- So after self-hosting a Known instance at https://coffeecode.ca for a number of months, I opted to pony up for a hosted version by the Known founders, partially for the sake of convenience, but mostly because I wanted to support a fledgling decentralized free software social media solution.
Shortly after they got http://thunk.coffeecode.ca running, I realized a problem: http, not https. And they have no way of enabling SSL for hosted sites. This is not an uncommon problem, as it turns out, but having not run much hosted software with custom domains before, it was new to me.
For that reason, I'm probably going to go back to self-hosting once my initial subscription runs out. Maybe I'll be able to get them (during my hosted time) to migrate my ancient s9y blog content over as well, and then make full use of the export/import facilities of Known to pull everything back to my self-hosted version.Show all 7 repliesSince coffeecode.ca is yours, can you obtain a cert for the subdomain and get them to set it up for you?
I do agree that it is more complicated and difficult than it should be, which is likely the reason why wordpress.com and squarespace also don't offer this.
- Freebase -> Wikidata
Per https://groups.google.com/d/msg/freebase-discuss/s_BPoL92edc/Y585r7_2E1YJ, Freebase is going away over the next six months and they're going to help transition as much of their data into Wikidata as possible.
Somewhere I feel the screams of millions of URIs dying... but I believe that data will be in better stewardship under the Wikimedia Foundation umbrella than it was under Google.
- Started trying out Known (http://withknown.com) as a microblog+ (supports longer-form content, photos, etc) POSSE service (syndicates to Twitter if you want), running at http://coffeecode.ca.
Microformats and WebMentions. This thing could work!
The code is licensed under the Apache License 2.0 and the maintainers have been very receptive to my initial patches.Show all 5 replies@firstname.lastname@example.org I haven't really had a chance to try out WebMentions yet, which appears to be its loosely-coupled form of federation. Given that I can't figure out how to reply to you here on identi.ca without agreeing to let your instance post updates on my behalf/modify my profile/see my list of followers, I'm not sure pump.io's form of federation is all that fantastic.
Also: I wasn't notified via email about your comment, so just happened to notice it running through my stream, and still can't use Firefox on Android to add comments. grumble grumble
Dan, giving permissions to other servers is not something related to the Pump.io federation, is only a detail of how the web interface works.
Using a client like Pumpa or Dianara you can follow anyone without ever giving those permissions.
I seriously recommend you check out the clients =)> giving permissions to other servers ... is only a detail of how the web interface works.
That is incorrect. To deliver events that interest you, they need to write to your inbox.
> Using a client like Pumpa or Dianara you can follow anyone without ever giving those permissions.
Without ever noticing that you are giving them those permissions.
email@example.com likes this.@dbs how's withknown working out for you 3 months in? I see lots of posts on your site, but can't tell for sure if that's only because you are mirroring twitter there (including not being sure if mirroring twitter is actually what you are doing) or because you're actively using and loving withknown. Thanks!
- Two interesting posts about Wikipedia "anon edits" tracking
People I know in the library community have been responsible for the https://github.com/edsu/anon/ code (licensed under CC0, okay) that tracks anonymous Wikipedia edits made from IP addresses belonging to the likes of the US Congress and the government of Canada.
Here are two blog posts that make for worthwhile reading about the project:
Stephen Michael Kellat likes this.
Stephen Michael Kellat shared this.
- I just built and installed a Tcl package from source. Which I had to retrieve from archive.org because the historical location of the package appears to have disappeared from the net.
Why? Because I just started doing some work on an old-school Perl script(hello, randomized indentation!) dating back to 1998 that is still in production, and it shells out to nsgmls + Tcl for part of its SGML processing.
Perl, Tcl, and SGML. In 2014. Amazing!
Yes, @bkuhn is right; Perl is the new COBOL. What does that make Tcl?
- I didn't realize Fedora had a badges thing running; looks like I'm either doing pretty well (https://badges.fedoraproject.org/user/dscott says top 5% of project participants), or very few other people knew about it either.
