| brolin_empey | Is it intentional that the marriage symbol = ⚭ = U+26AD appears to resemble the MasterCard logo? | 01:51 |
|---|---|---|
| brolin_empey | Or that the MasterCard logo appears to resemble the marriage symbol? | 01:52 |
| Maxdamantus | Makes me think of Venn diagrams. | 01:53 |
| Maxdamantus | I'm guessing the Venn interpretation is irrelevant and the other symbols are meant to indicate connectedness of chains. | 01:53 |
| brolin_empey | Yes, I thought that the MasterCard logo was a Venn diagram years ago. | 01:54 |
| brolin_empey | Maxdamantus: I finally got around to opening your links to the products to connect a normal SATA drive to an mSATA socket and the SATA host controllers on mini PCI-E cards. These products are neat. Apparently I should be able to connect at least three 2.5-inch or larger SATA drives via SATA to the small, fanless x86-64 computer I linked on AliExpress after all as long as Ubuntu 10.04.x LTS works on this computer that appears to be years newer than Ubuntu | 02:23 |
| brolin_empey | 10.04.x . | 02:23 |
| brolin_empey | I do not want to buy this computer if I have to upgrade from Ubuntu 10.04.x . | 02:25 |
| brolin_empey | https://www.amazon.co.uk/Fanless-Support-windows-usb3-0-Serial/dp/B01GBHEB3W | 03:04 |
| brolin_empey | This computer has the same Celeron J1900 model of CPU and one of the reviewers says it works well with Ubuntu 10.04 but apparently Amazon does not let me contact the reviewer even if I am logged in to the Amazon Web site so I cannot ask the reviewer if they truly meant Ubuntu 10.04 because they mentioned also upgrading to Ubuntu 14.04 but they would have to first upgrade from 10.04 to 12.04 so I do not know if they meant to say they used 10.04 or 12.04 . | 03:04 |
| DocScrutinizer05 | wikiwide: (or whatever nick you're just using ;-P) for supercaps polarity is of vital relevance | 12:31 |
| DocScrutinizer05 | if the mechanical dimensions fit, https://www.digikey.de/product-detail/en/elna-america/DSK-3R3H703T414-HRL/604-1148-1-ND/2171203 looks perfect for bupbat | 12:33 |
| DocScrutinizer05 | (polarity) so make damn sure you get the correct variant of component: *HR or *HL | 12:37 |
| DocScrutinizer05 | for soldering: first plate the PCB pads and the component pins' downsides with a moderately thin layer of solder, add high quality no-clean SMD flux, then place component on PCB and reflow-solder with quite hot soldering iron (I suggest >400°C) first the smaller pin, then the larger, by pressing soldering iron tip wetted with some solder on top of the component pin metal area | 12:43 |
| DocScrutinizer05 | the reflow process to mount the component should not take longer than 2 to 3 seconds max | 12:44 |
| DocScrutinizer05 | overheating the component will cause permanent damage resulting in loss of capacitance and probably leakage | 12:45 |
| DocScrutinizer05 | it's a tad counter-intuitive but the hotter the soldering iron, the less risk to overheat the component on reflow to mount to PCB | 12:46 |
| DocScrutinizer05 | brolin_empey: don't you install a completely fresh OS on any new computer you get, anyway? | 12:49 |
| KotCzarny | doc, no, why? unless its windoze | 13:01 |
| DocScrutinizer05 | I never feel confidence in an installation I didn't do myself | 13:17 |
| KotCzarny | i have one slack install that i just clone if i get new device | 13:17 |
| DocScrutinizer05 | and actually it's way easier to do a full fresh install than any update and cleaning -out | 13:17 |
| DocScrutinizer05 | KotCzarny: as I understand it, cloning an installation isn't always the optimal way to bring up a new hw platform, particularly when it differs significantly in hw features from the donor platform. You might miss out on relevant adjustments made during a regular installation process, like detecting which hw driver kernel modules are needed which which particular tweaks passed tot hem via parameters etc | 14:12 |
| DocScrutinizer05 | extreme example: when the CPU architecture differs completely, like intel -> ARM. No dice a cloned system would work | 14:13 |
| sixwheeledbeast | I generally just rip my /home and dump that in there after fresh start. | 14:14 |
| DocScrutinizer05 | on minor differences the target still may work but in a quite sub-optimal way in some aspects | 14:15 |
| DocScrutinizer05 | sixwheeledbeast: yeah, that's probably the canonical way | 14:31 |
| DocScrutinizer05 | though even that sometimes causes problems when the desktop changed in some way | 14:33 |
| DocScrutinizer05 | often updates run conversion scripts e.g. moving stuff from ~/.kde to ~/.local/kde | 14:36 |
| KotCzarny | sure, compile new kernel then | 14:36 |
| DocScrutinizer05 | actually ~/.local/share/kde5/ | 14:37 |
| DocScrutinizer05 | and a few other new locations | 14:38 |
