Re:PowerPC Linux users had compiled boot 'scripts'
on
Booting Linux Faster
·
· Score: 1
>My *cheap ass* Athlon 1700 gentoo box (total system value: 300$) can *reboot* gentoo in under 30 seconds, INCLUDING X. What else do you want?
Efficency, if it is there to be had.
Any improvements made to the boot times will affect all kinds of systems, from an old P100 (I have one runing Linux) to an Athlon.
Embedded Linux developers go through all sorts of hoops to clean up the boot system for each device. It's not so much that they need a less flexible system (tho that is true to a degree), but because the Linux/UNIX boot system is so damn bloated.
PowerPC Linux users had compiled boot 'scripts'
on
Booting Linux Faster
·
· Score: 5, Interesting
This was three years or more ago, but I remember one of the PPC Linux developers "converted" all his system boot scripts in init.d to compiled C.
Boot times went from about 2 minutes, to 35 seconds.
(It took "so long" because it was an old PPC 601 60MHz or something like that).
Distributions such as Mandrake and Gentoo claim they go the extra mile for "performance". I've wondered why neither has cleaned up their boot process.
You wouldn't think Bash is slow from interactive use, but it really it. Piggyback on that speed problem that too many "functions" (OK, *commands*) are standalone executables... greate sub-process, collect result, destroy, rinse repeat.
This is pretty interesting stuff, and I applaud this guys efforts. INIT script achitecture is pretty thankless stuff...no "glory". Fixing this would be like someone fixing fdisk... no one wants to touch the damn stuff...
>a person has, say, a 386 box with 8 megs of RAM and a 340 meg hard drive
I still do not see how my suggestion is less appropriate for really old systems.
Either way, you're going to have some "extra work" during or after the install. You'll hit old distro bugs and you won't even be able to get basic newsgroup or IRC support for it. And in the end, you'll likely a possibly insecure system that I wouldn't even trust behind a firewall.
Linux-On-A-Floppy, for your example, would be a starting point.
A better choice, since you suggest a specific hardware platform, is Debian. It's *very* easy to get a base system running at 300 megs even on the latest versions (or it was a couple of years ago... I tried it!).
Debian also provides boot-floppy images in half density and 5.25 inch disks. The tools to make these images are still maintained, so you could modify them if you have a really old setup no one makes images for.
Plus Debian has "diet libc" and "busybox" packages so you can *really* slim down applications.
Heck, I have a working Redhat 8.0 system that runs in 340 megs (OK, it's got 128 RAM but I'm positive it could handle 8 MB).
I don't really know Slackware well to be fair to it, but I don't buy the arguments that older distros are more appropriate than new ones.
Now, if we are talking XFree86 v4 drivers... that's another ballgame, but XFree86 3.x is still around and as far as I know it's compatible with 2.4x kernels (which run in 8MB).
Don't mess with one of these old distributions. Seriously... things were broken in the old days, and often you had a bear of a time even figuring out what was wrong.
And good luck getting any answers!
If you want to go through the pain of this for HISTORICAL value... do so if you really really want to. Just don't put it on the net, ok?:-)
If the intent is to squeeze some practical value out of an old system, then ignore these old distros and get something made for the job. One of the "Linux on a floppy" or "peanut" Linux distros would do nicely.
A really fun exercise would be "porting" all of today's "modern" Bash scripts to run on an embedded or stripped-down system.. nothing works because everyone uses the newer Bash coding styles (while still specifying the script as/bin/sh grrr...).
A system built around BusyBox and dietlibc is pretty minimal. Just expect to learn a lot of the "old" command switches, and other workarounds...
>>2. uClinux is not readily hackable, at least until you drift of it, and also know how to recover when this thing freezes. You can not just dive into it as if it was a linux PC.
>You have never worked on any embedded system, for pay or otherwise.
Be nice. I think what the author says is true... from the average Linux hacker's perspective, an embedded platform is NOT readily hackable.
Plus he will have no documentation from the manufacturer. The manufacturer likely modified uClinux and possibly BusyBox in some undocumented manner, to "save space".
I just finished a project where I added some CGI GUI's and a dynamic DNS client to a webcamera with embedded Linux. I couldn't have done it if it was not Linux inside, but it was a bear. When you don't know "what works" on an OS variant, trial and error gets old real fast...
The fact that there is no clear and obvious choice for "which" desktop a newbie will use is not the biggest problem with Linux.
(I use "problem with Linux" generally... personally I use one desktop environment, but do not want to see the other go away).
A bigger problem is the difficult of application developers SUPPORTING BOTH (gnome and kde) DESKTOP ENVIRONMENTS WELL. By that I mean registering with both environments, object/component re-use, drag & drop and other events, etc.
Because this is so difficult, almost no one does it.
Freedesktop.org has been good at *gradually* pushing for a merge of "standards" (like clipboards, shortcut icons etc) but not so quickly for my taste.
If I develop and install both KDE and GNOME applications, there should be NO REASON for me to care about petty things like file dialogs, for example. Microsoft sorted this out *years* ago... common dialogs take the appearance of the OS version you are on. I should be able to easily write a GNOME or KDE application (doesn't matter), and conform to some "file dialog API", and then it works best with whatever the end user wants.
Why do people get so caught up in trying to bash the other desktop, to the point where "cooperation" between GNOME and KDE is lip service (individual exceptions made here.. I'm saying there's no big push to cooperatively develop API's).
There is no point in talking about which desktop is "better", or "if there can be only one it should be X". No one is going to budge and everyone has good reasons for preferring one environment over another. If you NEVER want to see standardization, keep fighting over this point. I'm convinced that some folks against standardization actually advocate "all or nothing" approaches because they are unpopular ideas with no risk of getting what you wished for...
