With both RHEL6 and Debian Squeeze on their own versions of 2.6.32, as well as the last Ubuntu LTS 10.04, that version will effectively be the end of the 2.6 line for many places if version numbers jump like this. The kernel versions actively targeted by the -stable team are the only ones some people (including me) are interested in, and this cluster of distributions on 2.6.32 is a good thing in my book. The main issues I'm seeing in newer kernels that I'm concerned about backports of are the continued fixes to weird ext4 bugs happening in newer kernels. Keep those coming, backport drivers for the most common hardware out there, and the rest of the kernel development can march along without me being so worried about it. (Context disclaimer: I worry about PostgreSQL database servers for a living, so my customers are more paranoid about stability than most)
The eventual release of btrfs is one of the things I'd would be glad to see happen only in a kernel that's clearly labeled part of new, less stable development. New filesystems are one of the hardest things to get right, and there's no other class of bugs as likely to lead to major loss of data.
Your year of "Thing a Week" resulted in many great songs. With classics like "RE: Your Brains" on week 26 and "Code Money" on #29, from the outside and in retrospect it seems obvious you'd already reached serious momentum halfway through. Was this apparent to yourself, and did you ever consider ending the experiment early based on that progress? I think it's interesting to consider schedule vs. goal oriented development as something applicable to a self-improvement context.
Four of them here. That's how many systems I've converted from running Ubuntu to Debian Squeeze in the last two months. Ubuntu had a great opportunity to pick up users during the years when Debian released too infrequently to be viable for the desktop, and no other Linux distribution was built on that base and targeting the desktop well. At this point I see no reason to ever consider Ubunut's latest unstable bling when there's both a two-year Debian release cycle and more regular desktop releases from distributions like Linux Mint.
Next stop: migrating CentOS server systems, another distribution I no longer have any taste for, to Scientific Linux. That the developers were so clueless that dag is giving up on them, there's another distribution that's lost its way.
I'm not raising a family and have no plans to. In addition, I have contract obligations that it would take me a year to begin unraveling, such that I could consider that form of buy-out. My professional career is based on open source software solutions, and I have already gone hungry rather than take a job working as an Oracle DBA. I've also not worked at all when the main option available was working on Microsoft software. You can't call bullshit on me, coward.
I'm a PostgreSQL contributor. Oracle can't buy my copyright. There are dozens of other code contributors just like me in that regard, working for many companies. It was possible to buy MySQL because most of the code was developed by MySQL Ab, and copyright assignments required for contributions to be merged. See Some thoughts on Copyright Assignment for more details. That's not the case for PostgreSQL.
Furthermore, the PostgreSQL community has already been through the process of having a single corporation "buy" many of the top contributors, when a company named Great Bridge hired many of them. The disruption to the PostgreSQL community of Great Bridge failing was such that even starting in that direction is actively rejected now. So even if a company did start gobbling up developers one at a time, they would face increasing resistance at obtaining the remaining ones.
Looking through the Mono application screenshots, what I believe are the most popular programs impacted by Mono development slowing are Banshee, F-Spot, and Tomboy. Since this trio is easily replaced by Rhythmbox, gThumb, and Gnote, among other options, good riddance to the lot of them. In addition to the standard Stallman concerns, the high concentration of the development team within Novell was always a problem anyway. There are way too many similar applications within open-source operating systems, so culling out some of the weaker ones--from a development risk standpoint--is a net benefit as far as I'm concerned.
It depends on the market and price class. In areas where real estate is expensive, you'll be hard pressed to find a microwave, or even a refrigerated mini-bar, in hotels now. New York and London are examples where I've run into this recently.
There also seems to be a trend toward including less of these things in expensive hotels. If you're at a $200/night hotel, the presumption is you'll be ordering room service, not reheating your leftovers in the microwave. And even in the cases where that isn't true, it's not always in the hotel's best interest to provide you with a microwave. Yet another thing to clean and otherwise maintain, while working against sales of room service and in-hotel dining. It doesn't seem to be a competitive feature people demand anymore, and since it distracts from selling other services it's easy to see why they aren't included.
Too late for it to matter. BlackBerry seems to have lost any sort of forward vision for expansion a long time ago, instead being focused on extracting maximum profit from existing customers. They didn't do obvious things like release a free BES, which never should have been an important profit center, until after so much of their market was eaten by competitors that it was too late for that to help. They reacted similarly to the need to include a useful web browser on their phone--wait until competitors have had one for years before realizing it was important.
