Here are notes about topics I don't have time to fully research, and links I don't have time to fully explore, because I really need to spend more time implementing the portal to my new "cooperative" economy.

Special attention: CityofSunnyvale:BadHTML cutestkidever:FastLink GeorgeStephanopoulos:ForumBroken/Who?/BlogBroken PRI:BarterCensored N_Land_that_job:WebSiteBroken Jezza3110:LinkBroken Sheamus:WebSiteBroken Nierengarten:ManyComments PLPropertyClub:AlertCallback games2play4u:NonWorkingGameSite ScottKirsner:dlNlR/WebErrata/VC+Angels Natosha6193:Zooey politicorithmic:WSJpageBroken BigCarrots:prizes VirginMobileUS:CannotBlock Chris:VHS HDNR+jfredson:DAG AwOLonline:EllenMuthBadLink

ParseLynxEqualReport
[TWITTER DOWN 2009.Aug.06-07] [Draft Proposal for Twitter to meter usage to protect against DDoS attacks.]
[Priority event system] [projectIdeas]
[Description+explanation of terse notation presented (for cell-phones) by Portl1]
[Transcript of twitter.com login not working from VT100 term on Unix]
[Tribute to the late Les Paul]
[accessible-twitter down]
[ethnic-food page has no main content]
[Yahoo! Mail bad "html"]

Trying to find "Take Back the Land" (Max Rameau) on Twitter.
Did @advancnonprofit ever find Max Rameau on Twitter?
ow.ly/gZhJ indirects to Train game, saved locally as tmp-train.html:
- Really bad Web page as given, needs all the crap documented.
- Duplicate text needs to be found and flushed.
- But really good interview.
- Then I need to post the resultant clean version myself.
Trying to find
"Dawn Bushman" on Twitter
"Peter Sturdivant" on twitter
"John Lopata" on twitter undercounted: "JOHN LOPATA: I'm John Lopata. I'm an inventor with 38 United States patents. I've been out of work since January of 2008."
TinyURL.Com/BaiShel indirects to idea-BailoutMoneyLeaseForeclosedHomesForShelters but his blog posting form discriminates against low-income disabled people such as myself whose only from-home access to the net is via lynx.

New twitter links to organize: pbshawaii / joemiller1010 (PBS Sr Associate for Primetime Publicity and Awards)
Author (of bad Web page): 9brandon donaldknuth QuotdKllrQuotes Now_on_pbs / Picture the Homeless (pthny) on Twitter /
http://twitter.com/inthesetimesmag/status/2312366966 - hangs

Name to research:
Stina Katchadourian

Ponzi/Pyramid distinction: "Ponzi scam differs from pyramid selling in that individuals all tend to invest with the same person" - That link went bad, need to find new link to same article.
TinyURL.Com/PonPyr indirects to here: "Today the terms Ponzi scheme and pyramid scheme are often used to mean the same thing. However, there is a slight difference. A true Ponzi scheme usually promotes what appears to be a real investment opportunity which investors may contribute to without actually being an affiliate, distributor etc. A pyramid scheme, on the other hand, usually requires that participants make a payment for the right to recruit other people into the scheme, at which point they will receive money."
pyramids.htm

Newsgroups to research and post questions to:
comp.infosystems.www.authoring.html
.
usenetlinux.html contains list of Linux newsgroups, such as: comp.os.linux.misc
linux.dev.admin
linux.local.silicon-valley
linux.news.groups
linux.redhat.install

Additional ideas for NewEco not yet organized or developed;
- I should set up a DB table listing all the online barter systems and including consumer/critique feedback.

ImageMakers:
- Little Girls, Big Trouble - Big Girl - Samantha Weinstein is the gorgeous red-haired girl (role: nine-year-old Josephine) who deliberately lost the staring contest.
vote has complete list, but no descriptions or links, but then I just use Google with the exact episode title to find the info?
list of upcoming ImageMaker programs
Stand by Me - Wherever You Are - Lucia Jimenez, Enrique Alcides, Fernando Tielve, Maria Herze and Manuel de Blas. (Which are the little girl and 20-ish girl?)

Idea for blacklisting IP CIDRs because of spam:
For each nested CIDR containing the IP number under question, calculate the ratio between spam and good messages. Thus after one spam has been spent, any CIDR small enough not to contain any good IP number will be 100% bad, and from any good IP number any CIDR small enough not to include any spam will be 100% good. Larger CIDRs will then be between the extremes depending on how close the IP number clusters with spam vs. good sources. But how to balance the local (small CIDR) statistics against the global (up to /8 CIDR) statistics in deciding whether some new IP number is likely spam or likely clean? Just use the smallest CIDR that includes both the new IP number and at least one previously known and rated spam-or-good IP number, or go up from that to include a wider survey maybe 2 or 4 bits shorter (4 or 16 wider IP range)? If I have cached WHOIS info, then perhaps I should follow the hierarchy of officially managed and registered CIDRs, otherwise fallback to 0 2 or 4 extra bits above first known IP, or maybe use round numbers of bits i.e. instead of /23,21,19 I should use /20,16.

Strange quote, why do I have this?
i) data, data and data, but not your impressions. That is, no "narrative description" but instead an exact reproduction, by copy and paste with the mouse, of each and every datum that you are basing your ideas of what is going on on. Do not trust yourself to type! Use the mouse.

Mistakes on Web pages to use for Turing tests:
.
I mean, the government should never have expelled these [NOGs].
.
The government made clear while I was there that whatever they may have done in terms of expulsions, they do welcome NGOs on the ground, they do recognize the value they provide in [help] people.
.
KTVU: "riders generally need to arrive before 10 a.m. in the morning." (redundant)
.
If you're going to Sunday Streets today, stop [bye] and say hi at the KQED table!
.
monroescoop: "that is so sad it makes me cry ,i cant believe someone well i can is that sick enough to savagely beat there loved one almost to death whom labored there on children there own flesh and blood into the world a wife a mom whom was so innocent and defensles she loved him enough to have children by him and all she wanted was a decent husband to care for her ,i would make a decoy of my self and dig a foxhole and booby trap an address and when he showed"
(That has so many errors I don't have the energy to correct them all.)
.
If Israel acts [than] they can absorb the bombs
Remember what happened in [Ruasia] when Reagon used [strenght]
.
On 2009.Jun.12 KMTP says: "Next February the entire broadcast industry, including KMTP-TV, will complete a transition to a digital format."
.
KMTP: "Another Stories"

   srand(time());
   $random = (rand()%9);

   $random = (rand()%10);