Another opportunity for very close work between KDE and GNOME is internationalization. One team could span both desktops and probably offer some valid suggestopms and complaints to developers on how the desktop interfaces could be tightened and made more alike (without sacrificing the individuality that both desktops currently embrace). I could be totally ignorant on translation projects... it may well be that there are many people doing translation for both desktops.
There is just no much that could be standardized: file dialogs, toolbars, installation utilities and front ends, "services" (admittedly a lot of this belongs in INIT but we have some in the desktop and login managers anyways...), interfaces to device drivers (some KDE apps use "formerly GNOME" library gPhoto.. there's a success case), and so on..
Just because you feel your side is superior does not excuse one from working well with others who do not share your views.
from rc.local. I wouldn't even bother with an INIT script until you have something that works.
You could even make a system where you do: (untested bash code) if [ -f `ls/home/user/torrents/sharing/*.torrent` ]; then
for tor in `ls/home/user/torrents/sharing/*.torrent`
btdownloadheadless.py "${tor}" &2>1 >"/home/user/torrents/logs/${tor}.log"
fi; fi;
read as "for torrent, share it"
extra credit: Modify script to check.torrent file's create-date. If older than 14 days (etc), move file out of 'share' dir.
You could even do a wget on a web-based tracker, check for files that have 0 seeds, and if you hav that file then re-enable sharing.
( there's probably a much smarter, lighter way to detect unseeded shares than web scraping. I haven't done this myself so I'm not motivated to look. It will work either way. )
>IMO, what BT needs is an easy way to join a torrent as a sharer at boot time.
Why add an OS-level task to a user application?
Edit your/etc/rc.local to run bittorrent + your.torrent file.
If you're on Windows, it's not as easy. You can stick a.BAT or.vbs file in your StartUp items folder, but of course someone needs to login first (or enable AutoLogin... it's zero security but it does do the job).
>However, I pay $100/mo for my dsl (split with housemates, we all value having a 1.5/768 connection)
Ouch. FYI You can get 1.5/386 from Earthlink DSL for $49.
That's why I haven't switched to SpeakEasy. I know these guys are Linux friendly and people swear by their support. If they had a "no support" option that matched Earthlink's prices I would switch.
386 is more than most people these days and is plenty if you're not running servers that generate constant traffic (like game servers)
Re:OT: I bet your "realtime" MPEG is nowhere near.
on
Jaguar is Over
·
· Score: 1
>Just like Jobs claim that AACs sound BETTER than CDs. Less noticeable artifacts maybe - hell, very likely. Artifact free? No way in hell.
Maybe he meant "AAC ripped from vynal records":-)
OT: I bet your "realtime" MPEG is nowhere near...
on
Jaguar is Over
·
· Score: 0, Offtopic
[ OT ]
>My x86 PC does that trivially. It's a 2-year-old Athlon 1400. I record TV realtime in mpeg4 (2500kbit) and mp3 (160kbit) with 30-40% CPU to spare.
You seem to have lost the key point-- "without noticable artifacts".
Does your *realtime* MPEG4 encoder do realtime, without noticible artifacts? I think not.
For the record, it's possible to get "realtime" MPEG4 encoding that looks good.. but it still won't be artifact-free. That is what they are claiming here.
Lossless and "near lossless" codecs are extremely CPU-bound.
The Mac CPU has struggled to keep up with Intel in terms of MHz, but their "vector processor" is very highly regarded. It runs circles around Intel's SSE by a wide margin. Even with a clockspeed advantage, if SSE2 can theoretically process data as quickly as AltiVec, that would be news to me.
Of course a good VPU system does not automatically mean better quality. But like we saw when MMX arrived... software can all of a sudden do things performance-wise that just was not possible using integer of FPU code.
Existing codecs have been re-fitted and optomized for AltiVec. The next logical question was, if you start FROM SCRATCH... can you design a codec that targets and completely exploits AltiVec? If so, then you have the headroom to do things no one can try on today's CPU and software (for example: use interpolation to completely remove banding and artifacting).
The answer to better quality is not always to throw more MHz at it, or tweak existing codecs. Sometimes a new design is needed.
Having not seen the video quality or the codec in action, I can only guess. The claim to have eliminated artifacting however is very significant.
I was just thinking about this subject recently...
I "started out" at 14 years old with a 64K Atari and Atari BASIC. You guys with C64's, Timex etc. all know the story... you got your computer, traded code with friends and typed in every magazine listing you could find.
Back when there were more OS "choices" the vendors shipped decent languages... and even IMPROVED the languages after they shipped. The language was designed to add value to the computer, and hopefully drive sales. What "bundled" language does that today? Oh, that's right... monopolies do not depend on value;-)
Now we have, what? Microsoft Visual BASIC (for Windows only, not that I want it on Linux). VB's purporse is not to drive sales of computers (or Windows OS licenses), it's just a product. VB users are not quite happy with the MS market-roid decision to fold VB into the.NET either.
Best thing out there for curious coder folks is Python. It is clean, designed to be simple and very high level (more so than the language it is most compared to, Perl). Python favors an "application" approach as opposed to quick string-parsing tools. You have bindings for most popular GUI toolkits, such as GTK1/GTK2, Qt, wxWindows (Win32-ish but portable), Curses etc.
Assuming GTK is installed, I'm having a blast experimenting with GUI code that's at home on Linux or Windows. Yes, GTK runs great on Windows these days (assuming it's installed!:-)
It's still not quite like the 80's, but the open-source tools are the closest thing.
For graphics and sound, the *SDL libraries are nearly universally portable. For Python, you want "PyGame" not PySDL (which is unmaintained).