Sure; eating lunch together every day gets an extra chunk of time for discussing company business out of every person on the staff each day. Of course it's a great idea from the perspective of being good for the company. So is working an extra hour each day for the same salary, which is what company lunches are essentially doing.
Exactly--all you have to do in order to make Debian work with common hardware is install normally, edit/etc/apt/sources.list to enable the non-free and contrib repos, run apt-get update to make it recognize that change, then find the names of the additional packages you need so that you can install them! Everything Just Works!
Sheesh. I switched three systems from running Ubuntu to Debian recently. That's what I did on each of them, and the process was annoying but feasible. But it would be completely unreasonable to expect any normal desktop user to go through, and this all happens automatically on Ubuntu. At worst you run the little "Add a hardware driver" wizard and click to turn things on.
I also didn't like "If we modify these Terms of Service, we will post the modification on the Site or provide you with notice of the modification". *Or* provide me notification? So they basically can change the terms at any time just by posting a new set to the web site, and they aren't even promising I will be notified? Yeah, let me jump right on agreeing to that.
The main options I'm aware of for hosting your own bookmark server are:
Cutemarks. Pretty simple, but not maintained anymore I think.
Firefox Sync Weave server. Seems like it's trying to solve way more complicated problems than I really care about, with the associated complexity that comes with that.
Firefox Sync server on Google App Engine. Interesting proof of concept, adding another server source compatible with Firefox's API. I don't really trust Google's infrastructure either though, and this hasn't moved much beyond prototype yet.
Perhaps the terrible terms for the relocated Delicious site will finally kickstart more development in this area.
Look at the Davidson data breach class action lawsuit for a case extremely similar to this one. There's also the (still pending as far as I can tell) Citizens Financial Bank breach case. Not following the standards of the industry for securing this sort of data can absolutely lead to a class action settlement, even if there is no hard written security standard.
New terms and privacy policy Don't want to have your data switched? Worried about the transition? You can easily download all your bookmarks using their API with a command like this:
Another easy way to randomly delay users, show them the wrong data, and return 404 errors under load is to build your site with one of those trendy Rails frameworks. Make sure to use MySQL with MyISAM "for speed" if you want even more inexplicable service outages.
A third implementation would be to redirect the trolls over to the version of the site hosted in "the cloud".
That happened in the Bedford–Stuyvesant section of Brooklyn in the early 90's. Not a great neighborhood at the time, pretty famous as a place filled with crackheads in the 80's.
Regardless of the trivia around whether IP addresses are unique or not, accepting the idea that router == identity is a really bad idea. The obligation should be to prove that I committed an act, and that the act occurred at my home is not sufficient. I'm equally displeased at the owner == criminal implications of things like red-light and speeding ticket cameras. I've both sped and run a red light to avoid an accident before; I've also let other people use my car. Yet I'm exposed to tickets and presumed guilt in every case involving my vehicle.
The computing case has some very similar edge cases. I've had someone visit me with a computer filled with malware many times. I might have tried connecting it to my network presuming I removed all of it but been wrong. If it downloaded something illegal, as malware is apt to do, is it OK now for me to be on receiving end of a SWAT team because this happened at my house? It's automatically my fault as the homeowner for any act that occurs on my property, whether or not I did it? Next up: someone drops drugs in my backyard. So it's fine now if the police arrive with big guns and shoot me if I react badly, right? This isn't a theoretical question. Ever had someone break into your house, use your kitchen for a drug binge, then leave behind the mess? I've seen it happen. Way things are going, I'd be afraid to call the police the report the break-in now, lest I get accused of doing the crime committed where I live.
Given that wireless security is never absolute, I'm looking forward to the day when we get the opposite of this story: where you might want to leave just a little bit of insecurity around just to establish the possibility someone else used your network. I have a fixed IP address and a fairly locked down network. But if someone does manage to spoof as me, I will have very little ability to suggest it wasn't me who committed the act. Maybe I need a bit of deniability the way things are going. Lock things down too far, and it's easy for someone to prove to a non-technical jury that any network access that happened must have been me. Scary thought.
The alternative for PulseAudio was to wait one more release, so that it wasn't merged into a LTS release. Had it been pushed back to 8.10 instead, then a) the 8.04 LTS would have been much more stable, and b) the much improved PulseAudio in 8.10 wouldn't have caused as many complaints.