Since the tide was rising, I didn't [bothered] to push it back into the water.
.
Well, luckily the rain didn't hit me today. Hopefully I'll be just as [luck] tomorrow.
.
php-quake: "If the function is called with the given parameter the default value is [overwrite] with the value of the function call otherwise the default [values] is used."
.
washingtoninstitute: "... affiliate groups that work along the same lines as the parent group but have a wider degree [or] organizational latitude."
.
IMDB: "Sisko has made up with Kasidy [wih] a bunch of flowers"
.
france24" "When the French mini-submarine Nautile reaches the seabed thousands of [kilometers] underwater, it will be looking for small, hi-tech devices that are not actually black, but bright orange."
.
TinyURL.Com/n9o8x6 indirects to futuresmag:
Yesterday's trading volumes [where] not higher than in the previous session
[I t] is probable that the profit taking will hold
fund [manager s] have to show something better than three months ago.
get exceeded in [t he] next 48 hours
from [wh ich] the last sell off started.
[U ntil] that [happens ,] there is a chance that instead of moving higher
[O bviously] all the time that the index holds this [week 's] lows,
the other main markets, [l ed] by the Nasdaq
.
BBC "Investor Miriam Siegman's voice wavered as she admitted that she scavenged in rubbish trucks now **** her money is gone."
.
squidoo: "It is very detrimental to your relationship with your dog if you allow [you] dog to not adhere to [you] rules. You must maintain the alpha status at all times until [you] pet understands the importance of obedience. It may be difficult in the beginning but in the long run, [you] dog training will give your dog exactly what they're looking for These [tactic] and other k9 behavior training information are easily accessible so you don't have to leave [you] dog training knowledge to a minimal."
.
scam.com:
I was totally deceived by the people working this program and [it's] administrators.
The problem is [with] that if you [dont] have someone under you [thats] not able to bring in that much people [youre] basically out of luck.
.
KQED:
he sees her walking [passed] the window
.
askwisecounsel
[Your] are currently browsing this site with Internet Explorer 6 (IE6).
(No, I'm browsing with Lynx.)
.
@RuthEdeD bad URL (just a blank screen, not even a title.)
.
freerangetalk
The German company Desertech is [look] into building a solar farm ...
.
airliners.net
"at the first sign of icing [irregardless] of the outside temperature"
"the engine control computers [wills witch] to an alternate mode"
.
google
(paint-text:) "icing of the EPR probe [icing] would not have occurred"

Google Chrome is really stupid: It shows a box saying Adobe Flash plugin needs to be installed. But when I click on the link to try to install it, it says Chrome must be closed before I can do that. But if I close Chrome, I lose *all* my tabs, all the other things I'm working on, and then I have no idea how to install it when I'm not already in Chrome.

-IP- Google search for: stock screener

Marconi's Intelligent Vending technology = a solution that allows a consumer to pay for a drink using a GSM cellular phone.

Web forum for "This Week" is broken, details below:
File that you are currently viewing: GeorgeStephanopoulos (GStephanopoulos) on Twitter
Link that you currently have selected: http://bit.ly/PMNIW
(Click on that link, get to the page below:)
File that you are currently viewing: `This Week' Sunday: Health Care Reform Debate - George's Bottom Line
(I read some of the comments, and then scroll down to the submit-comment form, fill it out, get to this link:)
Link that you currently have selected
 Linkname: Post Comment
  Method: POST
  Enctype: application/x-www-form-urlencoded
  Action: http://www.typepad.com/t/comments
(When I click on that link to submit my comment, the Web browser flashes a "404 Not found" error, then brings me here:)
 File that you are currently viewing   Linkname: TypePad   URL: http://www.typepad.com/t/comments   Post Data:    static=1&entry_id=6a00d8341c4df253ef01157229d67c970b&author=Robert+Maas&email=&    url=twitter.com%2FCalRobert&text=If+you+have+less+government%2C%0d%0ayou+have+more+private+corporation.%0d%0aThe+same+amount+of+corruption%2C%0d%0ajust+a+different+*kind*.%0d%0aBOTH+kinds+must+be+reduced%21%21&postcomment=Post+Comment
  Post Content Type: application/x-www-form-urlencoded where I see the following WebPage content:

TypePad
Not Found
   We're sorry, the page you were trying to load cannot be found. If you
   believe you've reached this page in error please check the address and
   try again.
   Copyright 2009, Six Apart Ltd. All rights reserved.

Which subsets of these people, all with the same name "George Stephanopoulos", are the same person, and how many groups are different people by that same name: Known by Go.Com or ABC News / Described on WikiPedia / Known by Charlie Rose or Google / People who wrote books using that name / Using that name on Twitter / Listed on InterNet Movie Database: 3484354 0826888 / Registered on LinkedIn: 8bb 153 83b 994 / Married to Alexandra Wentworth according to People.Com

Catherine Galloway (F24)

Link I received by direct message from PRI: bartering-in-tough-economy
I posted this comment there:
I'm setting up a "cooperative" i.e. barter of labor over InterNet.
See http://TinyURL.Com/Portl1 for link to portal to current NewEco framework.
I'm looking for others to help me brainstorm the design and description documents and maybe help write the PHP/MySQL scripts.
It uses text-only Turing tests, instead of GIF captchas, so it works fine with text-only browsers such as Lynx.
It responded:
Your comment has been saved and will be visible after moderator's approval.
As of Jul.14 01 my comment hasn't yet appeared.
As of Jul.25 23 my comment still hasn't appeared.
I think it lied about saving my comment, or the moderator is censoring my speech like a dictator.

2009.Jul.26 08, new and wonderful idea for how to use "truth futures" to get answers from busy people: Post a poll like this:
"If [busyPerson] were ever to answer this question, what his/her answer be? [theActualQuestion]"
Then allow people to submit what they think are correct answers to that poll-question, i.e. their belief as to what [busyPerson] would answer to embedded question.
Then solicit buy/sellShort bids (at various percentage truth values) on each poll answer.
In theory the people who best know how [busyPerson] would answer the question will be more willing to "put their money where their mouth is" and hence will drive the fair-market truth-value of each survey answer towards an accurate estimate of how [busyPerson] really would answer that question, and thus effectively judge which answers are correct and which are incorrect.
But then the only way such expert on the [busyPerson] could cash out his/her investment would be to convince [busyPerson] to formally answer the question in a way that can be verified. Such expert under such financial pressure would find a way to get [busyPerson] to answer [theActualQuestion], thus solving the original problem of the person being too busy to bother. Finally, the expert will need to find where the answer is located, and report that location back to my system so that my staff can make a judgement on the actual truth value of each poll answer, whereupon each poll-answer future can finally be closed so that investors will then be paid accordingly.

Your blog is totally broken:
I was reading GeorgeStephanopoulos (GStephanopoulos) on Twitter and saw this link: http://bit.ly/sQCKk which redirects to Krugman: Recession is Over (Sort Of) - George's Bottom Line where I wanted to respond so I filled out the form, then scrolled to:

Link that you currently have selected
   Linkname: Post Comment
     Method: POST
    Enctype: application/x-www-form-urlencoded
     Action: http://www.typepad.com/t/comments

   Linkname: TypePad
        URL: http://www.typepad.com/t/comments
    Charset: utf-8
     Server: Apache
       Date: Sun, 26 Jul 2009 22:48:45 GMT
   Post Data:
static=1&entry_id=6a00d8341c4df253ef0115714272f7970c&author=Robert+Maas&email=a
bc1.3.CalRobert@SpamGourmet.Com&url=http%3A%2F%2FTwitter.Com%2FCalRobert&text=S
ee+twitter+@CalRobert+where+I+suggest%0d%0athat+since+the+business+recession+is
%0d%0aover+already%2C+with+only+8%25+of+stimulus%0d%0aspent+so-far%2C+but+unemp
loyment+lagging%0d%0aindicator+will+rise+for+1.5+more+years%0d%0aand+take+5+yea
rs+to+return+to+normal%2C%0d%0arest+of+stimulus+should+redirect+to+WPA.&postcom
ment=Post+Comment
   Post Content Type: application/x-www-form-urlencoded

TypePad
Not Found
   We're sorry, the page you were trying to load cannot be found. If you
   believe you've reached this page in error please check the address and
   try again.
   Copyright 2009, Six Apart Ltd. All rights reserved.

ImageMakers: (Nora Thomas, Belgium) Ellie, after the love
Second most beautiful woman in the world, after Alyson Hannigan.
LinkedIn
.
Playing For Change | Peace Through Music
.
getoutdoors.com:
"back to setting his sights [to] low."

