I posted about this in that article. Jean-Marc only buys boxes because he is too small to economically make his own. Once he has the revenues to support his own box-making factory, he will have one.
Are you just trolling? Congrats, you made it to +4.
Sen. John Kerry, D-Mass., introduced a bill in November requiring service representatives to disclose their physical location each time a customer calls to make a purchase, inquire about a transaction or ask for technical support. The proposal targets the increasingly popular decisions by companies to move their call centers overseas to capitalize on low labor costs.
I found this paragraph to be as interesting as the rest of the article. Can we please know who we are dealing with?
I use ATTWS. Their coverage is great. Their support was great. Their prices are higher than the other wireless services, but you get what you pay for.
Before DEC, every support call was handled efficiently and they tried to keep us happy. My sister told me it was a pleasure to talk to them., even when everything did not work out the way she hoped. The great support was one of the reasons to continue using the service.
She, I, and other friends have made several calls in the last few weeks. None of them were remotely close to making us happy; we were more upset after the call than before. Each time, the call was answered by a lady with an Indian accent. I do not KNOW that the calls are being routed to India; we all could have reached the same lady working in the US. I do know that instead of making us happy, she kept putting us on hold, then returned to tell us "The system does not allow for that." She told us information that contradicts the policies of the last 7 years. The big one was that switching between price levels within a plan requires contract extension. She told us that the levels listed on the web sites were not allowed to existing customers. She told us that the never-ending promotions for a given plan would be canceled if we switched price levels. Every tech before DEC told us contract extensions were only required when switching between the major plans, that they used the web site to look up the current price levels, and that promotions would only be lost if we switched major plans because the promotions are plan-specific.
She refused to tell me where she was located. She refused to tell me how long she had worked in this job. She refused to transfer me to a supervisor or anybody else.
She upset my sister enough that my sister will be changing carriers within weeks. (Thank you for number portability!)
ATTWS: Why did you replace your great support service with this customer-losing system? Are you trying to lose all your customers? Will you publicly declare you are changing it back really soon to keep us as customers?
If there was a law stating that they must identify their location, then I could at least know the support died because of offshoring. I will be in their American-staffed store tomorrow to see if the salespeople there can resolve our issues.
First you need a method for knowing how much throughput you need for your personal use. Then you need a method to find the total throughput you have used so far in the current billing period.
If the ISP will complain that you are exceeding their standards for throughput, they must give you a method for knowing how much throughput you have used in the current billing period so you can stay within their limits. If they will not provide this, they cannot ask you to monitor yourself.
The system: 1. Take total allowable throughput. 2. Subtract your average throughput and maybe a little extra for emergencies or for known extraordinary events, such as the imminent release of software you will want. 3. When using BitTorrent for your own use, leave it running until you get what you need. Keep it on a little longer if you know that the current files are in high demand, such as one extra day after the release of a new core. 4. Calculate how much extra throughput you have available without passing the limits. 5. Starting on your birth date (day of month), turn on your filesharing program for others to use. When the throughput reaches the number from step 4, turn it off. 6. Near the end of the month, turn it on to use any leftover throughput.
If everybody starts running it on their birth date, there should be a good distribution of people running the software for the first 28 days. The "use up what is left" will assist with the lower number of birthdays at the end of each month.
The only hole in this system is the first few days of the month, and that may be covered by people forgetting to turn the software off at the end of month. If you (accidentally) leave it on for the first 2 days of the month, then you should just have less throughput available at the end of the month.
I am assuming that your allowed throughput is enough to keep the software running for at least 3 days during your "assist others" phase. Even is you break the limit, it should not be by much, and you can tell your ISP you are trying to meet their standards. Hopefully there are enough people who do not have to worry about throughput issues to cover the low availability periods.
This system allows you to meet your needs while being a good filesharer without getting in trouble with your ISP. Maybe BitTorrent could include these recommendations so enough users would follow them to make them popular.
I found a program called "Real Alternative" that plays Real video files on MSWindows. I think I first found it from another Slashdot post. It works well without any of the bad stuff of the official players from Real.
From the installer notes: --- More information and updates can be found on the following websites:
Real Alternative will allow you to play RealMedia files. This way you can play RealMedia files without having to install RealPlayer/RealOne Player. You do need a player that is capable of playing RealMedia. The included Media Player Classic supports it and works very well.
Supported: - RealAudio (.ra.rpm) - RealMedia (.rm.ram.rmvb.rpx.smi.smil) - RealText (.rt) - ReadPix (.rp) - RealMedia embedded in webpages.smi and.smil files only play the first part of a clip. This is a limitation of the current Media Player Classic. ---
I tried to RTFify a medium-sized Word document that was less that 5 MB. The resultant rtf was over 200 MB.
MSWord bloats documents. I have seen 100KB text documents become 200KB RTF documents when adding formatting, then become 2MB documents when saved as a MSWord DOC file. The same text must exist in all three. The formatting adds some markup. MSWord adds tons of extra overhead.
I do not use MSWord enough to know if there are options for compression. Even if MSWord uses compression, what compression routine frees 97.5% of the size? I have seen databases compress 80%; I have never seen compression better than 95%.
I am not stating you are wrong, just asking how those numbers are possible. I look forward to your informative answer.
--- As far as using templates to create documents, nobody is suggesting you stop using them, just that you save/send the document in a more ubiquitous format than MSWord's DOC.
I work with a company that has IBM as its largest customer. Before IBM killed SmartSuite, the people who communicated with IBM were forced to use SmartSuite. Since IBM declared SmartSuite dead, using MSOffice has become acceptable, and sometimes required.
I would not call 123 or Wordpro "crappy programs". They both were easier to use than MSOffice, and had many features that MSOffice does not have or does not do well.
Those people still use Lotus Notes. This is a good thing because we are able to use Notes for other applications that improve the business but would not exist if the cost of development was not so low because Notes is already available. We are planning our upgrade to Notes6 because IBM is finishing their upgrade and we need to stay current with IBM.
I have been recommending they start using Linux for the past year. Their IT director only knows MS, and gets very upset at the thought that anything else is valid. He also said that he wanted to replace the Wintel-based IBM and Compaq servers because they were incompatible with his Wintel-based Dell servers; I have little respect for his opinion, but I also have little influence. I really hope that this move by IBM will force my client to consider Linux desktops.
Since early last year, I have predicted the fall of MS for the end of 2004. For that to happen, I predicted: 1. A major IT company will pick a Linux desktop environment and start investing until it far surpasses the others, making the other desktop environments into novelty programs. 2. A major non-IT company will announce the switch to Linux by March 2004, thus insuring that MS will crash by the end of 2004 3. On a different note, I feel IBM is attempting to kill Lotus Notes as part of a misguided attack on Sun.
This article states that IBM is fully moving to Linux desktops. One of the links at the bottom was that IBM will stop using Notes internally, although from personal knowledge I know IBM is still migrating to Notes6.
My current relevant thoughts are: 1. IBM using desktop Linux means they will need it to be fully functional for their needs. They have the resources to make it so. The article does not mention which desktop was chosen. Another post suggests IBM uses Gnome, but does anyone know the official policy? IBM was my first choice for the decision-maker. Once they make the decision, their chosen desktop will become the desktop of choice for the world.
2. With a Fortune 100 company like IBM switching, we can expect a large non-IT company to follow soon. There are only 2 months to go before my deadline.
3. I believe that most business application belong on the Notes platform. I do not believe the Notes client needs to be part of the solution; the benefits of Notes' rapid application development are still realized when using a browser as the client. The Notes client has better handling of RichText, but I do not believe business data should be stored in RichText. The Notes client is fully integrated with email, and the Notes6.5 client is fully integrated with Sametime, the Notes IM client; losing these abilities will cause difficulties for IBM.
The big issue is that the Notes development client (Notes Designer) has yet to be ported to Linux. Without Notes Designer, there is no development. The Notes server (Domino) was ported to Linux by a rogue, and that has turned out well. Hopefully there is still enough flexibility in the IBM-controlled Notes platform team that the Notes clients, including Notes Designer, will be ported soon.
I have difficulty believing that IBM can finish the move to the Linux desktop without a Notes client for Linux. Their current culture is based on the ubiquitous and constant use of Sametime. Without the Notes client on Linux, their non-technical employees will need a replacement. Even if the Notes client is not ported, they must port a Sametime client to allow management and everybody else who moves to Linux to continue to communicate with those who have not been migrated.
I dream about IBM porting the Notes client to Linux. With the current situation, if IBM does not do the port soon, then IBM will falter as their ability to communicate internally diminishes.
Yes, they lose some of the ownership, but the effects should not be evident for a few years.
Yes, they could pay for all those reports while staying private, but there are reasons that every company that has to file those reports go public. The SEC frowns on doing that. I think there are fees/fines/penalties that apply in that situation that do not apply to a public company.
Yes, issuing stock is about the worst way to get money. Almost all of the alternatives are better (except cousin Tony.) You then contradicted yourself saying they were doing it for the money.
With its proposed share issue (1/3 of the company), the owners will all of a sudden just get $100million now in income (instead of the $300m they were getting before).
No, and your math is bad. If they sell 1/3 of the company, they still have 2/3rds so they get $200M of the $300M, not $100M.
The owners get ZERO "extra" money. It's not as if the billions raised by google will go to the existing owners.
Where do you think that extra money goes? If they just issued more stock, then the money goes into the bank and they can give themselves bonuses. If it was their stock they are selling, then the money goes to them. From what I have read, it seems the first scenario is the usual. Does anybody know why the current owners would reduce their share if they do not directly gain from it? I know some companies go public to be better able to expand, but that does not seem to apply here, although I am certain a few people are trying to decide what to do with $12B. How did the dotcoms "cash out" by going public if the money does not go to them?
