Evil Spammers have found a new way to be The Spawn Of Satan. They are spamming blogger web site referral logs. You can find a blog discussion of it here at Blogroots
Go read the rest of this article in Wired for more on what new kind of evil The Spawn Of Satan have thought up:
Referral logs, intended to collect information on who visited a website and how they happened to arrive there, are being stuffed with bogus links. Curious bloggers who click on a logged link to see who visited their site are instead led to pornography or advertising sites.
Some bloggers publish a list that automatically updates links to sites that have linked to them. So visitors to spammed blogs who explore the link lists also find those sites full of porn and sales pitches.
In most cases the link spam appears to have been added to logs by one of several companies that are selling a service they describe as "referral marketing."
This page is a list of domains that are the known sources (at least so far) of the sites known to generate referral log spam. They do it by going around and reading pages many times for each web log and passing in a referral URL that is for porn or some other advertisement. You can tell this guy Philip Pearson if you see other domains showing up as the sources of web log spam and he will add new domains to his list.
In the year 2002 in a review comparing AOL 8 with MSN 8 we see what AOL brings out as a technological advance. I didn't realize that AOL customers had been suffering that much from using AOL:
Another long overdue fix addresses interrupted Internet connections. In the past, users whose connections were cut -- either because of an overloaded network or to make a telephone call -- lost the Web page they were visiting.
It was a big time waster. Users had to log back on and retrace their steps to their original location online.
Now, I'm pleased to see that Web pages remain onscreen after the telephone connection is cut. All I had to do was log back on before carrying on as if I had never left.
Tens of millions of AOL users have been putting up with this problem in previous versions. The mind boggles.
Mozilla is the open source browser development project that produces the source code that Netscape v6 and later are based upon. I personally use the Mozilla browser rather than the Netscape one. There are a couple of reasons for that choice:
- Netscape installs a bunch of annoying AOL stuff that I don't want to have popping up in my face.
- Mozilla is rev'ved more often and I like to stay up with the latest versions.
Mozilla has been my primary browser since about its v0.97 release. The latest release v1.2b is available here.
Given Sun's position selling their own OS and processor design in a market where higher volume lower cost competitors are coming up from the lower end market I wonder whether Sun can stay sufficiently financially viable to continue be the main developer of Java. Here's an article on their financial performance:
Merrill Lynch analyst Steven Milunovich said Sun may cut up to 8,000 jobs this quarter - Sun is due to announce its third-quarter results this Thursday. Milunovich joins Sanford C Bernstein analyst Toni Sacconaghi who is reported to have predicted cuts of between 4,000 to 8,000.
Anyone know whether Sun is laying off any of its Java developers?
Some people have published a Linux security kernel patch. But as The Register reports, they will not explain the vulnerability to you unless you certify that you are not an American:
So what's all this got to do with Red Hat? Well, non-qualifying people, we can't exactly tell you that. But when we asked Red Hat about it we got an official comment which at least partially explains it: "RHSA-2002-158 is an errata kernel which addresses certain security vulnerabilities. Quite simply, these vulnerabilities were discovered and documented by ppl outside of the US, and due to the Digital Millenium Copyright Act legislation in the US, it is potentially dangerous to disclose any information on security vulnerabilities, which may also be used in order to circumvent digital security - i.e. computer security. For this reason, RH cannot publish this security information, as it is not available from the community in the first instance. The www.thefreeworld.net site allows for accessing this information, but requires you agree to terms which protect the author and documenter of the patches from being accusations that they themselves have breached DMCA."
Check out this Red Hat page.
(thanks to Adam Flinton for the heads-up on this)
IBM would benefit greatly if they could make the PPC architecture a viable cost-effective alternative to the Intel x86 architecture at the lower end of the market. It seems to me that it will be very difficult for IBM to keep PPC competitive on the server high end if it doesn't have revenues from sales for the lower level machines to help fund processor research and development. This new chip has got to be aimed at the Macintosh and I would be very surprised if there was not already a secret business deal between IBM and Apple to use it for that purpose. However, given Macintosh's small market share it would help IBM even more if Linux developed a larger desktop and low-end server presence. In order to make that happen IBM needs to arrange for the manufacture of cheap motherboards for PPC to support this forthcoming chip. The question in my mind is whether this chip will be used to allow building of desktop machines as cheaply as can be done with x86 CPUs:
International Business Machines Corp. Monday announced a microchip for personal computers that will crunch data in chunks twice as big as the current standard and is expected by industry watchers to be used by Apple Computer Inc.