406 Not Acceptable
Not Acceptable
   An appropriate representation of the requested resource / could not be
   found on this server.
   Additionally, a 404 Not Found error was encountered while trying to
   use an ErrorDocument to handle the request.
     _________________________________________________________________
    Apache/2.2.11 (Unix) mod_ssl/2.2.11 OpenSSL/0.9.8e-fips-rhel5
    mod_auth_passthrough/2.1 mod_bwlimited/1.4 FrontPage/5.0.2.2635
    Server at www.land-that-job.com Port 80

;Given that lynx was at an ordinary link or menu or checkbox/radiobutton
; or plain-text page without any links, i.e. anything except a TextField
; or TextArea, then = command was pressed to bring up an info page showing
; current page URL and URL at target of link, then *that* info page was
; printed to a file;
;Given the name of that file:
;Load and parse the contents of that file to obtain the two URLs.
;Return value is list of two lists, each (name URL), first for viewed page,
; second for selected link you haven't clicked yet.
;Second element is NIL if no link on the viewable portion of the page.
;Second element is a keyword if something other than a link is selected.
;Viewed page info, if directory, nevertheless has final '/' only when
; given that way, which causes link to parent directory to be missing.
;Selected item info, if directory, never shows final '/' but I'll always
; append '/' in returned values when it's a directory.
(defun parse-printed-lynx-equal (filename)
  ...)

The morning of 2009.Aug.06 Twitter has been down for over an hour.
When I woke up this morning, I tuned in KCBS to check if it was going to rain or not today. I got the news that Twitter has been down for about an hour. I went online and tried Twitter and found it indeed wasn't responding. I did a Google search for [twitter down] and found status.twitter.com which said We are determining the cause and will provide an update shortly. I continued to listen to KCBS, and continued to check the status, and saw an update posted at 07:59 saying We are defending against a denial-of-service attack, and will update status again shortly. I then composed and posted an article to the newsgroup "misc.misc". A few minutes later, KCBS announced Twitter was back up, so I tried it, and indeed it was up, so I posted a tweet: Heard first on KCBS: Twitter is back up now! Shortly after that, the DOS-attack status was updated to add Update: the site is back up, but we are continuing to defend against and recover from this attack. but no separate status for it was put up, so there's no date for that update.
Then at 08:36 I heard the sound of rain running off the roof, looked outside, and saw indeed the parking lot was soaking wet, where it had been only slightly damp in low spots where night-condensation collects, and there was visible water dripping off the eave over my window. So at 08:37 I tried to tweet: At 08:36 here in Sunnyvale it's raining enough to make noticeable noise running off the roof. Arkadiy claimed it wouldn't rain today. (133 chars), but I got 408 Request Timeout instead. I continued to re-try my tweet about every minutes, but still I got 408 Request Timeout each attempt. So then I composed this pseudo-tweet (like a tweet but I can't post to twitter): When I tried to post that at 08:37 through 08:42 it gives me: 408 Request Timeout (i.e. Twitter is down again) (110 chars)
While I was typing that above into this pseudo-blog, a few minutes before 09:12 PDT, I noticed the sound of rain had stopped. At 09:13 I was finished typing so I looked out the window and indeed the rain has stopped.
I tried Twitter again at 09:19 PDT but connection times out when I try to see my timeline or post a new tweet via API. Turned KCBS back on at 09:20. Then I went back to my pending tweet and found that sometime during the past several minutes the connection finally succeeded. So then I posted the rest of my pending (already composed) tweets, which went very very slowly, probably because every one else is likewise trying to catch up with two hours of pending tweets.
cnn.com has typo: "Twitter users to post 140-character messages" (missing/bad verb!)
At 10:03 I tried to tweet: Catherine Galloway on @FRANCE24 changed her hair style, now pulled up in bun. but got 408 Request Timeout
Still getting that error when I re-tried at 10:12. Finally worked when re-tried again at 10:55.
Immediately after I successfully tweeted that, I tried a tweet about cheeky_geeky, but it hung trying to make connection. But at 10:58 it finally succeeded. Four more pending tweets went without delay, so at 11:01 my tweets were all caught up.
At 13:39 PDT I tried to tweet DW-TV "IN FOCUS" today: "The German language has a considerable choice of words to choose from." but the API gave me 408 Request Timeout. Eventually got it posted.

It's now Aug.07 02:49 PDT, trying to tweet Finished upgrading code to be able to parse any case of lynx '=' command when looking at directory. Now working on case of looking at file. but the API gave me 408 Request Timeout.
Started typing the new (below) proposal for metering access to Twitter to preclude dDOS attacks. At 03:33 PDT I tried the above tweet again, but again got 408 Request Timeout. At 05:16 PDT I tried the above tweet again, but again got 408 Request Timeout. It's been more than 13 hours since the last status update: Update (4:14p): Site latency has continued to improve, however some web requests continue to fail. This means that some people may be unable to post or follow from the website. Are Twitter staff totally unaware that their API has been down for hours? Or did they all go to bed for the night and not bother to maintain their very important utility? Or do they just not care to tell us they are aware of the problem whereby the API has been down for several hours? Either way I claim Twitter staff are seriously negligent!!
At 05:40 PDT I logged into Twitter from my cell-phone and successfully posted this tweet: API has been down for hours!
At 05:45 PDT I was finally able to post that tweet I first tried at 02:49 PDT.
At 07:29 PDT I tried to tweet I've finished first draft of proposal for Twitter design change to protect better against dDOS floods. See http://TinyURL.Com/TooBus but API getting 408 Request Timeout again. I'll try tweeting from cell-phone - SUCCESS: API down again - 408 Request Timeout

Aug.07 15:23 PDT I tried to tweet @ValliVoice I got inspired this past hour to write up http://TinyURL.Com/TooBus [Priority event system] but got 408 Request Timeout several times in a row, so I posted from cellphone that API is down. Then I went to sit on toilet for my daily defecation, and tweeted a few times from my cell-phone, but then cell-phone Twitter went down: First my 8 minutes of typing on 12-key-pad #IAmUnemployed http://TinyURL.Com/uh3t to see my r{e'}sum{e'} was lost when it gave 408 Request Timeout and then logged me out from Twitter and gave me login form instead of timeline. When I tried to log back in it said 406 Not Accepted. After several tries I got back logged in to see my 8 minutes not in my timeline and no way to get back my edit buffer. (A cell-phone is not like a desktop of laptop computer where you can use a text editor to compose your text and save to PC's hard disk then copy&paste across to Twitter. You have no choice but to type directly into Twitter's what-are-you-doing form and lose your text if Twitter barfs.) I then spent 4 minutes composing a complaint about how the 408 timeout error lost my 8 minutes of 12-key typing, but another 408 timeout caused that tweet text to also be lost, and 406 Not Accepted on log-back-in attempt persised until sometime between 16:16 and 16:21 PDT. As of 16:16 PDT both API and MobileWeb Twitter are still down.
Meanwhile: At appx. 16:09 PDT, WorldNews is claiming truth of the absurd story that "hackers" spammed e-mail containing links to tweets by the Georgian protester, and recipients of that spam clicked on the links enough to cause the DDoS attack that has lasted 33 hours already.
At 16:16 PDT WorldNews says a survey of Twitterers says that Bill Clinton's trip to North Korea is the biggest story of the week. But that's because with Twitter down, Twitterers can't tweet the new and bigger story of Twitter being down!!
At 16:21 PDT my cellphone showed my regular logged-in screen, so I asked MobileWeb also down? and it worked, so MobileWeb seems to be back up, but API (BasicAuth) may still be down.
At 16:24 PDT the API came back up too. Now I'm copying+pasting those pending tweets for belated posting. Done at 16:44 PDT.