Nottingham has a huge university (universities?)... there are tons of students and graduates (presumably), so it will be easier to find smart kids to run Linux.
Searching for employees (and replacements) is a big concern for management who entertain the idea of Linux: sure, Microsoft is expensive, proprietary, arrogant and unstable... but MSCE folks are much more common than knowledgable UNIX people.
Many small companies just want day-to-day IT operations like rebooting and virus removal... it is difficult to understand IT issues you cannot see, like proactive network security.
I imagine Nottingham U is on the boat and offering a good suite of UNIX programming classes.
Anyone here from Nottingham?
I worked for Serif (US) in the early 90's, but spent a summer at the Serif Notingham office. I swear... you need two livers to survive in that place.:-)
Yes, $20k is an amazing amount for a car in my position... it's money needed for a house downpayment (or if I had one, extra payments). A decent "new" commuter car is US$15k (with new Kia's sub-$10k)... a Kia + Harley makes more sense to me.:-)
Thanks to the wise folks at Honda and Toyota, I'll eventually see their hybrids at car auctions. I like the Toyota Prius hybrid's drivetrain, but the Honda Insight is a 2-seater (yum!).
> So what about all of those gas powered scooters tht are out there now? Do you see people trying to drive them on the interstate/freeway/highway? No, because it is illegal to have them there. Do people drive gas powered scooters from suburbia to metropolis? No, that isn't what they were designed for. How would thease scooters be any different? Would it suddenly become leagal to drive them on the interstate?
>This is sort of like whining about not being able to row boat to the moon. That isn't what it was intended for.
I don't disagree with what you stated -- but if you re-read my post and the responses, you'll see you took it out of context.
The issue isn't complaining you can't take underpowered vehicles on the freeway... the issue is that there is a dearth of "alternative vehicles" that DO have enough power for the highway.
Notable exceptions of course are the hybrid gas-electric cars that Detroit does not build, but at US$20,000.. are out of reach for most people.
There seem to be a number of "alternative" vehicles, but even if you ignore cost, refueling, etc. in the USA they are impractical for one major reason:
You can't drive them in enough places.
I DON'T expect these things to drive on the interstate highways at 55+ MPH, but...
AFAIK, most of the electric and other alt-vehicles will do 30MPH (48kph). That's great if you ONLY want to drive through your city center but it's useless anywhere else. Not just this vehicle, but lots of alt vehicles are governed to very low speeds, and then you have general cars/motorbikes/trucks. There's nothing in-between.
Most cities have flattened out into suburban sprawl. Here there are as many people (and jobs) AS the city proper. Good luck driving anything limited to 30MPH on a 40MPH road: it may be legal, but you'll probably be pulled over by an ignorant policeman (or one trying to protect you from the 30 cars tailgating and making illegal passes around you).
There are a lot of drivers who believe if you impede their progress, you are stealing their lifeforce and so you are attacking them and they must defend themselves by going into "road rage" mode.
I'm not kidding either. It's perfectly legal to pedal your bike on most non-interstate (highway) roads, but unless it has a wide paved margin/edge like a breakdown lane, good luck... you'll be run off the road. The police generally aren't interested unless you are seriously hurt or someone recently died. In my small city, they ticket bikes on the sidewalk but not hostile drivers.
Sorry for the rant. An alternative-powered scooter would be SWEET if it ran 40mph with a range of 100 miles. Then you could drive to work and back without feeling like you were "asking for trouble" on the road.
The bottom line is that for all users who don't need to stream MPEG directly to the DVD (which probably includes most Linux users), there is very little practical difference between the formats.
First, my post was a retort of some very negative DVD-R comments. A lot of the misinformation about DVD-R/-RW has been spread by the DVD+RW manufacturers, through their website dvdplusrw.org, a site that pretends to be independent and yet existed before the first models existed, and has spread quite a lot of lies.
lies? Yes, like saying:
DVD+RW is "2.4 times faster than DVD-R".
Sounds like a blanket statement to me, and deceptive to leave out DVD-RW 2X statistics
DVD-R cannot record variable bitrate movies (dvd+rw can).
Whups!! This one got pulled. Wish I still had the Google cache
DVD+RW drives will support DVD+R through a firmware upgrade.
Suure it will...:-) We all remember how THAT one was covered up at the request of those that paid dvdplusrw.org's bandwidth bills...
dvd-r lacks error management
Not sure what they're up to here... sounds like they are saying DVD-R doesn't have standard error-correcting bits, which is untrue.
And so on. The DVD Forum and the DVD-R community have no such need to lie: their product arrived on time and with full support of the DVD Forum, instead of rebelling and creating a new format, like the Circuit City/DIVX DVD fiasco.
DVD+RW, however was late even BEFORE the false-starts and further delays... they NEEDED to lie in order to slow DVD-R growth. I certainly held off on my DVD Recordable purchase for 6 months until I got enough information from cdinfo.com to make a judgement.
I stand by my statement -- DVD+RW and +R has wayyy less compatibility. Want proof? Just browse the CD compatability database of cdinfo.com.
Uncle Bob owns a $350 Sony DVD player from 2000 and is NOT going to switch to a $60 Apex DVD player. ANY player anyone can name that supports DVD+R, ALSO supports DVD-R. The reverse is NOT true.
Of course, if both formats work on everyone's you know's players... great! Just get a dual-format recorder that does both DVD-R and DVD+R. That way, when your friends buy blank discs for you to record things, they can vote with their wallet (which probably is not DVD+R at 300% more per disc!).:-)
>DVD+R/+RW is better technically, and doesn't require different discs for different purposes.