If Ubuntu were only a desktop distribution, the need for constant upgrades to get bug fixes wouldn't bother me. The point of my rant was that it's impossible to ever get a stable release from them. That means the server version is useless. I'm not going to adopt a desktop-only Linux; I have real work to get done, too, and having my desktop also be a deployable server install is very helpful.
As for what the other option is, I'm using Debian Squeeze now. Just got rid of all my Ubuntu systems with it shortly after it shipped. Debian now ships every 2 years, which is a very reasonable window--much faster than RHEL, which is quite out of date by the time the next rev comes out, while not being the constant rush that Ubuntu is under.
Your car analogy doesn't work because the first year of a car is normally the trouble-free part of its life, while the first year of a new piece of software is the most painful one. That's why companies upgrade auto fleets so often--get rid of them as they become more difficult to maintain.
Adjust the release schedule just because of bugs? Never again. Besides, Ubuntu is always in perpetual beta now. I think they're trying to be like Google or something.
For me a stable release implies a focus and quantity of backports into that version that I have never seen Ubuntu do. I'm curious since I wasn't following those two: were the "travesties" (agreed) you mention in LTS 8.04 fixed to your satisfaction at any point? Or were you forced onto a new version for things to work?
Ubuntu rushes everything out before it's ready; it's impossible for a 6 month release cycle to do anything else. This whole Unity experiment is no surprise to anyone who was using in Ubuntu in 2008, when the at the time barely working PulseAudio was integrated into the "Long-Term Release" 8.04. And by LTS, they mean "supported until the developers are whipped to start working on their next 6 month deadline the week after shipping".
The main issue I have is that the "Advanced Bash Scripting Guide" seems to be fascinated with listing every possible way to do the things it covers in bash. But missing from that tutorial is any notion of which of those approaches is the best one in any context. Certain shell constructs are best avoided, in favor of ones that are either more portable to other shells, more resilient in the face of bad data, or easier to maintain. But you get no sense of this from the guide; just a set of ways to hack the low-level code into place.
Slashdot Mirror
With both RHEL6 and Debian Squeeze on their own versions of 2.6.32, as well as the last Ubuntu LTS 10.04, that version will effectively be the end of the 2.6 line for many places if version numbers jump like this. The kernel versions actively targeted by the -stable team are the only ones some people (including me) are interested in, and this cluster of distributions on 2.6.32 is a good thing in my book. The main issues I'm seeing in newer kernels that I'm concerned about backports of are the continued fixes to weird ext4 bugs happening in newer kernels. Keep those coming, backport drivers for the most common hardware out there, and the rest of the kernel development can march along without me being so worried about it. (Context disclaimer: I worry about PostgreSQL database servers for a living, so my customers are more paranoid about stability than most)
The eventual release of btrfs is one of the things I'd would be glad to see happen only in a kernel that's clearly labeled part of new, less stable development. New filesystems are one of the hardest things to get right, and there's no other class of bugs as likely to lead to major loss of data.
Your year of "Thing a Week" resulted in many great songs. With classics like "RE: Your Brains" on week 26 and "Code Money" on #29, from the outside and in retrospect it seems obvious you'd already reached serious momentum halfway through. Was this apparent to yourself, and did you ever consider ending the experiment early based on that progress? I think it's interesting to consider schedule vs. goal oriented development as something applicable to a self-improvement context.
Four of them here. That's how many systems I've converted from running Ubuntu to Debian Squeeze in the last two months. Ubuntu had a great opportunity to pick up users during the years when Debian released too infrequently to be viable for the desktop, and no other Linux distribution was built on that base and targeting the desktop well. At this point I see no reason to ever consider Ubunut's latest unstable bling when there's both a two-year Debian release cycle and more regular desktop releases from distributions like Linux Mint.
Next stop: migrating CentOS server systems, another distribution I no longer have any taste for, to Scientific Linux. That the developers were so clueless that dag is giving up on them, there's another distribution that's lost its way.
He actually meant CHiPs.
I'm not raising a family and have no plans to. In addition, I have contract obligations that it would take me a year to begin unraveling, such that I could consider that form of buy-out. My professional career is based on open source software solutions, and I have already gone hungry rather than take a job working as an Oracle DBA. I've also not worked at all when the main option available was working on Microsoft software. You can't call bullshit on me, coward.