2009.Aug.07 02:55 - One reason why Twitter is vulnerable to dDOS attacks is that it is utterly and totally free for use by 'bots. Use by 'bots is unmetered, which IMO is a dire mistake, virtually inviting being afflicted by The Tragedy of the Unregulated&Unmetered Commons. (See in particular: 3 Application > 3.1 Modern commons > Public resources > Spam and Phishing which IMO ought to also include dDOS attacks.)
To fix this basic design problem, I now suggest a design change (several paragraphs below) for Twitter and similar systems.
Note: It's good that all applications, even those programmed and provided by Twitter itself, go through the API, because this implies the following change is clean to implement, with certain responsibilities of the API, and certain other responsibilities of the application that uses the API.

Main point: The API should require labor to be performed, or PayPal etc. payment to be transferred, from time to time, in proportion to amount of usage, thus effectively throttling usage. This prevents 'bots from making unlimited demands on services, but still allows worthwhile 'bots to be operated within limits of total usage, and avoids the hassle of deciding whether a given application should be excluded as a 'bot or allowed as a legitimate application. All users whether live humans or 'bots are treated equally, must pay for service exactly the same proportion, by labor or cash, choice of user. Offering payment by labor or cash allows people to decide which is more valuable, their money or their time, and allows everyone access regardless of income level. Presumably anyone without sufficient funds to make a nominal cash (PayPal etc.) payment must be unemployed, hence must have plenty of time on his/her hands, hence performing a little bit of labor from time to time won't be a burden.

The API consequently has an XML etc. reply available to say in effect request understood but refused because insufficient payment received, and also an associated reply that tells the application which specialized URL to go to in order to make payment by labor.

Note: Payment by PayPal etc. is supposed to be known a priori, not requiring any re-direct URL to remind the application how to do it, whereas Turing tests or other requests for performing labor are concocted on the fly whenever a user falls behind on payment, and that specific labor is required by that specific user in order to obtain further services. Even if PayPal etc. will be made promptly after this incident, because the user has already fallen behind in payments, that user is immediately required to perform one unit of labor to restore access priviledges before service is restored. It's the responsibility of cash-paying 'bots to keep track of current account balance and detect low balance before making a request that would overdraw the account. That avoids ever having to perform specific labor, interrupting normal access. The penalty for failing to maintain account balance is interruption of normal automated access, requiring manual intervention to re-establish service.

The application, upon seeing a refusal of service due to insufficient funds, with an accompanying URL for specific labor to be performed by this user, then automatically re-directs the user, or somehow queries the user whether to re-direct now or wait until later when the user may have more time available. But in any case, the specific labor must be performed before access is restored, and the application shouldn't repeatedly re-try the service request before the user has performed the specified labor and the application has gotten confirmation of acceptance of labor and consequent credit to to the account.

Note: Nothing in the above requires the API to provide a URL for which the specific labor has already been decided exactly. It's sufficient that a unique URL is generated for each instance of each user overdrawing the account, which therefore allows later concoction of the specific labor-task and association of that URL with that task, so that the user can't "cheat" by "forgetting" the task and asking for it again and getting a different task and hence being allow to pick and choose which task is "easiest" to perform.

Note: Nothing in the above requires that the user actually perform the labor him/herself personally. It's OK if the user contracts for somebody else to do the work, such as by posting a RFB (Request For Bid) on NewEco, accepting bids, paying the lowest bidder to do the work, then simply copying the work from NewEco back to the labor-task URL given by the Twitter API. Or paying actual cash to a laborer and copying the completed work back.

Note: For the purpose of metering the commons to prevent the tragedy, it doesn't matter whether the labor performed is some service actually useful to Twitter, or merely a "Turing test" to force a nominal amount of human time to be spent. For example, the user can simply be required to read a sentence with one word missing and key in that missing word, a task that is extremely difficult for 'bots but very easy for literate humans who understand the appropriate language. Examples of such "Turing tests", currently only in English (sorry), may be found in my NewEco portal, both at the main menu level to effectively meter access attempts by people not yet logged in and effectively preclude automated "spidering" to harvest my WebSite content en masse, and after login to maintain a positive credit balance. (At present, there is no charge for usage, but accounts that remain inactive more than 24 hours lose one second of labor-credit per full day inactive, which must be eventually restored by answering a randomly-selected missing-word question. This discourages spammers from setting up thousands of accounts and then using them at irregular intervals due to availability of botnets some days but not others. With my current system, a spammer might script the missing-word menus and thereby automate the creation of thousands of accounts, but if the accounts go idle too many days they will all lose their account balance and require the spammer to manually answer thousands of randomly-selected missing-word questions, one per account, to re-activate them all. Thus the spammer would be required to log into my system at least once every 24 hours, but that gives law-enforcement staff a guaranteed chance to observe their actions at least once per 24 hours, hence giving them better chance to track the service-request floods back to their sources and thus investigate the botnet on a regular basis and collect enough data to track the botnet controller back to the spammer and eventually put the spammer in prison.)

Furthermore: With respect to throttling/metering usage, there should be two kinds of applications, namely "trusted" and "untrusted". "Trusted" applications are guaranteed, by legal contract, to be honest and open (informative) in their dealings with Twitter, specifically to throttle their own users, and to share information with Twitter regarding IP address blocks that are flooding servers with too many requests for service, either when not logged in (attempting to guess passwords or requesting services that don't require login or just "ping"ing etc.), or while logged in but without a positive account balance. "Untrusted" applications are the contrary, any J. Random application somebody sets up for experimental usage of the API. "Trusted" applications are treated as part of Twitter itself, with usage metered on a per-user or per-user-IP-address-block basis, based on the user ID or not-logged-in IP number as reported by the "trusted" application, whereas *all* use from any given "untrusted" application is treated as if it came from the application itself at whatever IP number is hosting that application. Thus any "untrusted" application would be allowed to use the API only as much as a single regular user would, and it's the duty of the owner/operator/manager of said "untrusted" application to pay (labor or cash) directly to Twitter in exchange for that usage. By comparison, "trusted" applications would report which user is making each service request, and Twitter API would keep track of such usage on a per-user basis and allow users with positive credit balance to use the application even if other other users of the same "trusted" application have overdrawn their accounts. If an "untrusted" application overdraws its single account, that entire application is locked out of service until that application pays for further access by performance of labor. (If the manager of the "untrusted" application wishes to avoid need to perform labor for Twitter API at odd hours, the manager should be sure to pre-pay via PayPal etc. to keep the balance positive at all times.) Thus Twitter delegates the responsibility of managing sharing of a single "trusted" application among various IP address blocks and various users to the management of that application, while Twitter manages throttling use of "untrusted" applications directly itself.

Furthermore: I recommend that the Twitter API, and each "trusted" application, set up multiple primary servers for direct access, by applications or applications+user tuples, and by direct-users, respectively, and semi-permanently re-direct users and/or applications to randomly selected or optimally selected primary servers to distribute the load. Once a particular user or application has been re-directed to a given server, all other servers will normally refuse service, instead re-directing the user or application back to the one what was assigned, and penalizing that user or application for even attempting to violate the re-direction rules. Only that one server assigned to that one user or application will actually process service requests.

Note: In normal operation, each primary server will be aware of which other servers are up, by PINGing them at regular intervals if there isn't any other traffic already going on directly between the two particular servers. If a primary server goes down, other servers will quickly notice that it isn't responding to PINGs and/or normal traffic, and will automatically relax the requirement to abide by the per-user or per-application primary-server assignment. Applications will then switch to other known primary servers as backup, which other servers will accept their traffic and perhaps re-assign them temporarily to another primary server, and issue the appropriate re-direction reply. Users, and experimental applications, who are directly connecting to their assigned primary server, will need to manually re-configure their bookmark to switch to another server, for example by backing out of the current bookmark and picking another to try. Thus if one or more servers go down for any reason, the system will gracefully adapt to using fewer servers.