Sure DVD+R/+RW doesn't require different discs for different purposes... so long as you stick with DVD+R or DVD+RW. If that was your point, you had no point.
DVD-R/-RW doesn't require "different discs for different purposes" EITHER.
Or were you referring to the niche "DVD Authoring" format? This is irrelivent/FUD for 99.9999999% of DVD recorder owners, since DVD Authoring is a special authoring format not supported outside its small, vertical market.
I suppose in one sense tho that makes DVD-R/-RW "more complicated" than DVD+R/+RW... because there is *no* authoring format for the "+" writers. Sure, you can save some encoded files to DVD+RW... (and probably not proof it in your DVD player unless you just got a DVD player this year... then MAYBE). What your service provider will do is copy your DVD+RW onto a DVD-R Authoring disc because that is the master.
Personally, most people won't consider the advantage of being able to master on DVD-R(A) a negative... most people won't care actually. They just want their movies to work on Uncle Bob's DVD player, puchased 3 years ago... it ain't gonna work with DVD+RW.
It's a valid question... and, Google sucks for finding things like this. Sorry. Trust me.. Google is going to miss a lot of things, or perhaps you would filter them out by your keywords.
"Subject" matter like this can be grouped under a directory, like Open Directori, Yahoo (if they maintained theirs anymore which they don't), LDP, etc.
Anyways, the best place to "find these things" is the FootNotes (GNOME) website.
I think it's funny how people keep associating GTK and Glade with Linux... like BSD, Sun, and Windows don't exist (GTK2 runs great under Windows BTW.. I'm having fun experimenting with PyGTK under both Linux and win32)
You're right... it's pretty difficult to 'sell' something to a customer who doesn't see the need. And once you explain, they'll just fix it (or shrug).
LaughingBoy is also correct (child post)... any good developer realizes a tester with a Quality background is going to make a thorough and methodical pass through the software/site. Anyone can "test", but it's a specialized discipline so the SQE is going to find problems the developer never would.
I can see from my wording, how my meaning was lost. I think this is a MUCH much more probable service IF you can automate it.
Then send your tool out like a spider on the net and email results. You'd have to provide a "score" with encrypted details that only you could read. Maybe a downloadable version for people to run internal tests on.
I have done a bit of web-testing, and I can tell you every time you face a new project, it is a lot of work to automate tools for 'that' project. Designing a "killer" test tool might be a pipe dream... I dunno.. but one can dream.
>These two posts make me wonder: is there a market for "homebrewed website" quality checking? Say, a person with a lot of web development experience (definitely not me: I'm an embedded head:-) who accepts a small fee for testing your site just before you publish it to the world and finds bugs/makes usability suggestions/stress tests it. Anyone want to take this idea and run with it?
Yes, there is a market... but it is a LOT of work. Convincing the operator of this is not easy... you could point out example flaws and they would just say "typo" and fix it. You can't prevent this.
Now... if you can write a software package that does this -- you may have a better shot. Check links, check external links (including if they go offline and become a "search/shopping portal"...) checks images for stretching, estimate efficency improvement if you convert Frontpage-crap site to efficent HTML + style sheets, check for HTML compliance, submit to search engines (yeah, right), and auto-generate some "keywork fishhook" webpages for Google etc.
Automate this, and you can go after all the mom-and-pop websites on the net. How big that market is, I don't know...
> I have an 800 MHz Mini-ITX computer and it does not have the juice for a consistent smooth DivX/MPEG-4 or MPEG-2 playback.
This is certainly true if you use BUILT-IN VIDEO on many of these mini-PC's.
More CPU will help only a little.. faster video hardware's where it's at. A Shuttle mini-PC with an add-in GeForce2 makes all the difference. Built-in video is usually marginal for movie viewing (especially divx).
>>Everyone jumped to the same "genetic" conclusion (women make lousy hunters). It could be as simple as physical and chemical changes after having children (sometimes derridingly called 'placenta brain'): perhaps women's brains go into a rapid form of job-specialization (rearing) which translates into other disadvantages.
>Wouldn't this be caused by differences in genetics? Last I checked women weren't taught to change their brain chemistry growing up after child birth.
No, actually, it would be an environmental change rooted in genetics. Last I checked, women were not required to have children - at all - some do not because it is a choice.
Therefore you have two sub-groups of women, and no scientific data to support this finding applies to both groups.
>I honestly don't understand the problem. Men and women are physically different.
I think everyone is in agreement here.:-)
> Our brains are physically different (neuron density, halve independence, etc.). By default our abilities at birth are different.
But that's not to say our abilities are wholly defined by genetics. For example, environment shapes your abilities. Heck, even nutrition fulfillment shapes us.
You can't jump to scientific conclusion without accounting for and isolating all of the variables.
>That's not to say that the brain isn't an extremely flexible instrument that can't overcome quite a lot of the differences.
Irrelevent to the point, but no disagreement here..
>It just means that if you don't put the extra effort into overcoming the disparity, each gender on average will be better suited to certain roles.
Genetic "gender... roles" is hardly supported here, though people looking for reinforcement of those views will latch onto this.
All I'm saying is this is an interesting subject which is not getting good scientific treatment (or maybe it was, and MS finds the conclusion too controversial & doesn't want to lose sight that this is just a user-interface group).
No, they are not. They *are* expensive in stores.. even the big ones.
I know 4 people with DVDR, and not a single one of them is stupid enough to pay 400% markup in the stores. Quoting a store price is like quoting a car-dealer price, when you can get the service elsewhere. It's a meaningless statistic.
Everyone buys through an online store like supermediastore.com, etc. Online you can get a 25-pack of discs from $20 - $50, or a 100-pack of cheap PrinCo discs for $60 (I go a step up from Princo... Optidisc makes much more reliable discs.)