I'm a PostgreSQL contributor. Oracle can't buy my copyright. There are dozens of other code contributors just like me in that regard, working for many companies. It was possible to buy MySQL because most of the code was developed by MySQL Ab, and copyright assignments required for contributions to be merged. See Some thoughts on Copyright Assignment for more details. That's not the case for PostgreSQL.
Furthermore, the PostgreSQL community has already been through the process of having a single corporation "buy" many of the top contributors, when a company named Great Bridge hired many of them. The disruption to the PostgreSQL community of Great Bridge failing was such that even starting in that direction is actively rejected now. So even if a company did start gobbling up developers one at a time, they would face increasing resistance at obtaining the remaining ones.
Looking through the Mono application screenshots, what I believe are the most popular programs impacted by Mono development slowing are Banshee, F-Spot, and Tomboy. Since this trio is easily replaced by Rhythmbox, gThumb, and Gnote, among other options, good riddance to the lot of them. In addition to the standard Stallman concerns, the high concentration of the development team within Novell was always a problem anyway. There are way too many similar applications within open-source operating systems, so culling out some of the weaker ones--from a development risk standpoint--is a net benefit as far as I'm concerned.
It depends on the market and price class. In areas where real estate is expensive, you'll be hard pressed to find a microwave, or even a refrigerated mini-bar, in hotels now. New York and London are examples where I've run into this recently.
There also seems to be a trend toward including less of these things in expensive hotels. If you're at a $200/night hotel, the presumption is you'll be ordering room service, not reheating your leftovers in the microwave. And even in the cases where that isn't true, it's not always in the hotel's best interest to provide you with a microwave. Yet another thing to clean and otherwise maintain, while working against sales of room service and in-hotel dining. It doesn't seem to be a competitive feature people demand anymore, and since it distracts from selling other services it's easy to see why they aren't included.
When did they do that?
Too late for it to matter. BlackBerry seems to have lost any sort of forward vision for expansion a long time ago, instead being focused on extracting maximum profit from existing customers. They didn't do obvious things like release a free BES, which never should have been an important profit center, until after so much of their market was eaten by competitors that it was too late for that to help. They reacted similarly to the need to include a useful web browser on their phone--wait until competitors have had one for years before realizing it was important.
Sure; eating lunch together every day gets an extra chunk of time for discussing company business out of every person on the staff each day. Of course it's a great idea from the perspective of being good for the company. So is working an extra hour each day for the same salary, which is what company lunches are essentially doing.
I'd consider the use of high explosives.
I've found the explosives that aren't doing drugs are more reliable.
Exactly--all you have to do in order to make Debian work with common hardware is install normally, edit /etc/apt/sources.list to enable the non-free and contrib repos, run apt-get update to make it recognize that change, then find the names of the additional packages you need so that you can install them! Everything Just Works!
Sheesh. I switched three systems from running Ubuntu to Debian recently. That's what I did on each of them, and the process was annoying but feasible. But it would be completely unreasonable to expect any normal desktop user to go through, and this all happens automatically on Ubuntu. At worst you run the little "Add a hardware driver" wizard and click to turn things on.
I also didn't like "If we modify these Terms of Service, we will post the modification on the Site or provide you with notice of the modification". *Or* provide me notification? So they basically can change the terms at any time just by posting a new set to the web site, and they aren't even promising I will be notified? Yeah, let me jump right on agreeing to that.
The main options I'm aware of for hosting your own bookmark server are:
Perhaps the terrible terms for the relocated Delicious site will finally kickstart more development in this area.
Look at the Davidson data breach class action lawsuit for a case extremely similar to this one. There's also the (still pending as far as I can tell) Citizens Financial Bank breach case. Not following the standards of the industry for securing this sort of data can absolutely lead to a class action settlement, even if there is no hard written security standard.
New terms and privacy policy Don't want to have your data switched? Worried about the transition? You can easily download all your bookmarks using their API with a command like this:
curl --user username:password -o bookmarks.xml -O 'https://api.del.icio.us/v1/posts/all'
The XML format you get it out in is pretty easy to parse into something else.
Another easy way to randomly delay users, show them the wrong data, and return 404 errors under load is to build your site with one of those trendy Rails frameworks. Make sure to use MySQL with MyISAM "for speed" if you want even more inexplicable service outages.