Note: For every primary server to be PINGing every other primary server means that o(n**2) PINGs are required between N primary servers. If N is very large, a slightly different system may be used: 2**N primary servers are arranged in a two-in two-out butterfly (chaotic) graph, or any other suitable multi-but-sparsely-connected network. Each server pings only its immediate neighbors, and keeps track of whether that server is recently responding or not. Whenever one server notices any of its neighbors failing to respond to PINGs, that server sends a status update to its neighbors to "pass the word/rumor", whereupon the servers being informed try to PING the reportedly-down server to see whether the problem is truly that the server is down or contrarywise that for unknown/perverse reason the reporting server is unable to PING the reportedly-down server even though that server is not really down. If the third server (the one now PINGing) agrees that the first server really seems to be down, the "rumor" is passed to its neighbors, and so on and so on, until every server in the whole Twitter server Universe has personally verified that one server is really down. Meanwhile, each server upon itself seeing the other server down, immediately begins to accept any traffic that was supposed to be assigned to that other server. Thus any user or application that switches to a randomly selected altenate server will quickly find one that is both up and aware of the down-ness of the assigned server hence is already willing to accept traffic.

Result (how this kind of metering/throttling of usage, and semi-permanently assigning users/applications to various primary servers will defend against a dDOS attack):
If a dDOS flood of any kind hits any particular API or application server, that server immediately black-lists (at the API-server or application-server level) all IP CIDRs participating in the flood, and reports these CIDRs to management of the hosting service. If the flood is so severe that merely rejecting service requests from those CIDRs overloads the Apache server, management then re-configures the TCP/IP stack or even the router to discard all packets from the CIDRs most participating in the attack. If the pipe from any neighboring router gets so overloaded that even non-blacklisted packets are discarded because they won't fit in the pipe, then staff at this server telephone staff at the server at the other end of that pipe to request black-listing of packets there too. Working this back toward the source as many hops as needed, eventually black-listing of CIDRs will be close enough to the source that only routers and pipes owned/managed by the ISP hosting the botnets will be overloaded, the rest of the InterNet will be unaffected by the botnets.
Note that this is the same sequence of emergency measures I've proposed for several years in response to botnets sending SMTP spam. The only difference is whether the Twitter-API-server or Twitter-application-server or the SMTP-server performs the first step in black-listing. Once the offending CIDRs are reported to management of the hosting service, and CIDR-blacklisting at the TCP/IP stack level commences, the emergency procedures are identical except the for port number whose packets get discarded. And TCP-setup or PING packets would get dropped identically in all cases regardless of SMTP or Twitter-API or Twitter-application. And perhaps all packets for all ports of a particular CIDR should be discarded whenever that CIDR is flooding any port whatsoever. Like why should some ISP be allowed to host botnets that flood one service yet users of those same ISPs should be granted other services as if the ISP were otherwise clean? Black-list the entire ISP, deny all users of that ISP any services whatsoever, even [simple]-time-of-day, whenever any part of that ISP participates in a major dDOS attack, I say!

Phone tree useful links: How to Build a Phone Tree / What is a Phone Tree?

How my proposed alert system with priority levels will work:
You establish an account on NewEco alerts, and configure a "callback" method by which the alert system can immediately notify you whenever your profile sees an event with higher priority than your current silence threshold. This "callback" can be anything you want, such as custom URL to your own PHP script, or SMTP e-mail, or @you tweet, or Instant Message, or post to IRC channel, connection to custom TCP socket on your PC, etc
Next a test is made to see if that particular "callback" method really reaches you, not some "Joe" whom you are harassing. You must receive the test alert message and enter it back in the validation form.
Next you can begin actually configuring various kinds of events that are detectable either by callback from an event itself or by polling for an event to see if there's a new event since last polled. In the process of configuring your first event priority, you must specify both the priority-level generated by ths kind of event and what you want your initial silence threshold to be. If the slience threshold is below the level for this kind of event, then the first time this kind of event happens you'll be alerted and the silence level will be raised to prevent further alerts at the same level until and unless you manually lower it. But if the silence threshold is set at or higher than this kind of event, it means you are too busy to be bothered by this kind of event for the moment, and you won't ever be alerted to this kind of event until and unless you manually lower the silence threshold below this event's level.
If you don't presently have any reliable way to receive callbacks, then you can poll the event system at your discretion to see whether it *would* have alerted you if it could have, i.e. if there is any pending event at higher priority than your current silence threshold. When you poll you will get one of two kinds of interactive responses:
(1) There is nothing above your silence threshold which is N. The highest priority event currently pending is only level M. (where M <= N)
(2) There is|are pending event(s) of priority level M, namely #hashCode1 [#hashcode2,3,...], which is higher than your silence threshold of N (where M > N),
In the latter case you'll probably see a full-screen display like this:

7 (none, your house isn't on fire or being burgularized)
6 (none, your girlfriend isn't in the emergency room)
5 #emailFromGirlfriend #appointmentReminder
4 #majorNewsEventYouWantToKnowAbout #emailFromFriend
3 Silence threshold (you are too busy to respond to e-mail from your mother
                     or to catch up on lesser-interesting hot news stories)

Attempt to connect to the URL http://bit.ly/HLT3u spamvertised by Jezza3110 bombs out with a browser alert Error: Uncompressing temporary file. I saved a trace of the client/HTTP activities here

At twitter.com, after filling in my username and password in the login form, clicking the submit (login) button sets a session cookie but then just takes me back to a blank login form instead of showing my timeline and other usual stuff. Here is a VT100-output-transcript of:

1. manually invoking 'g' (Go to URL) command
2. entering twitter.com as the hostname
3. Lynx guessing how to turn that into a complete URL, and then submitting that URL
4. getting back the toplevel WebPage including login form
5. manually entering username and password
6. pressing submit/login button to send the form to the server
7. Lynx getting cookie and asking user whether to accept it
8. User (myself) confirming to accept all cookies from this domain
9. redirection back to original login form, re-loaded to erase username and password

tutorials4u.com
<dt> The UL element is the container for the entire Definition List
That doesn't look right. Maybe '<dt>' should read '<dl>' and 'UL' should read 'DT'? That's how it looks in the example later. The very next line has two more typos:
<dl> <dt> The DL DT inline element is used to create a definition term, before each <dd> tag Opt. compact.
Here's that example I mentionned:

  <dl>
      <dt>Term 1</dt>
      <dd>Definition 1</dd>
       <dt>Term 2</dt>
      <dd>Definition 2</dd>
  <d/l>

Note that when you try Portl1 you'll see:

skipSetup

For cellphones: to let you skip from here to after the setup portion of the output in a single keystroke (press OK button) instead of needing to scroll one line at a time.

[1249738369.38306800]

PHP universal time = number of seconds since start of 1970, to a precision of one micro-second, at start of script-running, so that at end of script I can subtract this from the ending timestamp to see how long the script took hence how much to charge user.

IP=198.144.192.42, port=4266.

This information is collected with each connection, and logged permanently. It's shown to you as a courtesy, in case you are trying to debug a proxy, or just check whether your IP number has changed since last time.

Con...sel...use...log...!