If you or anyone in the household has any medical condition, or there are children in the household, the government will pay for basic landline service if you cannot afford it. I know this is true for Pennsylvania, but believe it is nation-wide.
--- I stopped my landline service a few years ago. I have used the cell phone to call 911 several times.
The worst was for a car running head-on into the end of a cement median at over 50mph. I called it in within seconds of it happening. There was tons of paper flying around on the highway, and many cars were swerving. I tried to slow and my car slid for a second, so I headed for the exit ramp at Valley Forge. I happened to look left and saw the car hit and the front-end crumple. The medians were not lined up properly due to the construction on Route 202, and there were no orange bumpers in the area. I had the cell phone dialed before I reached the end of the ramp. 911 said they already knew about the accident that happened 5 minutes ago, but I was able to convince them that this was a new report since it had happened not more than 20 seconds before they answered. I would have had to go through at least 2 traffic lights to reach a landline. If I had waited to find a landline, they would not have believed this was a new report.
If you can afford it, then a cell phone is better for more emergencies than a landline, but the landline is free for the needy.
Google management will have 6 months after the IPO to find ways of keeping anybody important to their operation that is about to become rich. There are SEC rules that employees are not allowed to sell their shares for 6 months after the IPO. Here is a very good explanation. It is possible that people who "know too much" may not be allowed to sell their shares for years.
The other side is that while most full-timers at Google have either shares or options, they probably did not amount to much compared to their salary. If those shares become worth ten times their value, and the employee decides to cash out, they will probably gain a few years salary. That might be wrong in this case. With a market cap of $36 billion, even a few shares may be enough to retire. Most companies plan at least 10% of their stock to cover employee options. $3.6 billion / 650 employees gives an average of $5.5 million. On the good side (for us), maybe most of those options are not vested yet.
The big winners are the ones who started the company or invested cash for shares. The investors should not matter to operations, and the founders have already made enough to retire if that was their preference.
--- Here is a link to the story that Google might be forced to IPO that I should have included in my last post. 500 share or option holders and $10 million in assets forces an IPO.
Google has remained private as long as possible. If their VCs were looking to cash out, they could have done it before the crash. And everyone has been asking them to IPO for the last 2 years to kickstart the stock market. It was smart of them to wait until the DJ was above 10,000, but probably unnecessary.
They probably are not worrying about buying somebody either.
The reason they are going public is because SEC rules force companies with a certain number of owners to go public. The companies have to file all the costly paperwork as if they were a public company, and they lose most of the advantages of staying private, such as not releasing all that information about their activities. There is little reason to stay private, and the extra cash from the IPO is handy for paying for all that paperwork.
The famous case of this happening was Microsoft. Too many employees were exchanging shares privately, and the SEC forced them to go public. They did really well, and you cannot blame their decline on being a public company since the prior management is still running things. OTOH, because MSFT is public, the shareholders can insist on new management, but they will probably wait until the stock goes under $10, and that will be too late to save the company, if it isn't already.
Google is being forced into going public. There is no need to look for extra motives from their investors and management.
The whole webpage is a "recommendation", so it is defined as an opinion. It is MY opinion, but still just an opinion. I am glad you enjoyed it.
--- Operating System --- I was an avid gamer. (I have not had the time for a couple of years, since I am trying to launch several software products.) Windows98SE is still the best gaming platform. The PCs I build are for single-user, so the multi-user abilities of the NT line are more of an annoyance than a benefit.
A friend uses Win2K, and had problems with many games. We tried to troubleshoot them, but he eventually just gave them to me. We also had a few hardware issues. (He returned that stuff.)
I do not like WindowsXP. I do not like the interface, and even after changing everything back to "normal", it still performs differently in ways I do not like. (I also use several Linuxes, so just different would not bother me.) I also seem to have the ability to crash XP often. I know people who are using it, but I cannot recommend it.
My recommendations have not stopped several of my friends and family from running XP. It does mean they have to listen to me say "I told you so" every time they call me for software support, which is much more often than the Windows98SE users.
--- AMD vs. Intel --- I did start that recommendation with "I do not understand".
The savings is less than $100. According to SharkeyExtreme's DEC 31 CPU Price Guide, the P4 2.6Ghz 800Mhz bus is $162. The "equivalent" Athlon XP 2600+ (2.08GHz) 333Mhz bus is $90. So you save $72. - The P4-2.4 800bus is $153. Athlon XP 2400+ (2GHz) 200bus is $75, so the savings is $78, but the AMD bus is one fourth the Intel. - You are running a slower chip and get less than half the bus speed. Does the higher latency more than halve the speed of the bus? Is the equivalency valid when comparing to the 800Mhz bus? - AMD's fastest CPU is 2.2Ghz, but even the Athlon XP 3000+ (2.16GHz) costs more than the P4-2.6, so you cannot compare at the same CPU speed. According to AMD's chip speeds, every 80Mhz increase for AMD is equal to a 200Mhz increase for Intel. Can this be true? - Apple has been selling better quality but slower speed CPUs for years, so the "equivalency" system has tradition. It feels like AMDs equivalency system is a marketing ploy from when AMD had the faster bus and could prove the numbers, but it is still used because AMD fell behind in the 32-bit CPU technology and needs to appear better. - $72 is not much when the system still costs over $1000. If buying the Intel makes the system usable for a few more months, and gives better performance during its life, than it is worth it. My whole recommendation is about spending the money where it makes a difference.
If using an AMD gives a true perfomance increase, then please give me details. If it is to save a few bucks at the cost of any perfomance, then I would prefer to stay with Intel.
This article confuses terms. (BTW, I use the term "good" to mean anything that can be bought, usually referred to as "goods and services". Programming is a "good" when you only care about the final product.)
Outsourcing: Paying another company (to do a task that could be done internally.)
Offshoring: Hiring people from another country (to do a task that could be done by people in the same country as the company.)
You can combine the two by hiring a company in another country to do a task, but it is important to distinguish between them.
Offshoring takes advantages of differences in the markets in different countries for certain goods. If the logistics allow the product to be transported for less than the difference in cost of the goods, then it is economical to buy there and sell here. Traders have been doing this since before the invention of money. With the extremely low cost of moving electronic goods today, production can be done anywhere in the world. (As far as the quality of the software development, I believe there are very few great programmers, so find them wherever you can. If you are hiring code-monkeys, it may not matter except that the cost of logistics must include language translation and extra management.)
Outsourcing has very limited scope of usefulness. It is useful to buy goods from other companies when the cost of doing it in-house is greater than available funds. There are also issues where accountability is necessary, even when the work could be done in-house (such as accounting audits. You need someone who does not have a stake in the company.) Third, there are a few issues where the required expertise cannot be bought, or the cost of entry is so high that everybody buys the goods from others.
The article's example is wrong. Jean-Marc is buying boxes because he needs too few to make it worth the cost of building his own boxes. When he sells enough chocolate that the cost of buying boxes is more than the savings from owning his own box making factory, he will enter the box-making business. This is called vertical integration, and is one of the prime savings from economies of scale.
I have a (currently) small business. I "outsource" my legal needs because I do not need enough to hire a lawyer full-time. IBM requires more lawyers, and so has its own legal team. It does this because it is more economical to hire them than to pay the overhead of maintaining a separate business. Every customer is paying the overhead of any business from which they buy. I am willing to pay that overhead because it is less than the cost of maintaining my own legal team.
The third reason to "outsource" is evident with Operating Systems. The cost of writing your own, and making it compatible with existing systems, is much too high for any company. So everybody "buys" Operating Systems from the few "companies" that have built their own. I use quotes because companies could collaborate on services they need that they cannot afford alone. Linux is an example of companies investing in a good that is needed by many of them. Collaborative lobbying organizations also fit this role, although they also fit the (reverse) accountability reason (such as "Microsoft did not say it; it was the SBA!")
The confusion between "outsourcing" and "offshoring" is designed to make offshoring seem like an extension of the recent (and awful) trend for outsourcing. "You were already considering outsourcing. Offshoring will save you even more!" This will probably do more to kill offshoring than anything else. Outsourcing is always bad when you have the ability to avoid it; every company that outsources a major business function regrets it within a decade. Offshoring is fine if it is economical for your situation, but hire the people; do not add the overhead of maintaining another company.
--- I am a consultant. I am paid to provide expertise that cannot be gained from anybody else. I am
The fact that big business will start pushing money into toolkits doesn't mean that control will be lost.
Control would not (and cannot) be lost, but big business can push the priorities that are important to them.
ROI does not apply to big businesses as a group. IBM's investment can't be recouped by Sun, after all, so there's no reason to think that they shouldn't back different horses.
With OSS, IBM's investment can be recouped by Sun, IF SUN CHOOSES THE SAME PROJECT. If IBM puts a few million into Gnome, then Sun only benefits if they also choose Gnome. If Sun chooses KDE, they probably have to repeat the work that IBM has done. But Sun could choose Gnome, receive the benefits of IBM's work, and then continue with what is important to Sun. That is why I feel the first influential company to choose will have a great impact on all succeeding choices.
more importantly, control in an OS/FS project is going to tend to stick with the fork that produces the best code. Doesn't your reasoning mean that you think that big business, with its big bucks to spend, will be churning out more and better code, than OS coders can, and hasn't that long since been shown to be untrue?
The community will still decide the best code. But wouldn't the ability to spend a few million tend to increase the possibility of writing some decent code?
