by

Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
    Rob Reed
    @robreed_twitter

    I don't think there can ever be just 1 tool, which is a good thing. Try to go that route and you end up with Facebook or Amazon. I'm much more in favor of the small single purpose tools used in endless combinations strategy, as I'm sure you are. Integration becomes a problem, but I think integration is overrated, and may be a red herring.

    The right tool for me is a combination of email, and some kind of group IM, and an individual IM which might be the same as the group tool, and a forum, and personal blog, and group blog, (both with comments), and some kind of aggregator, and some form of document collaboration, and version control/version history like git, and… well that's a start.

    I like Slack because it behaves the way I except it too most of the time. I like the implementation. Git on the other hand, it wants to post every time I hit return. I really don't like the way they position Slack. Although I guess people are using it more and more as a public space, but I'm not sure I understand how that works.

    Steven Rosenberg
    @passthejoe
    I haven't joined any public Slacks yet, but I wonder why a Discourse forum isn't better for most things.
    Rob Reed
    @robreed_twitter
    I just think we should stop looking for single tools that work for every group and every kind of communication and start thinking in terms of valuing tools that work well together. IM is good. Email is good. Forums (especially Discourse) are good. But they're not all good for everything.
    Rob Reed
    @robreed_twitter
    New idea for Ode: Markup (standards compatible) to select a portion of a post that would be appropriate for a tweet. Also a special theme on the site that will present just that portion of each post along with a link to another theme (like the default theme) where the entire post is visible. So it's like a Twitter style UI for a site. What do you think?
    By the way do you see the other "conversations for Ode" - feature-requests etc? I'm afriad to use anything but the /Lobby because I'm afraid it's not visible.
    Steven Rosenberg
    @passthejoe
    Rob, I haven't been on here in a long time -- I looked at your site and don't see links where users can download Ode. I'm also wondering if you have the new theme packaged in there yet.
    Steven Rosenberg
    @passthejoe
    I just went to our Twitter DMs and found the version you sent me in February -- thanks. I'm getting ready to make my social-post site a separate one from my "traditional" site, and I wanted to use the new theme for one or both.
    Steven Rosenberg
    @passthejoe
    I thought I had ZIP copies of the Ode software and add-ins, but I can't find them.
    Rob Reed
    @robreed_twitter

    Sounds good and a lot of work, good job.

    "but individual users would need to "register" their apps with the Twitter API."

    I hate that. (I'm using the word hate here.) That is the point at which the open web started to die. That really made me upset and I'm still very sour about it. Anyway…

    I have my usual big (small) plans for Ode. I really want it to be a bulletproof local/remote writing and publishing solution and I want to use it to get a more diverse group of voices on the web. I've heard enough from the same people, especially in tech and web design and development. My attitude at this point is basically fuck them. We've been right about a lot of things over the years while they've gone with the flow. The flow sucks.

    Anyway, I like this path you're on.