| KotCzarny | i've tried different wm's, and i came to conclusion that all i need is solved by fluxbox, never looked back | 14:38 |
| DocScrutinizer05 | sucks when that happens | 14:38 |
| DocScrutinizer05 | and I admit I've yet to find a solution to that problem. Maybe there exists one involving special command parameters to apt-get, or something like `source /var/apt/postinstall/*` or somesuch | 14:46 |
| sixwheeledbeast | I have not experienced any of those issues with gnome flavour DE's | 15:12 |
| DocScrutinizer05 | completely unrelated: http://maemo.cloud-7.de/share-service/candle_with_pure_O2.mp4 | 15:23 |
| DocScrutinizer05 | really wow | 15:23 |
| DocScrutinizer05 | camera can't really catch it. The complete wax surface starts burning | 15:24 |
| DocScrutinizer05 | the "feed" is from a straw in a plastic bag I filled from a O2 generator | 15:25 |
| bencoh | do we have a copy of maemo-optify git repository somewhere? | 18:08 |
| bencoh | it was hosted on gitorious, but .... | 18:08 |
| brolin_empey | DocScrutinizer05: That video is neat but would be much better if the camera was held still. ;-) | 18:52 |
| brolin_empey | “don't you install a completely fresh OS on any new computer you get, anyway?” | 18:52 |
| brolin_empey | It depends upon the context/application/use case. One of the benefits of GNU+Linux over Windows is ease of moving an OS installation to a different hardware configuration but Windows seems to be improving in this ability since the Windows NT 6.x era. I often clone a bootable drive to continue using it in a different computer, often with a different hardware configuration. Sometimes I use a clone of a clone (of a clone). In this case, I think this Ubuntu | 18:52 |
| brolin_empey | installation started with 8.04.x LTS when it was current, which means this same OS installation has been in use for over a decade by now so it would be a huge chore to manually recreate the same configuration if I started over with a new OS installation. I want to use a computer to do things I enjoy or to get things done, not to spend weeks or even months on the same old chores over and over, such as starting over with a new OS installation. Another | 18:52 |
| brolin_empey | advantage of GNU+Linux over Windows is that upgrading the OS in place to a newer release is usable feasible. Theoretically, a Windows installation can be upgraded in place to a newer release as long as you continue using the same architecture but, in practice, I have never had this succeed. | 18:52 |
| brolin_empey | s/usable feasible/usually feasible/ | 18:54 |
| infobot | brolin_empey meant: advantage of GNU+Linux over Windows is that upgrading the OS in place to a newer release is usually feasible. Theoretically, a Windows installation can be upgraded in place to a newer release as long as you continue using the same architecture but, in pr... | 18:54 |
| brolin_empey | DocScrutinizer05: Also, I think the computer architecture has to be the same at least in some cases, not only the processor architecture. For example, and yes it is a mostly historical example by now, I do not know if the same m68k GNU+Linux installation can be used on both an Amiga and a Macintosh or some Atari computer or other computer using an m68k processor because the computer architecture is significantly different even if it uses the same processor. | 19:01 |
| brolin_empey | As another historical example, both the PlayStation 2 and SGI workstations use some form of the MIPS architecture and both can run Linux but I do not know if the same Linux installation can work on both of them. | 19:07 |
| brolin_empey | I guess in some cases you may be able to use the same userspace or at least part of the same userspace even if you need different kernels, boot loaders, firmware, and so on for the different computer architectures. | 19:12 |
| brolin_empey | Is it even feasible to use the same Linux kernel on different brands of the same ARM version, such as both the Qualcomm Snapdragon and Samsung Exynos variants of an ARMv7 computer from Samsung, such as the Galaxy Note 3? | 19:17 |
| brolin_empey | Oh, NeXT and Sun originally used m68k too. There used to be lots of different computer architectures using the same processor architecture, such as with the 6502 and m68k processor families. | 19:22 |
| DocScrutinizer05 | bencoh: it's just a few scripts afaik | 19:39 |
| DocScrutinizer05 | bencoh: and iirc it been community-provided anyway | 19:40 |
| DocScrutinizer05 | brolin_empey: >>...not to spend weeks or even months on the same old chores over and over,<< yeah, I try coping with this by more and more scripting of my OS customization | 19:42 |
