a thoughtful web.
Good ideas and conversation. No ads, no tracking.   Login or Take a Tour!
comment by goobster
goobster  ·  3070 days ago  ·  link  ·    ·  parent  ·  post: Pubski: June 29, 2016

In short, you can't. Without a lot of help.

The jerks will say "Ideas are like assholes: everybody has at least one." But the basic premise is right. The only difference between you and Mark Zuckerberg is that Mark Zuckerberg did what almost nobody else does: Followed through. If you had a unique idea (hint: you don't) and got a UX designer to mock it up, and pitched it to a venture capitalist, and they gave you money to pay for it to be built... ok, then you might have something someone might sell online for $5-15, if people don't just pirate it.

Oh. And Confluence already does what you want. In fact, what you want is a Content Management System, aka CMS, which can be as simple as a wiki of which there are oodles of them to choose from, or as complex as Savo.

I have personally designed a content management system (for Kaiser Permanente's printing services division, using a database called Helix), a Competitive Intelligence Library (for F5, using Confluence), and produced highly detailed and unique documentation library (in LaTeX for Applied Materials, the people who make the machines that make semiconductors), and am now working on content management for a new company.

In short, make a list of the three features you need. Then ask around. Download a lot of stuff and try it out. Find the two apps that meet those needs the best you can, and run one of them.

In a year the entire project will change anyway, and whatever you decide on won't meet the new (entirely unexpected) requirements. Nothing beats a well-organized directory on a file server, expect for meticulously managed wikis. That's the long and short of it.





Isherwood  ·  3070 days ago  ·  link  ·  

We're using Conflunce now and people are peeling away, first because it came fragmented, second because the first exodus left it outdated, and third because Google Sites were available and easier to create and edit. Core documentation is organized in a folder structure on drive but the bureaucracy around that removes the agency of individual contributors and leaves them feeling like their actions have no reactions and so they give up.

I'm looking for a CMS that follows CCMS standards, preferably DITA but with the ability to export to SCORM, so that we can implement a modular version of NPRs COPE model of content creation and publishing. All of that needs to be wrapped in a nice user interface, similar to a squarespace, that would allow individual teams to use the components to create team portals that presents only the information that team requires.

Of course I can and am making something work because, like I said, that's my job but I also choose to believe there's a better way and, like I said, I don't think I've found it yet.

Which is why I asked the question "how do I present an idea in a way that results in forward movement?"

goobster  ·  3070 days ago  ·  link  ·  

Ah! Right! Someone who knows what they are talking about. So, let's dig in...

    We're using Conflunce now and people are peeling away, first because it came fragmented, second because the first exodus left it outdated...

Same problem here with our installation of Confluence. People think if you build the tool, everyone will take ownership of their own part of it, and the total will become an amazing Wikipedia-like constantly churning source of the latest info and intel.

But it doesn't work that way due to one thing: People.

Every CMS I have worked with has failed because there is nobody tasked with its success. An Editor. A full-time person who cultivates content and content creators. Someone who gets daily aging reports and finds people to pick up abandoned sections and mentors them through the ups and downs of curation and cajoling.

I honestly believe - after building my first CMS in HyperCard at Apple up to the current Confluence installation I am editing today - that the tool is rarely ever the problem. No CMS on the market is just a failure. They all work to a greater or lesser degree. But the few that are successful are the ones that have someone paid, full-time, to manage the data.

Also, separating the data layer from the presentation layer (the promise of COPE and XML) always requires you to go with a least-common-denominator implementation that is neither pleasant to read or use.

Ultimately, you are right: There is a Better Way out there. Somewhere.

My money is on thegrid.io. Literally. I paid for it. (Still awaiting delivery, tho.)

The basic problem with CMS systems, as I see it, is that it tries to be the solution to anyone who wants access to the data contained within it. It's a scattershot, least-common-denominator approach that works well for very few people, marginally ok for most, and doesn't work at all for the edge cases.

I want a CMS that looks at the USER, and customizes the content according to the user's needs... not the needs or design of the CMS.

So until my computer knows more about my role, usage patterns, and personality - and can act on this information - I think the wiki + full-time editor model is the only real effective solution to content management.

Isherwood  ·  3069 days ago  ·  link  ·  

    separating the data layer from the presentation layer (the promise of COPE and XML) always requires you to go with a least-common-denominator implementation that is neither pleasant to read or use

I think this is the narrow path where I see a better way. A lot of people talking about this kind of thing are saying "I need to be producing a product either for the data layer or for the presentation layer", which is probably true, but I'm not finding someone who's creating a data layer that integrates seamlessly into a separate presentation layer.

I thought Google was going to do it with sites and their suite, but they always get so close to a complete product that it's infuriating. I thought atlassian could do it, but their wiki just bundles the two together. thegrid sounds interesting, but I think there's enough interest from management in a company that, if creating the site was easy enough, they would do it.

So yeah, I think that's a little closer to what I want to pitch.

goobster  ·  3069 days ago  ·  link  ·  

Well shit. Then count me in. I've designed many applications over the decades, and even got one funded by venture capitalists and then purchased by Deutsche Telekom. So I can help.

I also have worked with a fantastic Argentinian UX designer, who works cheap and fast. So I can bring her in when we have some functionality nailed down. Let me know if you are interested.

Isherwood  ·  3069 days ago  ·  link  ·  

Hmm, OK. I'm going into the mountains for a long weekend, but I can download the Dita documentation and put together some ideas. Let's talk next week and get on the same page.

goobster  ·  3069 days ago  ·  link  ·  

No rush. I'm getting married next weekend. So we have our things going on... let me know when/if I can be of help.

Isherwood  ·  3063 days ago  ·  link  ·  

Hey, congrats. I'm getting married at the end of August!

goobster  ·  3063 days ago  ·  link  ·  

Thanks man! And good luck with your planning!