* FatPhil hovers over the 'buy now' button... https://www.osta.ee/en/nokia-n900-125817110.html | 11:23 | |
KotCzarny | kinda expensive | 11:24 |
---|---|---|
FatPhil | does anyone have a spare TV-out cable. None of the ones that I've acquired have had theirs. | 11:24 |
KotCzarny | fatphil: any av cable will work | 11:24 |
KotCzarny | in the worst you will have to swap red/yellow cinch plugs | 11:24 |
FatPhil | It's not cheap, but the last one I bought was 50e, and that was about 4 years ago. | 11:25 |
KotCzarny | in .pl you can get almost-unused condition for that price | 11:25 |
bencoh | KotCzarny: that's not really the case in the rest of the world | 11:26 |
KotCzarny | :) | 11:26 |
bencoh | FatPhil: that sounds like a good deal I'd say | 11:26 |
KotCzarny | shall i start snagging them for #maemo folks? | 11:26 |
KotCzarny | oh, wow, nokia n-gage (original one) for ~170usd | 11:28 |
FatPhil | I'm very tempted. I will look at my stash of spare parts, and work out how many "whole" ones I have now. g/f will need replacing in about 2-3 years. I hope this new one will last 5, and I know I have at least one whole spare, so it's not a problem for another 5-7 years. | 11:28 |
KotCzarny | madness | 11:28 |
bencoh | KotCzarny: I guess some think of it as a collectable now a days | 11:28 |
KotCzarny | quite nice condition tho, almost new | 11:29 |
FatPhil | Some of the really old ones do seem to have a retro cult following, and prices that match | 11:29 |
FatPhil | 7110for 255e (the slide mouthpiece, oversized screen (for the day) and WAP) here | 11:31 |
KotCzarny | madness. | 11:32 |
FatPhil | KotCzarny: what are BL-5J prices like in .pl? All of ours have turned to pillows. | 11:32 |
KotCzarny | batteries are usually bought from china anyway | 11:33 |
KotCzarny | so you might refer to: | 11:33 |
KotCzarny | ~polarcell | 11:33 |
KotCzarny | ~ping | 11:33 |
KotCzarny | drat. bot missing again | 11:33 |
bencoh | inf^I^I damnbot :) | 11:33 |
FatPhil | I did pick up a few cheap clone batteries, but they didn't last very long. Probably better VFM than the real ones, as the price difference was 5:1. | 11:34 |
KotCzarny | many folks here bought polarcell brand and i havent seen any complaints, only positives | 11:34 |
FatPhil | oooh, /me reads https://talk.maemo.org/showthread.php?t=65568 | 11:35 |
KotCzarny | you should check dual 18650 mod for n900 too | 11:36 |
KotCzarny | ;) | 11:36 |
KotCzarny | 5200mAh baby, lasts for weeks ;) | 11:36 |
KotCzarny | http://talk.maemo.org/showthread.php?t=86571 | 11:37 |
KotCzarny | ;) | 11:37 |
KotCzarny | more if you can source good 18650 | 11:39 |
FatPhil | "OMG it looks like a bomb made to detonate if you place a call..." :-D | 11:42 |
FatPhil | I think i discovered a bug the other day. My old back lost the little metal part - I think it contained a magnet for the Hall sensor which should tell the system to unmount the uSD card. | 11:47 |
FatPhil | but my card was mounted anyway, even without the magnet. | 11:48 |
KotCzarny | sensor might be dead too | 11:48 |
KotCzarny | or simply something prevented card from umount | 11:49 |
KotCzarny | maybe swap? | 11:49 |
FatPhil | that's mostly from rusty memory, but I think it's correct. Alas there's not enough logging options available in stock code. | 11:49 |
KotCzarny | it's hackable shell script | 11:51 |
* FatPhil likes scripts. if your script isn't fast enough, optimise the interpreter/pcode runner, and magically you've made all scripts faster! | 11:59 | |
FatPhil | (assuming step 0 - ensure yor script isn't braindead in the first place) | 12:00 |
Maxdamantus | 21:49:10 < KotCzarny> or simply something prevented card from umount | 12:38 |
Maxdamantus | Very unlikely to be software at least. | 12:38 |
Maxdamantus | since the mmc driver itself just disables access based on the sensor. | 12:38 |
Maxdamantus | no usage checking or anything. | 12:39 |
KotCzarny | does it? i always thought it's just a gpio for os to trigger some actions | 12:40 |
Maxdamantus | It does. Unless you modify the omap_hsmmc driver, it's basically impossible to maintain access to the card. | 12:40 |
KotCzarny | does it at least finish pending writes? | 12:41 |