Sure, you'll wait a few days on the FREE SHIPPING, but it's kind of hard to run out of discs by surprise: they take so long to record, for one.
Slashdot Mirror
>My *cheap ass* Athlon 1700 gentoo box (total system value: 300$) can *reboot* gentoo in under 30 seconds, INCLUDING X. What else do you want?
Efficency, if it is there to be had.
Any improvements made to the boot times will affect all kinds of systems, from an old P100 (I have one runing Linux) to an Athlon.
Embedded Linux developers go through all sorts of hoops to clean up the boot system for each device. It's not so much that they need a less flexible system (tho that is true to a degree), but because the Linux/UNIX boot system is so damn bloated.
This was three years or more ago, but I remember one of the PPC Linux developers "converted" all his system boot scripts in init.d to compiled C.
.no "glory". Fixing this would be like someone fixing fdisk... no one wants to touch the damn stuff...
Boot times went from about 2 minutes, to 35 seconds.
(It took "so long" because it was an old PPC 601 60MHz or something like that).
Distributions such as Mandrake and Gentoo claim they go the extra mile for "performance". I've wondered why neither has cleaned up their boot process.
You wouldn't think Bash is slow from interactive use, but it really it. Piggyback on that speed problem that too many "functions" (OK, *commands*) are standalone executables... greate sub-process, collect result, destroy, rinse repeat.
This is pretty interesting stuff, and I applaud this guys efforts. INIT script achitecture is pretty thankless stuff..
>a person has, say, a 386 box with 8 megs of RAM and a 340 meg hard drive
I still do not see how my suggestion is less appropriate for really old systems.
Either way, you're going to have some "extra work" during or after the install. You'll hit old distro bugs and you won't even be able to get basic newsgroup or IRC support for it. And in the end, you'll likely a possibly insecure system that I wouldn't even trust behind a firewall.
Linux-On-A-Floppy, for your example, would be a starting point.
A better choice, since you suggest a specific hardware platform, is Debian. It's *very* easy to get a base system running at 300 megs even on the latest versions (or it was a couple of years ago... I tried it!).
Debian also provides boot-floppy images in half density and 5.25 inch disks. The tools to make these images are still maintained, so you could modify them if you have a really old setup no one makes images for.
Plus Debian has "diet libc" and "busybox" packages so you can *really* slim down applications.
Heck, I have a working Redhat 8.0 system that runs in 340 megs (OK, it's got 128 RAM but I'm positive it could handle 8 MB).
I don't really know Slackware well to be fair to it, but I don't buy the arguments that older distros are more appropriate than new ones.
Now, if we are talking XFree86 v4 drivers... that's another ballgame, but XFree86 3.x is still around and as far as I know it's compatible with 2.4x kernels (which run in 8MB).
Don't mess with one of these old distributions. Seriously... things were broken in the old days, and often you had a bear of a time even figuring out what was wrong.
:-)
/bin/sh grrr...).
And good luck getting any answers!
If you want to go through the pain of this for HISTORICAL value... do so if you really really want to. Just don't put it on the net, ok?
If the intent is to squeeze some practical value out of an old system, then ignore these old distros and get something made for the job. One of the "Linux on a floppy" or "peanut" Linux distros would do nicely.
A really fun exercise would be "porting" all of today's "modern" Bash scripts to run on an embedded or stripped-down system.. nothing works because everyone uses the newer Bash coding styles (while still specifying the script as
A system built around BusyBox and dietlibc is pretty minimal. Just expect to learn a lot of the "old" command switches, and other workarounds...
"Imagine a Beowulf cluster of these!"??
>>2. uClinux is not readily hackable, at least until you drift of it, and also know how to recover when this thing freezes. You can not just dive into it as if it was a linux PC.
>You have never worked on any embedded system, for pay or otherwise.
Be nice. I think what the author says is true... from the average Linux hacker's perspective, an embedded platform is NOT readily hackable.
Plus he will have no documentation from the manufacturer. The manufacturer likely modified uClinux and possibly BusyBox in some undocumented manner, to "save space".
I just finished a project where I added some CGI GUI's and a dynamic DNS client to a webcamera with embedded Linux. I couldn't have done it if it was not Linux inside, but it was a bear. When you don't know "what works" on an OS variant, trial and error gets old real fast...
The fact that there is no clear and obvious choice for "which" desktop a newbie will use is not the biggest problem with Linux.
(I use "problem with Linux" generally... personally I use one desktop environment, but do not want to see the other go away).
A bigger problem is the difficult of application developers SUPPORTING BOTH (gnome and kde) DESKTOP ENVIRONMENTS WELL. By that I mean registering with both environments, object/component re-use, drag & drop and other events, etc.
Because this is so difficult, almost no one does it.
Freedesktop.org has been good at *gradually* pushing for a merge of "standards" (like clipboards, shortcut icons etc) but not so quickly for my taste.
If I develop and install both KDE and GNOME applications, there should be NO REASON for me to care about petty things like file dialogs, for example. Microsoft sorted this out *years* ago... common dialogs take the appearance of the OS version you are on. I should be able to easily write a GNOME or KDE application (doesn't matter), and conform to some "file dialog API", and then it works best with whatever the end user wants.
Why do people get so caught up in trying to bash the other desktop, to the point where "cooperation" between GNOME and KDE is lip service (individual exceptions made here.. I'm saying there's no big push to cooperatively develop API's).
There is no point in talking about which desktop is "better", or "if there can be only one it should be X". No one is going to budge and everyone has good reasons for preferring one environment over another. If you NEVER want to see standardization, keep fighting over this point. I'm convinced that some folks against standardization actually advocate "all or nothing" approaches because they are unpopular ideas with no risk of getting what you wished for...