Another question: will this chip deliver competitive performance with less power use?
Linus does not see Microsoft declining any time soon. But he does expect Linux to play a role in eventually reducing the importance of Microsoft:
"I think, in the end, Linux will be a big part of why it happens but Microsoft is not going to shrink into nothingness in the foreseeable future," he said.
Instead it is more likely that Microsoft will become like IBM - "still huge but not the dominant force anymore", he said.
George Roettger of nthelp.com recommended that I try Crazy Browser. Well, I did and its not bad. It has the tabbed window support that Mozilla and Netcape 7.0 has but Crazy Browser uses the MS IE browser engine DLLs.
I've only been using it for an afternoon and so my impressions of Crazy represent a limited amount of use so far. Here's what I don't like about it:
- Right click on a link gives an "Open in New Window" that really opens in a new tab. One can't choose between opening in a new tab vs opening in a new window. With Mozilla that choice is available.
- Running on top of IE5.5 sp2 its threading does not feel as finely granular as Mozilla does. It gets into modes where for several seconds its not responsive while it does some operation. For instance in Mozilla if one page is loading slowly one can click quickly to switch to a different tab and read the page already loaded in the second tab while waiting for the first tab's new page to finish loading. Well, Crazy Browser will delay many seconds before making the switch. This poor ability to switch to view something else while an operation is happening is similar straight IE5.5. I haven't spent enough time yet with IE6 to say whether its ability to switch btw threads while doing tasks in one thread is any more finely grained.
- The "Close Current" button which closes the current tab is small and right next two buttons that close more than just the current tab. Its too easy to click the wrong button and lose a lot of tab windows. On the other hand its "Undo" functions will let you undo the close of tabbed window and even larger sets of undos of closed windows.
Its tabbed window behavior has some differences with Mozilla but nothing hard to adjust to. If you are currently using IE the Crazy Browser downoad is less than 700k and is worth a try. Its not going to replace Mozilla as my chief browser but I will use Crazy Browser whereever I currently use IE.
This has implications for how to stop viruses. Unless the immune machines are routers that can stop the viruses from spreading making more machines immune does not slow the rate at which viruses spread to the vulnerable machines
Until 1999, the standard way of modelling the Internet was to use randomly generated graphs, in which routers were represented by points and the links between them by lines. But it turns out that such random graphs are a poor approximation because they miss two important features. The first is that links in the net are “preferentially attached”: a router that has many links to it is likely to attract still more links; one that does not, will not. The second is that the Internet has more clusters of connected points than random graphs do. These two properties give the Internet a topology that is scale-free—in other words, small bits of it, when suitably magnified, resemble the whole.
In one of the services I use to look at web site access statistics for my 4 web logs I saw CPM. Huh? Does the CP/M operating system still exist? Or has some joker built his Linux or FreeBSD or other OS to report itself as CPM? Or is this joker doing this in his browser's configuration?
Clay Shirky thinks most online groups have a hard time making collective decisions:
These are really two related questions:
- Why is it hard to make decisions online?What is it about online forums that makes it hard for the users to make decisions? This is a question based on largely anecdotal evidence, but in my experience, only programmers have achieved any real success in participating in groups that meet mainly or solely online and still manage to get anything done. (The Open Source movement is particularly good at accomplishing things in distributed groups.)
The question is interesting to me partly because the mechanisms of decision making, whether democratic, consensus-based or some other form, are technically trivial to implement in online environments, and would seem to be a possibly useful tool for settling contentious issues or helping groups commit to a course of action. Yet groups online rarely pursue voting, or indeed any other formal and binding means of making a decision.
Even the Boy Scouts have formal procedures in their meetings. What is it about the online environment that leads to unstructured gatherings? Is asynchrony such a powerful force that it weakens the need for making decisions? Are people in online spaces so much more conscious of their individual identities and so much less conscious of group will that they never compromise? Is 'membership' in online spaces so tenuous that individuals simply refuse to enter into any sort of social contract that involves a commitment to act?