Why are you distinguishing between "big business" and "OS coders"? The whole SCO/IBM lawsuit is because some IBMers are OS coders. If a company is willing to pay (or hire) programmers to work on OSS, then those coders must become part of the OSS community to be effective. That their priorities are set by the company should not affect their coding abilities.
(Well, there is the part about scratching your own itch, but a paycheck is rather good incentive to scratch someone else's itch. And you would still want to write great code for the reputation/karma.)
Try not to disparage this too much. I am certain that many Slashdotters have been dreaming of the day when big business decides that OSS programming is important, and have their OSS resumes ready.
The posts so far are about Dell's business model, but the article is asking techies to ignore Dell and build their own PCs, then build a business and negotiate for better prices. Why isn't anybody discussing building PCs?
I build PCs for myself, my friends, and my family. Most of them have custom "Solprovider" machines. I pick the parts; I build it; I install the software; I support them. PCs I build remain usable for over 5 years, and I average about 1 support call per machine per year. (If you are interested in what I build, see my October recommendations from the last time I built a PC.)
I am not attempting to turn this into a business. I have a very successful career, and the effort is too high and the margin is too low for this business to be worth my time. I refuse any money, although I expect a home-cooked dinner for my trouble, but then I only build PCs for people I care about. The other side is that these people know that I will never interrupt my paying work to help them, so sometimes their problems can take over a month to solve.
But why aren't you trying to sell PCs? You are already technical. You probably understand what the hardware does. You can learn how to use a screwdriver. You already know how to install software. The bad side is that you might have to install MSWindows if the buyer insists, but you could install grub for dual-booting, so every time the buyers watch it boot there is a chance they may choose Linux.
(Use the Maxtor 160GB drive. Use only 10GB for Linux and they will not complain. Tell them it is there so you can troubleshoot easier, but they can try it if they want. I am actually installing a new hard drive this way this week.)
I usually spend several days to research my recommendations. One day is spent researching the new technologies; one day is spent researching the various products; and a few hours are spent checking prices. Someone who builds more than one PC every 6 months would spend much less time per PC because their knowledge would remain current.
When MS crashes, big business will pour tons of money into the Linux Desktop PC. They are going to pick only one desktop toolkit because it gives a better ROI than splitting the money to 2 or more projects. It is likely that the other desktops will survive for quite some time (reference people still using Amigas), but the others will fall behind unless they use software like the project in this article to borrow from the leader.
if you insist on unification, you destroy most of the features of Open Source / Free Software development that are meant to be its strengths. I know there is still strong development of the BSDs, but most of the corporate money for OSes is going into Linux. Is it "destroying the features of OSS/FS?" (Be careful how you answer. This is Slashdot and we do not tolerate criticism of the Linux development model.) The BSDs will survive just like the alternate desktops, but eventually the leader will far surpass them both in features and marketshare.
The key to the desktops is the toolkits. Eventually one will be chosen. We (all techies) will have little input into which is chosen. The first large software company that makes the choice and puts the money into one toolkit will make the choice for the entire world. It will probably be IBM, although Novell or SUN could take the role; Redhat has already given up. The final choice may be decided by one business manager that has been using Desktop Linux for some time and does not want to change.
Maybe Bruce Perens can influence the decision. As much as he likes KDE, he thinks Gnome is the better choice for a universal desktop (reference Slashdot article from last month.)
Yes, what you will get is a lot of people who are dissatisfied. Nobody says you cannot keep using Slackware 3 at home. But in a few years, all computer software jobs will require knowledge of the primary Linux desktop. Plan accordingly.
(This is a response to an AC that questioned my weight. I am assuming it was not a troll, but actually wanted information.)
I am 6'4" tall and weigh 235 lb. According to this, I am slightly overweight, but not obese.
I do not understand the judgement. I could lose a little from my belly, but the rest of me is muscular. I exercised constantly about 1.5 years ago. My belly was trim and started to show a six-pack. I initially lost weight to about 225 lb, then gained it back in muscles to reach 245 lb.
--- If you were questioning my ability to drink over a gallon of Pepsi or water each day, this site suggests drinking 1/2 oz of water daily per 1 lb of body weight. So at my current weight I should drink almost a gallon of water every day. But my intake has always been excessive.
Actually these numbers are hard to apply because my weight changed over the course of the caffeine addiction. When I started drinking Pepsi full-time, I weighed 175 lb. When I started the support job 8 years later, I weighed 185 lb. 8 months of being chained to a desk and I weighed over 260 lb (which was also a major factor in the acid reflux problems. Gaining 70 lb in <8 months is not good.) When I started the process of kicking the caffeine addiction 4 months later, I was around 250 lb. I stabilized around 240 lb a few years later. The major exercise program June-September 2002 is mentioned above, and as I lost muscle I have stabilized at 235 lb. I exercised some this Autumn, and am trying to make it a habit so I can lose a few inches off my belly.
In related news, I eat more than any other 2 people I know, and have since I was very young. I can easily eat a 30oz steak and be hungry a few hours later. My friends joke that I need to be rich just to supply myself with food. A few people have mentioned the possibility of diabetes. I keep getting checked; the doctors keep telling me that my eating habits are awful, but the only number that concerns them is my high cholesterol count (very likely related to my constant diet of cow.) I have changed my diet several times for several months; there is absolutely no correlation between the quantities I eat and my body weight. As mentioned earlier, I gained 10 lb over 8 years with a diet that varied (depending on my budget) between mostly pasta and mostly steak. In trying to lose weight, I did cut back for several months, but the only effect was fatigue. The only thing that affects my weight is the amount of exercise. (I usually lose weight when I have a girlfriend because I get more "exercise", but the girlfriends always gain weight from imitating my eating habits.)
76 percent of active Web surfers access the Internet using a non-browser based Internet application. Media players, instant messengers and file sharing applications are the most popular Internet applications... The top five applications are Windows Media Player, AOL Instant Messenger, Yahoo! Messenger, MSN Messenger Service and Real Player.
As best I can figure, they are counting unique IP Addresses. Then they counted how many IP Addresses used each "application". 24% of the addresses did not use any of these applications.
I would guess that email does not show since most email clients talk to a server on the local network. The number of IP Addresses of the email servers that cross the internet are not significant enough to show on this report.
They do not distinguish if the connections for Windows Media Player and Real Player were for plug-ins or for stand-alone usage. Both types of media are often embedded in web pages. Windows Media Player is constantly attempting to call home because it is confused about a codec. Real Player calls home for new ad content. I am surprised that Acrobat Reader did not make the list, probably because it uses standard HTTP for communication.
They also did not mention the total population using IM: 20% AOL, 19% MSN, 12% Yahoo could mean only 20% of internet users use IM if every Yahoo IMer also uses MSN, and every MSN IMer also uses AOL. I know that is unlikely, but the statistics are useless for learning the total IM population.
The same reasoning applies to the Media Players: 34% WindowsMediaPlayer, 20% RealPlayer. What is the intersection? It is very likely that 34% of internet users like video; they all use WindoiwsMediaPlayer, and sometimes use RealPlayer because WindowsMediaPlayer cannot play the Real format.
The only valid statistic is that 24% of active IP Addresses did not use Media players, instant messengers and file sharing applications that do not use HTTP. And even that number is suspect because they have not stated that they excluded email, web, and web-crawling robot servers that communicate over the internet. Web servers may rarely initiate connections, but robots could account for a significant portion of that 24%.
--- Using the number of unique IP Addresses is a very poor statistics. They did not mention how they captured the number of IP Addresses. We are behind a NAT, so we have 3 people using 1 IP Address. Most households with broadband use this setup. At companies using proxy servers, there could be thousands of people using 1 IP Address.
I agree that time used by each protocol could be a more useful statistic, but it would be difficult to capture. For example: I could request 10 web pages in one minute, or request the same web pages at one every 2 minutes. It still took me 20 minutes to read the pages. The first method would return time = 1 min; the second returns time = 18 min. A slightly better statistic would be number of requests, but that would be disrupted by web pages that have many graphics.
The only statistic that make sense is total bandwidth used per protocol. And that would be useless for learning the number of people using each protocol, which is the only number of interest outside the infrastructure community.
Thanks. Diabetes runs in my family. I get checked often, and have had a full physical since this incident. There are no indications that I am developing diabetes.
My doctors and family really approve of me using water as my primary drink because of the high probability that I will develop diabetes sometime in my life. Except for this Summer, water has been my primary drink for 5 years.
--- Changing from high sugar drinks to water is very difficult. Sugar controls so much of our somatic responses that many Americans feel hungry when they are dehydrated. The body knows that it had a drink because it tastes the sugar and the blood sugar-level rises. Changing my internal programming so that tasteless water would be recognized as valid input was difficult.
For the water overdose incident, I had not eaten recently. I was programming the computer with a Britta pitcher and a glass next to me. I refilled the pitcher at least 10 times. The drinking, refilling, and bathroom runs were on auto-pilot because my brain was busy with the computer programming. I finally noticed my throat hurt, but initially assumed that it was caused by acid reflux. It was another hour and 3 more bathroom runs before I realized that it was getting worse every time I had a drink. Then another hour passed before I stopped programming the computer and found some food and high sugar drinks to try to satisfy my body. It is amazing how much we can focus on computer programming to the exclusion of our corporeal needs.
The diet version does not have caffeine, but few restaurants have diet root beer, and I will never drink any beverage with "diet" or "lite" in the name.
--- Found this page about breaking caffeine addiction. It includes this list of withdrawal symptoms (as well as another product list that shows Barq's as having caffeine): Headaches Irritability Intensificati on of premenstrual symptoms Fatigue Generalized muscular tension Nausea Lack of appetite Constipation Lack of concentration Disorientation Forgetfulness
It sounds like caffeine withdrawal is PMS. Does "Generalized muscular tension" mean having cramps? Men, want to learn what PMS feels like? Just go through caffeine withdrawal!
