I have some code that we wrote GPL'd. It is under the GPL because it is easier from a business perspective to GPL it and let someone pay me for a more liberal license. I am likely to switch to the BSD anyways. Our corporate software license is based off the original BSD (with advertising clause) and the right to pay us to switch to a true BSD client. That's my open source involvement.
The Open Source community has written very little. The Mach microkernel was developed by researches at CMU. Like most research projects, some of their grant money was probably government money. Government research is NOT done to benefit whiny anti-corporate bordering on communist high school kids (which seem to dominate the slashdot posts). Government research is to advance national security interests or advance technology to benefit society and particularly the corporations that use it to power the economy. The US Government is interested in economic growth and security.
The BSD System was written by researchers at UCB. It was funded with some grants, and they developed a free implementation of Unix. That is made available for all Americans (and in this case, all people in the world) to use and advance the country.
My tax dollars should NOT be used to fund people whose objective is to derail one of the fastest growing sectors of the economy.
Linux borrowed heavily at various points from BSD. Linux also completely swiped the GNU toolset and now we have distributions, all but one of which don't acknowledge that they started as the GNU collection. The true irony is the RMS tirade about calling it GNU/Linux. I agree with his point that we want people interested in a Free Unix, not Linux in particular. If someone came and wrote this amazing new kernel, we would lose all the mindshare (and credibility and education done to the public) because it wasn't Linux, and they don't know GNU, they know Linux.
Major "Open Source" Milestones:
Kerberos: MIT Research Project
BSD: Berkeley Research Project
Mach: CMU Research Project
Apache: began as a set of patches to NCSA HTTPD (National Center for Supercomputing Applications, University of Illinois, Urbana-Champaign)
I'm sorry, but university research has spawned some impressive technologies that people have hacked in new ways and done some cool work with. But there is no "open source community" that owns this work. All tax-payers own this work that was created for the nation on grant money.
Just because you run Linux doesn't mean that you own CMU, MIT, UI, and UCB's work. Sorry.
I hate to sound biting, I like a lot of open source software. I respect RMS's beliefs. However, this is absurd.
The fact that RMS defines freedom one way doesn't make it so. Instead of spouting about Free Beer and Free Speech, why don't you think for yourselves for a moment. RMS declaring freedom one way is all well and good. Without a doubt, BSDL meets even RMS's definitions. His complaints about the advertising clause is ironic, because if the GPL included in, there'd be none of this GNU/Linux issue.
RMS: I'm a fan, and I respect what you've done here. You've done a lot of great work. Unfortunately, ESR's minions have made a mess.
SAMBA without WinBind still isn't a general purpose File Server. You need to go and create the Unix accounts. My understanding is that WinBind is Linux only right now, but that it is being merged into the core code. Either way, SAMBA has a great niche in small group servers. For example, if I want a file server for 5-10 people, SAMBA rocks. You can join the SAMBA server to your NT Domain, setup the usermap, and give users access to the shares. For that purpose, I love SAMBA.
We are starting to evaluate SAMBA 2.2, it's great. Our core infrastructure here is NT4 for various reasons, but our OpenBSD machines all share out the relavent portions via SAMBA 2.0. It's terrific, if I'm not at a Unix workstation and want to work on a developing website, I connect to the share in Windows and go nuts. It's terrifically powerful.
From where SAMBA was 4 years ago when I started playing with it, it's a new world. You used to need to use separate accounts and not have reasonable NT integration. This allowed you to share out files, but not seemlessly. WinBind promises to make SAMBA servers drop and forget, just administer everything from your NT groups.
For a reasonable sized office (>10 people) running Windows, I think you'll be happiest with an NT 4.0 Domain with a PDC and BDC, but you can then do your file sharing off your SAMBA boxes. This saves a LOT on NT CALs, as you don't need NT CALs for PDC/BDC access, just file/print sharing. Let SAMBA handle that, and use NT as a domain controller. That way you have central logons without the expense. The cost of two NT servers isn't bad, the cost of NT CALs as your office grows and multiple file servers as access speed matters makes SAMBA great.
Additionally, if you do any web development, share our your site tree and you can edit files from anything. It's great, whether you are using Linux w/ smbmount, Windows with CIFS networking, MacOS Class w/ Dave, or MacOS X w/ Sharity, you have all your files accessable and can be editted locally. That sure beats SSH/Emacs.
Our clients needed central control of the database and reporting, so we wrote a report generator. Our PHP System includes tools ror administering your database (the data side, not the DBA side) so anyone with a browser (and obviously a valid password) can do it. We wrote a report generator that you enter the SQL, it generates the report. If you want it printable, you click on the "export to excel" option, and it loads the thing up in Excel (gotta use IE for our admin, Netscape's busted CSS/DHTML support screws it, and Netscape 6/Mozilla isn't there yet), and they can print. As an added advantage, the Excel format is trivial.
Excel can read tables, you just set the content type. If you need fancy formating, plug it into the PHP source.
If you don't want to code, there are dozens of programming shops that will develop a system for you for a reasonable fee.
You don't want to spend money or effort? Well, you're in trouble. Labor isn't free. Do it in house or hire someone. Some of our baselevel systems (including our DB class, which can help you abstract your DB from the engine, you'd need to write your own if you aren't using Postgres or MySQL) are available on our website. If you want to hire us, e-mail myself or sales@feratech.com.
We all think of the unions as the semi-skilled laborer unions (auto-manufacturers, teamsters, etc). In those scenarios, the employees have a WEAKENED bargaining position because they have skills. Why is this?
As a machinist, my time working on machines is more valuable because of my skills. My opportunity cost is unskilled labor. Therefore, as long as my employer is willing to pay be slightly more than minimum wage, I will continue with my skill-based job.
That is the goal of those unions.
In tech, there is no monopsony (single hired, for example, the colluding big-three automakers, only factory in town, etc.) so unionizing directly isn't as useful, we can all jump ship.
However, we shouldn't look at the blue-collar unions as an the example, look at the unions for professional sports. In those cases, much better paid people than ourselves have unionized to fight monopsony hiring. In that scenario, the union negotiates for reasonable compensation, etc.
I still don't want to unionize, but I think that we are better off looking at pro-sports for the example than the mob-run extortion rackets that we are talking about.
