Beta
×

Welcome to the Slashdot Beta site -- learn more here. Use the link in the footer or click here to return to the Classic version of Slashdot.

Thank you!

We are sorry to see you leave - Beta is different and we value the time you took to try it out. Before you decide to go, please take a look at some value-adds for Beta and learn more about it. Thank you for reading Slashdot, and for making the site better!

Comments

top

Google's Nexus One, a Steal At $49 Unlocked?

Nevyn Re:Gah! What twisted logic! (311 comments)

"The plan" is a misnomer though, I currently have a Nexus One on a $10 a month plan.

more than 4 years ago
top

Push To End Online Gambling Ban Gains Steam

Nevyn Re:Online gambling is a bad idea. (206 comments)

"So, are you trying to ban etrade.com and "flipping houses"? Or is risk taking in general ok, and you just want to impose your peculiar morality about playing cards on others?"

They are not the same, Equities are investments, I don't know of any gambling site that pays dividends, do you?

Talk to the people who bought Bank of America stock at $40+. See if they think there is any "gambling" (or more commonly called risk in this context) in equities. Might want to ask them how their wonderful dividends are doing too.

more than 4 years ago
top

Android and the Linux Kernel Community

Nevyn Re:Yes, Redhat (354 comments)

"Redhat Enterprise Linux 5" is essentially a massive kernel fork at 2.6.18.

Backporting a lot of patches is not the definition of fork, for any sane person. Esp. when that same group of people are actively working upstream on the latest releases.

Depending on who you ask, RHEL can be more risky than mainline.

Sure, if you ask stupid people questions they can often give you stupid answers. But I've yet to see anyone intelligent run a vanilla upstream kernel in a production environment. Google are probably the closest, and they basically have a mini-RHEL kernel team that they employ ... I guess wasting huge amounts of money to do that isn't a big deal when you are Google.

I've definitely had RHEL panics take down production, only to later discover linux kernel bugs that had been fixed in mainline for a while, but that redhat hadn't backported to their ancient linux fork

Shocker, software having bugs. But here's a hint, if we play chess 30 times and you win 1 match ... that doesn't mean we are both winners, "depending on who you ask". Maybe you really are trying to say that RHEL vs. vanilla are going to have roughly the same number of bugs (and the same time to get them fixed) ... but you'd be a pretty small minority at that point.

more than 4 years ago
top

Is Programming a Lucrative Profession?

Nevyn Re:Really? (844 comments)

By default vim will make a backup copy of the file you are about to edit, which is why your HD was on fire. You can turn that feature off with -n (or in the config. file, see man vim) ... but then recovery after a crash is impossible. As the original poster said, vim does not load the entire file into memory.

more than 4 years ago
top

Is Getting Acquired Good For FOSS Projects?

Nevyn Re:I missed the memo about IBM (131 comments)

And how is this different from SUN? Oh, wait, SUNs OS is open-source, and, because the platform is open, it is easy enough to put Linux on.

Except Solaris isn't the same as OpenSolaris, in fact one of the biggest complaints I see from the few people still running Solaris is that they still can't get access to the source ... just to see wtf is going on. And OpenSolaris only happened as a last ditch effort before Solaris died to Linux, being forced to do something before you die because you didn't is hardly "leading".

NFS, NIS, Solaris, JAVA, OpenOffice, VirtualBox, MySQL

NFS and NIS can only be classed as open source projects if you are delusional. Yes, they dumped some code to try and make a de-facto std. ... but they were never an upstream. Solaris being open source is a lie. Java only because open source because Red Hat+IBM+etc. were going to create a real open source version, if they didn't. Open Office, you can have (although again, they were pretty much forced to open source it ... and there is no real community, because Sun are clueless about open source). MySQL was bought, complete with community (and only a significant part of Sun if you take their inflated price at face value). I know nothing about Virtual box, so maybe you can have that ... for what it's worth.

more than 4 years ago
top

Framerates Matter

Nevyn Re:The human eye can dectect 30 (521 comments)

Yeh, I do that with my foot too. But I'm pretty sure a lot of that is instinctual, because although I've done it to great success with class cups etc. ... I've also done it with knives and heavy metal objects (while not wearing shoes :).

more than 4 years ago
top

Microsoft Wants To Participate In SVG Development

Nevyn Re:Torpedo? (292 comments)

*BSD runs the BSD userland instead of the GNU userland. So it is just like linux with a worse UI.

