| wpwrak | whitequark: memory is somewhat slow, yes. i don't recall the exact numbers but at least your memory frequency sounds pausible. but i wasn't even referring to memory bandwidth but to memory size. | 00:11 |
|---|---|---|
| wpwrak | (good air) you're in china now ? :) | 00:11 |
| whitequark | wpwrak: mooooscowww | 00:14 |
| whitequark | but yes I did think of china for a minute | 00:14 |
| whitequark | (memory size) yeah, 32 is too low for realistic linux in 2012 :/ | 00:15 |
| whitequark | er 2013 already | 00:15 |
| wpwrak | it was even a great struggle to get urjtag to run on the ben. had to enable swap to compressed ram. | 00:16 |
| wpwrak | (urjtag uses a particularly memory-hungry approach for sending xilinx bitstreams) | 00:17 |
| whitequark | yeah I remember that | 00:17 |
| whitequark | right now there's a great deal of activity for integrating mips16 in llvm | 00:18 |
| whitequark | I wonder how much that would save on ben, considering it'd support mips16 | 00:18 |
| whitequark | (it's like thumb/arm, there's a trivial expansion step before instruction decoder) | 00:19 |
| whitequark | I also wonder who the fuck needs mips anymore | 00:19 |
| lekernel | hi wolfspraul | 14:37 |
| lekernel | wolfspraul, someone emailed me about buying a M1 and sharism.cc being down - what's your actual position on this? are you still selling M1s or not? | 15:14 |
| lekernel | I'm fine with either answer, and thanks for everything you have done for this project - I just want to know what I should tell him... | 15:15 |
| lekernel | if you don't reply within a week I will assume you don't. | 15:15 |
| whitequark | lekernel: he did sell an M1 here to someone, quite recently | 15:21 |
| lekernel | yes, but I got that email on Tuesday - after that IRC discussion. I forwarded it to wolfspraul, no reply yet... | 15:23 |
| kyak | viric: google reader is officially shutting down | 15:53 |
| kyak | probably it's a good chance to think about migrating to something i can really control.. i don't want to rely on another web-site | 15:54 |
| kyak | how is offrss doing? :) | 15:55 |
| viric | kyak: perfect :) | 16:06 |
| viric | kyak: that's what I use at least. | 16:07 |
| viric | it doesn't feature 'stars' or things like that though. | 16:07 |
| kyak | don't need this crap | 16:22 |
| kyak | do you "mark all as read" feature? :) i use this one very often | 16:23 |
| kyak | anyway, will try to build now.. not yet sure how i want it to run - all on my laptop and update feeds manually or have it run on my server... | 16:24 |
| kyak | i don't need to access it from other locations (or very rarely), so probably i will just run it locally for now | 16:25 |
| kyak | i wish it would understand opml format though.. | 16:43 |
| kyak | viric: for now i'm using the feedurls.txt like you shown in "Usage" | 16:46 |
| kyak | there is an error: offrss: offrss.c:128: open_meta: Assertion `f != ((void *)0)' failed. | 16:46 |
| viric | isn't opml a mess? | 16:46 |
| viric | auhm | 16:46 |
| viric | what a bad software. :) | 16:46 |
| kyak | and then "Wrong format version in files/dilbert/feedinfo" | 16:46 |
| viric | ha. I'll try | 16:46 |
| viric | do you use trunk? | 16:47 |
| viric | ah 1.3 is latest | 16:47 |
| kyak | viric: opml is basically is just xml.. perhaps it's a mess, i don't know - but it's an "industry standard" for RSS :) i don't see it a big problem, i can convert opml to offrss format though.. | 16:47 |
| kyak | im usin 1.3 | 16:47 |
| viric | ah you started "-w" without "-u" before? | 16:48 |
| kyak | ok, it updated | 16:48 |
| kyak | yes | 16:48 |
| kyak | this is what i did :)) | 16:48 |
| kyak | i then i pressed "update" in web ui | 16:48 |
| viric | ah | 16:48 |
| viric | and you got wrong format? | 16:48 |
| viric | hm | 16:48 |
| kyak | it's back to normal now | 16:49 |
| viric | ah | 16:49 |