A former customer of mine had that problem. They did the bridge loan (small investment to cover you between investment rounds), and got a decent piece of the company for cheap money. About 2 weeks later, the company closed on a much larger round at a higher valuation (as a bridge, the went in at the previous valuation), and the investment was looking good.
Unfortunately, they also chose a poor VC, and the VCs couldn't raise the money when one of the groups involved dropped out. The company that the client had invested in laid off 80% of its staff and was looking to sell its technology.
It's a shame. Slimy VCs screw a lot of businesses.
Id didn't choose OpenGL because it was cross-platform.
Id wanted to port Quake to the Voodoo chip, which was an exciting new and powerful 3D graphics processor. However, they had already done VQuake, and part of the contract was that they couldn't port the game to another graphics chip.
This meant they couldn't do another card specific port. They started doing Direct3D Quake, but D3D 3.0 was a total disaster, so he switched over to OpenGL and created the GLQuake we know and love.
As a result, MS made efforts to improve D3D (to the point that my understanding is that it is pretty powerful and not impossible to learn), but Id continued with Quake on OpenGL because everyone was supporting OpenGL for Quake, Carmack likes Linux and cross-platform, so if they can sell the copies, why switch.
However, OpenGL on Windows is ONLY because of Id. WinNT had OpenGL support because of the 3D Modelling, etc. WinNT needed OpenGL. The only game for OpenGL was Quake, and MS wanted to kill OpenGL gaming, refused to release the OpenGL subsystem for Win98 that was seen as a beta, forcing everyone to write either the Quake mini-GL drivers, or write a full ICD. The plugable OpenGL subsystem for NT was killed for Win98 to kill OpenGL gaming.
There was an article on the affect of losing company loyalty on the bottom line. Layoffs have a tremendous affect on loyalty, especially at a company like Intel. If you lay people off to avoid a short-term profit shortfall (and therefore are just propping up the stock price), you show people that they don't matter.
In this case, Intel isn't being a slime and rescinding offers (which can kill you at college recruiting, schools may ban you for a few years), and is offering to give them their signing bonus for NOT joining the firm. The recruit is happy (free money), the employees are happy (they aren't cutting staff to make way for the new people b/c of the intricacies of recruiting), and hopefully the company avoids burning money).
The short-term layoff strategy helps you in the short term, but alienates employees whose friends were let go. If you get fired for incompetance, etc., nobody feels for you. If you get let go because the company wants to keep its stock price up, that doesn't make the other employees feel valuable.
Regardless of Intel's other issues with employees, this move is a nice one.
Go Intel, you're being less slimy than a lot of companies I know.
My personal preference would be to drop OpenBSD on the web servers. The security by default makes tweaking it to your secure settings pretty simple. Yeah, the performance isn't up there with FreeBSD or Linux, but to me it is adequate. If I'm throwing beige x86s at the problem anyway, I can just throw one more if I want to compensate for the speed.
For the database, what are you doing? Are you deploying administrator entered data (in which case transactions are meaningless, go throw MySQL up there), or are you really getting and processing data.
One compromise (for added complexity) is to run two database servers. Put a FreeBSD or Linux Server with MySQL (I prefer PostgreSQL for development, but MySQL DOES Scream, Postgres is starting to annoy me on one of my platforms) hidden from the Public network. Load it with RAM and RAID 0 drive system. Just run the backup after you make changes, and you're good to go.
Then, for the live, interactive data, buy a real database server. Design it, pick your platform (DB2 or Oracle are both fine), and then pick your platform (IBM, Sun, and HP all make nice machines) and put it there. Hire someone who knows what they are doing to put this machine together and configure it, then DO NOT touch it.:)
This way you don't need to shell out big bucks for an Oracle Server for the content that MySQL is fine with. As you pay for the #Processors*MHz*RISC with the big boys for the database license, if MOST of your content is going to be "static dynamic" data (where you update once an hour/day/week, if that), there is no need for Oracle to touch it. Let Oracle process your transactions, and stick your "content" on MySQL.
Consider something, most database backed websites aren't really database systems. Most use a database as a convenient place to store info and retrieve it. With a "database system" I'm using relations to develop complex systems that can be cross-referenced with my keys, monitoring transactions, real-time changes, etc. If I am writing articles to dump on a server, yeah the DB is convenient, but Oracle isn't necessary.
As all of us here can admin a PHP/MySQL or PHP/Postgres solution, keeping the stuff that we work with there makes sense, then pay someone else for the real iron.
Grin, well despite being attacked for it being '94 and earlier, I'm glad to know that I hadn't lost my mind. The program was "cool" even if not totally useful. I liked the idea... but I'd spend 5 minutes as a slow moving Avatar, log off the Internet, and dial into a local BBS...:)
Neat idea, but the masses groked "channels" easier than people thought, at least once AOL started calling them chat rooms...:)
Note to self: pick up a copy of AOL for a free month one of these days... figure out why the masses seam to like it...
Note to others: if you are coding Open Source Stuff for end users, get copies of end user systems (Apple w/ MacOS 9, Apple w/ MacOS X, Windows 98, Windows ME, Windows XP, etc.) and see what people like. If you want to design for hackers, don't claim to be making things for the masses.
Either they bought the domain name, or this is the company that has been at this since '96/'97 (hard to tell, Slashdotted). They had cool Avatar chatting systems, etc.
Their stuff was really neat and novel, and probably deserving of protection. They brought a cool concept out to the marketplace years before others, and in doing so helped legitamize it.
You used to be able to go to Tucows and find a couple of chatting paradigms, and this one was really cutting edge back then.
That's very true, I don't get why they don't support foreign keys. However, the scenario you are describing is easier to work around. Given that you have to build a front end for MySQL, only presenting the options that are in that table largely removes the necessity.
You still have "foreign keys" (where Foreign Key is a reference to another table), you just don't have foreign keys, where the database makes certain that they point to an actual reference.
My point is that the MySQL busted/non-existing foreign keys only really kill you because it will let you delete fields referenced by others and other stupidity. MySQL isn't a relational database, it is just a giant file/data store accessable via SQL.
For starters, we know longer use MySQL, but we wrote a PHP generic DB class that lets us switch between PostgreSQL and MySQL (and will support other DBs that we add). As a DB designer, I hate MySQL. However, it isn't an embarassment.
