[00:30] rss is in xml =[ [00:32] yea, atom/etc are just xml files [00:33] fortunately there's stuff like http://pypi.python.org/pypi/feedreader/ for parsing/dealing with it [00:34] i want to create it [00:35] what i want to do is simple i've just never looked at rss before [00:35] http://www.dalkescientific.com/Python/PyRSS2Gen.html [00:36] your not going to stop until i'm using python are you? =] [00:37] hey, just saying :P [00:37] i was thinking i'd make the client in python at least [00:37] you didn't specify [00:37] easy and portable [00:37] yea because i hadn't decided [00:37] batteries included [00:56] http://paste.mitechie.com/raw/NimtS1TS0lcdPmJkExSg/ <- snort [00:57] Also, if you're hand-crafting XML files, you're either a masochist, or doing it wrong [00:57] snap-l: sweet, sounds like a great business deal [00:58] Yeah, I'll really put those ads on my site to good use [00:58] Get that Amazon Associates thing cooking [00:58] snap-l: so thinking the best way to do this talk is to slowly build up a Makefile from noting to useful, but this will require live coding. [00:58] I will be tempting the great angry live coding gods [00:58] I'll bring wifi router [00:58] ooh, I hear associates is where the $$ is at man [00:59] hah [00:59] and a stick to poke me as I typo each step of the way [00:59] but tried this a few ways and really think this will help 'click' with people best [00:59] I got the long antenna, so I think I could hit you from anywhere in the room [00:59] awesome [01:00] It looks like Gib's router that he used to bring, only cooler because it runs Tomato [01:02] rick_h: You might remember it. ;) [01:03] orly? [01:03] oh, an old wrt54gl? [01:03] Yeah [01:03] gotcha, cool [01:03] It was yours. :) [01:04] yea, recalling now [02:31] http://cdn.arstechnica.net/wp-content/uploads/2012/06/unreal4a.jpg [02:39] "I light my level by dropping a sun in." [02:40] "[the level is lit] by light being refracted off the dull stone walls and the colored carpet. And if the carpet changes color, so does the reflected light, in real time." [03:18] I feel like there used to be a feed icon in or near the address bar when you were on a page that provided a feed version [03:18] in mozilla [03:22] ok its in the bookmark menu [03:25] got to decide whether to make this ssd my main drive, or just put things on it strategically [03:30] make it / and then use another for /home if you need [03:33] I think that is legit... i could also make the slow drive the /, put /home on the ssd, map back media and download dirs [03:34] you want /usr/loca/bin /bin and such in ssd for speed in launching things [03:34] boot for boot speed etc [03:34] ive rebooted this thing all of like twice since i got it [03:34] honestly, these days just get a big enough ssd [03:35] 240... it is big enough actually [03:35] ah, well "it" isn't defined :) [03:35] i just figured to keep the hdd internal as well since i don't use the optical much [03:35] yea, then just put the whole thing as / and mount the slow drive as /slow [03:35] put music/videos on there and be done with it [03:36] sounds good [03:36] windows vm etc on /slow [03:36] if you can fit it on the main it'll feel a lot faster [03:36] and a backup of my home [03:36] but yea [03:39] http://www.metalsucks.net/2012/06/06/what-if-metal-album-covers-were-truthful/ [03:48] I think you can use wget to make a sort of static mirror of a page [03:48] I don't know how well it would work now, everything is so dynamic now [03:50] but, if so, you to grab a snapshot of every url that pops up on a feed [03:50] http://www.kchronicles.com/2012/06/05/kings-devils/ [03:50] s/you/wouldn't be hard [03:51] yea, there was a cool tool I was using at morpace for some stuff like that [03:51] of course now I can recall it...crap [03:51] and then generate a unique url for it, and stick it as static content, and put the url to the archived copy next to the actual url [03:51] geeze, I watch 123 repos in github now...missed that getting so high [03:51] https://github.com/cpinto/python-webpage-inliner [03:52] that works pretty well and makes for great offline copies of sites in single files [03:52] doesn't get some things like images defined in the css, but pretty good copies [03:52] Sure we all watch a lot of repos when we're high. [03:53] eventually I want to add support for that into bookie so that you get both the readable pretty content, but also a single file I can load up if you want a more full featured view of the page you bookmarked [03:54] oh you do grab some of the content [03:54] I totally missed that [03:54] yea, if you click the 'eye' icon on the left side it loads a readable parsed