My guess is on the latter, much as my ego would like to think the former :)
- Mmm. The Islandora digital assets management system appears to use a broad CLA (http://islandora.ca/sites/default/files/islandora_cla.pdf) rather than inbound == outbound licensing. :/
- At the second day of the Supporting Cultural Heritage Open Source Software (SCHOSS) Symposium (https://foss4lib.org/schoss/symposium) -- lots of library, museum, and archives people here trying to figure out how to run free software and open source projects in a sustainable way. Governance models, funding models, foundations... it's good to have Karl Fogel here, along with Chris Cormack from Koha (a GPL library system that has been in existence since 1999, and from which many of the new projects that want to impose a heavy top-down approach have a lot to learn).
Evan Prodromou likes this.
- It's alive!
My article Open source library system Evergreen rewards the community (not my choice of title FWIW) is now available at opensource.com.
In the article, I provide some examples of how software freedom enables us to tailor our offerings for our users, maximizes our investment in development, pool the library community's limited resources, and provide grounded education opportunities both for the computer science students at our university and for our community in general.
Also, see the bonus mention of the Software Freedom Conservancy as our trusted third party for holding community assets. I now realize I said "neutral third party" which of course @bkuhn will point out is not true, given the SFC's mission to promote software freedom; but I suspect that (important!) nuance would be lost on most readers.
Anyway, I'm happy to see this out in the world. As an aside, the editorial team at opensource.com was a breeze to work with: I continue to hold the copyright over my article, which is made available to opensource.com to publish under a CC-BY-SA licence.
2013-12-07T16:41:46Z via Identi.ca Web To: CC: PublicAs I watch Fedora 20 struggle with the Intel graphics on this new Haswell-based laptop, I begin to think that there's more to Bradley Kuhn's strategy of sticking with old used laptops than just environmentalism and ethics. It's entirely pragmatic; most driver issues should be sorted out on older laptops, so you can focus on getting core work done instead of getting distracted by filing bug reports (or fixing driver issues, if you have the skills)!
Mike Linksvayer shared this.
@Dan Scott, indeed, I have many reasons for doing that, and being sure that drivers are available is among my reasons.
Amusingly, I just got a "real upgrade" today: I had a 2GB RAM for a T61 sitting around unused, and so now I have 3GB RAM instead of 2. Mairix indexing doesn't cause my machine to thrash anymore. :)
(Side note on my side note: thrashing is in the tradition of Computer Science to use terms that sound really bad to refer to things that are somewhat bad. I'm thinking of things like abort, fail, crash, etc.)
- I donated to archive.org (second time I've done that now) and wikimedia.org (first timer, sad but true) yesterday. It felt good!
I anticipate donating to more orgs in the free software / creative commons sphere of things before the end of the year, and I encourage you to do the same.
- Thanks to the power of wget, I mirrored a chunk of the Library of Congress standards documentation before it went dark due to the US government shutdown and tossed it onto http://stuff.coffeecode.net/www.loc.gov/
Thanks also to the Internet Archive for providing an awesome service, as always. I had to point to the IA WayBackMachine mirrors for some of the standards that were in non-standard places (hah).
Greg Grossmeier likes this.
2013-08-07T19:32:45Z via Identi.ca Web CC: PublicJust created a new project, https://gitorious.org/postgresql-full-text-search-engine to demonstrate the use of PostgreSQL as a full-text search engine in 200 lines or less.
- It uses Python 3 and Flask.
- I suppose I cheated because the jinja2 templates push the thing past 200 lines, but the core code is less, so I'm happy.
- There's lots more I could demonstrate in terms of PostgreSQL capabilities, but this is a decent starting point.
- Wow, PyCon Canada (http://pycon.ca) sold out just days after they released the schedule of speakers!
I'm feeling lucky and excited to be on the slate (giving a twenty minute talk on building a full-text search application using Flask and PostgreSQL).
Evan Prodromou likes this.
- "With increasing Perl usage, I realized I had to learn it." http://momjian.us/main/blogs/pgblog/2013.html#June_22_2013