Remember programming little BASIC toys? If you wanted to keep info you openned a random-access or sequential datafile? MySQL is a set of fast random-access datafiles. It is accessable via a subset of SQL, because people who are comfortable using databases with SQL find it easy.
If all you are doing is supporting a website (no delete operations) then the lack of foreign keys, etc., doesn't matter.
MySQL can be tricked into being useful. You just have to write EVERYTHING (and therefore QA A LOT) in the database. Unfortunately, it is reinventing the wheel.
PostgreSQL is a reasonable database. I don't know why MySQL gets all the credit. But if you have real database logic in your website, it is worth looking into PostgreSQL.
It was a good thing. MIT wrote code to support their infrastructure. Because of the MIT License, corporations were able to expand and adapt that system to build systems for their network. As a result, we have other Kerberos networks, and Kerberos is used.
Had MIT put it under the GPL, there would be no Kerberos. Without the ability to get machines supporting Kerberos, MIT would have scrapped the project. If they couldn't run Athena off Kerberos, no reason to have developers there.
Their is the theoretical notion that Sun would have GPL'd Solaris to support Kerberos. Fat chance. The companies will ignore GPL'd code, while BSD code creates standards and growth that benefits real users.
The GPL theoretically benefits users. The BSD actually does.
Kerberos is under the MIT License. It was developed as part of a research project. It was released for a good reason, they let the OS Vendors port it to their systems.
MIT WAS NOT going to handle the Solaris, Ultrix, HP-UX, and Irix (all Athena systems in time) ports, and they needed it to all play nicely. As MIT had a HUGE Unix system, vendors played nice to get the sales, which both got a large Unix installation and their system in front of MIT students.
Kerberos was developed to provide a secure environment for MIT Computing. Lots of University research is government funded. Tax-payer research financed should be available for EVERYBODY, not just those that you consider "worthy." Sorry, Microsoft and its investors pay their share of taxes (and probably more than most of us do), and they should be able to reap the rewards of the tax-payer research, just as the Free Software/Open Source crowd do.
MIT Released Kerberos because they built it for their needs, MIT IS and MIT LCS are not business trying to maximize profits, releasing it furthered their needs, and releasing it bettered mankind.
MS extended Kerberos using a portion designed for vendor extensions. They did this for backwards compatibility for their systems.
Their older domain system was proprietary, this one is too. So what that they used Kerberos code. None of YOU wrote Kerberos, why should you have a say how it was used. MIT's research project turned platform is benefiting Microsoft customers, who happen to be people. MIT's licensing choice benefitted a class of people that all of you trolls would like hung out to dry.
Know your history, TCP/IP did get chosen because of the BSD availability and as a result, it ran on all the machines of the era. DARPA used it because Berkeley wrote it, released it, and Berkeley's version of the OS ran on the prevalent machines, so the Unix OS and TCP/IP became big.
If you are making improvements to a community supported project, you will give back the code anyway. Otherwise, you need to maintain your own fork and roll the improvements in. It is far better to get your changes into the final product.
Companies with BSD software are most likely to either: A) start with BSD code and morph it into something very different and proprietary, or B) release a value-add to the BSD code, and release the improvements to the BSD stuff to keep synced.
In case A, they could start from scratch (but it is a waste of effort and therefore inefficient, regardless of your holy war), or B, you get the code. In case B, if it is GPL'd, they are likely not to start the project, and therefore, you don't get more code.
Pre-web days, the GPL solved a problem because there was no concept of open source communities. Stallman's free software position had tremendous benefits, and did result in companies GPLing code to comply. However, in this day and age, the open source community creates enough of an incentive to get code released, without the formality of the GPL.
Example: SAMBA. Apple knows that good NT integration would go a long way towards making OS X a good corporate citizen. SAMBA has a solid implementation. However, as it is GPL'd, they can't build a SAMBA based login, etc., system and release it with the OS. Someone would have to package it up as an add-on, distributed separately. This benefits nobody.
I support free software, I try to release what I can. I provide all my clients with the source code to the systems that they paid me to write. I feel that it is the right thing to do. We even negotiate a transferability license if they want, I'm willing to provide all the freedoms of the GPL (even GPL'd code) for a fee.
However, I feel like the GPL is a mixed blessing. It does create an open-source playground where companies have an incentive to enter the GPL world.
At the same time, however, I feel that it is keeping valuable software from existing (benefitting nobody). Additionally, the concept of keeping the code free is silly. If I release code into the Public Domain or with the BSD license, that code is perpetually free as long as SOMEBODY hosts it somewhere. Improvements may not always be free, but the code is free.
The tradeoff is: is it more valuable to (potentially) get more free software out there by coercing companies, or will I get more value from propreitary add-ons, or even, will I get more free software by allowing proprietary add-ons and getting the patches released.
As a software author, you need to make a decision for your code.
One thing you should definitely do, however, is get people to assign copyright (like the FSF does). That way, if you think that their is an advantage to releasing under a new license, you can do so. Just realize that if you go BSD->GPL, the old code remains under BSD. If you go the otherway you are fine.
So the holywar compromise might be: release under the GPL, maintain copyright (and get updates assigned), then if a compelling case is made, you can proprietarily license it to a company making the add-ons (perhaps distributing or donating some of the fees, something to discuss in the assign copyright place), release under the BSD license, etc.
Both have their place, but you have to decide which will generate more benefit.
This doesn't let you break the law and remain anonymous. It says that you can't get the identity so that you can prove the crime. This is VERY important.
Previously, you could file a BS lawsuit against John Doe, then subpeona the identities to identify John Doe. This allowed harassment.
What this does mean, however, is that if the SEC wanted to investigate, they could get the names.
Realize this case: the shareholders are suing that the stock ran up for a company on the verge of bankrupcy. None of the actions of these individuals caused the company to go bankrupt. The company's inability to have Cash>0 did so.
The shareholders aren't suing that the price dropped. And if the company's business plan required a certain stock price, well, that's their problem. Public companies should be able to function regardless of share price.
The judge didn't rule that you can't reveal names, merely that you can't without a compelling reason. If there was evidence of a crime, they'd be released.
For example, if the company compiled a list of short sellers, they could probably give the list to the chat room operators and demand the identities of anyone matching that list.
The burder is on those wishing to reveal the names, as it should be.