Be fair. They have a worse kernel too.

more than 4 years ago
top

Angry AT&T Customers May Disrupt Service

Nevyn Re:Should be (572 comments)

Welcome to reality. The amount of electricity to your house is not limited

Not true, even ignoring the physical problems there's a main breaker (which you can't play with without the electric companies approval) which has a limit. But I've also never seen the electric company advertise "unlimited" electricity.

And I think network usage has to go the same way. The reason that's scary to a lot of people is because most "per-megabyte" rates are way, way too high.

That is a worry ... but the biggest problem is still "we sold you unlimited, but have now realized we can't give it to you".

more than 4 years ago
top

Saying No To Promotions Away From Tech?

Nevyn Re:You can't say NO (410 comments)

The thing is the only thing overseas sub-contracting has to offer are lower costs. So it makes sense that the overseas companies will "optimize" everything for cost, this implies the cheapest labour you can get. This probably works "acceptably" for a level 1 call centre, not so much for knowledge workers.

By the same token, I've never seen a small .us contracting company primarily optimize for cost (they may exist, and I just haven't worked with/for them).

more than 4 years ago
top

Google Chrome Extensions Are Now Available

Nevyn Re:Take on AdBlock? (291 comments)

As someone who makes his living selling content through the Internet, I want people to think several times before building a tool like AdBlock.

You are speaking to the wrong people, IMO. NoScript/AdBlock/etc. are there because users want them. So you are saying "please don't do what users want, because it'll make me more money". Well, sucks to be you then.

If you really want to make a difference and see Ads be viable on the web. N years from now ... you need to speak to "most" of the large content providers. For instance my wife is pretty clueful and had happily not been using NoScript/etc. ... until recently when she hit usatoday and a giant popup came up and refused to go away (so she couldn't read the content). Now she has NoScript installed and only approved sites can run any JS.

It's the same with TV, 5 minutes of commercials every 10 minutes (33%) is just way too much and their real customers fought back ... so now they get 0% ads from a growing percentage of people. Yeh, that implies bad things for the future of TV, but then in many ways nothing really is better than what was there before.

more than 4 years ago
top

Google Tries Not To Be a Black Hole of Brilliance

Nevyn Re:What a coincidence (322 comments)

They recently stopped buying third party maps, and are now reportedly paying mobile providers to put their google maps app. on the phones (which they can only do because they aren't locked into the map provider duopoly). Which speaks to, a least, a significant amount of forward planning on their side.

But, yeh, brilliance is hard to measure. Some of the DNS tricks they just released, and more, I've wanted some Linux DNS software to do for at least 10 years ... so it's not "genius". But the number of useful things they've actual got out implies they are way above average (as a company), IMO.

more than 4 years ago
top

Scientology Charged With Slavery, Human Trafficking

Nevyn Re:That's pretty evil. (802 comments)

However, even the most jaded, cynical atheist cannot deny that organized religions have also done some amazing things for the societies in which they existed.

Yeh, I'm sure Scientology has done some good. So what? Are you suggesting we try and keep some kind of score card, we can put "imprison children and malke them slaves" on one side and "feed homeless" on the other? Not sure what we do when they "convert" a follower with those "selfless acts" though.

Of course Atheists never do any charitable works on their own, so as long as the nutjobs "help" roughly as many as they screw over, it's all good.

is every bit as ignorant and superficial as claiming that someone cannot drive because their plumbing is different than mine ...

That's right, I forgot that most "organised" religions fought so hard for equal rights and civil rights, oh wait...

more than 4 years ago
top

Wal-Mart, Amazon Battle For Online Retail's Future

Nevyn Re:Stereotypes much? (272 comments)

And for anyone else who is not an economist "negative income elasticity" means "prefer cheaper, lower quality, goods".

more than 4 years ago
top

Google Betas Chrome 4, Touts 30% Speed Boost

Nevyn Re:Cheating on my first love - Firefox (383 comments)

Depends, I don't see many people using wordpad/gedit instead of MS-word/OO-writer. Features always matter, and they often matter more than speed/security/usability (all of which have their fans who ask "why can't all apps. consider X, and stop being bloated") ... but sometimes, if you do _really_ well on one of those three and you are "close enough" feature wise, then people can live without the features you don't have. But it's a big gamble, much bigger than just adding features as fast as possible.

about 5 years ago
top

ZFS Gets Built-In Deduplication

Nevyn Re:Building it in makes no sense (386 comments)

leaving you with a junk file

That's why the minimum recommendation for updates like that is to "write to tempfile, close, check that close is happy and then rename". Unless you can deal with someone pulling the power at any point as you update.