Another opportunity for very close work between KDE and GNOME is internationalization. One team could span both desktops and probably offer some valid suggestopms and complaints to developers on how the desktop interfaces could be tightened and made more alike (without sacrificing the individuality that both desktops currently embrace). I could be totally ignorant on translation projects... it may well be that there are many people doing translation for both desktops.
There is just no much that could be standardized: file dialogs, toolbars, installation utilities and front ends, "services" (admittedly a lot of this belongs in INIT but we have some in the desktop and login managers anyways...), interfaces to device drivers (some KDE apps use "formerly GNOME" library gPhoto.. there's a success case), and so on..
Just because you feel your side is superior does not excuse one from working well with others who do not share your views.
Just my $0.02. Flame away if you like.
>I don't know how to start bittorrent from a script and a .torrent file.
/home/user/torrents/sharing/mytorrent.torrent
/home/user/torrents/sharing/*.torrent` ]; then /home/user/torrents/sharing/*.torrent`
.torrent file's create-date. If older than 14 days (etc), move file out of 'share' dir.
I think you can just do:
btdownloadheadless.py
from rc.local. I wouldn't even bother with an INIT script until you have something that works.
You could even make a system where you do:
(untested bash code)
if [ -f `ls
for tor in `ls
btdownloadheadless.py "${tor}" &2>1 >"/home/user/torrents/logs/${tor}.log"
fi;
fi;
read as "for torrent, share it"
extra credit:
Modify script to check
You could even do a wget on a web-based tracker, check for files that have 0 seeds, and if you hav that file then re-enable sharing.
( there's probably a much smarter, lighter way to detect unseeded shares than web scraping. I haven't done this myself so I'm not motivated to look. It will work either way. )
>IMO, what BT needs is an easy way to join a torrent as a sharer at boot time.
/etc/rc.local to run bittorrent + your .torrent file.
.BAT or .vbs file in your StartUp items folder, but of course someone needs to login first (or enable AutoLogin... it's zero security but it does do the job).
Why add an OS-level task to a user application?
Edit your
If you're on Windows, it's not as easy. You can stick a
>However, I pay $100/mo for my dsl (split with housemates, we all value having a 1.5/768 connection)
Ouch. FYI You can get 1.5/386 from Earthlink DSL for $49.
That's why I haven't switched to SpeakEasy. I know these guys are Linux friendly and people swear by their support. If they had a "no support" option that matched Earthlink's prices I would switch.
386 is more than most people these days and is plenty if you're not running servers that generate constant traffic (like game servers)
>Just like Jobs claim that AACs sound BETTER than CDs. Less noticeable artifacts maybe - hell, very likely. Artifact free? No way in hell.
:-)
Maybe he meant "AAC ripped from vynal records"
[ OT ]
>My x86 PC does that trivially. It's a 2-year-old Athlon 1400. I record TV realtime in mpeg4 (2500kbit) and mp3 (160kbit) with 30-40% CPU to spare.
You seem to have lost the key point-- "without noticable artifacts".
Does your *realtime* MPEG4 encoder do realtime, without noticible artifacts? I think not.
For the record, it's possible to get "realtime" MPEG4 encoding that looks good.. but it still won't be artifact-free. That is what they are claiming here.
Lossless and "near lossless" codecs are extremely CPU-bound.
The Mac CPU has struggled to keep up with Intel in terms of MHz, but their "vector processor" is very highly regarded. It runs circles around Intel's SSE by a wide margin. Even with a clockspeed advantage, if SSE2 can theoretically process data as quickly as AltiVec, that would be news to me.
Of course a good VPU system does not automatically mean better quality. But like we saw when MMX arrived... software can all of a sudden do things performance-wise that just was not possible using integer of FPU code.
Existing codecs have been re-fitted and optomized for AltiVec. The next logical question was, if you start FROM SCRATCH... can you design a codec that targets and completely exploits AltiVec? If so, then you have the headroom to do things no one can try on today's CPU and software (for example: use interpolation to completely remove banding and artifacting).
The answer to better quality is not always to throw more MHz at it, or tweak existing codecs. Sometimes a new design is needed.
Having not seen the video quality or the codec in action, I can only guess. The claim to have eliminated artifacting however is very significant.
I was just thinking about this subject recently...
;-)
.NET either.
:-)
I "started out" at 14 years old with a 64K Atari and Atari BASIC. You guys with C64's, Timex etc. all know the story... you got your computer, traded code with friends and typed in every magazine listing you could find.
Back when there were more OS "choices" the vendors shipped decent languages... and even IMPROVED the languages after they shipped. The language was designed to add value to the computer, and hopefully drive sales. What "bundled" language does that today? Oh, that's right... monopolies do not depend on value
Now we have, what? Microsoft Visual BASIC (for Windows only, not that I want it on Linux). VB's purporse is not to drive sales of computers (or Windows OS licenses), it's just a product. VB users are not quite happy with the MS market-roid decision to fold VB into the
Best thing out there for curious coder folks is Python. It is clean, designed to be simple and very high level (more so than the language it is most compared to, Perl). Python favors an "application" approach as opposed to quick string-parsing tools. You have bindings for most popular GUI toolkits, such as GTK1/GTK2, Qt, wxWindows (Win32-ish but portable), Curses etc.
Assuming GTK is installed, I'm having a blast experimenting with GUI code that's at home on Linux or Windows. Yes, GTK runs great on Windows these days (assuming it's installed!
It's still not quite like the 80's, but the open-source tools are the closest thing.
For graphics and sound, the *SDL libraries are nearly universally portable. For Python, you want "PyGame" not PySDL (which is unmaintained).