Trace of setting up the PHP/MySQL interface:
Con... means it's trying to connect to the database server.
If that suceeds, next:
sel... means it's trying to select the database.
If that succeeds, next:
use... means it's transmitting the SQL command to use the database, meaning it doesn't need to repeat the name of the database with every later SQL command.
If that succeeds, next:
log... means it's making a permanent log entry in a table of connections, showing the timestamp and your IP number and port number, and a text field for comments, initially empty, but to eventually have a log of what you actually did during this connection, such as put some text in the TextField and try to submit it.
If that succeeds, next:
! just means that last item succeeded and there's nothing remaining to do for setup.

skipInstructions

For cellphones: Another button to skip a bunch of stuff in just one step if you have already seen it and don't need to see it again before proceeding with the action you need to do. Thus two presses to the OK button on cellphone skip both setup trace and instructions and take you directly to the main menu. Then a single press of DOWN button on cellphone scrolls one more line to hide the main-menu title and show the text field and submit button.

twittercism.com survey question can't be answered without JavaScript, because there's no HTML submit button:

   Would you pay for Twitter?
     * (*) Yes, I'd pay for cool extra features
     * ( ) No, I'd never pay, no matter what
   [BUTTON]

     * (*) Yes, I'd pay real $money$ for cool extra features
     * ( ) No, but I'd be willing to exchange labor by barter
     * ( ) No, I'd never pay, nor even exchange, no matter what

/--------------------------------------------------------------------------\
 |    | _   _| _ ._ _  _ (~| _  _|_|_  _     _  __|_  o._ _|_ _._  o _
  \/\/ }_  (_|(_|| | |(_| _|}_   | | |}_  \/(_|_\ |   || | | }_| \/|}_\/\/o
\----(Rendered by means of <http://www.schnoggo.com/figlet.html>)----------/

http://twitter.com/DomainsDomains/status/3231895924 ContainsLink-> http://bit.ly/pUdKi RedirectsTo-> http://www.washingtonpost.com/wp-dyn/content/article/2009/08/10/AR2009081001970.html
Your browser's settings may be preventing you from commenting on and viewing comments about this item. See instructions for fixing the problem.
Additionally, entries that are unsigned or contain "signatures" by someone other than the actual author will be removed.
How am I supposed to sign something that's just keystrokes on my computer then uploaded to Web site??

Shortly after I heard that Les Paul had died, I searched Twitter to find mention of his death, found several, including Jimmy Stevenson who wrote: Trying to write a goodbye tweet to Les Paul. None of them sound like me. So goodbye Les Paul. Rest in Peace. Thanks for the guitar.3:44 PM Aug 13th from web
I decided to take up the challenge by citing some song he sang (with Mary Ford). I searched online, and in my own tape listings, to try to find a song title I could cite, but couldn't find anything that would be a suitable tribute to him.
Then I came up with an alternate idea, writing my own "ditty" in honor of Les Paul. Shortly I came up with just one line, to the first stanza of the tune for "Kiss me Once Again", and tweeted it (see first line in next paragraph below). Then I went on with other tweeting, only to start to get ideas for additional lines continuing from my first line, and soon I had three more lines to fit that same tune, one complete verse:

• 2009.Aug.13 15:59 PDT
  Parting ditty tweet to Les Paul
  "Play guitar, and play guitar, and play it once again,
   you've got an eight, track, tape."
• 2009.Aug.13 16:09 PDT
  "Rock and roll, wouldn be the same, without your steel guitar,
   it's been a grand long time."
• 2009.Aug.13 16:11 PDT
  "You'll never know how many times I've loved your music,
   and just how much I'll miss you now you're gone."
• 2009.Aug.13 16:13 PDT
  "So, let's all play your songs again, and play them once again,
   you made some great, great, sounds."
  (commas inadvertantly omitted in original tweet, added here)

afip.org typos:
"[Not] the long, individualized villi."
"The cause of this is the ongoing lymphocytic inflammation in the [ntestine]"
(Also has good refutation of various rumors circulating on the InterNet.)

2009.Aug.20 09 PDT I connected to accessibletwitter for the first time in days, and got this page:

Accessible Twitter - web accessibility for the Twitter web application
Aw, man! You have reached the limit on the Twitter API hit rate.
   Please wait a little while and return. Your reset time is .
   You may:
     * Try again (be sure to wait at least a few minutes).
     * Contact Twitter and complain about the limited API access! (It's
       currently 100 per user per hour.)
     * Go to Accessible Twitter Home page or About page.
     * Sign Out

                       Error 500 - Internal server error
   An internal server error has occured!
   Please try again later.

Mike Nierengarten posted on Twitter the same URL (bit.ly/Future_Newspaper) six times in a row (1 2 3 4 5 6) which link re-directs to socialsearchmarketer.com which has a lot of good ideas about how a future InterNet-based newspaper should be. But early in that article, there's a link to another page (Y Combinator: Startup Ideas We'd Like to Fund) which lists several problems for which they'd like to fund a solution. I have lots of responses to that want-to-fund page here. After I wrote those responses, then I finally quickly responded to items in the main page that weren't already covered. Those final responses are here.

Responses to Y Combinator...:
A cure for the disease of which the RIAA is a symptom. Something is broken when Sony and Universal are suing children.
Idea from years ago, but need non-disclosure agreement and meet in person before I discuss it.
.
Simplified browsing. There are a lot of cases where you'd trade some of the power of a web browser for greater simplicity.
NewEco could "pay" people to select "best" of various topics, and NewEco could "pay" people to quickly answer questions or set up new FAQ links and then truth-futures market could filter those answers so that only the best are seen by the novice who wants an answer.
.
Outsourced IT.
This is already planned as a major feature of NewEco, in particular much of NewEco itself will be bootstrapped by outsourced contract work.
.
Enterprise software 2.0.
NewEco software will be reusable components that interact with each other, so you can buy just exactly the pieces you need. I particular if you have most of what you want and just want to plug a gap, you post a RFB for somebody to spend a few minutes creating exactly that one gap-filling piece you need.
.
interactions with customers
Combination of reverse-tree and FAQ engine+experts.
.
Something your company needs that doesn't exist.
You don't need to quit your job to start a new company to produce the thing your (old) company needs. Instead, you join NewEco to cooperate with others to brainstorm the details of the design, and then share the workload via RFBs, so you personally need spend only as much time as you really can afford, and you spend that time *where* you want, such as making sure the design team stays fixed on your need rather than wandering to something you don't need.
.
Dating.
See HotNot for my major ideas for how to handle the matchmaking itself, namely #1 face rating without seeing text yet (A,B,C,D,F), which gives more honest rating than HotOrNot gives, then see text only if/when you need to see it, #2 separate face/text matching with double match *only* if both match. See LinkII for establishing trust networks to elminate fakers on dating sites and to advertise by e-"word of mouth".
.
Photo/video sharing services.
Slightly off-topic, but see my photo-clustering proposal, which should give much better results than Google Images.
.
Fix advertising.
See my proposal I tried to pitch to Yahoo and Google but neither would allow me to talk with them in confidence; one didn't return my call, other said I should just write up my proposal and snail-mail it to them and thereby forfeit all my intellectual property. Require meet you in person and verbally agree to non-disclosure before I say *anything* about my invention. If you like the basic idea, we sign written non-disclosure agreement before I discuss details.
think what the goal of advertising is ...
You want users (of online service) to learn about your products/services and remember that information, and in particular learn your company/product name and associate that name with what you provide. My proposal fully addresses this, much better than traditional singing ads on TV/radio, much better than pop-up/banner ads on Web, using technology that is possible *only* in an interactive medium such as the Web.
.
Online learning.
There are several different kinds of learning. One essential skill, for many tasks (reading, writing, spelling, foreign languages, technical subjects with jargon, even learning how to sing particular songs/vocals) is to *memorize* the vocabulary/jargon/lyrics. My algorithm, invented circa 1970, installed on Macintosh 1995, ported to Web 1992, solves that problem, making memorization very very efficient, and totally monitoring progress. Once the vocabulary/jargon has been learned, then deeper kinds of learning become possible.
I also have prototypes of technology for teaching people how to set up "word problems", namely assign symbols for each concept, then write various relations between those symbols to express the problem statement.
.
Another, particularly for younger kids, would be to let them learn by watching one another (anonymously) solve problems.
I like that idea. Never heard of it before. I suggest first do the automated teaching for setting up word problems. If the student has trouble, then watch another student do the same task. Or for absolute beginners, offer a video of somebody solving that kind of problem before starting the problem-setup lesson.
.
The most important of all is the defining flaw of large organizations: you can't tell who the most productive people are.
NewEco payment based on micro-contracts fixes the problem. If somebody doesn't submit bids, or doesn't fulfill bids, that person makes no income, and will resign from the game, no need to formally fire that person.
.
Off the shelf security. I've proposed to several people recently my idea for a worldwide network linking SecurityWebCams with piece-work watchers, with "fake crimes" inserted randomly to make sure the watchers really are paying attention.
.
"A form of search that depends on design." (Please define that better.)
.
New payment methods.
** "cooperative" economy = "laboralism" = http://TinyURL.Com/NewEco ** My proposed replacement for "capitalism". Labor, not money, is unit of exchange, so *everyone* can be an employer, no money needed to hire somebody else.
.
The WebOS.
I consider the appropriate basic design to be RPC/RMI/SOUP embedded in public-key cryptosystem, whereby *each* component has its own ""personal"" crypto-key, thus assuring reliable operation because bogus requests for services are either not recognized as valid at all or are recognized as valid for somebody who is not currently entitled to services. I'm currently building that into NewEco. See for example CalRobert/status/ 3420913448, 3420956056, 3421019398, 3423636232.
.
More open alternatives to Wikipedia.
I've already proposed an alternative, basically using a reverse tree and truth-futures market for version control, so that only the best edits get high rating and hence show in the default view. Make example, from Wiki, with drop-down menus.
.
A buffer against bad customer service.
I've already proposed truth-future-market reverse-tree mated with FAQ engine as solution to this. Customers can get most answers without ever needing to talk to human, because volunteers (with the incentive of getting paid for good work) produce more comprehensive answers than the regular company staff has time to set up and organize. Common questions are answered that way rather than requiring time of company staff. Thus company staff aren't overworked, and can pay serious attention to the very few questions that get all the way to the root of the reverse-tree. In fact I've been planning to set up exactly this for Virgin Mobile as soon as I have some spare time.
.
Fixing email overload.
Reverse tree again, coupled with my proposed event-priority-alert system. In particular, an over-eager spam filter can send all suspected spam to reverse tree for evaluating whether it's really spam or not, as well as "consumer reports" tagging of ads as to whether they are probably fraudulent, explaining *why* they are scams.