Maxdamantus | Probably not, depending on what you mean by "pending writes" | 12:41 |
KotCzarny | the writes that went from os into driver | 12:42 |
Maxdamantus | Don't think so. | 12:42 |
Maxdamantus | https://android.googlesource.com/kernel/omap/+/glass-omap-xrw14b/drivers/mmc/host/omap_hsmmc.c#1429 | 12:44 |
Maxdamantus | So it just sets that `protect_card` flag, which causes any requests to fail. | 12:47 |
Maxdamantus | I think I remember ke-recv or something also responding to it in userspace, but ultimately the card is blocked in the kernel anyway at that point. | 12:49 |
Maxdamantus | so it might be able to additionally force unmount it, but it can't do a clean unmount. | 12:50 |
Maxdamantus | https://talk.maemo.org/showpost.php?s=e1199d2319f601c0662ec53d481f22bd&p=1457369&postcount=4 | 12:51 |
KotCzarny | it should try to umount card with 1-2s timer before block | 12:52 |
KotCzarny | and signal with flashing red led while doing so | 12:52 |
KotCzarny | that would be the best course of action | 12:53 |
Maxdamantus | I imagine the feature is pretty much completely pointless anyway. | 12:53 |
KotCzarny | nah, it has it's uses | 12:53 |
Maxdamantus | afaict, the driver is able to reset the state of the controller anyway. | 12:54 |
KotCzarny | yeah, but you lose data | 12:54 |
Maxdamantus | so the only thing I can see it being useful for is allowing the memory card itself to try and persist its writes to non-volatile storage before it loses power, but the stuff in the card's memory is only one part of the chain. | 12:54 |
KotCzarny | with umount and visual notification for the user it allows for safe eject | 12:54 |
Maxdamantus | Sure, I'm not arguing against a safe unmount. I'm arguing against a driver-level block of the card. | 12:55 |
Maxdamantus | because there's no point in time where it makes sense to prevent an unsafe unmount. | 12:55 |
Maxdamantus | if the card is still physically in the slot, why not give it some further chance to unmount? | 12:56 |
Maxdamantus | Personally, I don't even really like the automatic unmounting functionality. I basically never take my card out anyway. | 12:56 |
Maxdamantus | (and I have swap on my card) | 12:57 |
KotCzarny | you dont know how fast user ejects, or if he/she just wants to blow the dust off | 12:57 |
FatPhil | but I bet you don't take the back off often either | 12:57 |
Maxdamantus | (my device doesn't unmount when I take the back cover off; I did that modification to the driver as mentioned in the TMO post, and I might have done something with the mount to make it not automatically unmount, can't remember) | 12:57 |
FatPhil | has that propagated to the power kernel? might that be why the hall sensor did nothing? | 12:58 |
KotCzarny | the only time when i take the back off is when i change battery or clean the insides | 12:58 |
KotCzarny | nah, it's his personal modification | 12:58 |
Maxdamantus | er, no point in time where it makes sense to prevent a safe unmount* | 13:04 |
FatPhil | at no point does it make sense to prevent an atomic write, alas it's the non-atomic ones that are the problem. | 13:11 |
FatPhil | but at that point, you're trying to prevent the user from harming himself when he does something a bit dumb. perhaps training him not to be dumb is better. | 13:12 |
sixwheeledbeast | search "wannasee electronic" re: polarcell bl5j | 13:12 |
FatPhil | sync on cover removal seems more sensible than do nothing on cover removal | 13:13 |
sixwheeledbeast | 13euro | 13:13 |
sixwheeledbeast | IIRC | 13:13 |
FatPhil | sixwheeledbeast: someone here's claiming to have real nokias for that price | 13:13 |
sixwheeledbeast | "real" ones have less capacity | 13:13 |
FatPhil | which comes as a surprise, they used to be ~35 here | 13:14 |
FatPhil | real ones are sometimes sanyo, and sometimes samsung, so not guaranteed to be completely equal. | 13:14 |
FatPhil | (and probably other OEMs, I've not peeled the labels off many) | 13:15 |
sixwheeledbeast | polarcell are a good source of new batteries. I can't imagine Nokia make them any more and while they have 10 year shelf life I would prefer to promote the slightly higher capacity still manufactured product | 13:19 |
