I have to buy a new PC for my parents. I want to reuse an existing 19″ screen, so I am only looking for a desktop. My dad has been using a PC for a couple of years and don’t feel like switching to Mac. So a Mac Mini is not an option. So I started looking for a PC that was as simple and beautiful as that, at the same or smaller price. I was in for a disappointment…
The baseline
First off, the original Mac Mini
sleek, square, white, with just a minimal CD slot and no buttons or logo on the frontpanel.
If evolution would build a PC, this is what it would look like. Now let’s take a look at what ‘intelligent design’ has to offer:
Continue reading ‘Who makes a pretty PC?’
If you're new here, you may want to subscribe to my RSS feed or receive updates via email. Thanks for visiting!
Published by Peter on January 24, 2006
in spam.
I was reading about a technique to discourage spammers: let an organised mob fill in thousands of fake submissions so that there is no way telling how to distinguish them from real responses. They targeted a known spammer, Alex Polyakov, currently #8 in Spamhaus top 10 and he did feel the pain.
During the 13-minute call, Polyakov claims that his “interest is only to make honest dollars.” As a peace offering, Polyakov proposes to create a global opt-out list, “the anti list of all anti lists.” Polyakov says he has no interest in sending spam to people who don’t want to receive it, and he guarantees that he will persuade all his spam-business associates to clean their mailing lists.
from Spamkings blog via digg.com
Let’s consider such a global opt-out list:
DISTRIBUTED OPT-OUT LIST
- let’s say it would be something like 1 million addresses (just a ballpark figure). All in lower case, with no funny characters.
- In order to make sure the list is not used as a spamming list itself (since these guys are not known for playing by the rules), it should be communicated not as email addresses, but as a list of hashes (e.g. MD5/SHA-1) of email addresses. (Which means you cannot get back the email addresses from the hash)
- SHA-1 is 160 bits or 20 bytes per address. MD5 is 128 bits or 16 bytes per address. MD5 is less secure but for this purpose, who cares (false positives are not a big issue).
- The size of the list would be 16 bytes x 1 million = 16MB, which is manageable for daily/weekly updates.
- One could accept domain wildcards (*@example.com) but since Hotmail, Yahoo, Gmail … would want to add a wildcard for their users, this would kill the spammers’ lists so no one would use it. Plus, some people might object to the fact that they are not kept up-to-date with the latest Ci@lis/Vi@gr@ prices.
- Let’s say a spammer would use a 100-million addresses target list. This means 100 million emails of something like 30 bytes on average (high estimate, I know). So he would need to calculate the MD5 for 100.000.000 x 30 bytes or 3GB. Looking at some MD5 throughput stats (20MB/s) this is a matter of minutes, not hours.
- Then the spammer has to remove all addresses that feature in the opt-out list. This can easily be done as a merge of 2 sorted lists. The overhead is negligible.
- If the opt-out list grows to 100 mio addresses, and the size to 1.6 GB, download is still done in less than 1 hour over ADSL.
- HOWEVER: dictionary attack! I am ruthless spammer and I just got a list of 1 million hashes? Mmm … I could create a dictionary of probable email addresses and see if they actually exist! An email consist of the letters [a-z], numbers [0-9] and the characters [-._] before the ‘@’ sign. So all combinations up to 10 chars are around 40^10 (gross simplification, I know) or 10^16, and if I filter out the incorrect ones (44444444444@) and use the billion most probable ones (e.g. “jill.jackson@” is more probable than “a77..-_-8@”), combined with the postfixes hotmail.com, yahoo.com, comcast.com, … I could probably find some addresses of notorious anti-spammers, send them loads of email and destroy the credibility of the opt-out list immediately.
EMAIL SERVICE PROVIDER
- someone that sends email on behalf of spammers, that always uses the opt-out list, and that because of this admirable behaviour gets treated more leniently by anti-spam software.
- Advantage: the opt-out list never has to be sent to spammers, and no mails go to the opt-outers.
- Disadvantage: ain’t never gonna happen. Spammers would have to pay for this service and they won’t, the service would have to be operated by a trusted 3rd party but who would want to do that?
SELF-REGULATION
The American Direct Marketing Association (DMA) has the e-Mail Preference Service (e-MPS), the Belgian Direct Marketing Association has the Robinson-list. As I recall from my Direct Marketing days, the Robinson list was always used to clean up addresses.
But getting the emailers in the DMA to use a global opt-out list will only help very little. They’re not the real problem. The real problem are the Russian/American vilains on the Spamhaus top 10.
Conclusion
I would have to agree with Spamhaus:
1. For-a-fee Address Remove Lists are operated by conmen.
2. No legitimate marketing firm sends Unsolicited Bulk Email in the first place.
3. Can you imagine spammers doing this?
4. All spammers believe their junk is different from the junk other spammers send.
from spamhaus.org
Technorati: spam - opt-out - spammer - spamhaus
#1: Feels great, but don’t want any pictures caught doing it.
#2: Nobody quite sure what it is, but everyone wants to try.
#3: First learned how to do it at [foo | bar | summer] camp.
#4: Lots of parties, alcohol, and women with big hair.
#5: Can fool most people if you can just do [ajax | the hustle].
#6: More about having fun than doing something useful.
#7: Open source, free love, & fashion from the 70’s.
#8: People are remixing it all the time.
#9: More popular it gets, more people trash it, more popular it gets.
and last but not least:
#10: Done best when you don’t give a damn what anyone else thinks.
from blog.simplyhired.com
You weren’t around in the seventies? No idea what the cultural significance of Saturday Night Fever is? Find out what you missed on Wikipedia or Jahsonic!