--- Yes, I also notice when I have had much sugar during times when water is my main drink. I was high the first week after discovering Tropical Sprite until my body became accustomed to the higher sugar level. And I felt the energy loss when I switched back to water. Again, my body returned to normal in about a week.
I actually switched back because the supermarkets are always sold out of Tropical Sprite. I checked 2 supermarkets twice a week for 3 weeks, and even asked a store manager to stock more. Coca-Cola lost me as a customer because they did not stock enough. From working in supermarkets, I know the soft drinks are stocked by the vendors. I think Tropical Sprite would be a great success if they stocked enough to meet demand.
Summary: I kicked the caffeine addiction. The lack of caffeine had me sleeping incredible amounts of time during the withdrawal. Tylenol kept me functional for the months of headaches. I substituted other drinks with varying degrees of success until water became my main drink.
--- I was a major caffeine addict. I sometimes drank a few pots of coffee at night, but my major addiction was to Pepsi. I drank at least a gallon of Pepsi each day for 8 years, and sometimes finished 4 2-liter bottles in one day. Sometimes I drank Jolt instead, but still in large quantities. I usually slept about 2 hours per day, with a crash for 12 hours of sleep once a fortnight.
Pepsi destroyed my stomach. I did not have an ulcer, but I did have constant acid reflux. This may also have become a problem because before the support job, I rarely had to speak much. As phone support, I was talking for 10 hours per day. The new pattern of my mouth constantly moving may have caused my stomach to believe that food was coming, and I rarely ate while doing support. (Anybody doing phone support should have some kind of snack to keep the stomach happy. If you are over-weight, make the snack something like Pirate's Booty that has no value except to give your stomach something to work on.)
I was transferring from support to administration. My first day as an administrator would not be for 2 months, but I definitely had the job, so I was not worried about my performance in support. It seemed the perfect opportunity to kick the addiction.
I first switched from Pepsi to Mountain Dew, thinking that the lemon-lime drinks had less caffeine. (You can laugh now.) After reading the label and realizing my mistake, I switched to Gatorade. No caffeine, but tons of sugar to match the Pepsi.
I slept 10 hours every day during withdrawal, and woke up still tired. I had headaches for the first time in my life. I learned the joys of Tylenol, which was necessary so I could think while learning my new job.
Withdrawal lasted almost 3 months. After the first 2 months, the sleep I required started to reduce until by 6 months I was sleeping 4 hours a day, which is what I needed before the addiction.
After about a year, I switched from Gatorade to Sprite. That lasted another year, then I switched to water. I actually overdosed on water. The lack of sugar meant that my tastebuds did not recognize that I was drinking, so I was constantly thirsty. After pouring about 4 gallons of water through my system in 5 hours, my throat was stripped (and I was sick of running to the bathroom every 20 minutes.) I had to alternate Sprite and water for a week. Then I managed to stay with water with an occasional Sprite until this April, when Tropical Sprite (sold under the silly name "Sprite Remix") was released. I really like it, and it became my primary non-alcoholic drink for the Summer, after which I switched back to water.
--- After-effects: If I have any caffeine, I feel it immediately. About half an hour afterwards, I crash; it becomes almost impossible to stay awake. I will not drive a car for the hour after I have caffeine. (This happens because many restaurants have awful-tasting water and do not serve alcohol or Sprite, so I try the root beer. Waitresses usually insist it is not caffeinated, but they are often wrong.)
--- The parent post's advice seems good. I wish I had read it before my attempt. One week was not enough for my withdrawal, but YMMV.
I do not know. Does the Linux2.4 DRM include the BIOS locking ability? Was there any BIOS with that capability when 2.4 was released?
A Linus quote: I want to make it clear that DRM is perfectly ok with Linux! He does not like DRM, but feels it may have positive uses.
But could DRM in Linux ever work? "Making DRM in Linux secure would be like winning a hand of poker against someone who can change all the playing cards at will," wrote [Tony] Mantler. DRM would require proprietary binary modules. How long until someone in the free world (somewhere without the DMCA) released a module that returned "passed" for every function in the verification API?
These quotes are from April 2003, so the kernel developers may still be debating it. I doubt any implementation available today has all the abilities we fear.
You are contrasting the "number of internet users using Google" to the "quantity of money that Microsoft received in a specific period." There should be no correlation. Those Windows95 and WindowsXP surfers should have needed to pay only one time for their OS. MS should not make any more money from them even if they continued using the software for decades.
For the first time, MS has made less money in a 3-month period than the previous 3-month period. Their growth has stalled, and the decline is starting. Because of their fun financial practices, the fall will be fast. Just wave goodbye as they go past.
Unless you are dependent on MS software for part of your income. Then you need to examine your budget for things you can cut, or look to replace that income by learning a new career.
--- Google, which is the geek's best friend, which would have naturally higher numbers than many other sites. Google is great for geeks, but geeks are still considerable outnumbered by non-geeks, and many non-geeks like Google too. Your proof sounds like "I found some shells on the beach, therefore the beach must be made of shells." Please read about writing proofs. (Your local college should have a course in algebra or symbolic logic that will cover the important concepts.) If you want statistics on what geeks use, get your statistics from a geek-only site like Slashdot.
In the near future we will have two kinds of platforms. One platform will be a fully integrated appliance that runs Windows in DRM-nightmare mode with BIOS lockin.
No. The point of the article is that MS has announced they are dying. Bill knows there is no path for MS to survive in any recognizable form, and has given up. He already gave himself a billion dollars in dividends, so he will not starve. I expect their stock to be halved by this time next year.
Without MS leading the way into DRM-nightmare mode, I doubt Apple will even give it lip service.
Do you want to suggest to Linus that DRM would be a great addition for Linux 2.7?
--- Welcome to the new world of computers. - MS is dying. Soon it will be dead. - Every major company with any MS products will at least be researching how to get rid of them by the end of next year. - More money and developer time will be spent on the Linux desktop in the next 2 years than has been spent so far. The windowmanager that gets the early money will become the only windowmanager. - Computer technologists who were smart enough to avoid MS technologies will reap major pay increases as companies fight for these human resources. - Computer technologists who depended on MS technology will flip burgers. Some may attempt to learn non-MS technologies, but few will understand how operating systems and programs are supposed to work after years of MS indoctrination.
Hiring a code maintainer is exatly what a company does not want to do.
Every company should have someone who is responsible for tracking what software is used, who is using it, and what releases are used. They should check at least monthly to see if there are any updates, what the updates do, and whether the updates are critical or desirable for the company. This person is usually an administrator with no programming skills. This applies whether the software is all proprietary, all OSS, or mized. It is REQUIRED if proprietary software is involved to track the per-person and per-server and per-CPU licenses.
Maintainers are very expensive, and you can't hire them in pieces.
I cannot be the only consultant who works hourly. Companies hire me when they need me. I go away when my task is done. They know I am available if they need more assistance. Does that fit your definition of "hiring in pieces?
The closest you can come to this is to hiring a company which specializes in maintaining the package you wanted.
Does it matter if it is a company or an individual? 1099s are 1099s regardless of how many people do the work. Hiring a company does mean you can have a better SLA than an individual can deliver. I had 4 companies want me in their offices for the same day in May; I was able to keep them all happy, but that is because my relationships are based on my controlling the schedule while making them happy, and they all know and accept that. I am hired through consulting companies, so the customers could ask for someone else from the consulting companies if there was anyone else who could handle the tasks.
Why do they need to specialize in THIS package? I am often shown the systems where some critical application is running and asked to fix something. I have to figure out where the files are, what language they are in, where the issue is, write the fix, and attempt to test it without destroying the live system. (The last time this happened, they pointed me to the wrong server!) A good programmer can work on anything. If you know the programming language, then you can narrow your search to people/companies that specialize in the language, but why limit yourself to somebody that knows THIS product? Maybe it was on the resume because they did some little fix for another customer. Check the list of active contributors to the package and see if one of them is willing to accept money from you in return for guaranteeing that your needs are filled. If not, then either find someone willing to become a (paid) contributor, or just find someone who is really good at programming.
In that case, you've lost the price advantage you had, and you've not gained the source.
It does not stop being OSS just because you hire someone else to look at it. I believe support for most OSS is equal or less expensive than equivalent support for proprietary software. It may have something to do with people being able to read the code. Or it might be that people who work with OSS tend to like software much more than people who just do it as a job.
You don't have people on site who can actually read the source,
Depends on your contract. Did you ask for someone onsite 24x7? Do you really need that? If you are hiring a company for support, then get a 4-hour SLA for critical issues. If the system is mission-critical, it should have enough redundancy that there is never a critical issue.
and if your service provider goes under, you're left without an escape.
Do you understand OSS? If your service provider goes under, you choose another service provider.
What's the advantage of this over buying from Microsoft?
When Microsoft goes under, there will be no updates. No bug fixes. No security fixes. With OSS, there is always somebody who can fix it. Your worst case is that you hire the best programmer you can find and give him much money to work on the software. But your company does not fall apart because you cannot fix/patch your productivity apps and web servers and every other piece of software you depend on.
Slashdot Mirror
I posted about this in that article. Jean-Marc only buys boxes because he is too small to economically make his own. Once he has the revenues to support his own box-making factory, he will have one.
Are you just trolling? Congrats, you made it to +4.
Sen. John Kerry, D-Mass., introduced a bill in November requiring service representatives to disclose their physical location each time a customer calls to make a purchase, inquire about a transaction or ask for technical support. The proposal targets the increasingly popular decisions by companies to move their call centers overseas to capitalize on low labor costs.