We may actually be running 7.1, I code, not sysadmin. Lots of our code was written before that upgrade. Thanks for the tips, but lots of people have told me.
At my old job, we hired Rackspace to do our hosting based on a banner ad that I saw on Slashdot. Of course I did my research, but that's how I learned about them. Before the dot-com went under a few months later, they made maybe $10,000 off the account (we had two decked out servers).
Now, with my consulting firm, we are a Rackspace reseller. We are delivering them a contract for $14,000. Of course this was because of my good experience before, but it worked.
When looking for Linux servers, I know to look at VA Linux and Penguin Computing. The latter I only know about from Banner Ads and the occaisional mention on Slashdot, the former I know about from tech rags AND their constant presence on Slashdot.
Banner Ads work for creating brand awareness. On the Internet, where competition is fierce, you either need to own keywords on search engines, or use banner ads to generate awareness.
Trying to stick with the stable releases. But yes, the improvements have been there, but I don't want to go and find all the SQL that needs improvement.
I make my living promoting Open Source tools. For most of our systems, PostgreSQL is adequate. The lack of Left and Right joins makes me want to shoot myself occaisionally, and we are limited in our ability to scale, but the product works, and works well.
We can even support our tools on MySQL, but we'd have to make a lot of changes because we have database appications, and we'd have to change them to MySQL applications because MySQL isn't really a database, it's a storage format that is retrievable via SQL.
However, Oracle kicks the shit out of these low-end toys.
Hands down, they are number 1, by far.
Open Source databases are not ANYWHERE close to catching up. MySQL is a simple database whose entire reason for existance is powering websites. If you are a programmer, NOT a database designer, and you think that a database will help you store your data better than files and AWK, you use MySQL.
PostgreSQL is at least a properly designed system, but it is limited.
However, despite NEITHER database approaching Oracle, we have decided that Open Source will overtake them?
This is arragance beyond belief.
I don't get it, what Open Source product has been SO successful that it dominates everything?
The closest is Apache, but while it is on a lot of servers, it isn't as well represented in the top traffic sites. Apache is the closest to leading its field of any package.
Linux? Not a shot in hell. Linux doesn't lead in ANY market. Desktop, Microsoft Owns that. Unix Server market: owned by Sun. Unix Workstation Market: MAYBE won by Linux, but Apple OS X is likely to own that by the end of the year.
Let's be real people. Open Source Products right now are getting better. Whenever they reach your "good enough" level, then you can use them. I need certain features, PostgreSQL provides them. As much as I acknowledge that Oracle is better, PostgreSQL meets my needs and is cheaper.
However, to suggest that Open Source will win and dominate everything is kind of silly. This Manifest Destiny, that we will in the long run own the world is kind of silly. It seems that the areas Open Source has "won" have been small servers/daemons (sendmail, bind, etc.) where the open source version does the trick, the needs are limited, and therefore there is limited value added options.
However, let us understand where open source packages are good and where they need work. The big-three software titans will be remembered as the Carnegies, Rockafellers, etc., of the early computer era. However, to be deifying Linus this early is kind of silly, don't you think?
This anti-VC rhetoric on Slashdot is REALLY childish. In most cases, the VCs don't come in and take the shots, they usually provide a board member and try to guide the company, but if they wanted to run one, they'd start there own.
The problem here, there is a HUGE market downturn. The VCs try to get money out of each investment. If one looks like it is doing well (gonna IPO, etc.) they leave it alone. If one is not doing that well but they believe that they can extract money from it, they will try to get some money out.
In this case, the company has a lot of silly side projects. Greenspun was teaching at MIT, including a class that taught the ACS system. Additionally, his Arsdigita University was teaching ACS. With the products being Open Source, training hundreds of people how to be ACS consultants probably didn't make the VCs happy.
The company was run like it had the value of Microsoft with it's side projects. The VCs realized that aD didn't have the goal of maximizing revenue and minimizing costs. They had a goal of becoming famous.
I believe the VCs saw that wall street wasn't going to reward them, because they were running PROGRAMS designed to DESTROY their OWN competitive advantage. Think about it, as a consulting company around a program you have released Open Source, you get hired because you know it best. With everyone learning it, that's not the case.
I've had potential clients approach me asking to hire me for ACS projects. I know others from MIT that get the same.
The VCs have a right to be pissed.
However, the VC takeover is unlikely to work. Completely reinventing a start-up isn't that bad, but aD might be a BIT too big for that.
Umm, NTFS4 and NTFS5 fail. The version of NTFS in Win3.51 I believe fails in 2080. Or perhaps it's old Windows code that dies then. Something kills NT3.1-3.51 in 2080, and IIRC, it also kills Win3.11...
So I am on my cel phone with an important client, and my battery dies. I dart to the nearest pay phone and start dropping change. Went through a 35-cent startup, and 3 5-cent increases.
Without a payphone, I'd be SOL.
I can't tell you how many times you are in an airport, and you see people on the pay phone. With the digital cel phones, the rates are good but the service is choppy.
Cel phones won't kill the payphone.
The "paper" disposable cel phones that I saw on a science portion of the news a few weeks ago will. Payphones as they are will probably be vending machines for these paper cel phones, and there no longer is a reason to not have an emergency out-going only cel phone on you.
Okay, thanks, I couldn't get the details out of the site, it was/.'d.
I see your point. That's very salesy.
My, "sales" announcement was too the extent: we have our new discussion forums up. If you have feedback, e-mail: (email address and site hidden).
Yeah, I didn't see the Kozmo Ad at the bottom, okay, that is very sketchy.
I see what you mean regarding the administrative e-mail being excessive, also.
What do you guys think (not legal advice, ethical advice, the cost of $50 judgements is kinda insignificant regardless) is appropriate in a technical announcement.
Obviously, people opt-ing out shouldn't get routine e-mail. However, if there is a technical problem, you can obviously e-mail them. Is mentioning new features after the technical announcement WAY over board? Or is it only rude if you have a sales pitch there? I mean, once you send the e-mail...
Slashdot Mirror
Rant that agrees with you...
I have some code that we wrote GPL'd. It is under the GPL because it is easier from a business perspective to GPL it and let someone pay me for a more liberal license. I am likely to switch to the BSD anyways. Our corporate software license is based off the original BSD (with advertising clause) and the right to pay us to switch to a true BSD client. That's my open source involvement.