| DocScrutinizer05 | brolin_empey: actually I'm pretty much in a similar situation like you, just coming from opposite extreme, where I had to realize a lot of my e.g. /etc/* customizations like for init jobs starting my "proprietary" stuff not only didn't magically appear in a new OS installation but actuaööy it wasn't even possible to just copy them from my old system since damn systemd messed up stuff completely | 19:46 |
| sixwheeledbeast | bencoh: I don't see maemo-optify https://web.archive.org/web/20130514233332/http://gitorious.org/community-ssu | 20:00 |
| sixwheeledbeast | bencoh: https://web.archive.org/web/20120423141006/http://maemo.gitorious.org/maemo-af/maemo-optify maybe helpful? | 20:04 |
| DocScrutinizer05 | wow, even perl: https://web.archive.org/web/20110817091119/http://maemo.gitorious.org/maemo-af/maemo-optify/blobs/master/maemo-optify | 20:35 |
| DocScrutinizer05 | anyway | 20:38 |
| DocScrutinizer05 | ~optification | 20:39 |
| infobot | optification is a inventive duct tape workaround to reclaim space in fs root, done due to the fact the systeminit *and* partitioning is FUBAR, http://wiki.maemo.org/Documentation/Maemo_5_Developer_Guide/Packaging,_Deploying_and_Distributing/Installing_under_opt_and_MyDocs, or ""OMG - I wish they looked into FHS and moved /usr to eMMC"", http://www.pathname.com/fhs/pub/fhs-2.3.html#PURPOSE2 bullet1,2 and fhs-2.3.html#PURPOSE16 dot3" | 20:39 |
| DocScrutinizer05 | my very own take on optification ^^^ | 20:40 |
| DocScrutinizer05 | funny enough systemd / freedesktop.org ran into a similar "problem" and solved it by the most idiotic and and ignorant way they could come up with: declaring /usr as obsolete and moving it to root /, the unifying it with /(s)bin | 20:42 |
| DocScrutinizer05 | theN | 20:42 |
| KotCzarny | cool beans | 20:43 |
| KotCzarny | and there are people saying 'systemd is ok because distros use it' | 20:43 |
| KotCzarny | which means, distro maintainers are.. well.. | 20:43 |
| DocScrutinizer05 | 15 trillion flies can't be wrong - shit is delicious | 20:55 |
| jrayhawk | why would you want a separate /usr these days | 21:01 |
| KotCzarny | because it had an idea behind it | 21:02 |
| jrayhawk | it has like three ideas behind it, all of which are fulfilled better by modern initramfses and now cause complication for distro maintainers and vendors | 21:04 |
| KotCzarny | and now you will say they are fullfilled by systemd alone | 21:09 |
| jrayhawk | AFAIK systemd isn't particularly relevant to initramfs | 21:09 |
| jrayhawk | Maybe the udev component. | 21:09 |
| bencoh | sixwheeledbeast: I eventually retrieved it from the repository source package(apt-get source), but thanks :) | 21:12 |
| bencoh | what bothers me is that part of maemo-related git projects are no longer available along with their full git history | 21:13 |
| bencoh | I vaguely recall someone dumped pretty much eveyrthing back then, though | 21:14 |
| Maxdamantus | I just have maemo's root in /home | 21:32 |
| Maxdamantus | so it's all on eMMC | 21:32 |
| DocScrutinizer05 | how's initramfs relevant for partition size? | 21:49 |
| DocScrutinizer05 | http://www.pathname.com/fhs/pub/fhs-2.3.html#PURPOSE18 | 22:13 |
| DocScrutinizer05 | http://www.pathname.com/fhs/pub/fhs-2.3.html#PURPOSE2 >> Rationale: The primary concern used to balance these considerations, which favor placing many things on the root filesystem, is the goal of keeping root as small as reasonably possible. [...] << | 22:16 |
| DocScrutinizer05 | Poetterings argument to ignore this: "who cares, there are no systems anymore that have less than $[whatever randomly picked ridiculous number, like 1GB] sized root partition" | 22:19 |
| DocScrutinizer05 | Poettering further: >>systems like embedded which might actually have a smaller root partition are simply not the target of systemd<< | 22:20 |
| bencoh | at least those would be free of this horror | 22:22 |
| DocScrutinizer05 | he clearly says he simply doesn't care if a standard linux, which according to him (and alas de facto) now comes with systemd, can get installed on those embedded systems. Systemd is *only* caring about deslktop systems, actually not even laptops are getting tested for any problems introduced by systemd to their rather different usage patterns | 22:24 |
| DocScrutinizer05 | which resulted in a guy not being able to boot his laptop at restaurant since the WLAN was different to the one found at home | 22:25 |
| bencoh | :D | 22:25 |
| KotCzarny | they are not the target, yet, being cluelessly used by distros for everything | 22:29 |
| KotCzarny | supah | 22:29 |
| KotCzarny | lies. | 22:29 |
| DocScrutinizer05 | it's this mindset and attitude that makes me hate this guy, the attitude of re-arranging and restructuring large parts of the system core with a bulldozer in a totally careless way, leaving it to a hundreds of more skilled and experienced developers to pick up the pieces and fix the damage done to all the domains Mr Poettering was not interested in | 22:35 |