I found this paragraph to be as interesting as the rest of the article. Can we please know who we are dealing with?
I use ATTWS. Their coverage is great. Their support was great. Their prices are higher than the other wireless services, but you get what you pay for.
Before DEC, every support call was handled efficiently and they tried to keep us happy. My sister told me it was a pleasure to talk to them., even when everything did not work out the way she hoped. The great support was one of the reasons to continue using the service.
She, I, and other friends have made several calls in the last few weeks. None of them were remotely close to making us happy; we were more upset after the call than before. Each time, the call was answered by a lady with an Indian accent. I do not KNOW that the calls are being routed to India; we all could have reached the same lady working in the US. I do know that instead of making us happy, she kept putting us on hold, then returned to tell us "The system does not allow for that." She told us information that contradicts the policies of the last 7 years. The big one was that switching between price levels within a plan requires contract extension. She told us that the levels listed on the web sites were not allowed to existing customers. She told us that the never-ending promotions for a given plan would be canceled if we switched price levels. Every tech before DEC told us contract extensions were only required when switching between the major plans, that they used the web site to look up the current price levels, and that promotions would only be lost if we switched major plans because the promotions are plan-specific.
She refused to tell me where she was located. She refused to tell me how long she had worked in this job. She refused to transfer me to a supervisor or anybody else.
She upset my sister enough that my sister will be changing carriers within weeks. (Thank you for number portability!)
ATTWS: Why did you replace your great support service with this customer-losing system? Are you trying to lose all your customers? Will you publicly declare you are changing it back really soon to keep us as customers?
If there was a law stating that they must identify their location, then I could at least know the support died because of offshoring. I will be in their American-staffed store tomorrow to see if the salespeople there can resolve our issues.
First you need a method for knowing how much throughput you need for your personal use. Then you need a method to find the total throughput you have used so far in the current billing period.
If the ISP will complain that you are exceeding their standards for throughput, they must give you a method for knowing how much throughput you have used in the current billing period so you can stay within their limits. If they will not provide this, they cannot ask you to monitor yourself.
The system:
1. Take total allowable throughput.
2. Subtract your average throughput and maybe a little extra for emergencies or for known extraordinary events, such as the imminent release of software you will want.
3. When using BitTorrent for your own use, leave it running until you get what you need. Keep it on a little longer if you know that the current files are in high demand, such as one extra day after the release of a new core.
4. Calculate how much extra throughput you have available without passing the limits.
5. Starting on your birth date (day of month), turn on your filesharing program for others to use. When the throughput reaches the number from step 4, turn it off.
6. Near the end of the month, turn it on to use any leftover throughput.
If everybody starts running it on their birth date, there should be a good distribution of people running the software for the first 28 days. The "use up what is left" will assist with the lower number of birthdays at the end of each month.
The only hole in this system is the first few days of the month, and that may be covered by people forgetting to turn the software off at the end of month. If you (accidentally) leave it on for the first 2 days of the month, then you should just have less throughput available at the end of the month.
I am assuming that your allowed throughput is enough to keep the software running for at least 3 days during your "assist others" phase. Even is you break the limit, it should not be by much, and you can tell your ISP you are trying to meet their standards. Hopefully there are enough people who do not have to worry about throughput issues to cover the low availability periods.
This system allows you to meet your needs while being a good filesharer without getting in trouble with your ISP. Maybe BitTorrent could include these recommendations so enough users would follow them to make them popular.
I found a program called "Real Alternative" that plays Real video files on MSWindows. I think I first found it from another Slashdot post. It works well without any of the bad stuff of the official players from Real.
.rpm) .ram .rmvb .rpx .smi .smil) .smi and .smil files only play the first part of a clip. This is a limitation of the current Media Player Classic.
From the installer notes:
---
More information and updates can be found on the following websites:
http://www.freecodecs.com
Now a "Coming Soon. But domains from us" page
http://mirror.edskes.com
Redirects to http://home.hccnet.nl/h.edskes/mirror.htm that has downloads available
Real Alternative will allow you to play RealMedia files. This way you can play RealMedia files without having to install RealPlayer/RealOne Player. You do need a player that is capable of playing RealMedia. The included Media Player Classic supports it and works very well.
Supported:
- RealAudio (.ra
- RealMedia (.rm
- RealText (.rt)
- ReadPix (.rp)
- RealMedia embedded in webpages
---
The player says it is GPL by "Gabest". He has programs at http://sourceforge.net/projects/guliverkli/. He wrote the "Media Player Classic" that RealAlt extends.
Your post confused me.
I tried to RTFify a medium-sized Word document that was less that 5 MB. The resultant rtf was over 200 MB.
MSWord bloats documents. I have seen 100KB text documents become 200KB RTF documents when adding formatting, then become 2MB documents when saved as a MSWord DOC file. The same text must exist in all three. The formatting adds some markup. MSWord adds tons of extra overhead.
I do not use MSWord enough to know if there are options for compression. Even if MSWord uses compression, what compression routine frees 97.5% of the size? I have seen databases compress 80%; I have never seen compression better than 95%.
I am not stating you are wrong, just asking how those numbers are possible. I look forward to your informative answer.
---
As far as using templates to create documents, nobody is suggesting you stop using them, just that you save/send the document in a more ubiquitous format than MSWord's DOC.
I work with a company that has IBM as its largest customer. Before IBM killed SmartSuite, the people who communicated with IBM were forced to use SmartSuite. Since IBM declared SmartSuite dead, using MSOffice has become acceptable, and sometimes required.
I would not call 123 or Wordpro "crappy programs". They both were easier to use than MSOffice, and had many features that MSOffice does not have or does not do well.
Those people still use Lotus Notes. This is a good thing because we are able to use Notes for other applications that improve the business but would not exist if the cost of development was not so low because Notes is already available. We are planning our upgrade to Notes6 because IBM is finishing their upgrade and we need to stay current with IBM.
I have been recommending they start using Linux for the past year. Their IT director only knows MS, and gets very upset at the thought that anything else is valid. He also said that he wanted to replace the Wintel-based IBM and Compaq servers because they were incompatible with his Wintel-based Dell servers; I have little respect for his opinion, but I also have little influence. I really hope that this move by IBM will force my client to consider Linux desktops.
Since early last year, I have predicted the fall of MS for the end of 2004. For that to happen, I predicted:
1. A major IT company will pick a Linux desktop environment and start investing until it far surpasses the others, making the other desktop environments into novelty programs.
2. A major non-IT company will announce the switch to Linux by March 2004, thus insuring that MS will crash by the end of 2004
3. On a different note, I feel IBM is attempting to kill Lotus Notes as part of a misguided attack on Sun.
This article states that IBM is fully moving to Linux desktops. One of the links at the bottom was that IBM will stop using Notes internally, although from personal knowledge I know IBM is still migrating to Notes6.
My current relevant thoughts are:
1. IBM using desktop Linux means they will need it to be fully functional for their needs. They have the resources to make it so. The article does not mention which desktop was chosen. Another post suggests IBM uses Gnome, but does anyone know the official policy? IBM was my first choice for the decision-maker. Once they make the decision, their chosen desktop will become the desktop of choice for the world.
2. With a Fortune 100 company like IBM switching, we can expect a large non-IT company to follow soon. There are only 2 months to go before my deadline.
3. I believe that most business application belong on the Notes platform. I do not believe the Notes client needs to be part of the solution; the benefits of Notes' rapid application development are still realized when using a browser as the client. The Notes client has better handling of RichText, but I do not believe business data should be stored in RichText. The Notes client is fully integrated with email, and the Notes6.5 client is fully integrated with Sametime, the Notes IM client; losing these abilities will cause difficulties for IBM.
The big issue is that the Notes development client (Notes Designer) has yet to be ported to Linux. Without Notes Designer, there is no development. The Notes server (Domino) was ported to Linux by a rogue, and that has turned out well. Hopefully there is still enough flexibility in the IBM-controlled Notes platform team that the Notes clients, including Notes Designer, will be ported soon.
I have difficulty believing that IBM can finish the move to the Linux desktop without a Notes client for Linux. Their current culture is based on the ubiquitous and constant use of Sametime. Without the Notes client on Linux, their non-technical employees will need a replacement. Even if the Notes client is not ported, they must port a Sametime client to allow management and everybody else who moves to Linux to continue to communicate with those who have not been migrated.
I dream about IBM porting the Notes client to Linux. With the current situation, if IBM does not do the port soon, then IBM will falter as their ability to communicate internally diminishes.
Yes, they lose some of the ownership, but the effects should not be evident for a few years.
Yes, they could pay for all those reports while staying private, but there are reasons that every company that has to file those reports go public. The SEC frowns on doing that. I think there are fees/fines/penalties that apply in that situation that do not apply to a public company.
Yes, issuing stock is about the worst way to get money. Almost all of the alternatives are better (except cousin Tony.) You then contradicted yourself saying they were doing it for the money.
With its proposed share issue (1/3 of the company), the owners will all of a sudden just get $100million now in income (instead of the $300m they were getting before).
No, and your math is bad. If they sell 1/3 of the company, they still have 2/3rds so they get $200M of the $300M, not $100M.
The owners get ZERO "extra" money. It's not as if the billions raised by google will go to the existing owners.
Where do you think that extra money goes? If they just issued more stock, then the money goes into the bank and they can give themselves bonuses. If it was their stock they are selling, then the money goes to them. From what I have read, it seems the first scenario is the usual. Does anybody know why the current owners would reduce their share if they do not directly gain from it? I know some companies go public to be better able to expand, but that does not seem to apply here, although I am certain a few people are trying to decide what to do with $12B. How did the dotcoms "cash out" by going public if the money does not go to them?