Or, taken from a working example, what characteristics let programmers actually design and build software in online groups? Do programmers have a different (and more achievement-oriented) set of expectations about collaboration? Are programming problems more easily partitioned among individuals, or more amenable to "Try both options and see what works?" types of decisions? Does the compiler act as a neutral referee, thus short-circuiting the tendency for interminable argument?
My guess is that not all distributed open source groups make decisions in an efficient manner. Its just that the ones that do it well are such visible successes. For example, in Linux's case Linus Torvalds has the final say and he commands the respect of his peers as making the right decisions enough of the time to be worth following. Also, programmers in a project like the Linux kernel are very bright and can follow complex discussions and most can come to understand the reason for a decision. When there is a greater intellectual disparity between the members of a group it is more likely that some members will not understand - or more likely will even misunderstand - the reasoning cited for any given decision.
While technical debates can be very heated it is often the case that as long as a given module ends up working many team members can afford to be neutral about what ultimate architectural decisions get made about it. There are also implementers who enjoy implementng who do not care as much exactly what the design is. It is easier to make decisions when there are few architects and were most of the non-architects do not have aspirations to be architects. Think of a large building being built. Most of the welders are indifferent to the floor layout choices. They have their jobs and they are mostly concerned only about their immediate work environment and count on the architects to make wise decisions.
Then there is also the nature of the sorts of tasks that are having decisions made about them. Try bringing a group to a consensus about some political issue where individual differences in knowledge and values are likely to be much greater than is the case with programming questions. There are just too many factors to juggle and attempts to reach online consensus for political decisions are an exercise in futility. Part of the reason its hard to come to mutually agreeable decisions on a matter of politics is that often times what is required is trading. Engineering trade-offs are not in the same league in terms of what gets traded off in politics. A fair amount of political trading has to get done non-verbally so that the various actors will leave no written record of their motives. That way each can portray to their supporters contradictory reasons for some compromise.
There is a company called Modex that makes Epidex which is a product for chronic skin ulcers. They take cells from a person's own hair follicles and use them to grow skin grafts for that person. Modex has set up a site dedicated to Epidex and the Epidex site provides some nice examples of things not to do in a web site design.
Check out the text on this page. Its a JPEG image. Since its an image you can't scale up the text to make it easier to read. But if you just happen to put your mouse over the small hazy text you will notice that you can click on the text image and get a larger copy of the same text table image that is readable. What were the designers of this site thinking? Why do such a thing? It would take less than an hour to type in the image's text into an HTML table. Why not do it that way?
But lets go back to the Epidex home page. Put your mouse over the Product Information box. See the pop-down list? Bet you thought you could move your mouse down to choose one of the items on the list, didn't you? Well, I did and I was wrong. Try that and the list then disappears. You have to click on the Product Information box to get another page that then lets you chose one of those options.
I do not hold myself out as an expert in web site design. But the Epidex site seems so obviously wrong that one doesn't need to be an expert to find it seriously wanting.
Most people think their current machines are fast enough and don't see any greate benefits from upgrading. While Intel touts some future processor hungry applications such as voice recognition and video editing as reasons people will want to upgrade I don't see it. Its going to take some other application that everyone wants that needs a lot of processor power in order to get people to upgrade. Intel expects half of its sales growth to be in emerging markets. So its clear they don't have that much faith in new application catergories at growth generations. Meanwhile, ARM is experiencing rapid sales growth in embedded processors. That's where the action is shifting to. This NY TImes article requires free registration to access.
That new attitude is shown clearly in a recent national opinion survey by Odyssey Ventures, a San Francisco market research firm. Among households with PC's, the intention to buy a new computer in the next six months has fallen to just 11 percent from 21 percent in early 2000 and the lowest level in five years. And half of PC owners now have home computers that are at least two years old — more than at any time since 1994, when Odyssey began keeping track. The pace of upgrades is crucial because, according to the Gartner market research organization, they account for 80 to 85 percent of new computer sales.
Here's an idea for a sales generator: Software that does voice and face and body recognition for home security. Put a box in the basement or closet that is the house server computer. Connect a bunch of cameras to it. It could dial out to a security agency to report a video feed if it saw something suspicious. Or it could call your cell phone and let you know what it sees. Maybe a small display device on the cell phone to show you what it saw would let you decide whether its fears were justified. If you weren't sure you could tell it to send you a series of pictures.