A third implementation would be to redirect the trolls over to the version of the site hosted in "the cloud".
One senator is already writing them nasty notes.
That happened in the Bedford–Stuyvesant section of Brooklyn in the early 90's. Not a great neighborhood at the time, pretty famous as a place filled with crackheads in the 80's.
Regardless of the trivia around whether IP addresses are unique or not, accepting the idea that router == identity is a really bad idea. The obligation should be to prove that I committed an act, and that the act occurred at my home is not sufficient. I'm equally displeased at the owner == criminal implications of things like red-light and speeding ticket cameras. I've both sped and run a red light to avoid an accident before; I've also let other people use my car. Yet I'm exposed to tickets and presumed guilt in every case involving my vehicle.
The computing case has some very similar edge cases. I've had someone visit me with a computer filled with malware many times. I might have tried connecting it to my network presuming I removed all of it but been wrong. If it downloaded something illegal, as malware is apt to do, is it OK now for me to be on receiving end of a SWAT team because this happened at my house? It's automatically my fault as the homeowner for any act that occurs on my property, whether or not I did it? Next up: someone drops drugs in my backyard. So it's fine now if the police arrive with big guns and shoot me if I react badly, right? This isn't a theoretical question. Ever had someone break into your house, use your kitchen for a drug binge, then leave behind the mess? I've seen it happen. Way things are going, I'd be afraid to call the police the report the break-in now, lest I get accused of doing the crime committed where I live.
Given that wireless security is never absolute, I'm looking forward to the day when we get the opposite of this story: where you might want to leave just a little bit of insecurity around just to establish the possibility someone else used your network. I have a fixed IP address and a fairly locked down network. But if someone does manage to spoof as me, I will have very little ability to suggest it wasn't me who committed the act. Maybe I need a bit of deniability the way things are going. Lock things down too far, and it's easy for someone to prove to a non-technical jury that any network access that happened must have been me. Scary thought.
You recycled it? That was supposed to into the compost pile! Don't you know anything about how to treat the environment?
The alternative for PulseAudio was to wait one more release, so that it wasn't merged into a LTS release. Had it been pushed back to 8.10 instead, then a) the 8.04 LTS would have been much more stable, and b) the much improved PulseAudio in 8.10 wouldn't have caused as many complaints.
If Ubuntu were only a desktop distribution, the need for constant upgrades to get bug fixes wouldn't bother me. The point of my rant was that it's impossible to ever get a stable release from them. That means the server version is useless. I'm not going to adopt a desktop-only Linux; I have real work to get done, too, and having my desktop also be a deployable server install is very helpful.
As for what the other option is, I'm using Debian Squeeze now. Just got rid of all my Ubuntu systems with it shortly after it shipped. Debian now ships every 2 years, which is a very reasonable window--much faster than RHEL, which is quite out of date by the time the next rev comes out, while not being the constant rush that Ubuntu is under.
Your car analogy doesn't work because the first year of a car is normally the trouble-free part of its life, while the first year of a new piece of software is the most painful one. That's why companies upgrade auto fleets so often--get rid of them as they become more difficult to maintain.
Adjust the release schedule just because of bugs? Never again. Besides, Ubuntu is always in perpetual beta now. I think they're trying to be like Google or something.
For me a stable release implies a focus and quantity of backports into that version that I have never seen Ubuntu do. I'm curious since I wasn't following those two: were the "travesties" (agreed) you mention in LTS 8.04 fixed to your satisfaction at any point? Or were you forced onto a new version for things to work?
Ubuntu rushes everything out before it's ready; it's impossible for a 6 month release cycle to do anything else. This whole Unity experiment is no surprise to anyone who was using in Ubuntu in 2008, when the at the time barely working PulseAudio was integrated into the "Long-Term Release" 8.04. And by LTS, they mean "supported until the developers are whipped to start working on their next 6 month deadline the week after shipping".
And members of the band are encouraged to spend 20% of their time on their solo albums.
The main issue I have is that the "Advanced Bash Scripting Guide" seems to be fascinated with listing every possible way to do the things it covers in bash. But missing from that tutorial is any notion of which of those approaches is the best one in any context. Certain shell constructs are best avoided, in favor of ones that are either more portable to other shells, more resilient in the face of bad data, or easier to maintain. But you get no sense of this from the guide; just a set of ways to hack the low-level code into place.