Nottingham has a huge university (universities?)... there are tons of students and graduates (presumably), so it will be easier to find smart kids to run Linux.
:-)
Searching for employees (and replacements) is a big concern for management who entertain the idea of Linux: sure, Microsoft is expensive, proprietary, arrogant and unstable... but MSCE folks are much more common than knowledgable UNIX people.
Many small companies just want day-to-day IT operations like rebooting and virus removal... it is difficult to understand IT issues you cannot see, like proactive network security.
I imagine Nottingham U is on the boat and offering a good suite of UNIX programming classes.
Anyone here from Nottingham?
I worked for Serif (US) in the early 90's, but spent a summer at the Serif Notingham office. I swear... you need two livers to survive in that place.
-Scott
Nice post; thanks.
:-)
Yes, $20k is an amazing amount for a car in my position... it's money needed for a house downpayment (or if I had one, extra payments). A decent "new" commuter car is US$15k (with new Kia's sub-$10k)... a Kia + Harley makes more sense to me.
Thanks to the wise folks at Honda and Toyota, I'll eventually see their hybrids at car auctions. I like the Toyota Prius hybrid's drivetrain, but the Honda Insight is a 2-seater (yum!).
> So what about all of those gas powered scooters tht are out there now? Do you see people trying to drive them on the interstate/freeway/highway? No, because it is illegal to have them there. Do people drive gas powered scooters from suburbia to metropolis? No, that isn't what they were designed for. How would thease scooters be any different? Would it suddenly become leagal to drive them on the interstate?
>This is sort of like whining about not being able to row boat to the moon. That isn't what it was intended for.
I don't disagree with what you stated -- but if you re-read my post and the responses, you'll see you took it out of context.
The issue isn't complaining you can't take underpowered vehicles on the freeway... the issue is that there is a dearth of "alternative vehicles" that DO have enough power for the highway.
Notable exceptions of course are the hybrid gas-electric cars that Detroit does not build, but at US$20,000.. are out of reach for most people.
There seem to be a number of "alternative" vehicles, but even if you ignore cost, refueling, etc. in the USA they are impractical for one major reason:
You can't drive them in enough places.
I DON'T expect these things to drive on the interstate highways at 55+ MPH, but...
AFAIK, most of the electric and other alt-vehicles will do 30MPH (48kph). That's great if you ONLY want to drive through your city center but it's useless anywhere else. Not just this vehicle, but lots of alt vehicles are governed to very low speeds, and then you have general cars/motorbikes/trucks. There's nothing in-between.
Most cities have flattened out into suburban sprawl. Here there are as many people (and jobs) AS the city proper. Good luck driving anything limited to 30MPH on a 40MPH road: it may be legal, but you'll probably be pulled over by an ignorant policeman (or one trying to protect you from the 30 cars tailgating and making illegal passes around you).
There are a lot of drivers who believe if you impede their progress, you are stealing their lifeforce and so you are attacking them and they must defend themselves by going into "road rage" mode.
I'm not kidding either. It's perfectly legal to pedal your bike on most non-interstate (highway) roads, but unless it has a wide paved margin/edge like a breakdown lane, good luck... you'll be run off the road. The police generally aren't interested unless you are seriously hurt or someone recently died. In my small city, they ticket bikes on the sidewalk but not hostile drivers.
Sorry for the rant. An alternative-powered scooter would be SWEET if it ran 40mph with a range of 100 miles. Then you could drive to work and back without feeling like you were "asking for trouble" on the road.
The bottom line is that for all users who don't need to stream MPEG directly to the DVD (which probably includes most Linux users), there is very little practical difference between the formats.
First, my post was a retort of some very negative DVD-R comments. A lot of the misinformation about DVD-R/-RW has been spread by the DVD+RW manufacturers, through their website dvdplusrw.org, a site that pretends to be independent and yet existed before the first models existed, and has spread quite a lot of lies.
lies? Yes, like saying:
Sounds like a blanket statement to me, and deceptive to leave out DVD-RW 2X statistics
Whups!! This one got pulled. Wish I still had the Google cache
Suure it will...
Not sure what they're up to here... sounds like they are saying DVD-R doesn't have standard error-correcting bits, which is untrue.
And so on. The DVD Forum and the DVD-R community have no such need to lie: their product arrived on time and with full support of the DVD Forum, instead of rebelling and creating a new format, like the Circuit City/DIVX DVD fiasco.
DVD+RW, however was late even BEFORE the false-starts and further delays... they NEEDED to lie in order to slow DVD-R growth. I certainly held off on my DVD Recordable purchase for 6 months until I got enough information from cdinfo.com to make a judgement.
I stand by my statement -- DVD+RW and +R has wayyy less compatibility. Want proof? Just browse the CD compatability database of cdinfo.com.
Uncle Bob owns a $350 Sony DVD player from 2000 and is NOT going to switch to a $60 Apex DVD player. ANY player anyone can name that supports DVD+R, ALSO supports DVD-R. The reverse is NOT true.
Of course, if both formats work on everyone's you know's players... great! Just get a dual-format recorder that does both DVD-R and DVD+R. That way, when your friends buy blank discs for you to record things, they can vote with their wallet (which probably is not DVD+R at 300% more per disc!). :-)
>DVD+R/+RW is better technically, and doesn't require different discs for different purposes.
Sure DVD+R/+RW doesn't require different discs for different purposes... so long as you stick with DVD+R or DVD+RW. If that was your point, you had no point.
DVD-R/-RW doesn't require "different discs for different purposes" EITHER.