The Open Source community has written very little. The Mach microkernel was developed by researches at CMU. Like most research projects, some of their grant money was probably government money. Government research is NOT done to benefit whiny anti-corporate bordering on communist high school kids (which seem to dominate the slashdot posts). Government research is to advance national security interests or advance technology to benefit society and particularly the corporations that use it to power the economy. The US Government is interested in economic growth and security.
The BSD System was written by researchers at UCB. It was funded with some grants, and they developed a free implementation of Unix. That is made available for all Americans (and in this case, all people in the world) to use and advance the country.
My tax dollars should NOT be used to fund people whose objective is to derail one of the fastest growing sectors of the economy.
Linux borrowed heavily at various points from BSD. Linux also completely swiped the GNU toolset and now we have distributions, all but one of which don't acknowledge that they started as the GNU collection. The true irony is the RMS tirade about calling it GNU/Linux. I agree with his point that we want people interested in a Free Unix, not Linux in particular. If someone came and wrote this amazing new kernel, we would lose all the mindshare (and credibility and education done to the public) because it wasn't Linux, and they don't know GNU, they know Linux.
Major "Open Source" Milestones:
Kerberos: MIT Research Project
BSD: Berkeley Research Project
Mach: CMU Research Project
Apache: began as a set of patches to NCSA HTTPD (National Center for Supercomputing Applications, University of Illinois, Urbana-Champaign)
I'm sorry, but university research has spawned some impressive technologies that people have hacked in new ways and done some cool work with. But there is no "open source community" that owns this work. All tax-payers own this work that was created for the nation on grant money.
Just because you run Linux doesn't mean that you own CMU, MIT, UI, and UCB's work. Sorry.
I hate to sound biting, I like a lot of open source software. I respect RMS's beliefs. However, this is absurd.
The fact that RMS defines freedom one way doesn't make it so. Instead of spouting about Free Beer and Free Speech, why don't you think for yourselves for a moment. RMS declaring freedom one way is all well and good. Without a doubt, BSDL meets even RMS's definitions. His complaints about the advertising clause is ironic, because if the GPL included in, there'd be none of this GNU/Linux issue.
RMS: I'm a fan, and I respect what you've done here. You've done a lot of great work. Unfortunately, ESR's minions have made a mess.
Alex
SAMBA without WinBind still isn't a general purpose File Server. You need to go and create the Unix accounts. My understanding is that WinBind is Linux only right now, but that it is being merged into the core code. Either way, SAMBA has a great niche in small group servers. For example, if I want a file server for 5-10 people, SAMBA rocks. You can join the SAMBA server to your NT Domain, setup the usermap, and give users access to the shares. For that purpose, I love SAMBA.
We are starting to evaluate SAMBA 2.2, it's great. Our core infrastructure here is NT4 for various reasons, but our OpenBSD machines all share out the relavent portions via SAMBA 2.0. It's terrific, if I'm not at a Unix workstation and want to work on a developing website, I connect to the share in Windows and go nuts. It's terrifically powerful.
From where SAMBA was 4 years ago when I started playing with it, it's a new world. You used to need to use separate accounts and not have reasonable NT integration. This allowed you to share out files, but not seemlessly. WinBind promises to make SAMBA servers drop and forget, just administer everything from your NT groups.
For a reasonable sized office (>10 people) running Windows, I think you'll be happiest with an NT 4.0 Domain with a PDC and BDC, but you can then do your file sharing off your SAMBA boxes. This saves a LOT on NT CALs, as you don't need NT CALs for PDC/BDC access, just file/print sharing. Let SAMBA handle that, and use NT as a domain controller. That way you have central logons without the expense. The cost of two NT servers isn't bad, the cost of NT CALs as your office grows and multiple file servers as access speed matters makes SAMBA great.
Additionally, if you do any web development, share our your site tree and you can edit files from anything. It's great, whether you are using Linux w/ smbmount, Windows with CIFS networking, MacOS Class w/ Dave, or MacOS X w/ Sharity, you have all your files accessable and can be editted locally. That sure beats SSH/Emacs.
Alex
Our clients needed central control of the database and reporting, so we wrote a report generator. Our PHP System includes tools ror administering your database (the data side, not the DBA side) so anyone with a browser (and obviously a valid password) can do it. We wrote a report generator that you enter the SQL, it generates the report. If you want it printable, you click on the "export to excel" option, and it loads the thing up in Excel (gotta use IE for our admin, Netscape's busted CSS/DHTML support screws it, and Netscape 6/Mozilla isn't there yet), and they can print. As an added advantage, the Excel format is trivial.
Excel can read tables, you just set the content type. If you need fancy formating, plug it into the PHP source.
If you don't want to code, there are dozens of programming shops that will develop a system for you for a reasonable fee.
You don't want to spend money or effort? Well, you're in trouble. Labor isn't free. Do it in house or hire someone. Some of our baselevel systems (including our DB class, which can help you abstract your DB from the engine, you'd need to write your own if you aren't using Postgres or MySQL) are available on our website. If you want to hire us, e-mail myself or sales@feratech.com.
Alex
We all think of the unions as the semi-skilled laborer unions (auto-manufacturers, teamsters, etc). In those scenarios, the employees have a WEAKENED bargaining position because they have skills. Why is this?
As a machinist, my time working on machines is more valuable because of my skills. My opportunity cost is unskilled labor. Therefore, as long as my employer is willing to pay be slightly more than minimum wage, I will continue with my skill-based job.
That is the goal of those unions.
In tech, there is no monopsony (single hired, for example, the colluding big-three automakers, only factory in town, etc.) so unionizing directly isn't as useful, we can all jump ship.
However, we shouldn't look at the blue-collar unions as an the example, look at the unions for professional sports. In those cases, much better paid people than ourselves have unionized to fight monopsony hiring. In that scenario, the union negotiates for reasonable compensation, etc.
I still don't want to unionize, but I think that we are better off looking at pro-sports for the example than the mob-run extortion rackets that we are talking about.
A former customer of mine had that problem. They did the bridge loan (small investment to cover you between investment rounds), and got a decent piece of the company for cheap money. About 2 weeks later, the company closed on a much larger round at a higher valuation (as a bridge, the went in at the previous valuation), and the investment was looking good.