Responses to future-newspaper:
"NewsPaper" is misnomer. Better would be "NewsPages" where "page" is used in the Web sense rather than the wood-pulp sense.
.
It makes zero financial sense to compete on breaking news.
That can best be handled by volunteer witnesses, such as Twitter did well with events in Iran, except with NewEco payment for good work, and with LinkII references to prove all sources are what they claim to be.
.
Investigative journalism should be the heart of any newspaper.
Perhaps much of that can be replaced by an integration between volunteer (but paid) witness reports, trust network to cite who said what and who is generally honest, truth-futures market to give more accurate judgement of correctness, and e-interviews between back-at-home journalists and on-the-scene witnesses to re-tell story with more details analagous to Dianetics "auditing" (therapy) and courtroom cross-examination. In some cases a witness may be able to step back from the situation and sit at a computer to directly write up a well-organized report, in collaboration with ournalist back-at-home who will purchase it upon satisfactory completion.
.
Editor(s)
Most of this labor can be done by reverse-tree and futures market, basically each person tries to predict how well the final editor will like the submitted material. Those who best understand the final editor's criterion will predict best and thus be automatically promoted to near the root.
.
Mobile
The most important needs are very terse output (to fit small screens, and to avoid overcharging for downloads) and extremely few keystrokes to achieve navigation. See http://TinyURL.Com/Portl1 for my own attempt to target cellphones with one-inch screens.
.
Future newspapers will take advantage of ad networks to supply their advertising.
See my note on other page about my new idea I want to discuss with you confidentially.
.
I would definitely appreciate any comments.
Well here they are, above and on other page. Please let me know that you've read my comments.

I filled out your form, and clicked:

   Linkname: Submit
     Method: POST
    Enctype: application/x-www-form-urlencoded
     Action: http://www.socialsearchmarketer.com/wp-comments-post.php

@PLPropertyClub posted a link to an advertisement for Google Alerts, here.
.
I responded: Google Alerts are crap! Instead of flooding my InBox, allow me to specify custom HTTP/SOAP callback.
.
PLPropertyClub responded: Please explain How do I allow you to specify custom HTTP/SOAP callback?
.
My next reply is here:
First some group designs a simple XML-based protocol by which the central Google alert handler can communicate with each of the various user-agents. From user agent to Google we can re-configure our personal account. From Google to user agent we can receive alerts.
Next, Google must set up its own official public-key, so that it can send out public-key-signed alerts, and so that it can receive encrypted reconfiguration requests.
Next, each user wishing to join this system must set up his/her own public-key, so that such user can send public-key-signed reconfiguration requests, and received encrypted alerts.
Next, each user must actually set up a CGI or PHP etc. "script" that handles signed+encrypted SOAP messages from Google Alerts.
Next, each user uses the userAgent-to-Google protocol to set up a GoogleAlertsSOAP account, directing it to the user's "script".
Finally, whenever Google finds an event for which some user wishes to receive an alert, Google composes the alert in the agreed-upon XML notation, signs it, encrypts it for the particular user, connects to the user's "script", and transmits it.

diamondsecretPK has link to tinyurl.com/cbdetg which indirects to theme_healthy_ethnic_foods.shtml which has some links but doesn't have any main content. This is all I see there.

New idea (2009.Aug.23 02) for how to use NewEco to bootstrap ways to contact hard-to-reach people. On NewEco I could have a form like this:
.
I'm desperately trying to establish communication with each of these people:
  (list of checkboxes)
If you know how to contact any of those people, and get a meaningful reply back, not just an auto-ack by their mail system, please check the corresponding box(es) then press the SUBMIT button.
If you don't know how to contact one of those people, but you think you know how to contact somebody else, a "middle person", who can then contact the person, please contact that "middle person" and ask that "middle person" to please log in here and respond just as you did, except hopefully with a better result.
If you don't know how to contact any of those people, and you have no idea who might specifically know how to contact any of them, but you have contact with somebody who just generally is expert at tracking down how to contact lots of hard-to-reach people, please contact that "how-to-contact-people expert" and ask that "how-to-contact-people expert" to please log in here and respond just as you did, except hopefully with a better result.

http://twitter.com/diamondsecretPK/status/3477663071 contains link: http://tinyurl.com/ph38ml which indirects to: http://www.healthcastle.com/unusual_health_foods.shtml "Although it is much less common [that] cow's milk" (should read "than") "Sunchoke or Jerusalem artichoke ... is not related to artichoke." (wrong! Every living thing on Earth is related by common ancestor to every other living thing on Earth.) http://twitter.com/diamondsecretPK/status/3472217540 contains link: http://tinyurl.com/krwboy which indirects to: http://www.dmconfidential.com/blogs/column/Digital_Thoughts/2377/ "Assuming that [the the] availability of de -averaging can free up inventory" (should omit one of the two "the" copies) http://battellemedia.com/archives/000063.php "all kinds of intent, [contreversies], politics, relationships." (misspelled)

@games2play4u tweeted me statuses/3648534932 containing the shortened link bit.ly/zw79l which redirects to thron.biz where the first six screens are essentially blank except for scattered repeats of [EMBED]. Then there's a listing of games with links, in alphabetical order, so I clicked on: Checkers There's just a bunch of text such as [EMBED], and one mention of checkers: Checkers: Koala Checkers: The classic board game. With koala pictures on the pieces! and then just one link on that page: Dark Orbit 728x90 Everything there is in German, so I can't understand what it says.

