- From: James Fuller <jim@webcomposite.com>
- Date: Mon, 13 Nov 2017 21:23:51 +0100
- To: Norman Walsh <ndw@nwalsh.com>
- Cc: XProc Dev <xproc-dev@w3.org>
If you have not had the pleasure and are able - I would highly recc. coming to XML Prague if you want to meet up with ~200 practitioners of the black arts ... attendees come from a broad range of domains - you will not be bored and I guarantee you will meet like minded folks or at least people who challenge your ideas on the status quo of general computing, it is a conference about markup that goes far beyond markup itself. I cannot provide much insight in W3C - I do know other orgs (IETF comes to mind) are more immune from the necc constraints of time/money/people/popularity. I agree with Norm's observations though I will note (w/o bitterness) that the WG work on XML Processor Profiles delayed initial work on XProc itself. With hindsight I personally would have opted to spend that time working directly on XProc - that however would not have guaranteed success - I think the current effort that is going on proves out the continuing interest people have with processing pipelines. The problems XProc is trying to solve (using pipelines) will not go away and I will be excited to see where things are going with XProc at XML Prague. XML is a failure if you consider its reason for existence to be the default markup language in the browser. HTML5 struck out on a different direction, less interesting for me personally, as an application platform (mobile, etc) - some specs later and the web stack and the APIs it defines seems daunting to this programmer of some years. Perhaps the modularity of HTML5 hides a lot of the complexity I perceive but I still remember the negative comments of 'too many specs and too complicated' in reference to XML from developers who then went on to win the complexity war with 50 specs. Things always look easier when you are starting from scratch, but the hard earned nuggets of computing wisdom stand the test of time. I begrudge no one if they have a job to do and want to get it done - though personally i am always distracted by the funny languages that challenge me to learn more about programming. Lets wait a few years for json to cook - it is grappling with the same compromises XML had to make ... namespaces will emerge as well as a lot of familiar approaches. I can use many of the idioms honed with xml technologies against these other markup/data formats, so it is kind of win win for me. XML is successful in a lot of scenarios - and predicated a lot of change to make a richer programming environment ... knowing where and when to apply a technology is always the challenge in computing. It is worth doing your own analysis and come to your own conclusions on XML though only you will know if it is fit for purpose for the problem you want to solve. J On 13 November 2017 at 17:01, Norman Walsh <ndw@nwalsh.com> wrote: > Andreas Mixich <mixich.andreas@gmail.com> writes: >> I am aware of the damaged state of XML, especially on the Web (HTML5, >> JSON, etc.), but has XML lost its authoritative decisions body? > > I don’t think there’s anything damaged about the state of XML. It’s > not seriously threatened by HTML5 or JSON or anything else for the > kinds of problems I’m interested in: mostly prose documents written > largely by human beings for consumption by other human beings and/or > machines. > > The fact that I’m now constrained by circumstances to deal with JSON > for mostly-data APIs is a minor inconvenience. JSON is just like XML, > really, if you ignore most of the useful features of XML. > > HTML5 isn’t really about markup anymore, an inflamatory observation, I > know. It’s part of the application development platform for the web. > It’s probably good enough for that, I guess. And web components means > it’ll have something that resembles XML, with less rigor and no > mechanism for avoiding name collisions. But I’m sure that’ll never be > a problem. > >> Or are these things now being done over at OASIS? > > Some of them are being done at OASIS. If you invent a new XML > vocabulary and you think it needs to be standardized, OASIS is a > perfectly fine place to do that. > > The core XML specs were defined at W3C so that’s where some of the > core language extensions (stylesheets, query, linking, pipelines) got > defined. > > However, standards organizations live and die by what their members > are willing to fund. When XML stopped being exciting, and could be > viewed as largely “finished”, member interest waned. If you’ve got > limited staff and limited resources, as the W3C does, you have to > apply them to the problems that the membership is interested in > funding. > > Consequently, the XML activity at W3C has wound down. And it’s > important to note that membership interest isn’t limited (or perhaps > even principally) about member $$$. The working groups dwindled as > well. Companies stopped participating; they have limited budgets as > well. It became harder and harder to achieve quorum at Working Group > meetings or maintain sufficient membership to actually accomplish > anything. > > Before we abandoned our attempts to standardize the next version of > XProc as a W3C Working Group, we were down to about four or five > members in the working group; that’s just not enough. And we were all > complete volunteers with no direct support from our employers. > > If you want more XML standardization, you have to participate and/or > lobby for your employers to participate. > >> Is there some clarifying article/blog post, I could learn more about the >> current state of affairs? One of the worst things, that could happen to >> XML, would be, if, having become a niche-product, organizations would >> start implementing and extending “in-house”, fragmenting what never >> ought to become fragmented. > > Maybe that will happen. If it does, and if the fragments start to > catch on, there may be interoperability problems. And maybe, just > maybe, that will eventually generate sufficient interest to reopen the > question of standardization. At that point, someone will invent “the > next standard thing.” It won’t be XML. Whether it’s a useful successor > is something we’ll just have to wait and see. > > It’s also important to put this all in historical context. In the > mid-1990’s when work on XML started, there wasn’t ubiquitous high > speed internet. There wasn’t github. There wasn’t npm. Companies > didn’t routinely build products with a tower of random dependencies > slurped down from the internet. (I just peeked at a project I’m > involved in, even de-duped, there are more than *a thousand* dependent > modules; and it’s a reasonably small, utterly straightforward web > application.) > > If we invented XML today, would it need to be standardized? Or would > it just need to be an NPM module? > > Be seeing you, > norm > > -- > Norman Walsh > Principal Engineer > MarkLogic Corporation > Phone: +1 512 761 6676 > www.marklogic.com
Received on Monday, 13 November 2017 20:24:15 UTC