| KotCzarny | and idiots all over the world quaffing that shit happily | 22:42 |
| KotCzarny | 'because it's in distro's default\ | 22:42 |
| sicelo | spinal has made systemd version of Maemo Leste and reports faster bootup | 22:44 |
| brolin_empey | Should I open a still sealed can of tuna if its “best before” date is almost five years ago? | 22:48 |
| KotCzarny | sicelo: sure, any numbers? | 22:49 |
| KotCzarny | also, how many times do you boot your phone? | 22:50 |
| sicelo | i never got to test it. | 22:52 |
| brolin_empey | A stationary/desktop computer may be used on/in a vehicle, which makes it a mobile computer in some sense, and a portable computer, such as a notebook computer, is often used mostly or even always in the same place, which makes it a stationary computer in practice. | 23:00 |
| brolin_empey | I think I already asked a while ago but I did not receive an answer or I forgot the answer: why does the N900 have multiple flash memory parts instead of using only eMMC? | 23:06 |
| DocScrutinizer05 | "faster boot" is the least argument I could think of, for a init system. Maybe only overtrumped by "shorter init files" | 23:07 |
| brolin_empey | I mean parts as in physical devices such as ICs on the motherboard. | 23:07 |
| DocScrutinizer05 | brolin_empey: the NAND on PoP/SoC is instantly available for boot. eMMC is not | 23:08 |
| DocScrutinizer05 | IOW your bootloader HAS TO go to nand (well, this is a massively simplified picture that actually isn't really true but nevertheless hives best large picture of the concept behind it) | 23:10 |
| brolin_empey | DocScrutinizer05: OK but what about a computer design such as an ARM or x86 Chromebook that uses eMMC as far as I know? Does it still have a separate flash memory part for booting? | 23:12 |
| brolin_empey | I thought that some SoCs can boot directly from eMMC but I may be misremembering? | 23:13 |
| brolin_empey | DocScrutinizer05: Linux is embedded in some products whose end users do not even know that they are actually booting a Linux computer when they power on the device. In this case, it is desirable to have the cold boot time for the OS be as short as is feasible. | 23:17 |
| DocScrutinizer05 | brolin_empey: maemo evolved from 770 to N800/810 to finally N900, and from Bora to Chinook to D...iforgot. And a lot is legacy, you even can find a initfs partition on NAND which is completely unused | 23:17 |
| DocScrutinizer05 | for embedded with a need for ultra low boot time, you boot up a system image basically like resume from suspend2disk | 23:19 |
| brolin_empey | DocScrutinizer05: Diablo. | 23:19 |
| DocScrutinizer05 | Diablo :-D Thanks! | 23:20 |
| brolin_empey | Some computers, such as the Up Board, Latte Panda, and some notebook computers, now run Windows 10 from eMMC. Do these computer still have a separate flash memory part for booting? I have a Latte Panda and Up Board in my office. | 23:23 |
| DocScrutinizer05 | most boot from the one storage device they got, nowadays | 23:31 |
| brolin_empey | I guess the computer boots the motherboard firmware from NVRAM then the motherboard firmware is able to boot an OS from the eMMC. So, in a strict or literal sense, the computer still does not directly boot from the eMMC. | 23:38 |
| DocScrutinizer05 | re maemo with systemd, I doubt this has much in common with maemo then, since in maemo another system occupied cgroups and I'm pretty sure sharing them between systemd and the maemo version of cgroup management won't fly | 23:39 |
| Wizzup | someone has most of leste working with systemd already | 23:39 |
| DocScrutinizer05 | let me guess which are the few negligible parts that are not working yet... Is audio management one of them? OOM another? | 23:41 |
| DocScrutinizer05 | and modem/dialer the third? | 23:42 |
| Wizzup | none of those relate to init system at all | 23:42 |
| Wizzup | modem works fine with openrc and systemd, oom is not relevant, audio management same, really | 23:43 |
| DocScrutinizer05 | and yet I guessed those | 23:43 |
| Wizzup | but sure, you can cherry-pick closed parts ;) | 23:43 |
| Wizzup | closed source parts* | 23:43 |
| Wizzup | I'm sure that the old browser also heavily depends on systemd and that's the reason it's not running ... not the fact that it's closed source, ancient, not worth bringing up in that form, etc | 23:44 |
| Wizzup | heavily depends on upstart* | 23:44 |
| DocScrutinizer05 | hint: the relation to systemd-initsystem is... *cgroups* | 23:44 |
| Wizzup | hint: it doesn't really matter | 23:44 |
| DocScrutinizer05 | sure | 23:45 |
| DocScrutinizer05 | nothing really matters | 23:45 |
| DocScrutinizer05 | seems it helps to adopt a poettering approach to system architecture to adopt systemd | 23:51 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!