bencoh | oh and, finding real nokias can be a pain anyway | 13:38 |
bencoh | I first bought two "genuine" batteries for my n900, I never managed to charge them over ~800mAh | 13:39 |
bencoh | then moved to polarcell | 13:39 |
Maxdamantus | 23:11:08 < FatPhil> at no point does it make sense to prevent an atomic write, alas it's the non-atomic ones that are the problem. | 13:59 |
Maxdamantus | No hardware guarantees collections of writes to be atomic though afaik | 13:59 |
Maxdamantus | to the extent that any hardware guarantees atomicity of writes, it's basically just writes of sectors on the device. | 14:00 |
Maxdamantus | the primitives you normally have to work with is basically just that (blocks are written atomically), and you usually also have some way of knowing that writes up until a particular point in time have been persisted on the device. | 14:01 |
Maxdamantus | (the latter is basically the hardware version of fsync) | 14:02 |
FatPhil | anyone got any cranky funky old browser on their phone, and wanna help my webpage be maximally readable? | 14:07 |
KotCzarny | install netscape navigator 3.x | 14:09 |
KotCzarny | and have fun | 14:09 |
sicelo | FatPhil: you have MicroB .. old enough :) | 14:09 |
KotCzarny | sicelo: you havent seen old unless you've seen nn3 | 14:10 |
KotCzarny | (there are older, but i've spent lots of time with nn) | 14:10 |
FatPhil | I would like arrows rather than +/=/- signs in the change indicators here: http://fatphil.org/beer/stats/Germany_2019-08-01_total.html | 14:10 |
KotCzarny | use utf? | 14:11 |
FatPhil | at the moment that's a mess, it has both the unicode arrows and the ascii. | 14:11 |
FatPhil | if arrows work everywhere (not just rendered, but rendered and readable and meaningful) then I'll ditch the +/=/-. Works on w3m (pc) and lynx (n900) | 14:12 |
KotCzarny | i've found to not bother | 14:13 |
KotCzarny | if you want them to work everywhere, use buttons with images and proper alts | 14:13 |
FatPhil | I do like the pointing arrow (or triangle) motif in such contexts. | 14:14 |
brolin_empey | KotCzarny: Netscape Navigator 3.x is not released for GNU+Linux on ARM. Note the “on their phone” part of the question asked by FatPhil. Yes, a desktop computer can be used as a telephone too but I think “phone” as used by FatPhil in this case means a handheld computer. | 14:14 |
FatPhil | I'd like to avoid images, if possible (even data url ones) | 14:14 |
FatPhil | oh, if you've got an ancient cranky pc browser, feel free to try on that too. | 14:15 |
FatPhil | I reckon that if lynx copes, everything will, but I might be wrong. | 14:15 |
FatPhil | it totally screws up links! | 14:16 |
FatPhil | the browser /links/, that is | 14:16 |
KotCzarny | links need proper terminal and settings | 14:16 |
KotCzarny | and if you used any js you need elinks | 14:17 |
FatPhil | Friends don't let friends do any JS ;-) | 14:18 |
brolin_empey | KotCzarny: The oldest Web browser is WorldWideWeb.app from 1990 because it is the original Web browser. | 14:19 |
sixwheeledbeast | single arrows would be clearer than double | 14:24 |
sixwheeledbeast | css arrows would be better | 14:32 |
sixwheeledbeast | symbol and number together probably makes it look too busy | 14:34 |
FatPhil | good suggestion. I've tweaked the generator to make swapping in new =+- triplets trivial, I'll try single arrows and triangles, as I've seen triangles a lot | 14:40 |
sixwheeledbeast | FatPhil: was imagining this https://pastebin.com/SnKXr75r re arrows | 15:02 |
FatPhil | looks great in a graphical browser, but my g/f will be using lynx on her n900, and the arrows don't come out at all, some alt text might help, I'll add some. | 15:10 |
sixwheeledbeast | hmm | 15:11 |
sixwheeledbeast | I am surprised it's pretty basic css, alt text could be + - = I suppose | 15:12 |
FatPhil | I'll get a blind mate to send her screenreader over some of the alternatives too. I want up and down semantics. | 15:12 |
sixwheeledbeast | oh I forgot lynx is text only | 15:14 |
FatPhil | it maps lots of unicode things onto plain old text, and this is the best of the batch so far: http://fatphil.org/beer/stats/Germany_2019-08-01_total.html | 15:28 |