| viric | :) | 16:49 |
| kyak | but i think i can reproduce it.. i'll try to remove files and run again | 16:49 |
| kyak | so yeah, just run offrss -w in a clean directory and press "Update feeds" in web ui | 16:50 |
| viric | I can't see the format thing. But the Assert yes | 16:50 |
| kyak | i don't see the format thing now as well :) | 16:50 |
| kyak | ok | 16:50 |
| kyak | you need to press "use remote images" | 16:51 |
| kyak | then you will see it | 16:51 |
| viric | Ok | 16:51 |
| viric | on dilbert? | 16:52 |
| viric | can't see it | 16:52 |
| kyak | yep | 16:52 |
| kyak | on dilbert | 16:52 |
| whitequark | uh rss fetcher in C | 16:52 |
| kyak | Error parsing url http://img72.imageshack.us/img72/1486/angstvorclowns.jpg: curl error 28 | 16:52 |
| whitequark | I wonder how much security holes it has | 16:52 |
| kyak | this is another message | 16:52 |
| viric | whitequark: noone knows | 16:52 |
| whitequark | viric: :( | 16:53 |
| viric | will fix once reported. :) | 16:53 |
| kyak | don't use rss feeds you don't trust :) | 16:53 |
| viric | kyak: ah yes, fails to download it I think. | 16:53 |
| viric | I've seen much more security holes in php than in offrss. ;) | 16:54 |
| viric | even ruby. | 16:54 |
| kyak | viric: sometimes it just sits in "Updating..." forever.. iguess it's because of errors | 16:54 |
| viric | kyak: network timeouts | 16:54 |
| viric | the timeout is hardcoded somewhere, if you don't have patience. | 16:54 |
| kyak | ok.. anyway.. let's see how it goes.. i've packaged offrss for arch linux in the meantime.. https://aur.archlinux.org/packages/offrss/ | 16:55 |
| viric | Ah brave | 16:55 |
| viric | o/ | 16:55 |
| viric | are you an arch maintainer? or anyone can package that? | 16:55 |
| kyak | it's user repository | 16:55 |
| viric | ok | 16:55 |
| viric | haven't you added podofo? :) | 16:56 |
| kyak | nope :) | 16:56 |
| viric | It's a nice feature. | 16:56 |
| kyak | i decided to go one bit at a time | 16:56 |
| viric | :) ok | 16:56 |
| kyak | btw, i'm building as make PDF_CFLAGS= PDF_LDFLAGS= PDF_OBJECTS= | 16:56 |
| viric | fine | 16:56 |
| kyak | probably easir than to comment those things | 16:56 |
| viric | up to you | 16:56 |
| viric | Maybe I should use some USE_PODOFO=1 or so | 16:57 |
| kyak | oh yeah, would make sense | 16:57 |
| kyak | viric: is it possible to use spaces in "feedid"? | 17:01 |
| kyak | gives me some errors | 17:01 |
| viric | kyak: no. spaces separate from url | 17:01 |
| kyak | hmm | 17:02 |
| viric | kyak: updated the Makefile thing in trunk | 17:02 |
| kyak | thanks! | 17:02 |
| kyak | i see.. you use that as a file name | 17:03 |
| viric | yes | 17:03 |
| kyak | and you are blacklisting many symbols :) | 17:04 |
| viric | for people like whitequark | 17:04 |
| viric | :) | 17:04 |
| viric | I don't remember much of the code now. hehe | 17:05 |
| viric | it just worksf or me. | 17:05 |
| kyak | luckily, you don't have to remember the code for it to keep working :) | 17:05 |
| viric | nah, it's C. | 17:06 |
| viric | It's not like .. the server upgraded from php4 to php5, or python2 to python3, ... | 17:06 |
| kyak | viric: luckily, you missed the '\t' | 17:15 |
| kyak | which works fine instead of the space :) | 17:15 |
| kyak | now i have "dilbert?arse" directory (or dilbert^Iarse) | 17:16 |
| kyak | no, probably not so fine | 17:17 |
| kyak | the tab gets swallowed in web interface | 17:17 |
| viric | ha | 17:20 |
| viric | :) | 17:20 |
| kyak | what google does.. is it uses the feed url in address bar, not feed name | 17:20 |
| viric | google? | 17:21 |
| kyak | gogole reader | 17:21 |
| viric | address bar? | 17:21 |
| kyak | yes.. | 17:21 |
| kyak | i mean, you could use http://localhost:8090/http://feeds.feedburner.com/DilbertDailyStrip instead of http://localhost:8090/dilbert | 17:22 |
| kyak | of course, url encoded | 17:22 |