Unfortunately, they also chose a poor VC, and the VCs couldn't raise the money when one of the groups involved dropped out. The company that the client had invested in laid off 80% of its staff and was looking to sell its technology.
It's a shame. Slimy VCs screw a lot of businesses.
Id didn't choose OpenGL because it was cross-platform.
Id wanted to port Quake to the Voodoo chip, which was an exciting new and powerful 3D graphics processor. However, they had already done VQuake, and part of the contract was that they couldn't port the game to another graphics chip.
This meant they couldn't do another card specific port. They started doing Direct3D Quake, but D3D 3.0 was a total disaster, so he switched over to OpenGL and created the GLQuake we know and love.
As a result, MS made efforts to improve D3D (to the point that my understanding is that it is pretty powerful and not impossible to learn), but Id continued with Quake on OpenGL because everyone was supporting OpenGL for Quake, Carmack likes Linux and cross-platform, so if they can sell the copies, why switch.
However, OpenGL on Windows is ONLY because of Id. WinNT had OpenGL support because of the 3D Modelling, etc. WinNT needed OpenGL. The only game for OpenGL was Quake, and MS wanted to kill OpenGL gaming, refused to release the OpenGL subsystem for Win98 that was seen as a beta, forcing everyone to write either the Quake mini-GL drivers, or write a full ICD. The plugable OpenGL subsystem for NT was killed for Win98 to kill OpenGL gaming.
Speak for yourself, nobody else has.
:)
They're traitors through and through...
Comment from a friend at Univ. or Maryland (she was a local):
Maryland, Southern Efficiency with Northern Charm
There was an article on the affect of losing company loyalty on the bottom line. Layoffs have a tremendous affect on loyalty, especially at a company like Intel. If you lay people off to avoid a short-term profit shortfall (and therefore are just propping up the stock price), you show people that they don't matter.
In this case, Intel isn't being a slime and rescinding offers (which can kill you at college recruiting, schools may ban you for a few years), and is offering to give them their signing bonus for NOT joining the firm. The recruit is happy (free money), the employees are happy (they aren't cutting staff to make way for the new people b/c of the intricacies of recruiting), and hopefully the company avoids burning money).
The short-term layoff strategy helps you in the short term, but alienates employees whose friends were let go. If you get fired for incompetance, etc., nobody feels for you. If you get let go because the company wants to keep its stock price up, that doesn't make the other employees feel valuable.
Regardless of Intel's other issues with employees, this move is a nice one.
Go Intel, you're being less slimy than a lot of companies I know.
Alex
My personal preference would be to drop OpenBSD on the web servers. The security by default makes tweaking it to your secure settings pretty simple. Yeah, the performance isn't up there with FreeBSD or Linux, but to me it is adequate. If I'm throwing beige x86s at the problem anyway, I can just throw one more if I want to compensate for the speed.
:)
For the database, what are you doing? Are you deploying administrator entered data (in which case transactions are meaningless, go throw MySQL up there), or are you really getting and processing data.
One compromise (for added complexity) is to run two database servers. Put a FreeBSD or Linux Server with MySQL (I prefer PostgreSQL for development, but MySQL DOES Scream, Postgres is starting to annoy me on one of my platforms) hidden from the Public network. Load it with RAM and RAID 0 drive system. Just run the backup after you make changes, and you're good to go.
Then, for the live, interactive data, buy a real database server. Design it, pick your platform (DB2 or Oracle are both fine), and then pick your platform (IBM, Sun, and HP all make nice machines) and put it there. Hire someone who knows what they are doing to put this machine together and configure it, then DO NOT touch it.
This way you don't need to shell out big bucks for an Oracle Server for the content that MySQL is fine with. As you pay for the #Processors*MHz*RISC with the big boys for the database license, if MOST of your content is going to be "static dynamic" data (where you update once an hour/day/week, if that), there is no need for Oracle to touch it. Let Oracle process your transactions, and stick your "content" on MySQL.
Consider something, most database backed websites aren't really database systems. Most use a database as a convenient place to store info and retrieve it. With a "database system" I'm using relations to develop complex systems that can be cross-referenced with my keys, monitoring transactions, real-time changes, etc. If I am writing articles to dump on a server, yeah the DB is convenient, but Oracle isn't necessary.
As all of us here can admin a PHP/MySQL or PHP/Postgres solution, keeping the stuff that we work with there makes sense, then pay someone else for the real iron.
Grin, well despite being attacked for it being '94 and earlier, I'm glad to know that I hadn't lost my mind. The program was "cool" even if not totally useful. I liked the idea... but I'd spend 5 minutes as a slow moving Avatar, log off the Internet, and dial into a local BBS... :)
:)
Neat idea, but the masses groked "channels" easier than people thought, at least once AOL started calling them chat rooms...
Note to self: pick up a copy of AOL for a free month one of these days... figure out why the masses seam to like it...
Note to others: if you are coding Open Source Stuff for end users, get copies of end user systems (Apple w/ MacOS 9, Apple w/ MacOS X, Windows 98, Windows ME, Windows XP, etc.) and see what people like. If you want to design for hackers, don't claim to be making things for the masses.
Either they bought the domain name, or this is the company that has been at this since '96/'97 (hard to tell, Slashdotted). They had cool Avatar chatting systems, etc.
Their stuff was really neat and novel, and probably deserving of protection. They brought a cool concept out to the marketplace years before others, and in doing so helped legitamize it.
You used to be able to go to Tucows and find a couple of chatting paradigms, and this one was really cutting edge back then.
That's very true, I don't get why they don't support foreign keys. However, the scenario you are describing is easier to work around. Given that you have to build a front end for MySQL, only presenting the options that are in that table largely removes the necessity.
You still have "foreign keys" (where Foreign Key is a reference to another table), you just don't have foreign keys, where the database makes certain that they point to an actual reference.
My point is that the MySQL busted/non-existing foreign keys only really kill you because it will let you delete fields referenced by others and other stupidity. MySQL isn't a relational database, it is just a giant file/data store accessable via SQL.
Alex
For starters, we know longer use MySQL, but we wrote a PHP generic DB class that lets us switch between PostgreSQL and MySQL (and will support other DBs that we add). As a DB designer, I hate MySQL. However, it isn't an embarassment.
