to use other > programs' functionality through Intents, rather than code functionality > themselves (this is microservicing architecture in action, albeit an > elementary implementation of it). Programs can also act as "Content > Providers" in the public space, laying the foundations for common data > access. Most importantly, programmers can /choose/ to use these > features, or choose not to, and slowly, we're seeing that they actually > do prefer to use them (despite the fact that each programmer is seeking > to monetize only his or her app). > > I don't deny that this idea has been popular in fantasy for a long time. > My argument is more one of context: Facebook and MySpace were > fundamentally equivalent (and they both had a much earlier ancestor, > Geocities, which was also more or less functionally equivalent), but the > context in which Facebook eventually flourished was different, and > decidedly right. > > In the same way, I think every idea has its context, and frankly, you're > right about one thing: 2016 is not the year to expect GnuCash and > Quickbooks to work together. But 2026 is, and if we can eventually agree > on that, then the early adopters among us can begin to lay the > infrastructure (again, funding pipelines, data standards, etc.) that > will eventually lead to such a world. > > >> Because collaboration infrastructure would provide very low barriers > >> to standardization > > No, it does not. > The problem is not that the infrastructure doesn't work, the problem is > that the infrastructure doesn't exist. What's been probably the most > surprising fact about virtual communities to me is that you can scream > at the top of your lungs without anyone ever hearing you. While we're > highly connected right now, visibility and penetration are extremely > low. In other words, it's possible to have great ideas without anyone > ever knowing about them! This seems almost incomprehensible, given our > connectivity, but it's painfully obvious when you look at how the system > works, and it's a symptom that suggests an infrastructural problem. > (Mass communication infrastructure like TV, radio and eventually the > internet itself were the first step to solving this problem. The next > step is to build tools /on top of/ these communication channels that > work to address the issue of idea visibility. Google, unfortunately, > falls short, because it can't yet separate discrete ideas from the CMB.) > > > >> and because culture would provide high pressure to comply > > No, it does not. > Asana, Trello, Wunderlist, Basecamp.... these are all for-profit task > managers that have mature APIs that allow users to cross system > boundaries. It's obvious that very soon, it will make more sense > economically for them to develop common back-end protocols, rather than > maintaining unique APIs that overlay distinct back-ends. They could all > just as easily have continued in the vein Apple's system lock-in, but > they didn't. > > > >> (already indicated by the Convergence of the Web Browsers and the > >> availability of APIs on paid services like the above), > > Note: "APIs", plural. > For those of us with 5-minute attention spans, this is a big deal. The > rest, however, can recognize that change takes time and work (and help: > this is where an external organization like IEEE or ISO could step in > and help to unify things). The existence of multiple APIs is equivalent > to the use of the "Adapter" pattern in program design, and seems related > to the lack of visibility that I spoke of earlier. > > > >> a world where programs are siblings (i.e., interchangeable units) in > >> a "functionality hierarchy" seems perfectly conceivable to me. > > Such a world is unadulterated fantasy. We've been talking about that > > since the early 1990s. Interoperability is very very hard beyond some > > lousy least-common-denominator that makes nobody happy. > > > > Notice that the world is trending exactly away from inter-changeable > > components to emphatically more monolithic and proprietary systems: > > Office365, GMail+GoogleDocs+GoogleDrive, ... all tightly integrated > > ***within themselves***. APIs/Hooks provided for whatever bit of glitz > > you want/need to hang off the side. > > > > To define "the world" in terms of Microsoft is a joke that people got > tired of laughing at a decade ago ;). You're right that as of today, > there are good examples of traditional monolithic systems, but their > days are numbered. Just this week, there's been a flurry of activity on > the OpenCloudMesh mailing list for the development of the underlying > protocols for Federated Cloud Sharing -- a HUGE step toward unifying > cloud back-ends, including Google Drive. (Incidentally, they seem to > have raised the necessary funds for this development in 5 days flat -- > an inspiring and suggestive achievement.)
Sadly, often only underdogs care about interoperability. The Federated Cloud Sharing protocol was a great idea and I'll soon do a blog about where we'd like to take federated sharing. Bjoern & Frank have been working on a global, optional lookup-server which imho really brings it to a whole new level ;-) So Pyd.io joined and perhaps, some day, Seafile would. Or one of the small proprietary ones. But Google? Microsoft? What incentive do they have to support a protocol which breaks their lock-in on users? I think that this is the issue: Microsoft and other large companies will support standards when they are in the minority (browser wars for example - MS is now touting standards support, something they cared little about in the IE5 and 6 times). They will not want open standards when they own the market and do what they can to kill them (Office - openxml vs ODF). > And as an aside, I think it's interesting and positive to note that > Microsoft document formats have at least been published, semi-open > standards since 2007. While they clearly remain committed to the old > world, they're at least recognizing the business advantage that open > standards represent. Hell no, that whole format was nothing more than a strategic move to prevent a REAL open format to be forced upon them. Lots of governments, esp in Europe and South America, had decided they wanted documents to be stored in an open format; and the ODF format was on track of becoming that format. Microsoft wanted to stop that in its track and they did that by releasing openXML, a open standard to the extend that marketing goes but, in reality, so full of undefined or referring-to-closed-non-standards format it is not much easier to support than the old binary formats. > Finally, I do agree that the "functionality hierarchy" I spoke of has > been in the dreams of programmers since the beginning of time, and out > of reach for just as long. Remember, though, we spent 500 years trying > to figure out how to fly, and now it's a given. Impossible is nothing :). The markets weren't against flying. Here, you ARE fighting the needs of the big players: keep the market under control at all costs because they know all too well what their advantage depends on: lock-in, nothing more. Even if governments put their foot down, the big players often manage to avert most of the danger, sadly. I love ambition, don't get me wrong - that's why I do what I do ;-) But we have to do it ourselves. Don't expect the big players to join us or do anything other than try to slow us down. Cheers, Jos > Kael > -- Disclaimer: Everything I do and say is based on my view of the world today. I am not responsible for changes in the world, nor my view on it. Everything I say is meant in a positive and friendly way, unless explicitly stated otherwise. find me on blog.jospoortvliet.com _______________________________________________ Osdc-list mailing list | This is a place for our readers, writers, moderators and artists to discuss matters concerning Opensource.com and otherwise do the work that makes this a community practicing the open source way. Sign-up for our weekly newsletter: http://opensource.com/email-newsletter Send a message: [email protected] Change preferences: https://www.redhat.com/mailman/listinfo/osdc-list Unsubscribe: https://www.redhat.com/mailman/options/osdc-list