| viric | ah | 17:22 |
| viric | tricky google. | 17:22 |
| kyak | or even could use some internal auto-generated identifer | 17:22 |
| kyak | but these are fantasies.. | 17:23 |
| viric | is this mess in order to get a unique string from opml? | 17:23 |
| kyak | i thought dilbert%20arse would work.. but it displays the literal string. ok | 17:24 |
| kyak | viric: i want to have spaces in feed names :) | 17:24 |
| kyak | likes "Films 2013" | 17:25 |
| viric | hm ok | 17:25 |
| viric | fair | 17:25 |
| viric | filesystems can handle spaces | 17:25 |
| kyak | not "Films_2013" | 17:25 |
| viric | I use underscore | 17:25 |
| viric | let me check | 17:25 |
| viric | I can reproduce the wrong forma | 17:34 |
| kyak | meanwhile, i was trying to exclude space from banned chars and use | as a separator of feed name and feed url.. but it breaks something - i see empty feeds | 17:35 |
| viric | kyak: https://viric.name/cgi-bin/offrss/info/91f9fc3e38 | 17:36 |
| kyak | lemme try this | 17:37 |
| kyak | what about char bad[]? | 17:38 |
| viric | annoying in bash | 17:38 |
| viric | let me check that blacklist :) | 17:39 |
| kyak | perfect, it works! | 17:39 |
| viric | the title_to_filename thing is not for your feed names | 17:40 |
| viric | it's for the posts | 17:40 |
| kyak | ah, ok | 17:40 |
| viric | kyak: https://viric.name/cgi-bin/offrss/info/539a77f2df and this is for the 'wrong file version' thing | 17:40 |
| kyak | viric: did you accumulate enough changes to bump a new version? :) | 17:41 |
| viric | hehe | 17:41 |
| viric | I'll make a new version if you test it more :) | 17:41 |
| kyak | or probably it's better to try it on some real data.. | 17:41 |
| kyak | yeah, that;s what i meant | 17:41 |
| kyak | so now i'm off to convert opml file to offrss format | 17:42 |
| kyak | Error: not authorized to clonet: 0 received: 0 | 17:44 |
| viric | no? ouch | 17:44 |
| kyak | when trying to fossil clone | 17:44 |
| viric | permissions updated | 17:45 |
| viric | you are good at finding problems | 17:45 |
| viric | thank you :) | 17:45 |
| kyak | interesting | 17:45 |
| kyak | the first time i see fossil.. it returned me some kind of "password" | 17:45 |
| kyak | should i keep it? | 17:45 |
| viric | nah | 17:46 |
| viric | you can override it whenever you want | 17:46 |
| viric | if you run "fossil serve", that's the password for the Admin | 17:46 |
| viric | in fossil, each system has users, passwords, permissions, web, ... | 17:47 |
| viric | if you run "fossil ui", you have the whole offrss website offline :) | 17:47 |
| kyak | that's interesting.. you like to use things noone seems to be using :) i mean, fossil, podofo, nix.. Affero GPL | 17:47 |
| kyak | i never heard of these before | 17:47 |
| viric | haha | 17:48 |
| kyak | unison | 17:48 |
| viric | I can't understand the trends | 17:48 |
| viric | kyak: if you have a webserver, remember offrss works as CGI | 17:51 |
| viric | (if you don't want it locally) | 17:51 |
| kyak | i guess i have no other options but to write opml to offrss script using... i don't what what.. D? :) | 17:52 |
| viric | I write most in Go nowadays :) | 17:52 |
| kyak | yeah, that's not mainstream :) | 17:52 |
| viric | offrss would be much less lines in Go | 17:53 |
| viric | but if written in go, I'm not sure if it would work in the nanonote. It's tricky on its virtual memory. | 17:53 |
| viric | kyak: podofo is a lib; not to be known by people other than programmers. That's the easiest thing I could find for the task. | 17:54 |
| kyak | viric: got it about CGI.. | 17:54 |
| kyak | there's a mupdf | 17:54 |
| kyak | it kind works on Nanonote | 17:54 |
| viric | isn't that for reading pdfs? | 17:54 |
| kyak | oh.. | 17:54 |
| kyak | it probably is.. yeah, you need creating | 17:55 |
| whitequark | viric: I'd be interested in results of running go code on NN | 17:55 |