content view of the page [03:54] the search is a fulltext search that goes across tags, title, and content [03:55] I found that once I went to look. [03:55] https://bmark.us/bmark/readable/42742d9ffa6e7a [03:55] it's why I've been working on a better readable parser and the seperate service: http://readable.bmark.us/view/http%3A%2F%2Fwww.mail-archive.com%2Flxc-devel%40lists.sourceforge.net%2Fmsg00151.html [03:56] to get bookie to use it as a service vs doing the work itself [03:56] scale better, etc [03:56] i see [03:57] do you want a version thats true to the original or more like instapaper/read-it-later [03:57] now known as pocket [03:57] so the original vision was instapaper [03:57] but I can see the value to wanting the images/etc (more true original view) [03:57] thus the python-webpage-inliner tool [03:57] so that all the JS/CSS/images get inlined into a single content file that I can store as well [03:58] but you don't want that to be searchable because it's too messy [03:58] might include ads, etc [03:58] a third option is actually rendering it to an image [03:58] yea, but then you lose copy/paste, no JS interactions, etc [03:58] i guess like the google preview [03:58] I agree you wouldn't want it as the only view [03:58] but that's what I was doing at morpace, using the webpage inliner tool to generate a single .html file [03:59] and then sending that through wktopdf to generate a pdf image of a site [03:59] https://github.com/antialize/wkhtmltopdf [03:59] Yea I had made a one-liner to take my bookmarks file (back when it was just a .html file) and create a structure of offline mirrors of all the pages [03:59] so that would runa webkit browser at the generated single file .html and then give me a pdf I could serve to clients as a 'printable' view of a dynamic JS page [03:59] heh, remember what I said about how I should have made everything a script and saved it? [04:00] :) [04:01] why not render from the original? [04:01] maybe i'm just presuming that the inliner would get something wrong. [04:01] because you were too slow/network bound [04:02] it was meant to go througha series of queues [04:02] one queue would fetch urls and inline them, cache them [04:02] then another would go through those and pdf them, and shoot off the pdfs to the requestor [04:02] you could 'request pdf' and get an email, we could cache/store the inline pages, etc [04:03] I think there's always going to be some sort of issue. javascript popover "take our survey" or something. [04:03] oh yea, the inliner wasn't perfect for sure [04:03] however since we were using it against our own sites, we had more control and it worked better [04:04] we were doing some heavy JS driven content stuff and when they hit 'pdf me' we had to generate monster urls of settings so that we got the right content back on the page [04:04] Personally I think that being able to get and the primary content of a site through a text only browser should be a requirement. [04:04] heh, yea we do that at launchpad [04:05] depending on what that content is of course [04:05] http://readable.bmark.us/view/http%3A%2F%2Fblog.launchpad.net%2Fgeneral%2Flaunchpad-for-textual-graphical-and-interactive-browsers [04:05] was from just this week [04:06] but you can do some awesome stuff very fast with pure client side rendering/handling of things [04:06] and it even works in IE8 :) [04:07] nice [04:09] ok I'm conking out [04:09] night [04:09] laters [04:10] So I have an idea, and like most ideas lots of others have had the same idea or a very close. [04:10] This is not too far off from bookie as it happens. [04:10] oh do tell :) [04:11] if it's interesting to you I always say do it, worst case you learn alot [04:11] bookie moves pretty slowly and I've given up on ever making it a big product, but I'm learned a ton and it's great to point employers to [04:12] Yea I think this should be pretty strait forward to do, and maybe it'd even work to start with bookie as a base... but its simple enough I'm not sure [04:12] well simple until you need it to scale of course, but if that ever happens that'll be a good problem [04:13] cool, well if you want we can chat at CHC or something. I'm always up for talking code and if it's related can help with things [04:13] you authenticate to the service, and you can post a url to a feed/channel [04:13] that you have post access to [04:13] reddit? [04:13] anyone can create new feeds (that have that permission on the service) [04:14] I've never used [04:14] But probably like a simpler version with a different focus [04:14] first thing that comes to mind from that initial pitch [04:14] Not so much a chat. [04:14] Yea