about 5 years ago
top

Lost Northwest Pilots Were Trying Out New Software

Nevyn Re:It's a tough job (518 comments)

According to that page on that site, _minimum_ pay for a pilot is 67k (which I would put a lot of money on being BS). Of course a different page (http://www.avjobs.com/salaries-wages-pay/historical-aviation-wages.asp) says 16k-60k or 23k-250k (depending on if you are "regional" or "national"). And I assume for "average" they are using mean and not median which, from all I've heard of the industry, will make the numbers higher.

about 5 years ago
top

Film Studios May Block DVD Rentals For One Month

Nevyn Re:hey, it beats (545 comments)

People want free, anytime they can get it. Not a good business model.

I think the PS3 "video rental" is a good counter example, I'm a happy netflix subscriber ... and would be happy to give money to sony instead, for a comparable product. But the product is significantly worse and a lot more expensive. I assume they think the fact I can have it "instantly" should make up for that, but it doesn't.

If you then look at "buying" instead of renting, the product is a little cheaper but a lot worse -- and there are even weirdos with the price (like their "sales" which are much more arbitrary than any real shop could get away with). Plus the "instant download" feature isn't as big a hook.

I use the Amazon mp3 store all the time, I hear iTunes is doing great ... I'll likely never use the PS3 store again though. So just saying "people want free" is outright ignoring reality.

about 5 years ago
top

Can Nintendo Really Be Planning Another DS Variant?

Nevyn Re:Err, why? (187 comments)

When it was launched, the DS was an experimental console, Nintendo's so-called "3rd leg". Nintendo had no significant faith in it, but threw it out there anyhow as an experiment while working on a proper Game Boy.

One consequence of this is that the components of the DS weren't necessarily picked as they would have been for a handheld designed to match the long life of a Game Boy.

It was an "experiment" sure, they didn't know it would succeed and likely had some backup plans. But to pretend it wasn't meant to succeed is just insanity. Nintendo haven't made a "top of the current gen. specs." console since the N64, so the fact they do that for the DS means nothing.

Meanwhile in Sony-land, manufacturing technology has finally caught up with the ridiculously overbuilt PSP, which was an absolute brick when launched. The Go has some pricing/design issues

hahaha ... you mean like the "issue" that it isn't backwards compatible AT ALL. So you can't buy any cheap used games, and PSP owners are basically forced to not upgrade. And combine that with the fact that it's digital only, so your gamefly subscription is useless and you need to buy two games if you have a spouse/brother/roommate/friend who likes to play occasionally (compared to the DS which not only let you share disks but often let you play limited multiplayer on one disk) and they are the same price as original PSP games (which have none of those restrictions).

I was interested in the PSP Go, when I first heard about it 3-6 months ago (mainly due to the size), but there's no way I'm getting one atm.

about 5 years ago

Submissions

Nevyn hasn't submitted any stories.

Journals

top

Wii friend code: 2790-4876-4470-0181

Nevyn Nevyn writes  |  more than 7 years ago

So, like all the other old gits ... I got a Wii on launch day. Not really said much about it as I've spent too much time playing on it. If you want to XP my Mii's or whatever, friend code is: 2790-4876-4470-0181.

top

UI for 80 cores

Nevyn Nevyn writes  |  more than 8 years ago

So you've probably all seen the 80 core CPUs in 5 years story.

One thing that I immediately thought (after, OMG WTF!) was "what the hell is the UI going to be like for that". For instance top currently has a summary line for each core, but having over 3 terminal screens (80x24 is as god intended) just for the summary seems unlikely to be appreciated. On the other hand having a program with a single task have CPU usage with an upper bound of 1.25% CPU seems less than useful. This seems really bad (Ie. can't treat that many cores seperately and you can't treat them as a single unit).

There's also the (probably huge) set of minor problems with things like And-httpd starting sysconf(_NPROCESSORS_ONLN) number of tasks by default. Which works well for the 1-4 CPU/core case, but is likely to be annoying with 80 cores.

Any thoughts?

top

Fools Friday?

Nevyn Nevyn writes  |  more than 9 years ago

So, I'm not sure if this is all tomhudsons fault ... but it seems like any random moron can get stuff posted to /. nowadays. Hmmm, maybe I'm just getting old and cranky.

Exibit A, so we're going to do a "DB benchmark" to see how well the different OSes compare to each other. Ok, I can buy that ... so you're using MySQL, which isn't that tastful, but I guess PostgreSQL isn't stupid enough to use threads and maybe that's what you want to test ... and people do use MySQL, so I can forgive that, I guess.