| whitequark | but it doesn't compile to mips iirc | 17:55 |
| viric | whitequark: gccgo does. I use it on mips64. | 17:55 |
| kyak | viric: is it safe to remove the .fossil file now? | 17:55 |
| viric | kyak: no | 17:56 |
| kyak | i thought so.. | 17:56 |
| viric | well, safe for what? | 17:56 |
| viric | You'll have to clone again if you want the reposotiry | 17:56 |
| viric | repository | 17:56 |
| kyak | well, do i need that file to update the repo? | 17:56 |
| viric | yes | 17:56 |
| viric | it's "the repo" :) | 17:56 |
| viric | mercurial and svn use a directory, and in it, .hg or .svn. But fossil stores it to a file apart. | 17:57 |
| viric | you can put it wherever you want though. | 17:57 |
| kyak | how would fossil find that file if i move it? | 17:57 |
| viric | you'll have to open it again | 17:57 |
| viric | :) | 17:57 |
| viric | or checkout | 17:57 |
| kyak | ok.. it's clear | 17:58 |
| viric | the local checkout has a ".fossil" file that points to it | 17:58 |
| kyak | .fslckout in my case | 17:58 |
| viric | ah that. | 17:58 |
| viric | Some relevant people use fossil: sqlite and fossil of course, tcl/tk, ... and some people try to move the openbsd/netbsd/pkgsrc too | 18:00 |
| kyak | ..and i'll write this script in tcl | 18:03 |
| viric | kyak: the pdf generation works nice with the firefox pdf.js :) | 18:03 |
| viric | hehe | 18:03 |
| kyak | what is it? | 18:03 |
| viric | kyak: I mean the in-firefox pdf viewer | 18:04 |
| kyak | but still, i would need to compile with podofo | 18:04 |
| kyak | is it right? | 18:04 |
| viric | yes | 18:04 |
| viric | couldn't be easier | 18:04 |
| viric | :) | 18:04 |
| kyak | podofo is so cool that it's not even in arch user repos | 18:05 |
| kyak | so would have to package it first :( | 18:05 |
| viric | time for 70:CA:0 | 18:05 |
| viric | no? oh. :) | 18:05 |
| viric | see? you found what needs to be fixed in the world, and you can fix it! | 18:06 |
| kyak | yeah, "the world" :) | 18:07 |
| whitequark | uh pdf.js | 18:16 |
| whitequark | "because you totally need 2G of RAM and four-core CPU to open a three-page pdf" | 18:16 |
| lindi- | whitequark: 2 GB of RAM is very cheap, I'll take that any day over having to use some non-free reader | 18:18 |
| whitequark | lindi-: now multiply that by number of pages | 18:21 |
| lindi- | whitequark: that doesn't sound very realistic | 18:21 |
| whitequark | lindi-: nevertheless, the fact that it eats orders of magnitude more resources than it should, is | 18:22 |
| lindi- | whitequark: well it's a tradeoff of course, if it was easy to make it use less resources I'm sure they would :) | 18:23 |
| whitequark | lindi-: that's the point. there is some weird notion going on right now that everything should be written in html5+js | 18:23 |
| whitequark | whereas that's a good choice or not | 18:23 |
| lindi- | whitequark: well, they wrote it. you can write something else if you want? ;) | 18:24 |
| whitequark | lindi-: that is a well-known fallacy ;) | 18:24 |
| lindi- | whitequark: pdf.js in chromium seems to only have RSS of 58 MB with 14-page pdf | 18:25 |
| whitequark | "you don't need to be a chef to recognize burnt food", etc. | 18:25 |
| whitequark | lindi-: yeah, I'm looking at their last version right now | 18:25 |
| lindi- | or hmm, wrong tab maybe | 18:25 |
| whitequark | it's far better than it was. still, their demo page ate 500MB of RAM | 18:25 |
| lindi- | it's probably this one that uses 217 MB | 18:25 |
| whitequark | I also wonder how would it perform on something slower than my top-notch i7 from 2012. | 18:26 |
| whitequark | and with less than my 10GB of RAM, etc. | 18:26 |
| whitequark | lindi-: don't forget the "GPU Process" worth 350MB | 18:27 |
| whitequark | and "Browser" of 400MB | 18:27 |
| viric | that pdf.js doesn't take that much | 18:56 |