If you or anyone in the household has any medical condition, or there are children in the household, the government will pay for basic landline service if you cannot afford it. I know this is true for Pennsylvania, but believe it is nation-wide.
---
I stopped my landline service a few years ago. I have used the cell phone to call 911 several times.
The worst was for a car running head-on into the end of a cement median at over 50mph. I called it in within seconds of it happening. There was tons of paper flying around on the highway, and many cars were swerving. I tried to slow and my car slid for a second, so I headed for the exit ramp at Valley Forge. I happened to look left and saw the car hit and the front-end crumple. The medians were not lined up properly due to the construction on Route 202, and there were no orange bumpers in the area. I had the cell phone dialed before I reached the end of the ramp. 911 said they already knew about the accident that happened 5 minutes ago, but I was able to convince them that this was a new report since it had happened not more than 20 seconds before they answered. I would have had to go through at least 2 traffic lights to reach a landline. If I had waited to find a landline, they would not have believed this was a new report.
If you can afford it, then a cell phone is better for more emergencies than a landline, but the landline is free for the needy.
Google management will have 6 months after the IPO to find ways of keeping anybody important to their operation that is about to become rich. There are SEC rules that employees are not allowed to sell their shares for 6 months after the IPO. Here is a very good explanation. It is possible that people who "know too much" may not be allowed to sell their shares for years.
The other side is that while most full-timers at Google have either shares or options, they probably did not amount to much compared to their salary. If those shares become worth ten times their value, and the employee decides to cash out, they will probably gain a few years salary. That might be wrong in this case. With a market cap of $36 billion, even a few shares may be enough to retire. Most companies plan at least 10% of their stock to cover employee options. $3.6 billion / 650 employees gives an average of $5.5 million. On the good side (for us), maybe most of those options are not vested yet.
The big winners are the ones who started the company or invested cash for shares. The investors should not matter to operations, and the founders have already made enough to retire if that was their preference.
---
Here is a link to the story that Google might be forced to IPO that I should have included in my last post. 500 share or option holders and $10 million in assets forces an IPO.
Here is a link to the actual rules. See "Corporate Reporting".
Google has remained private as long as possible. If their VCs were looking to cash out, they could have done it before the crash. And everyone has been asking them to IPO for the last 2 years to kickstart the stock market. It was smart of them to wait until the DJ was above 10,000, but probably unnecessary.
They probably are not worrying about buying somebody either.
The reason they are going public is because SEC rules force companies with a certain number of owners to go public. The companies have to file all the costly paperwork as if they were a public company, and they lose most of the advantages of staying private, such as not releasing all that information about their activities. There is little reason to stay private, and the extra cash from the IPO is handy for paying for all that paperwork.
The famous case of this happening was Microsoft. Too many employees were exchanging shares privately, and the SEC forced them to go public. They did really well, and you cannot blame their decline on being a public company since the prior management is still running things. OTOH, because MSFT is public, the shareholders can insist on new management, but they will probably wait until the stock goes under $10, and that will be too late to save the company, if it isn't already.
Google is being forced into going public. There is no need to look for extra motives from their investors and management.
The whole webpage is a "recommendation", so it is defined as an opinion. It is MY opinion, but still just an opinion. I am glad you enjoyed it.
--- Operating System ---
I was an avid gamer. (I have not had the time for a couple of years, since I am trying to launch several software products.) Windows98SE is still the best gaming platform. The PCs I build are for single-user, so the multi-user abilities of the NT line are more of an annoyance than a benefit.
A friend uses Win2K, and had problems with many games. We tried to troubleshoot them, but he eventually just gave them to me. We also had a few hardware issues. (He returned that stuff.)
I do not like WindowsXP. I do not like the interface, and even after changing everything back to "normal", it still performs differently in ways I do not like. (I also use several Linuxes, so just different would not bother me.) I also seem to have the ability to crash XP often. I know people who are using it, but I cannot recommend it.
My recommendations have not stopped several of my friends and family from running XP. It does mean they have to listen to me say "I told you so" every time they call me for software support, which is much more often than the Windows98SE users.
--- AMD vs. Intel ---
I did start that recommendation with "I do not understand".
The savings is less than $100. According to SharkeyExtreme's DEC 31 CPU Price Guide, the P4 2.6Ghz 800Mhz bus is $162. The "equivalent" Athlon XP 2600+ (2.08GHz) 333Mhz bus is $90. So you save $72.
- The P4-2.4 800bus is $153. Athlon XP 2400+ (2GHz) 200bus is $75, so the savings is $78, but the AMD bus is one fourth the Intel.
- You are running a slower chip and get less than half the bus speed. Does the higher latency more than halve the speed of the bus? Is the equivalency valid when comparing to the 800Mhz bus?
- AMD's fastest CPU is 2.2Ghz, but even the Athlon XP 3000+ (2.16GHz) costs more than the P4-2.6, so you cannot compare at the same CPU speed. According to AMD's chip speeds, every 80Mhz increase for AMD is equal to a 200Mhz increase for Intel. Can this be true?
- Apple has been selling better quality but slower speed CPUs for years, so the "equivalency" system has tradition. It feels like AMDs equivalency system is a marketing ploy from when AMD had the faster bus and could prove the numbers, but it is still used because AMD fell behind in the 32-bit CPU technology and needs to appear better.
- $72 is not much when the system still costs over $1000. If buying the Intel makes the system usable for a few more months, and gives better performance during its life, than it is worth it. My whole recommendation is about spending the money where it makes a difference.
If using an AMD gives a true perfomance increase, then please give me details. If it is to save a few bucks at the cost of any perfomance, then I would prefer to stay with Intel.
[Yes, I posted this on the Forio site too.]
This article confuses terms. (BTW, I use the term "good" to mean anything that can be bought, usually referred to as "goods and services". Programming is a "good" when you only care about the final product.)
Outsourcing: Paying another company (to do a task that could be done internally.)
Offshoring: Hiring people from another country (to do a task that could be done by people in the same country as the company.)
You can combine the two by hiring a company in another country to do a task, but it is important to distinguish between them.
Offshoring takes advantages of differences in the markets in different countries for certain goods. If the logistics allow the product to be transported for less than the difference in cost of the goods, then it is economical to buy there and sell here. Traders have been doing this since before the invention of money. With the extremely low cost of moving electronic goods today, production can be done anywhere in the world. (As far as the quality of the software development, I believe there are very few great programmers, so find them wherever you can. If you are hiring code-monkeys, it may not matter except that the cost of logistics must include language translation and extra management.)
Outsourcing has very limited scope of usefulness. It is useful to buy goods from other companies when the cost of doing it in-house is greater than available funds. There are also issues where accountability is necessary, even when the work could be done in-house (such as accounting audits. You need someone who does not have a stake in the company.) Third, there are a few issues where the required expertise cannot be bought, or the cost of entry is so high that everybody buys the goods from others.
The article's example is wrong. Jean-Marc is buying boxes because he needs too few to make it worth the cost of building his own boxes. When he sells enough chocolate that the cost of buying boxes is more than the savings from owning his own box making factory, he will enter the box-making business. This is called vertical integration, and is one of the prime savings from economies of scale.
I have a (currently) small business. I "outsource" my legal needs because I do not need enough to hire a lawyer full-time. IBM requires more lawyers, and so has its own legal team. It does this because it is more economical to hire them than to pay the overhead of maintaining a separate business. Every customer is paying the overhead of any business from which they buy. I am willing to pay that overhead because it is less than the cost of maintaining my own legal team.
The third reason to "outsource" is evident with Operating Systems. The cost of writing your own, and making it compatible with existing systems, is much too high for any company. So everybody "buys" Operating Systems from the few "companies" that have built their own. I use quotes because companies could collaborate on services they need that they cannot afford alone. Linux is an example of companies investing in a good that is needed by many of them. Collaborative lobbying organizations also fit this role, although they also fit the (reverse) accountability reason (such as "Microsoft did not say it; it was the SBA!")
The confusion between "outsourcing" and "offshoring" is designed to make offshoring seem like an extension of the recent (and awful) trend for outsourcing. "You were already considering outsourcing. Offshoring will save you even more!" This will probably do more to kill offshoring than anything else. Outsourcing is always bad when you have the ability to avoid it; every company that outsources a major business function regrets it within a decade. Offshoring is fine if it is economical for your situation, but hire the people; do not add the overhead of maintaining another company.
---
I am a consultant. I am paid to provide expertise that cannot be gained from anybody else. I am
The fact that big business will start pushing money into toolkits doesn't mean that control will be lost.
Control would not (and cannot) be lost, but big business can push the priorities that are important to them.
ROI does not apply to big businesses as a group. IBM's investment can't be recouped by Sun, after all, so there's no reason to think that they shouldn't back different horses.
With OSS, IBM's investment can be recouped by Sun, IF SUN CHOOSES THE SAME PROJECT. If IBM puts a few million into Gnome, then Sun only benefits if they also choose Gnome. If Sun chooses KDE, they probably have to repeat the work that IBM has done. But Sun could choose Gnome, receive the benefits of IBM's work, and then continue with what is important to Sun. That is why I feel the first influential company to choose will have a great impact on all succeeding choices.
more importantly, control in an OS/FS project is going to tend to stick with the fork that produces the best code. Doesn't your reasoning mean that you think that big business, with its big bucks to spend, will be churning out more and better code, than OS coders can, and hasn't that long since been shown to be untrue?