However, when running a "real world" test on different OSes it seems like crack to install an upstream MySQL tarball ... no sane person is going to be doing that, if they can get an errata supported MySQL package locally. Sure, there are going to be differences apparent from the different versions ... but that's going to be true if you deploy them. And if any of the OS vendors did any work on their package, that's now lost too.

And then to give a "methodology" that says:

I installed each operating system "stock" -- with default options -- and recompiled a kernel here and there when necessary, to add, for instance, SMP support in FreeBSD or increase memory allocation limits. For the most part, I didn't do any special OS tweaking beyond what was specifically recommended for MySQL. Any changes to the OS I made are documented in their respective sections below.

Oh, a recompile "here and there" and some random MySQL tweaking that doesn't appear to be documented and for the "non-special OS tweaking in Linux":

I ended up going with Gentoo 2004.3 ... With Gentoo it was also relatively easy to install NPTL for 2.6 ... (although they [ed: NPTL] didn't make any difference) ... For these tests, I chose ReiserFS version 3.

hahahahahahahahahahahahahahahahahahahaha. Ahhh, I have to laugh or I'll cry.

top

Vstr testing and daemons

Nevyn Nevyn writes  |  more than 10 years ago

Well I released Vstr-1.0.13 last night, it's now "fully tested". I've posted some stats. etc. on my LJ "blog", or you can just install the example programs including a http server.

I'm pretty sure it's the most tested open source code, so one the one hand it's hard not to brag ... but on the other hand I can't help thinking to myself, "Well done, you've written a HTTP/1.0 server with quite a few options missing and it only took you about eight years :o).".

Still I'm running the example httpd on my ADSL, and I'm pretty sure it's secure :).

top

Written to be read by noone...

Nevyn Nevyn writes  |  more than 10 years ago

Is it me, or does noone actually read the web. One giant hypertext inverted WORM device.

Maybe, I'm getting cynical in my old age ... I mean I've seen random people write things that "I know" has been written about as being false for months or years. But I kind of expected IBM research ... to at least have done, you know, a little research before posting articles.

But hey, I guess Fox is allowed to say it's a "NEWS" channel ... so maybe I've just not been cynical enough.

top

"Scalable" network IO

Nevyn Nevyn writes  |  more than 11 years ago

So, as some of you may know, I'm pretty interested in scalable network IO. So I was pretty excited about the recent /. article. Alas, it was written by Felix "I write broken code" von Leitner of dietlibc and libowfat "fame".

Apart from the fact it mainly talked about dealing with network IO events, it completely screwed up the definition of the poll() API ... to the point where all the poll() related data is worthless (see socket_poll, which does what he says you can't with poll ... and faster). This includes the point about fork() being faster then poll() on Linux (which I'd put money on not being true). I was also sad to see alarm() even as a consideration for a timeout API, IMO do not use alarm() ... ever.

It only touched on iovec's and scatter gather IO from the application POV, which is probably due to zero support for that from the DJB designed buffer_* string IO APIs. Also the connect() calls were all blocking, and I think the data traveled using httpbench never used more than a single socket at once (Ie. there weren't ever multiple sockets seeing data at once). Sigh, I guess I'll have to finish my text for all that stuff ... and here I was hoping someone would do it all for me :).

On the other hand, there were a couple of interesting points I didn't know...

  • epoll now does Level triggered events by default (stupid, stupid, stupid...). People should just learn to deal with edge triggering properly (of course I'm slightly biased as Vstr helps out a lot here).
  • DJB has an IO API
  • DJB wants to do non-blocking IO, I'd assumed DJB was firmly in the fork() and blocking IO model after qmail.
  • Felix has written a read only LDAP server ... it's very possible that even being Felix this will be better to salvange than anything from OpenLDAP anytime soon.
  • The description of the O(1) schedular was pretty good.
  • The description of level triggering events and edge trigger events, was also pretty good.
  • It's probably always good to have advice out there telling people not to use threads, even if I'm not sure the numbers are right.
  • It is nice to have a simple reference for all the IO event mechanisms, this would be better somewhere on c10k really though, IMO. esp. as it's more likely to be accurate.
  • I didn't realise you could make BSD TCP_NOPUSH actually do anything useful, although TCP_CORK is vastly nicer it appears you might be able to get generic code to use TCP_NOPUSH on *BSD with most of the gain.

Slashdot Login

Need an Account?

Forgot your password?