| viric | I bet acroread takes much more | 18:56 |
| kyak | viric: can it NOT download images? | 19:55 |
| kyak | i mean, i'm fine using remote images.. but when i do -u, it just downloads the images | 19:56 |
| viric | ahm | 19:56 |
| viric | I never thought of that | 19:56 |
| viric | :) | 19:56 |
| kyak | btw ive converted my opml file | 19:56 |
| viric | I'll try to add some option... | 19:56 |
| kyak | now just waiting for the feeds to update.. these images take a long of time | 19:57 |
| kyak | i'll cleanup the conversion script and give it to you, if you interested | 19:57 |
| viric | add it to the ticket about opml | 19:57 |
| viric | if you don't mind | 19:58 |
| kyak | ok, i'll do that | 19:58 |
| kyak | btw | 20:00 |
| kyak | it seems that OPML is free | 20:00 |
| kyak | http://dev.opml.org/ | 20:00 |
| kyak | i don't know why you stressed it out in the ticket about "non-free" :) | 20:01 |
| viric | me? nonfree? | 20:01 |
| viric | I didn't open that ticket | 20:01 |
| kyak | ah, ok, you were refering to the service | 20:01 |
| kyak | but opml is not only used by google | 20:01 |
| viric | :) | 20:01 |
| kyak | viric: how many last items does it download? | 20:02 |
| kyak | or does it download the whole feed? | 20:02 |
| viric | whole available | 20:02 |
| viric | do people publish long feeds? | 20:03 |
| kyak | it's just that i got >100 of feeds | 20:03 |
| viric | It seems that your rss experience is quite different than mine :) | 20:03 |
| kyak | i need to cleanup... probably :) | 20:03 |
| viric | But it's very refreshing | 20:03 |
| kyak | i think it's ok downloading the whole feed.. but downloading images should be optional | 20:04 |
| viric | agreed | 20:05 |
| viric | maybe optional per-feed? | 20:05 |
| kyak | i don't need that kind of flexibility :) | 20:06 |
| kyak | but why not.. if it's easy to implement | 20:06 |
| kyak | are you thinking about some kind of flag in feedurls.txt? | 20:06 |
| viric | yes | 20:07 |
| viric | what character you don't need' | 20:07 |
| viric | ? | 20:07 |
| viric | :) | 20:07 |
| kyak | | | 20:07 |
| viric | hehe | 20:07 |
| viric | ok, can you give *some*? | 20:07 |
| kyak | i found out i have name having "ยป" in it :) | 20:07 |
| viric | utf8 ftw | 20:08 |
| kyak | i'll get rid of it | 20:08 |
| kyak | yeah | 20:08 |
| kyak | it's utf-8 | 20:08 |
| kyak | not sure.. | 20:08 |
| kyak | so you are asking about the separator between url and the flag" | 20:09 |
| kyak | ? | 20:09 |
| kyak | what if you decide to add another option in a future? you would have to find on more character :) maybe better to keep space and parse string with strrchr from the end? | 20:10 |
| kyak | viric: i added the script to the ticket. Sorry i couldn't find how to attach files | 20:31 |
| kyak | time to sleep now | 20:31 |
| whitequark | viric: who cares about acroread? | 20:33 |
| viric | kyak: no, I mean that a char could introduce keywords. like *noimages* | 20:34 |
| viric | kyak: ahm right, attach disabled to anonymous | 20:35 |
| viric | enabled. | 20:36 |
| kyak | viric: "*noimages*" is OK for me | 20:37 |
| kyak | viric: just found a problem in the script; i would need to replace those & and similar shit from XML with something valid | 20:46 |
| kyak | i'll check it later, need to go now | 20:46 |
| viric | :) ok | 20:46 |
| viric | thank you | 20:46 |
| kyak | np! and thank you! | 20:47 |
| wolfspraul | lekernel: thanks for the heads up. the shop goes down quite a bit lately, I just reset it and it's working now | 23:40 |
| wolfspraul | as for the m1 buyer, does h:d still have stock? I tried to contact them for about a half year or so on various channels, but they were amazingly consistent in not replying :-) so I don't know whether they still have stock and/or are selling | 23:41 |
| --- Sat Mar 16 2013 | 00:00 | |
Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!