sixwheeledbeast | much better than before | 15:40 |
FatPhil | Thanks for the feedback. I already sent that one to my accesibility hit-squad for their opinion on it. Nothing else coped on the n900 apart from that. | 15:42 |
sixwheeledbeast | ▬ for none movers? | 15:43 |
FatPhil | blank says everything that absolutely needs to be said! | 15:43 |
FatPhil | now trying n900 ssh to home PCs, and w3m/elinks, as that's something I quite often seem to do... | 15:44 |
sixwheeledbeast | i suppose | 15:45 |
FatPhil | It's incredible how many weird platform users I know. I have 3 other IRC channels I'll now go and pester with the same question. I'm sure someone will tell me it's terrible in BeOS or OS/2 Warp! | 15:51 |
FatPhil | as n900 users (or at least fans) I gave you priority :) | 15:51 |
bencoh | priority for? | 15:53 |
brolin_empey | BeOS and OS/2 are operating systems, not Web browsers. | 15:58 |
sunshavi | FatPhil: I have w3m running on maemo and also within emacs23 on maemo | 16:12 |
DocScrutinizer05 | oldest web app, thought that was Mozaic | 16:14 |
bencoh | same | 16:15 |
DocScrutinizer05 | https://en.wikipedia.org/wiki/Mosaic_(web_browser) | 16:16 |
DocScrutinizer05 | 92 | 16:16 |
bencoh | Mosaic was preceded by WorldWideWeb, the lesser-known Erwise[6] and ViolaWWW. | 16:16 |
bencoh | well :) | 16:16 |
DocScrutinizer05 | https://en.wikipedia.org/wiki/Web_browser#History | 16:17 |
DocScrutinizer05 | >>The first web browser, called WorldWideWeb, was created in 1990 by Sir Tim Berners-Lee.[5]<< | 16:18 |
Elleo | https://worldwideweb.cern.ch/browser/ <-- you can play with an emulation of that here | 16:18 |
bencoh | :) | 16:36 |
FatPhil | bencoh: for being guineapigs who are likely to have weird browsing capabilities! | 17:01 |
bencoh | :] | 17:02 |
bencoh | how thoughtful! we really appreciate it :) | 17:03 |
FatPhil | brolin_empey: occasionally, OSes ship with browsers. They even get taken to court for doing this, if they do it with too much lock-in. | 17:03 |
FatPhil | criminey - the .ee auction site has one of these on sale for 120e - I've (fortunately) never seen it before: https://en.wikipedia.org/wiki/Nokia_7280 . I think I've just thrown up a little in my mouth. | 17:04 |
bencoh | wtf o.O | 17:09 |
KotCzarny | bencoh, lol, you seriously never seen it before? | 17:10 |
bencoh | I don't think I had until now | 17:13 |
FatPhil | we humans are blessed with something called "selective amnesia" - tomorrow you will have purged it from your memory. | 17:25 |
FatPhil | or at least I hope I will have | 17:25 |
sicelo | ~techstaff | 17:30 |
sicelo | hah, i also never saw that 7280 in person, although i've seen the pics. | 17:31 |
sicelo | xes: in page, http://maemo.org/downloads/Maemo5/desktop/ , clicking Next Page or a specific number page results in "connection time out" - this happens for the other category pages as well. | 17:33 |
sicelo | ok, scratch that ... if i'm on a socks5 proxy it works. so i guess problem is in my network (no idea how/why) | 18:43 |
bencoh | looks like we have another garage/builder issue though ... I pushed a packet a few days ago and forgot about it, but it doesn't show up | 18:53 |
bencoh | (xdg-utils) | 18:53 |
KotCzarny | sicelo: maybe your operator drops some known packets? | 19:29 |
sicelo | i think the network is just not stable. actually even with the socks proxy, that page is still very slow - so there is some problem on the site iteself | 19:31 |
KotCzarny | is it only one particular site? | 19:33 |
* DocScrutinizer05 muses about a new law yet to get implemented, that makes code publish mandatory after 7 years no matter what the code for | 19:55 | |
sicelo | KotCzarny: well i do have odd issues from time to time with different websites, including google sites. | 19:56 |
sicelo | KotCzarny: does this work well for you? https://maemo.org/packages/source/view/fremantle_extras_free_source/advanced-clock-plugin/0.15.3/ | 20:20 |
KotCzarny | it asks for login | 20:21 |
KotCzarny | which i apparently forgotten | 20:21 |
sicelo | alright. it's one of the pages which don't seem to be working OK. | 20:22 |
sicelo | oh, plain HTTP seems to be working | 20:23 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!