that came to me [04:15] Then the feeds are of course available as rss / atom [04:15] collections around tags then? where tags = feed/channel aroud a topic [04:15] also, you can create private feeds which only you can reed, its all ssl so it can be secure (as secure as the server) [04:15] right, but where a primary tag is required [04:15] yea, I've been meaning to add rss to bookie views but not gotten there yet [04:15] also I guess by that analogy you'd have where one group of people control access to a tag [04:16] So a group or a channel is a better term. [04:16] I'd not get too hung up on that part. People will naturally build community around their interest/tag/feed [04:16] and ignore the ones they're not inerested in [04:17] and it makes life much easier to start when you're not bogged down in security implications/et [04:17] why bookie only does public urls right now [04:17] concentrate on getting the other bits right vs the lock down issues [04:17] So that is the technical description of the service... and right its something that I think is simple enough that people might / will find other uses for [04:17] build the social part, and only limit/lock when it becomes a requirement [04:17] just a tool [04:18] sure, I could make it all public first as a proof of concept, and then evolve it into a all public only version and a secure by default version [04:18] right [04:19] so bookie has some good stuff to help with that. First, we've got the infrastructure for saving urls, users, api, etc [04:19] we also have the concept of command tags [04:19] which is? [04:19] !toread marks the bookmark as toread for you [04:19] ah [04:19] the idea was that private bookmarks will be implemented as a command tag !private [04:19] so that the interface stays clean, but you can expand on special actions that occur when a command tag is hit [04:20] thats not a bad idea [04:20] to #somegroup could be an indication that this bookmark is for a certain feed/group [04:20] /to/so [04:20] right, bookie is more like twitter [04:20] gives it more meaning thatn just taggins somegroup [04:20] you don't have the follow timeline thing, not that I'm sure you need it [04:21] well, it's potential to grow into some of that. The groundwork is there, but it's not fleshed out yet [04:21] Don't go throwing in follow or timeline just because another popular service has it [04:21] your right to look for ways to keep it simple [04:21] but yea, then you could basically biuld api calls that filter/generate rss feeds based on the #something tags only [04:22] exactly [04:22] and allow users to 'subscribe' to those and get them dropped into their bookmark lists [04:22] search/etc [04:22] and I think behind the scenes thats what how the channels would work [04:22] Man I'm having trouble with words tonight. [04:22] anyway, just saying that yea, there's some potential for bookie cross over there if you're interested. Some low level stuff supports some of the ideas [04:23] Cool [04:23] but however you decide to try it out, I always say try to build it and see if it 'works' and worst case you learn a lot about the problem, development, etc [04:23] the initial proof of concept of bookie was written during pycon in one week 1.5yr ago [04:25] Here was the original idea. I sometimes come accross a link to a file, and I want that file, but its useless to me on my phone. So I had this idea of a "download it later" app on the phone, and i'd implement it as a private rss feed of links [04:25] then of course ala podcasting, you could have a service running on your computer all the time if you wanted, so its more of a "download-to" and/or "download later" [04:25] ah, that's cool. One of the ideas on the bookie todo was to better support bookmarking files and providing interesting tools around it [04:26] like a gallery of your bookmarked images, etc [04:26] so you could filter in filetype, maybe do interesting things around that [04:27] so, exactly, then I thought of other stuff you could do, like say a feed of images, and maybe you'd want to share that or maybe now (now its more like reddit, but still simpler) [04:28] yea, cool stuff [04:28] yea.... so basically all this could be implemented by extending bookie. [04:28] sounds like it, at least initially [04:28] but would that make bookie better? [04:29] yea, well some of it is on the todo list already [04:29] so, for instance, I think the first step would be implementing rss feeds of tags/searches [04:29] sure [04:29] I can see that [04:29] so you could start out the simplest case, tag someting 'community' and get that rss feed [04:30] then work up to where users could 'subscribe' to