    Steven Rosenberg
    @passthejoe
    Re the Twitter API, registering your app (or in my case my instance of my app) is a pain point, to be sure. A better idea is leveraging the Twitter authentication system to let the app post. I arrived where I did due to Ruby's Twitter gem. CPAN's Net::Twitter works the same way: https://metacpan.org/pod/Net::Twitter
    Steven Rosenberg
    @passthejoe
    The Twitter API can do a lot of things, and I'm only doing one of them (posting updates). It's likely possible to post to Facebook and other social sites in a similar manner, but my emphasis is on creating and posting that kind of update to my own sites. The Twitter posting code is really a bolt-on -- it's optional all the way through. What prompted me to include it was dlvr.it limiting the number of updates on free accounts to 10 per day. I moved my "automatic" posts from the blog to IFTTT.com and now get the Ruby script to do the social updates. I wasn't excited about paying $99.50 a year for "unlimited" dlvr.it posting from the blog to Twitter.
    Dlvr.it allows you to exclude a directory from posting to Twitter. In IFTTT, I had to code that in JavaScript. With separate Ode sites for social and not-social, I won't have that issue.
    Rob Reed
    @robreed_twitter
    Yeah I've played around with Net::Twitter. It's fine. I had a nice Twitter posting addin. It matained a loose connection to the Tweet by storing all of the meta-data for the tweet and allowed the user to delete the tweet and stuff like that after the fact. You could pick the part of the post you wanted to tweet. Lots of stuff like that. I'd have to dig it up.
    I think going directly with the service's API is way better than something like dlvr.it or IFTTT anyway. The fewer dependencies the better. I love the idea of having your own site and pushing content to social networks. Thats exactly how it should work. We should have independent websites and aggregators. Full stop.
    Steven Rosenberg
    @passthejoe
    That would be great as an add-in. My idea was to do all post creation outside of Ode and have the app be a blogging client
    Rob Reed
    @robreed_twitter
    I see a lot of potential in the Ode way of doing things as far as a CMS is concerned. I think you know I've started doing some teaching - I'm up to 3 classes and a weekly meetup. It's too much honestly especially considering I get paid for none of it. But it's giving me a lot of opportunity to think about stuff as has everything that's happening and not happening with social media these days. I have a good read on the situation I think. (After something like 15 years why not, right?) It doesn't mean what I think is right will "win". In fact I'm pretty sure the righter something is the less of a chance it has at "winning" (which is why I don't think we can make everything a competition).
    Steven Rosenberg
    @passthejoe
    I started hacking away at the old theme -- literally hacking, as in removing bits. More of that to come. Here is the site with content: http://updates.stevenrosenberg.net/cgi-bin/ode.cgi
    Rob Reed
    @robreed_twitter
    The old theme is rough.
    You've really got to care about everything or it starts out ehhh and later it's like ugghhh.
    Ha.
    Steven Rosenberg
    @passthejoe
    I forgot to turn on Markdown, I think. I'll get to that later. The new theme is so huge, I will tackle that one later.
    Rob Reed
    @robreed_twitter
    Huge meaning filesize?
    Page weight I mean?
    I still want to make my own version of markdown. I don't like all of the crap surrounding Markdown and the different dialects. I just want a simple lightweight markup format. I'm calling it unmarkup. I've probably told you that already.
    Steven Rosenberg
    @passthejoe
    Re the new theme -- I mean it's a lot of files, and there are a lot of areas/sections on the page, and I will want to make it more simple.
    Rob Reed
    @robreed_twitter
    Gotcha. As soon as you say that I want to go make it simpler. Ha. No time at the moment but as soon as I hear that anyone is doing something with Ode, that suddenly all I want to do.
    Steven Rosenberg
    @passthejoe
    Rob, while I was looking for .htaccess hacks, I found some things on caching. Would this work for Ode due to its dynamic nature? https://www.askapache.com/htaccess/speed-up-sites-with-htaccess-caching/
    Rob Reed
    @robreed_twitter
    I'll have to read that thing carefully. It's pretty confusing the way it's written. Also, I think I'll have to play around with it. Most if not all of them may work for Ode depending on how clients handle expiry dates. Ode is dynamic but doesn't use query strings mostly, so if it reads an address as a page then I don't see why it wouldn't work. But one of the suggestions is to generate static pages and that's what I want to do anyway. That will work with all caching methods AND even when the cache expires visitors will hit the static pages not the script anyway. The performance for a busy site should be just about on par with a purely static site with none of the drawbacks.
    Steven Rosenberg
    @passthejoe
    Sounds great.
    Rob Reed
    @robreed_twitter

    Hey Steven, I'm thinking of renaming Ode, "Ebeling". Any thoughts about that? It's because of this guy Robert Ebling. https://www.npr.org/sections/thetwo-way/2016/01/28/464744781/30-years-after-disaster-challenger-engineer-still-blames-himself

    He kind of epitomizes how I feel about technology – Going along with what everyone is doing because everyone else seems to be OK with it is in no way OK.

    I also like that the name is easier to pronounce and kind of sounds like an open source project.

    I'm just looking for a little feedback.

    jgbrwn
    @jgbrwn
    Hello, I remember looking at Ode many years ago, and stumbled back upon ode.io, but looks like it's not available to download currently?
    Rob Reed
    @robreed_twitter

    @jgbrwn Hi. Thanks the message. Out of curiosity, what are you thinking about using Ode for?

    Yes it's not generally available at the moment. I'm working on a significant update. I'll be honest that's it's long overdue but I am committed to the project and making it more of a priority so that it gets done sooner rather than later.

    The next version will still be a Perl CGI. After that I may rewrite it for JavaScript . Not much will change about the existing functionality because honestly it works really well, and I like it. But there can be some refinements for sure and there is some much needed functionality. The goal is a platform that is (or continues to be):

    • Approachable for people new to programming
    • Bullet proof for server and local installations
    • Syncs seamlessly
    • Generates and caches pages in response to requests on the fly
    • Works with git for version control, backup, migration, and collaboration
    • Does all of the things Ode does as well or better than everything including themes for example

    Would something like that work for you?

    Thanks again! Rob

    Steven Rosenberg
    @passthejoe

    Hey Steven, I'm thinking of renaming Ode, "Ebeling". Any thoughts about that? It's because of this guy Robert Ebling. https://www.npr.org/sections/thetwo-way/2016/01/28/464744781/30-years-after-disaster-challenger-engineer-still-blames-himself

    Sounds good to me.

    Rob Reed
    @robreed_twitter
    @passthejoe OK thanks.
    john henry
    @john_henry_twitter
    hi Rob, I just saw a link to ode.io on your twitter and had somehow totally missed it until now. I just started working on a django blog engine for myself after not being excited about hugo, jekyll, wordpress or any of the hosted solutions. I strongly agree with your thoughts on the open web vs. platforms and while I'm not sure there are easy answers, I'm very interested to see where you're going with it.
    rob-for-gitter
    @rob-for-gitter

    (I cleaned this up some and I'm reposting it).

    Glad you found the gitter community. It's nice to know gitter is sitting here serving this purpose and doing a decent job of it.