Remember programming little BASIC toys? If you wanted to keep info you openned a random-access or sequential datafile? MySQL is a set of fast random-access datafiles. It is accessable via a subset of SQL, because people who are comfortable using databases with SQL find it easy.
If all you are doing is supporting a website (no delete operations) then the lack of foreign keys, etc., doesn't matter.
MySQL can be tricked into being useful. You just have to write EVERYTHING (and therefore QA A LOT) in the database. Unfortunately, it is reinventing the wheel.
PostgreSQL is a reasonable database. I don't know why MySQL gets all the credit. But if you have real database logic in your website, it is worth looking into PostgreSQL.
Alex
It was a good thing. MIT wrote code to support their infrastructure. Because of the MIT License, corporations were able to expand and adapt that system to build systems for their network. As a result, we have other Kerberos networks, and Kerberos is used.
Had MIT put it under the GPL, there would be no Kerberos. Without the ability to get machines supporting Kerberos, MIT would have scrapped the project. If they couldn't run Athena off Kerberos, no reason to have developers there.
Their is the theoretical notion that Sun would have GPL'd Solaris to support Kerberos. Fat chance. The companies will ignore GPL'd code, while BSD code creates standards and growth that benefits real users.
The GPL theoretically benefits users. The BSD actually does.
Alex
Kerberos is under the MIT License. It was developed as part of a research project. It was released for a good reason, they let the OS Vendors port it to their systems.
MIT WAS NOT going to handle the Solaris, Ultrix, HP-UX, and Irix (all Athena systems in time) ports, and they needed it to all play nicely. As MIT had a HUGE Unix system, vendors played nice to get the sales, which both got a large Unix installation and their system in front of MIT students.
Kerberos was developed to provide a secure environment for MIT Computing. Lots of University research is government funded. Tax-payer research financed should be available for EVERYBODY, not just those that you consider "worthy." Sorry, Microsoft and its investors pay their share of taxes (and probably more than most of us do), and they should be able to reap the rewards of the tax-payer research, just as the Free Software/Open Source crowd do.
MIT Released Kerberos because they built it for their needs, MIT IS and MIT LCS are not business trying to maximize profits, releasing it furthered their needs, and releasing it bettered mankind.
MS extended Kerberos using a portion designed for vendor extensions. They did this for backwards compatibility for their systems.
Their older domain system was proprietary, this one is too. So what that they used Kerberos code. None of YOU wrote Kerberos, why should you have a say how it was used. MIT's research project turned platform is benefiting Microsoft customers, who happen to be people. MIT's licensing choice benefitted a class of people that all of you trolls would like hung out to dry.
Know your history, TCP/IP did get chosen because of the BSD availability and as a result, it ran on all the machines of the era. DARPA used it because Berkeley wrote it, released it, and Berkeley's version of the OS ran on the prevalent machines, so the Unix OS and TCP/IP became big.
Alex
If you are making improvements to a community supported project, you will give back the code anyway. Otherwise, you need to maintain your own fork and roll the improvements in. It is far better to get your changes into the final product.
Companies with BSD software are most likely to either: A) start with BSD code and morph it into something very different and proprietary, or B) release a value-add to the BSD code, and release the improvements to the BSD stuff to keep synced.
In case A, they could start from scratch (but it is a waste of effort and therefore inefficient, regardless of your holy war), or B, you get the code. In case B, if it is GPL'd, they are likely not to start the project, and therefore, you don't get more code.
Pre-web days, the GPL solved a problem because there was no concept of open source communities. Stallman's free software position had tremendous benefits, and did result in companies GPLing code to comply. However, in this day and age, the open source community creates enough of an incentive to get code released, without the formality of the GPL.
Example: SAMBA. Apple knows that good NT integration would go a long way towards making OS X a good corporate citizen. SAMBA has a solid implementation. However, as it is GPL'd, they can't build a SAMBA based login, etc., system and release it with the OS. Someone would have to package it up as an add-on, distributed separately. This benefits nobody.
I support free software, I try to release what I can. I provide all my clients with the source code to the systems that they paid me to write. I feel that it is the right thing to do. We even negotiate a transferability license if they want, I'm willing to provide all the freedoms of the GPL (even GPL'd code) for a fee.
However, I feel like the GPL is a mixed blessing. It does create an open-source playground where companies have an incentive to enter the GPL world.
At the same time, however, I feel that it is keeping valuable software from existing (benefitting nobody). Additionally, the concept of keeping the code free is silly. If I release code into the Public Domain or with the BSD license, that code is perpetually free as long as SOMEBODY hosts it somewhere. Improvements may not always be free, but the code is free.
The tradeoff is: is it more valuable to (potentially) get more free software out there by coercing companies, or will I get more value from propreitary add-ons, or even, will I get more free software by allowing proprietary add-ons and getting the patches released.
As a software author, you need to make a decision for your code.
One thing you should definitely do, however, is get people to assign copyright (like the FSF does). That way, if you think that their is an advantage to releasing under a new license, you can do so. Just realize that if you go BSD->GPL, the old code remains under BSD. If you go the otherway you are fine.
So the holywar compromise might be: release under the GPL, maintain copyright (and get updates assigned), then if a compelling case is made, you can proprietarily license it to a company making the add-ons (perhaps distributing or donating some of the fees, something to discuss in the assign copyright place), release under the BSD license, etc.
Both have their place, but you have to decide which will generate more benefit.
Alex
This doesn't let you break the law and remain anonymous. It says that you can't get the identity so that you can prove the crime. This is VERY important.
Previously, you could file a BS lawsuit against John Doe, then subpeona the identities to identify John Doe. This allowed harassment.
What this does mean, however, is that if the SEC wanted to investigate, they could get the names.
Realize this case: the shareholders are suing that the stock ran up for a company on the verge of bankrupcy. None of the actions of these individuals caused the company to go bankrupt. The company's inability to have Cash>0 did so.
The shareholders aren't suing that the price dropped. And if the company's business plan required a certain stock price, well, that's their problem. Public companies should be able to function regardless of share price.
The judge didn't rule that you can't reveal names, merely that you can't without a compelling reason. If there was evidence of a crime, they'd be released.
For example, if the company compiled a list of short sellers, they could probably give the list to the chat room operators and demand the identities of anyone matching that list.