Or were you referring to the niche "DVD Authoring" format? This is irrelivent/FUD for 99.9999999% of DVD recorder owners, since DVD Authoring is a special authoring format not supported outside its small, vertical market.
I suppose in one sense tho that makes DVD-R/-RW "more complicated" than DVD+R/+RW... because there is *no* authoring format for the "+" writers. Sure, you can save some encoded files to DVD+RW... (and probably not proof it in your DVD player unless you just got a DVD player this year... then MAYBE). What your service provider will do is copy your DVD+RW onto a DVD-R Authoring disc because that is the master.
Personally, most people won't consider the advantage of being able to master on DVD-R(A) a negative... most people won't care actually. They just want their movies to work on Uncle Bob's DVD player, puchased 3 years ago... it ain't gonna work with DVD+RW.
> It's called google you dumb shit.
It's a valid question... and, Google sucks for finding things like this. Sorry. Trust me.. Google is going to miss a lot of things, or perhaps you would filter them out by your keywords.
"Subject" matter like this can be grouped under a directory, like Open Directori, Yahoo (if they maintained theirs anymore which they don't), LDP, etc.
Anyways, the best place to "find these things" is the FootNotes (GNOME) website.
I think it's funny how people keep associating GTK and Glade with Linux... like BSD, Sun, and Windows don't exist (GTK2 runs great under Windows BTW.. I'm having fun experimenting with PyGTK under both Linux and win32)
You're right... it's pretty difficult to 'sell' something to a customer who doesn't see the need. And once you explain, they'll just fix it (or shrug).
LaughingBoy is also correct (child post)... any good developer realizes a tester with a Quality background is going to make a thorough and methodical pass through the software/site. Anyone can "test", but it's a specialized discipline so the SQE is going to find problems the developer never would.
I can see from my wording, how my meaning was lost. I think this is a MUCH much more probable service IF you can automate it.
Then send your tool out like a spider on the net and email results. You'd have to provide a "score" with encrypted details that only you could read. Maybe a downloadable version for people to run internal tests on.
I have done a bit of web-testing, and I can tell you every time you face a new project, it is a lot of work to automate tools for 'that' project. Designing a "killer" test tool might be a pipe dream... I dunno.. but one can dream.
>These two posts make me wonder: is there a market for "homebrewed website" quality checking? Say, a person with a lot of web development experience (definitely not me: I'm an embedded head :-) who accepts a small fee for testing your site just before you publish it to the world and finds bugs/makes usability suggestions/stress tests it. Anyone want to take this idea and run with it?
Yes, there is a market... but it is a LOT of work. Convincing the operator of this is not easy... you could point out example flaws and they would just say "typo" and fix it. You can't prevent this.
Now... if you can write a software package that does this -- you may have a better shot. Check links, check external links (including if they go offline and become a "search/shopping portal"...) checks images for stretching, estimate efficency improvement if you convert Frontpage-crap site to efficent HTML + style sheets, check for HTML compliance, submit to search engines (yeah, right), and auto-generate some "keywork fishhook" webpages for Google etc.
Automate this, and you can go after all the mom-and-pop websites on the net. How big that market is, I don't know...
> I have an 800 MHz Mini-ITX computer and it does not have the juice for a consistent smooth DivX/MPEG-4 or MPEG-2 playback.
This is certainly true if you use BUILT-IN VIDEO on many of these mini-PC's.
More CPU will help only a little.. faster video hardware's where it's at. A Shuttle mini-PC with an add-in GeForce2 makes all the difference. Built-in video is usually marginal for movie viewing (especially divx).
>>Everyone jumped to the same "genetic" conclusion (women make lousy hunters). It could be as simple as physical and chemical changes after having children (sometimes derridingly called 'placenta brain'): perhaps women's brains go into a rapid form of job-specialization (rearing) which translates into other disadvantages.
:-)
>Wouldn't this be caused by differences in genetics? Last I checked women weren't taught to change their brain chemistry growing up after child birth.
No, actually, it would be an environmental change rooted in genetics. Last I checked, women were not required to have children - at all - some do not because it is a choice.
Therefore you have two sub-groups of women, and no scientific data to support this finding applies to both groups.
>I honestly don't understand the problem. Men and women are physically different.
I think everyone is in agreement here.
> Our brains are physically different (neuron density, halve independence, etc.). By default our abilities at birth are different.
But that's not to say our abilities are wholly defined by genetics. For example, environment shapes your abilities. Heck, even nutrition fulfillment shapes us.
You can't jump to scientific conclusion without accounting for and isolating all of the variables.
>That's not to say that the brain isn't an extremely flexible instrument that can't overcome quite a lot of the differences.
Irrelevent to the point, but no disagreement here..
>It just means that if you don't put the extra effort into overcoming the disparity, each gender on average will be better suited to certain roles.
Genetic "gender... roles" is hardly supported here, though people looking for reinforcement of those views will latch onto this.
All I'm saying is this is an interesting subject which is not getting good scientific treatment (or maybe it was, and MS finds the conclusion too controversial & doesn't want to lose sight that this is just a user-interface group).
>Are DVDs really that expensive in the states?
No, they are not. They *are* expensive in stores.. even the big ones.
I know 4 people with DVDR, and not a single one of them is stupid enough to pay 400% markup in the stores. Quoting a store price is like quoting a car-dealer price, when you can get the service elsewhere. It's a meaningless statistic.
Everyone buys through an online store like supermediastore.com, etc. Online you can get a 25-pack of discs from $20 - $50, or a 100-pack of cheap PrinCo discs for $60 (I go a step up from Princo... Optidisc makes much more reliable discs.)
Sure, you'll wait a few days on the FREE SHIPPING, but it's kind of hard to run out of discs by surprise: they take so long to record, for one.