    Anyway, Ode, I mean ebeling… (I am committed to the name change but it's going to take a while for it to stick. It already sounds less strange to me than it used to.)

    I am planning on a next release. I've been saying that for a while. I need to figure out how to work on it consistently. The goal for ebeling is for it to be essentially what it is - a self-hosted, super simple, yet very functional, bulletproof platform with a minimum of dependencies.

    It's not intended to be the app the cool kids are using for 6 months. That does create some problems. For one, I think there is a minimum viable community. I'm a little torn about that because I do think ebeling is something uniquely good and deserves to exist.

    As for specific plans, I have lots, all of them very doable.

    There are small things… I need to put together a basic authentication mechanism. Currently you can create posts by connecting to the server and savings posts directly with a text editor, which I highly recommend because it's the best possible experience. (Web forms suck for writing.) There is also an addin that allows posts to be created through a browser, but it has only a basic password mechanism and depends on TLS for security (which is technically fine in terms of security but not ideal). There are quite a few little things like that. Oh, I need to write a search addin for one thing.

    Beyond those things I'd like to focus on making sure it can run absolutely everywhere and although I see a lot of value in an (improved) manual install, it would be nice if it were as simple to get an ebeling site up and running as creating an instance of a container or something.

    I'd love to have an Ebeling site on a flash drive that you could plug in and work on locally and then sync to a remote server to publish (utilizing git ideally). I'd feel good about that.

    I see it as a publishing platform, and also a local writing / note-taking / presentation app too. I want someone with unreliable internet access to be able to maintain an ebeling site and sync it when possible. That's the potential I see, a few of the billions of people who can't or don't want to use a centralized service with a monthly fee. How is someone supposed to pay for a WordPress site if they live in a part of the world where $30 USD a month is a significant portion of avg income?

    I can also see rewriting it as a JavaScript app. I know JS well enough to do that. I also think JS is stable enough now to justify the effort. Having said that there's no real reason to do it. Perl is perfectly fine and in a lot of ways I think the perfect platform. I'll take Perl's regular expression support over JavaScript's for one thing, and you can get far with regular expressions for what ebeling needs to do.

    Beyond all of that, I'd also like to teach people how to use it, or use it as a platform to teach web development. I've started doing a little teaching, and I've put together hundreds of pages of notes covering the fundamentals of web design and development.

    I see it all as the same effort. It's my futile attempt to push back against the current state of the internet, in a constructive way.

    There also needs to be some work done on the default theme. I can do that too. It's just a matter of time (and daily motivation).

    Steven Rosenberg
    @passthejoe
    Rob, this is interesting, including the JS part. One thing that Perl CGI has going for it is that it's easy to deploy on "standard" shared hosting. One thing that I think would be interesting is deploying Ode/Ebeling in a Docker/Kubernetes/whatever-you-call-it container. That would make for easy deployment on cloud services.
    One thing that JS would do for the project would be to attract more developer interest. However good or bad it is, Perl = !sexy. But I still think the project can thrive with Perl. Go (aka golang) is another language to look at. Ease of deployment is a big part of that language.
    Steven Rosenberg
    @passthejoe
    The problem, though, remains that dynamic apps are difficult to deploy (and difficult to deploy cheaply). That's a sweet spot for CGI and even PHP apps.
    rob-for-gitter
    @rob-for-gitter
    All good points @passthejoe. Thanks for the thoughts!
    rob-for-gitter
    @rob-for-gitter
    I think you pretty much nailed most of the pros and cons. I was thinking JS would reduce the surface area of what people need to know to work with the project. If someone has to know HTML, CSS, and JS anyway for the client-side development, maybe it's a shorter leap to JS on the backend. But as I type that and think about what you've written, server-side JS really does come with a lot of overhead. I don't think anything will attract attention to the project, and that's fine. I just want to do the right thing by the project.
    Steven Rosenberg
    @passthejoe
    So I decided to play around with Gatsby.JS, which is a static site generator that uses React to build its pages. I started the Hello World demo, and I kid you not, I had a directory with more than 28,000 files in it.
    Re: attracting attention to the project. If it was done in JavaScript, whether that was a sound choice or not, it would attract attention.
    rob-for-gitter
    @rob-for-gitter
    @passthejoe About Gatsby… It's like every web project is a fancy upscale restaurant. It's very very opinionated and I'm not sure "JAMstack" is a stack at all. Sorry I think there's supposed to be a line over the T in JAMSTACK and that the JAM is supposed to be bold and the T slightly shorter than all of the other characters. That kind of fussiness over simplicity seems to be at the heart of a lot of this kind of stuff. 🙂
    Steven Rosenberg
    @passthejoe
    At this point, I'm still getting over the 28K file shock. That's only on the "build" client. I'll have to get to making a site and seeing what it generates for the server. I can tell you that most of the static-site generators that I've tried (Go-based Hugo, Racket-based Frog, Java-based JBake, Clojure-based Cryogen) are pretty good in terms of what they have you ship over to the server.
    rob-for-gitter
    @rob-for-gitter
    I just don't think the juice is worth the squeeze. I appreciate the info though, thanks!