Technorati: disco
Headlines from 1981:
- Prince Charles got married
- Liverpool crowned soccer Champions of Europe
- Australia lost the Ashes tournament
- Pope died
Headlines from 2005:
- Prince Charles got married
- Liverpool crowned soccer Champions of Europe
- Australia lost the Ashes tournament
- Pope died
In the future, if Prince Charles decides to remarry, somebody should warn the pope.
(via haveadaydotcom)
Technorati: charles - pope - 2005
When Apple reinvented the photofeed, they actually were a bit sloppy. Instead of building upon standard RSS and the Media RSS extensions backed by Yahoo!, Feedburner et al., they decided to do what Microsoft has always been accused of: they made a different, non-compatible RSS format.
cf http://static2.podcatch.com/blogs/gems/snedit/rss.xml
It’s pretty bad. There are lots of errors, the date formats are wrong, there are elements that are not in RSS that aren’t in a namespace.
via scripting.wordpress.com (Dave Winer)
from static2.podcatch.com/blogs/gems/snedit/rss.xml
- First of all, it looks like they made a ‘wallpaper-cast’ instead of photocast. The RSS extensions are called www.apple.com/ilife/wallpapers.
- The RSS feeds are only accessible with a specific UserAgent, i.e. only with Apple Safari. Try to open it in any browser and you get an error message. (Update: actually, while I was writing this, the behaviour seems to have been changed to delivering the RSS with
Content-Type: application/octet-stream. So this is more or less fixed - application/rss+xml would have been better)
- The dates are not conform the RFC822 standard: “2006-01-11 16:43:22 -0800″ should be “Wed, 11 Jan 2006 16:43:22 -0800″. Most RSS parsers will have no problem with this, but if there’s an official RSS specification, why not follow it.
- They put the image URL in the
link field, which does not allow extra attributes like type or size. Why not use enclosure?
- For all the date related metadata (photoDate, cropdate), why not use Dublin Core dcterms?
Continue reading ‘Apple creates RSS the Microsoft way’
Published by Peter on January 13, 2006
in spam.
For some reason, if I search for “baeyens” on Technorati (sorry, John), all I get is a list of splogs (spam blogs). The first ‘real’ result is somewhere at #50, drowned between WEBCAM, CAMERA and PHONE CARD splogs.
They all have the same characteristics:
- all on Blogger’s blogspot.com
- post title is up to three spam words in upper case
- blog title is up to three spam words in lower case
- blog post contents is a sequence of words without any meaning (apparently ‘baeyens’ has become part of a standard splog dictionary)
- at the end of the blog post is an
iframe part
- the
iframe inserts code from www.webs-search.com in the page that also redirects the browser to e.g. http://www.webs-search.com/search.php?key=guns (if the blog topic was ‘guns’)
- that page is filled with ads that go through www.peakclick.com, an Austrian PPC site
What I mean is: Dave, you guys should be able to filter this scum out! And Matt, can’t you give the Blogger team a hand in attacking the splog problem from their side? We don’t want Technorati installing a if (domain ends in "blogspot.com") {/* treat as splog */ ... } rule, do we? Or do we?
webs-search.com
- domain is registered by a “Anrev, Kovacz contact@mwayc.com - 1003 Star Street - Novambark, na 88737363 - CA”
- the registration address for mwayc.com is “41 State Street - New York, NY 12345 - US”
- domain is hosted on an EV1 server: ev1s-67-15-104-73.ev1servers.net [67.15.104.73]
- the page title is ‘Licht und tonanlage’, which could mean that either the above Kovacz speaks German, or -more probable- that the site’s code was delivered by the Austrian PPC site.
Technorati: technorati - splog - blogspot - detection
Recent posts + comments in Blogger
One of the disadvantages of using Blogger is that by default, you don’t have categories and recent comments on your blog. There is a trick to put recent comments on this Blogger help page. It involves introducing a new
<Blogger> ... </Blogger>loop in the blog template. A nice trick, but I don’t like the fact that they only show the comment dates.So I developed the idea a bit further: on the main and archive pages of my blog, you now see the normal “Recent Posts” in the sidebar, but the posts that have comments also have those listed (see image at the right). On the individual post pages, I have the normal list of recent posts. I put each comment in a
divwith a height of 25 pixels and withoverflow: auto(defined in the CSS file) so that long comments don’t take too much space. The layout of the thing (e.g. each comment starts with a clickable “>”) is easy to adapt.This is the Blogger template code (in the standard templates, put this instead of the existing code that is much like the part here between the
<ItemPage> ... </ItemPage>tags). Feel free to use it if you want to.<ItemPage>
<h2 class="sidebar-title">Previous posts</h2>
<ul id="recently">
<BloggerPreviousItems>
<li><a href="<$BlogItemPermalinkURL$>" mce_href="<$BlogItemPermalinkURL$>"><$BlogPreviousItemTitle$></a></li>
</BloggerPreviousItems>
</ul>
</ItemPage>
<MainOrArchivePage>
<h2 class="sidebar-title">Previous posts</h2>
<ul>
<Blogger>
<li><a href="<$BlogItemPermalinkURL$>" mce_href="<$BlogItemPermalinkURL$>"><$BlogItemTitle$></a>
<BlogItemCommentsEnabled>
<BlogItemComments>
<div class="recent_comment">
<a href="<$BlogCommentPermalinkURL$>" mce_href="<$BlogCommentPermalinkURL$>">></a> <$BlogCommentBody$>
</div>
</BlogItemComments>
</BlogItemCommentsEnabled>
</li>
</Blogger>
</ul>
</MainOrArchivePage>
You will need a CSS class (in your CSS file, or in the first part of your Blogger template) with something like the following code:
.recent_comment { overflow: auto; border-bottom: 1px #999 dashed; font-size: .8em; height: 25px; }