The burder is on those wishing to reveal the names, as it should be.
Alex
We may actually be running 7.1, I code, not sysadmin. Lots of our code was written before that upgrade. Thanks for the tips, but lots of people have told me.
Alex
Rackspace, they advertise on Slashdot, A LOT.
At my old job, we hired Rackspace to do our hosting based on a banner ad that I saw on Slashdot. Of course I did my research, but that's how I learned about them. Before the dot-com went under a few months later, they made maybe $10,000 off the account (we had two decked out servers).
Now, with my consulting firm, we are a Rackspace reseller. We are delivering them a contract for $14,000. Of course this was because of my good experience before, but it worked.
When looking for Linux servers, I know to look at VA Linux and Penguin Computing. The latter I only know about from Banner Ads and the occaisional mention on Slashdot, the former I know about from tech rags AND their constant presence on Slashdot.
Banner Ads work for creating brand awareness. On the Internet, where competition is fierce, you either need to own keywords on search engines, or use banner ads to generate awareness.
Alex
Trying to stick with the stable releases. But yes, the improvements have been there, but I don't want to go and find all the SQL that needs improvement.
Alex
I make my living promoting Open Source tools. For most of our systems, PostgreSQL is adequate. The lack of Left and Right joins makes me want to shoot myself occaisionally, and we are limited in our ability to scale, but the product works, and works well.
We can even support our tools on MySQL, but we'd have to make a lot of changes because we have database appications, and we'd have to change them to MySQL applications because MySQL isn't really a database, it's a storage format that is retrievable via SQL.
However, Oracle kicks the shit out of these low-end toys.
Hands down, they are number 1, by far.
Open Source databases are not ANYWHERE close to catching up. MySQL is a simple database whose entire reason for existance is powering websites. If you are a programmer, NOT a database designer, and you think that a database will help you store your data better than files and AWK, you use MySQL.
PostgreSQL is at least a properly designed system, but it is limited.
However, despite NEITHER database approaching Oracle, we have decided that Open Source will overtake them?
This is arragance beyond belief.
I don't get it, what Open Source product has been SO successful that it dominates everything?
The closest is Apache, but while it is on a lot of servers, it isn't as well represented in the top traffic sites. Apache is the closest to leading its field of any package.
Linux? Not a shot in hell. Linux doesn't lead in ANY market. Desktop, Microsoft Owns that. Unix Server market: owned by Sun. Unix Workstation Market: MAYBE won by Linux, but Apple OS X is likely to own that by the end of the year.
Let's be real people. Open Source Products right now are getting better. Whenever they reach your "good enough" level, then you can use them. I need certain features, PostgreSQL provides them. As much as I acknowledge that Oracle is better, PostgreSQL meets my needs and is cheaper.
However, to suggest that Open Source will win and dominate everything is kind of silly. This Manifest Destiny, that we will in the long run own the world is kind of silly. It seems that the areas Open Source has "won" have been small servers/daemons (sendmail, bind, etc.) where the open source version does the trick, the needs are limited, and therefore there is limited value added options.
However, let us understand where open source packages are good and where they need work. The big-three software titans will be remembered as the Carnegies, Rockafellers, etc., of the early computer era. However, to be deifying Linus this early is kind of silly, don't you think?
Alex
This anti-VC rhetoric on Slashdot is REALLY childish. In most cases, the VCs don't come in and take the shots, they usually provide a board member and try to guide the company, but if they wanted to run one, they'd start there own.
The problem here, there is a HUGE market downturn. The VCs try to get money out of each investment. If one looks like it is doing well (gonna IPO, etc.) they leave it alone. If one is not doing that well but they believe that they can extract money from it, they will try to get some money out.
In this case, the company has a lot of silly side projects. Greenspun was teaching at MIT, including a class that taught the ACS system. Additionally, his Arsdigita University was teaching ACS. With the products being Open Source, training hundreds of people how to be ACS consultants probably didn't make the VCs happy.
The company was run like it had the value of Microsoft with it's side projects. The VCs realized that aD didn't have the goal of maximizing revenue and minimizing costs. They had a goal of becoming famous.
I believe the VCs saw that wall street wasn't going to reward them, because they were running PROGRAMS designed to DESTROY their OWN competitive advantage. Think about it, as a consulting company around a program you have released Open Source, you get hired because you know it best. With everyone learning it, that's not the case.
I've had potential clients approach me asking to hire me for ACS projects. I know others from MIT that get the same.
The VCs have a right to be pissed.
However, the VC takeover is unlikely to work. Completely reinventing a start-up isn't that bad, but aD might be a BIT too big for that.
Alex
Umm, NTFS4 and NTFS5 fail. The version of NTFS in Win3.51 I believe fails in 2080. Or perhaps it's old Windows code that dies then. Something kills NT3.1-3.51 in 2080, and IIRC, it also kills Win3.11...
Alex
So I am on my cel phone with an important client, and my battery dies. I dart to the nearest pay phone and start dropping change. Went through a 35-cent startup, and 3 5-cent increases.
Without a payphone, I'd be SOL.
I can't tell you how many times you are in an airport, and you see people on the pay phone. With the digital cel phones, the rates are good but the service is choppy.
Cel phones won't kill the payphone.
The "paper" disposable cel phones that I saw on a science portion of the news a few weeks ago will. Payphones as they are will probably be vending machines for these paper cel phones, and there no longer is a reason to not have an emergency out-going only cel phone on you.
Alex
Okay, thanks, I couldn't get the details out of the site, it was /.'d.
/. thread, someone stop this...
I see your point. That's very salesy.
My, "sales" announcement was too the extent: we have our new discussion forums up. If you have feedback, e-mail: (email address and site hidden).
Yeah, I didn't see the Kozmo Ad at the bottom, okay, that is very sketchy.
I see what you mean regarding the administrative e-mail being excessive, also.
What do you guys think (not legal advice, ethical advice, the cost of $50 judgements is kinda insignificant regardless) is appropriate in a technical announcement.
Obviously, people opt-ing out shouldn't get routine e-mail. However, if there is a technical problem, you can obviously e-mail them. Is mentioning new features after the technical announcement WAY over board? Or is it only rude if you have a sales pitch there? I mean, once you send the e-mail...
That's for the comments...
Wow, a useful
Alex