The community will still decide the best code. But wouldn't the ability to spend a few million tend to increase the possibility of writing some decent code?
Why are you distinguishing between "big business" and "OS coders"? The whole SCO/IBM lawsuit is because some IBMers are OS coders. If a company is willing to pay (or hire) programmers to work on OSS, then those coders must become part of the OSS community to be effective. That their priorities are set by the company should not affect their coding abilities.
(Well, there is the part about scratching your own itch, but a paycheck is rather good incentive to scratch someone else's itch. And you would still want to write great code for the reputation/karma.)
Try not to disparage this too much. I am certain that many Slashdotters have been dreaming of the day when big business decides that OSS programming is important, and have their OSS resumes ready.
The posts so far are about Dell's business model, but the article is asking techies to ignore Dell and build their own PCs, then build a business and negotiate for better prices. Why isn't anybody discussing building PCs?
I build PCs for myself, my friends, and my family. Most of them have custom "Solprovider" machines. I pick the parts; I build it; I install the software; I support them. PCs I build remain usable for over 5 years, and I average about 1 support call per machine per year. (If you are interested in what I build, see my October recommendations from the last time I built a PC.)
I am not attempting to turn this into a business. I have a very successful career, and the effort is too high and the margin is too low for this business to be worth my time. I refuse any money, although I expect a home-cooked dinner for my trouble, but then I only build PCs for people I care about. The other side is that these people know that I will never interrupt my paying work to help them, so sometimes their problems can take over a month to solve.
But why aren't you trying to sell PCs? You are already technical. You probably understand what the hardware does. You can learn how to use a screwdriver. You already know how to install software. The bad side is that you might have to install MSWindows if the buyer insists, but you could install grub for dual-booting, so every time the buyers watch it boot there is a chance they may choose Linux.
(Use the Maxtor 160GB drive. Use only 10GB for Linux and they will not complain. Tell them it is there so you can troubleshoot easier, but they can try it if they want. I am actually installing a new hard drive this way this week.)
I usually spend several days to research my recommendations. One day is spent researching the new technologies; one day is spent researching the various products; and a few hours are spent checking prices. Someone who builds more than one PC every 6 months would spend much less time per PC because their knowledge would remain current.
When MS crashes, big business will pour tons of money into the Linux Desktop PC. They are going to pick only one desktop toolkit because it gives a better ROI than splitting the money to 2 or more projects. It is likely that the other desktops will survive for quite some time (reference people still using Amigas), but the others will fall behind unless they use software like the project in this article to borrow from the leader.
if you insist on unification, you destroy most of the features of Open Source / Free Software development that are meant to be its strengths.
I know there is still strong development of the BSDs, but most of the corporate money for OSes is going into Linux. Is it "destroying the features of OSS/FS?" (Be careful how you answer. This is Slashdot and we do not tolerate criticism of the Linux development model.) The BSDs will survive just like the alternate desktops, but eventually the leader will far surpass them both in features and marketshare.
The key to the desktops is the toolkits. Eventually one will be chosen. We (all techies) will have little input into which is chosen. The first large software company that makes the choice and puts the money into one toolkit will make the choice for the entire world. It will probably be IBM, although Novell or SUN could take the role; Redhat has already given up. The final choice may be decided by one business manager that has been using Desktop Linux for some time and does not want to change.
Maybe Bruce Perens can influence the decision. As much as he likes KDE, he thinks Gnome is the better choice for a universal desktop (reference Slashdot article from last month.)
Yes, what you will get is a lot of people who are dissatisfied. Nobody says you cannot keep using Slackware 3 at home. But in a few years, all computer software jobs will require knowledge of the primary Linux desktop. Plan accordingly.
(This is a response to an AC that questioned my weight. I am assuming it was not a troll, but actually wanted information.)
I am 6'4" tall and weigh 235 lb. According to this, I am slightly overweight, but not obese.
I do not understand the judgement. I could lose a little from my belly, but the rest of me is muscular. I exercised constantly about 1.5 years ago. My belly was trim and started to show a six-pack. I initially lost weight to about 225 lb, then gained it back in muscles to reach 245 lb.
---
If you were questioning my ability to drink over a gallon of Pepsi or water each day, this site suggests drinking 1/2 oz of water daily per 1 lb of body weight. So at my current weight I should drink almost a gallon of water every day. But my intake has always been excessive.
Actually these numbers are hard to apply because my weight changed over the course of the caffeine addiction. When I started drinking Pepsi full-time, I weighed 175 lb. When I started the support job 8 years later, I weighed 185 lb. 8 months of being chained to a desk and I weighed over 260 lb (which was also a major factor in the acid reflux problems. Gaining 70 lb in <8 months is not good.) When I started the process of kicking the caffeine addiction 4 months later, I was around 250 lb. I stabilized around 240 lb a few years later. The major exercise program June-September 2002 is mentioned above, and as I lost muscle I have stabilized at 235 lb. I exercised some this Autumn, and am trying to make it a habit so I can lose a few inches off my belly.
In related news, I eat more than any other 2 people I know, and have since I was very young. I can easily eat a 30oz steak and be hungry a few hours later. My friends joke that I need to be rich just to supply myself with food. A few people have mentioned the possibility of diabetes. I keep getting checked; the doctors keep telling me that my eating habits are awful, but the only number that concerns them is my high cholesterol count (very likely related to my constant diet of cow.) I have changed my diet several times for several months; there is absolutely no correlation between the quantities I eat and my body weight. As mentioned earlier, I gained 10 lb over 8 years with a diet that varied (depending on my budget) between mostly pasta and mostly steak. In trying to lose weight, I did cut back for several months, but the only effect was fatigue. The only thing that affects my weight is the amount of exercise. (I usually lose weight when I have a girlfriend because I get more "exercise", but the girlfriends always gain weight from imitating my eating habits.)
76 percent of active Web surfers access the Internet using a non-browser based Internet application. Media players, instant messengers and file sharing applications are the most popular Internet applications ... The top five applications are Windows Media Player, AOL Instant Messenger, Yahoo! Messenger, MSN Messenger Service and Real Player.
As best I can figure, they are counting unique IP Addresses. Then they counted how many IP Addresses used each "application". 24% of the addresses did not use any of these applications.
I would guess that email does not show since most email clients talk to a server on the local network. The number of IP Addresses of the email servers that cross the internet are not significant enough to show on this report.
They do not distinguish if the connections for Windows Media Player and Real Player were for plug-ins or for stand-alone usage. Both types of media are often embedded in web pages. Windows Media Player is constantly attempting to call home because it is confused about a codec. Real Player calls home for new ad content. I am surprised that Acrobat Reader did not make the list, probably because it uses standard HTTP for communication.
They also did not mention the total population using IM: 20% AOL, 19% MSN, 12% Yahoo could mean only 20% of internet users use IM if every Yahoo IMer also uses MSN, and every MSN IMer also uses AOL. I know that is unlikely, but the statistics are useless for learning the total IM population.
The same reasoning applies to the Media Players: 34% WindowsMediaPlayer, 20% RealPlayer. What is the intersection? It is very likely that 34% of internet users like video; they all use WindoiwsMediaPlayer, and sometimes use RealPlayer because WindowsMediaPlayer cannot play the Real format.
The only valid statistic is that 24% of active IP Addresses did not use Media players, instant messengers and file sharing applications that do not use HTTP. And even that number is suspect because they have not stated that they excluded email, web, and web-crawling robot servers that communicate over the internet. Web servers may rarely initiate connections, but robots could account for a significant portion of that 24%.
---
Using the number of unique IP Addresses is a very poor statistics. They did not mention how they captured the number of IP Addresses. We are behind a NAT, so we have 3 people using 1 IP Address. Most households with broadband use this setup. At companies using proxy servers, there could be thousands of people using 1 IP Address.
I agree that time used by each protocol could be a more useful statistic, but it would be difficult to capture. For example: I could request 10 web pages in one minute, or request the same web pages at one every 2 minutes. It still took me 20 minutes to read the pages. The first method would return time = 1 min; the second returns time = 18 min. A slightly better statistic would be number of requests, but that would be disrupted by web pages that have many graphics.
The only statistic that make sense is total bandwidth used per protocol. And that would be useless for learning the number of people using each protocol, which is the only number of interest outside the infrastructure community.
Thanks. Diabetes runs in my family. I get checked often, and have had a full physical since this incident. There are no indications that I am developing diabetes.
My doctors and family really approve of me using water as my primary drink because of the high probability that I will develop diabetes sometime in my life. Except for this Summer, water has been my primary drink for 5 years.
---
Changing from high sugar drinks to water is very difficult. Sugar controls so much of our somatic responses that many Americans feel hungry when they are dehydrated. The body knows that it had a drink because it tastes the sugar and the blood sugar-level rises. Changing my internal programming so that tasteless water would be recognized as valid input was difficult.
For the water overdose incident, I had not eaten recently. I was programming the computer with a Britta pitcher and a glass next to me. I refilled the pitcher at least 10 times. The drinking, refilling, and bathroom runs were on auto-pilot because my brain was busy with the computer programming. I finally noticed my throat hurt, but initially assumed that it was caused by acid reflux. It was another hour and 3 more bathroom runs before I realized that it was getting worse every time I had a drink. Then another hour passed before I stopped programming the computer and found some food and high sugar drinks to try to satisfy my body. It is amazing how much we can focus on computer programming to the exclusion of our corporeal needs.
Barq's root beer has caffeine. So if a restaurant has Coca-Cola products, the root beer is caffeinated.
i on of premenstrual symptoms
The diet version does not have caffeine, but few restaurants have diet root beer, and I will never drink any beverage with "diet" or "lite" in the name.