that community tag, and see it in their own bookmark list [04:30] no matter who stored it [04:30] and then work up from there by maybe adding the concept of ownership of that tag [04:30] at that point you want it to be a magic tag, vs just generic, so @community becomes a bookie command tag [04:31] and that can be owned, but a list of users [04:31] and move on from there [04:31] at that point, you've got links, grouped, and subscribe/protected [04:32] from there you can build out apis then to provide special ui views of those @commnity tags [04:32] if I'm following you at all here, it is getting late :) [04:33] I think that would basically work [04:33] Its coming at it from a different starting point [04:33] right, and it could probably go differently. Just trying to break it into a series of 'smallest stepping stones to work' [04:34] I find that's the only really good way to do these things, else they get too big and hard to move forward [04:34] smaller goals that show more progress and don't require buying in all the way ftw [04:35] more room to pivot mid-stream (hah, get some agile wordage in there for you) [04:36] ug... I'm just leaving an agile gone wrong situation [04:36] hah [04:37] well this is OSS, so we basically mess up all that stuff to suit our needs in a more practical way [04:37] don't be agile, increase your agility! [04:37] sorry, how about 'lean' :P [04:37] my hobby is replacing agile terms with religious terms. [04:38] sweet [04:38] don't be christian, be more christ-like! [04:38] we need more chanting in our meetings [04:38] haha [04:38] gregorian chant is pretty rad [04:38] but if you haven't learned to do that you can just all hmmmmm as a group [04:43] what is del api? [04:44] man sunday I should make an android client [04:44] its time I got some projects out there [04:48] oh look, docs =] [04:49] del api? oh delicious api [04:49] not sure that still works tbh [04:49] originally wanted to be compatible, but it was limiting/holding me back from the things i wanted to make unique [04:50] I _just_ read the same thing in the doc [04:50] heh [04:51] heh, yea spent a bunch of time on the docs, especially the api [04:52] http://docs.bmark.us/en/latest/api.html [04:52] anyway, way past my bed time, the boy will be up in 6 hours if I'm lucky. [04:52] night [04:53] thanks, see ya [14:31] Good morning [14:59] its like gentoo is trying so hard to help to avoid dependancy hell. .. and failing [15:02] all the time [15:02] I wonder what I could buy download.to for [15:28] uh oh, snap-l is talking with mlinksva re patents :) [15:29] he meant s/protects/restricts/ [15:29] there is no actual protection going on. protection is a doublespeak word [15:31] I mean, saying "patent pending" isn't actually a physical shield (or in his words, a physical shelter) [15:34] greg-g: Again, it's a matter of perspective [15:35] if you're the one filing the patent, it's protection [15:35] if you're very much against patents, it's restriction [15:35] I get rather irritated when someone tries to choose my vocabulary for me. :) [15:37] And my point that the patent office has outlived it's initial function as a clearinghouse for keeping ideas from being kept as trade secrets is more relevant [15:38] I can see them being useful for physical manefestations, but I'm also talking myself out of that perspective the more I think about it [15:40] words are important, they perpetuate themselves through ideas. If you want people to think that patents/copyright protect them, use protect. If you want people to understand that patents/copyright are restrictions on the public good, use restrict. :) [15:41] I should have said: ideas are important, they perpetuate themselves through words, but both ways of saying it is probably right :) [15:46] "No Trespassing" is enforced because I protect my land with a 12 gauge [15:47] The restriction (no trespassing) is because of protection. ;) [15:48] so patents are a 12 gauge shotgun? I'm more for gun laws than I thought! ;) [15:58] ;) [16:21] meh, I'd argue that's not hte enforcement of no tresspassing. It's more a reminder of a social norm 90% of the time and if me walking beyond the sign doesn't enable enforcement by 12 guage [16:46] rick_h: Thank you for doing the presentation in June. Looks like the Cloud Presenter isn't going to be able to make it after all [16:46] Also: has anyone had trouble sending me mail? [18:15] snap-l: yea np, started working on it. So should I be shooting at 45min then? [21:50] rick_h: Yeah, if that wouldn't be any trouble [22:24] snap-l: yea not a problem at all, just want to make sure I don't run too much/short if I can help it