---
Found this page about breaking caffeine addiction. It includes this list of withdrawal symptoms (as well as another product list that shows Barq's as having caffeine):
Headaches
Irritability
Intensificat
Fatigue
Generalized muscular tension
Nausea
Lack of appetite
Constipation
Lack of concentration
Disorientation
Forgetfulness
It sounds like caffeine withdrawal is PMS. Does "Generalized muscular tension" mean having cramps? Men, want to learn what PMS feels like? Just go through caffeine withdrawal!
---
Yes, I also notice when I have had much sugar during times when water is my main drink. I was high the first week after discovering Tropical Sprite until my body became accustomed to the higher sugar level. And I felt the energy loss when I switched back to water. Again, my body returned to normal in about a week.
I actually switched back because the supermarkets are always sold out of Tropical Sprite. I checked 2 supermarkets twice a week for 3 weeks, and even asked a store manager to stock more. Coca-Cola lost me as a customer because they did not stock enough. From working in supermarkets, I know the soft drinks are stocked by the vendors. I think Tropical Sprite would be a great success if they stocked enough to meet demand.
Summary: I kicked the caffeine addiction. The lack of caffeine had me sleeping incredible amounts of time during the withdrawal. Tylenol kept me functional for the months of headaches. I substituted other drinks with varying degrees of success until water became my main drink.
---
I was a major caffeine addict. I sometimes drank a few pots of coffee at night, but my major addiction was to Pepsi. I drank at least a gallon of Pepsi each day for 8 years, and sometimes finished 4 2-liter bottles in one day. Sometimes I drank Jolt instead, but still in large quantities. I usually slept about 2 hours per day, with a crash for 12 hours of sleep once a fortnight.
Pepsi destroyed my stomach. I did not have an ulcer, but I did have constant acid reflux. This may also have become a problem because before the support job, I rarely had to speak much. As phone support, I was talking for 10 hours per day. The new pattern of my mouth constantly moving may have caused my stomach to believe that food was coming, and I rarely ate while doing support. (Anybody doing phone support should have some kind of snack to keep the stomach happy. If you are over-weight, make the snack something like Pirate's Booty that has no value except to give your stomach something to work on.)
I was transferring from support to administration. My first day as an administrator would not be for 2 months, but I definitely had the job, so I was not worried about my performance in support. It seemed the perfect opportunity to kick the addiction.
I first switched from Pepsi to Mountain Dew, thinking that the lemon-lime drinks had less caffeine. (You can laugh now.) After reading the label and realizing my mistake, I switched to Gatorade. No caffeine, but tons of sugar to match the Pepsi.
I slept 10 hours every day during withdrawal, and woke up still tired. I had headaches for the first time in my life. I learned the joys of Tylenol, which was necessary so I could think while learning my new job.
Withdrawal lasted almost 3 months. After the first 2 months, the sleep I required started to reduce until by 6 months I was sleeping 4 hours a day, which is what I needed before the addiction.
After about a year, I switched from Gatorade to Sprite. That lasted another year, then I switched to water. I actually overdosed on water. The lack of sugar meant that my tastebuds did not recognize that I was drinking, so I was constantly thirsty. After pouring about 4 gallons of water through my system in 5 hours, my throat was stripped (and I was sick of running to the bathroom every 20 minutes.) I had to alternate Sprite and water for a week. Then I managed to stay with water with an occasional Sprite until this April, when Tropical Sprite (sold under the silly name "Sprite Remix") was released. I really like it, and it became my primary non-alcoholic drink for the Summer, after which I switched back to water.
---
After-effects:
If I have any caffeine, I feel it immediately. About half an hour afterwards, I crash; it becomes almost impossible to stay awake. I will not drive a car for the hour after I have caffeine. (This happens because many restaurants have awful-tasting water and do not serve alcohol or Sprite, so I try the root beer. Waitresses usually insist it is not caffeinated, but they are often wrong.)
---
The parent post's advice seems good. I wish I had read it before my attempt. One week was not enough for my withdrawal, but YMMV.
Isn't DRM already available in the 2.4 series?
I do not know. Does the Linux2.4 DRM include the BIOS locking ability? Was there any BIOS with that capability when 2.4 was released?
A Linus quote:
I want to make it clear that DRM is perfectly ok with Linux! He does not like DRM, but feels it may have positive uses.
But could DRM in Linux ever work?
"Making DRM in Linux secure would be like winning a hand of poker against someone who can change all the playing cards at will," wrote [Tony] Mantler. DRM would require proprietary binary modules. How long until someone in the free world (somewhere without the DMCA) released a module that returned "passed" for every function in the verification API?
These quotes are from April 2003, so the kernel developers may still be debating it. I doubt any implementation available today has all the abilities we fear.
You are contrasting the "number of internet users using Google" to the "quantity of money that Microsoft received in a specific period." There should be no correlation. Those Windows95 and WindowsXP surfers should have needed to pay only one time for their OS. MS should not make any more money from them even if they continued using the software for decades.
For the first time, MS has made less money in a 3-month period than the previous 3-month period. Their growth has stalled, and the decline is starting. Because of their fun financial practices, the fall will be fast. Just wave goodbye as they go past.
Unless you are dependent on MS software for part of your income. Then you need to examine your budget for things you can cut, or look to replace that income by learning a new career.
---
Google, which is the geek's best friend, which would have naturally higher numbers than many other sites.
Google is great for geeks, but geeks are still considerable outnumbered by non-geeks, and many non-geeks like Google too. Your proof sounds like "I found some shells on the beach, therefore the beach must be made of shells." Please read about writing proofs. (Your local college should have a course in algebra or symbolic logic that will cover the important concepts.) If you want statistics on what geeks use, get your statistics from a geek-only site like Slashdot.
In the near future we will have two kinds of platforms. One platform will be a fully integrated appliance that runs Windows in DRM-nightmare mode with BIOS lockin.
No.
The point of the article is that MS has announced they are dying. Bill knows there is no path for MS to survive in any recognizable form, and has given up. He already gave himself a billion dollars in dividends, so he will not starve. I expect their stock to be halved by this time next year.
Without MS leading the way into DRM-nightmare mode, I doubt Apple will even give it lip service.
Do you want to suggest to Linus that DRM would be a great addition for Linux 2.7?
---
Welcome to the new world of computers.
- MS is dying. Soon it will be dead.
- Every major company with any MS products will at least be researching how to get rid of them by the end of next year.
- More money and developer time will be spent on the Linux desktop in the next 2 years than has been spent so far. The windowmanager that gets the early money will become the only windowmanager.
- Computer technologists who were smart enough to avoid MS technologies will reap major pay increases as companies fight for these human resources.
- Computer technologists who depended on MS technology will flip burgers. Some may attempt to learn non-MS technologies, but few will understand how operating systems and programs are supposed to work after years of MS indoctrination.
Hiring a code maintainer is exatly what a company does not want to do.
Every company should have someone who is responsible for tracking what software is used, who is using it, and what releases are used. They should check at least monthly to see if there are any updates, what the updates do, and whether the updates are critical or desirable for the company. This person is usually an administrator with no programming skills. This applies whether the software is all proprietary, all OSS, or mized. It is REQUIRED if proprietary software is involved to track the per-person and per-server and per-CPU licenses.
Maintainers are very expensive, and you can't hire them in pieces.
I cannot be the only consultant who works hourly. Companies hire me when they need me. I go away when my task is done. They know I am available if they need more assistance. Does that fit your definition of "hiring in pieces?
The closest you can come to this is to hiring a company which specializes in maintaining the package you wanted.
Does it matter if it is a company or an individual? 1099s are 1099s regardless of how many people do the work. Hiring a company does mean you can have a better SLA than an individual can deliver. I had 4 companies want me in their offices for the same day in May; I was able to keep them all happy, but that is because my relationships are based on my controlling the schedule while making them happy, and they all know and accept that. I am hired through consulting companies, so the customers could ask for someone else from the consulting companies if there was anyone else who could handle the tasks.
Why do they need to specialize in THIS package? I am often shown the systems where some critical application is running and asked to fix something. I have to figure out where the files are, what language they are in, where the issue is, write the fix, and attempt to test it without destroying the live system. (The last time this happened, they pointed me to the wrong server!) A good programmer can work on anything. If you know the programming language, then you can narrow your search to people/companies that specialize in the language, but why limit yourself to somebody that knows THIS product? Maybe it was on the resume because they did some little fix for another customer. Check the list of active contributors to the package and see if one of them is willing to accept money from you in return for guaranteeing that your needs are filled. If not, then either find someone willing to become a (paid) contributor, or just find someone who is really good at programming.
In that case, you've lost the price advantage you had, and you've not gained the source.
It does not stop being OSS just because you hire someone else to look at it. I believe support for most OSS is equal or less expensive than equivalent support for proprietary software. It may have something to do with people being able to read the code. Or it might be that people who work with OSS tend to like software much more than people who just do it as a job.
You don't have people on site who can actually read the source,
Depends on your contract. Did you ask for someone onsite 24x7? Do you really need that? If you are hiring a company for support, then get a 4-hour SLA for critical issues. If the system is mission-critical, it should have enough redundancy that there is never a critical issue.
and if your service provider goes under, you're left without an escape.
Do you understand OSS? If your service provider goes under, you choose another service provider.
What's the advantage of this over buying from Microsoft?
When Microsoft goes under, there will be no updates. No bug fixes. No security fixes. With OSS, there is always somebody who can fix it. Your worst case is that you hire the best programmer you can find and give him much money to work on the software. But your company does not fall apart because you cannot fix/patch your productivity apps and web servers and every other piece of software you depend on.