@ScottKirsner posted status/3555430562 which contains the link bit.ly/dlNlR which redirects to boston.com where I see the following utterly inappropriate text shoved at me:
... Invalid email address ... Invalid email address ... Sending your article ... Your article has been sent. ... Sorry, we could not find your e-mail or password. Please try again, or click here to retrieve your password. ... Your comment is subject to the rules of our Posting Policy
All I did was click on a link to see what was there, presumably an article that somebody else posted for the public to read. I didn't try to submit an e-mail address, nor send/post an article, nor post a comment, so why is it harassing me with all those irrelevant error/warning messages?

@ScottKirsner posted status/3517025297 which says New business idea: online grammar/spelling lessons for the web's most dumbest [sic] commenters. (Whoops.)
In fact this file you're looking at now already contains many examples of gross spelling etc. mistakes not just by commenters but in Web pages of many types, with the intention of publicly humiliating these authors until they clean up their stupid mistakes. Also, DiscrimBad.html and DrivingDirections.html contain many complaints about Web sites/pages that violate anti-discrimination laws such as the ADA by deliberately imposing the extra burden of viewing images or running JavaScript in order to use their online services.
Furthermore to more actively go after Web authors who never bother to look here, or who ignore my critiques here, as well as to enlarge the volume of such public humiliation, I currently am planning as part of NewEco a registry of such errata gripes, not just those I find myself, but also Web-mistakes found by my users. Furthermore, I plan to add an extra twist of paying users to directly petition Web authors who have so-far refused to fix their mistakes, and also paying people to provide info about how to contact the people to more effectively petition them to fix their mistakes!
And of course there will be a mechanism for any Web author to post a notice that such-and-such mistake has been fixed, and then for my users to check the updated Web page to see if the mistake has indeed been fixed, and to use TruFut to evaluate these mistake-fixed notices and consequently automatically update my database to show which mistakes have now been fixed so those Web authors should be congratulated and no-longer petitionned.

@ScottKirsner posted status/3515060182 which was a re-tweet from @mike_feinstein which said My rebuttal to @ScottKirsner and his desire to blow up Mount Money in Waltham http://bit.ly/o259c
My reply:

• I could really uses some extremely small amount of venture capital or angel funding already, just to help brainstorm my ideas and alpha-test my software.
• My plans for NewEco+TruFut could include evaluating startup ideas and business plans etc. as a guide to help Venture-Capitalists/"Angels" decide where best to put their money for maximum likelihood of helping a good new technology get developed to public benefit and possible financial return.

http://twitter.com/ScottKirsner/status/3542527241 (vacation tweet)
.
http://twitter.com/mike_feinstein/status/3604590363 (Priority would fix that need-to-check, while TruFut within RevTre would handle all regular business during meeting and have it all ready at end-meeting for the CEO to quickly read-and-approve.)
.
mike_feinstein/status/3603923408 (TruFut will be fun, trying to out-guess others about hard-to-find information about current facts and hard-to-predict future events.)
.
mike_feinstein/status/3493200682 (That article is such a lie! College graduates with Bachelor of Science degrees in mathematics can't get even minimum-wage jobs. The claim that we need *more* such college graduates is a gross lie, and if people actually believe it then it does great harm to those already unemployed, making even less likely jobs will ever be available for them. Let's hire the ones we already have before crying about not enough of them competing for the very few jobs currently available.)
.
http://twitter.com/ScottKirsner/status/3534468202 LinkTo: http://bit.ly/32oBEt RedirectTo: http://www.boston.com/business/technology/innoeco/2009/08/blogging_about_failure.html Waving the white flag after spending years trying to get a business [of] the ground is never pleasant. ... Abby and Ivan Kirigin ... shutting down their start-up, TipJoy. TipJoy had aimed to make it easy for you to leave a small tip -- say, 50 cents -- for a blogger whose site you read daily or as a thank-you for downloading some free music from a musician. Maybe NewEco-LinkII-RevTre would help? ... Scott Rafer ... "commencing an 'orderly shutdown'" of Lookery, ... building ahead of the market (December 2008). Chicken-and-egg problem if you don't build any capability until there's a market but no market develops until after you have something to interest them.

@politicorithmic posted statuses/3880993554 which contains the short link bit.ly/KwBci which redirects to wsj.com
That Web site is grossly broken. Before I see any of the intended content, I first see these stupid remarks thrown at me:

1. Thank you for registering.
2. We sent an email to:
   Please click on the link inside the email to complete your registration
3. An account already exists for the email address entered.
4. This service is temporary unavailable due to system maintenance.
   Please try again later.
5. The username entered is already associated with another account. Please enter a different username
6. The email address you have entered is already in use. Please re-enter the email address.
7. Your login is either a username or an email address.

http://www.investopedia.com/university/shortselling/shortselling3.asp - Good article about dangers of short-selling anything with no upper bound.
.
http://www.investopedia.com/university/shortselling/shortselling2.asp [Shorter] sellers use an endless number of metrics should read "Short"
.
http://www.investopedia.com/ask/answers/06/squeezingtheshorts.asp he was forced to [declared] bankruptcy. should read "declare"

I downloaded the page that I get at the very top of Yahoo! Mail after previously logging in, here. Then I passed it to the W3C validator, with this result: 229 Errors, 47 warning(s)

Regarding Random Occurrence of Recognizable Images:
I have a better idea: Rather than compare with a priori image, such as Mona Lisa, have users watching the screen saver click a button/mouse whenever they see *anything* whatsoever recognizable. Because of delay, they click late, but with pseudo-random-number generator you can easily run backwards to the exact point when the image was most recognizable to that user. Then upload the image and ask others to describe it, and (analagous to Google's image-tagging game) if several different people describe it the same way then it is tentatively approved as a "really recognizable" image and then fed into http://TinyURL.Com/RevTre for more people to check if they also agree. Winners would be determined by whatever reaches the root of the tree, with multiple winners rated in a (popularity) survey.

Alternately, to test evolution by mutation and selection, have a way to sense when anyone is watching the screen, and reward whatever is showing during that time and just before they are drawn to it and punish whatever causes them to stop looking, using random number generator to drive mutations but having selection of mutated genomes determine what gets most displayed. (This is the same idea I proposed to my girlfriend Jean Conner in 1988, after we attended a SIG-Graph meeting, where the speaker said anything like that (computer generation of "art" that is attractive to people) would be forever impossible, and I disagreed and proposed this kind of genetic algorithm, demonstrated by setting up a computer-controlled vending machine, such as Coke or Pepsi, that attracts customers by a computer-generated display, but Jean sided with the speaker saying it would be forever impossible.)

Regarding Green Generator Challenge:
Dry brush on hillsides is readily available and needs to be thinned to reduce ability of wildfires to spread rapidly. See my plan for a system of gravity-powered brush-trimmers-collectors to work their way down the hillside from the crest, ending up with full load at bottom to empty into efficient combustion-to-energy unit such as my four-channel reverse-flow heat exchanger.

Regarding The Babel Prize:
Hey, my existing flashcard technology can do that, depending on what you actually mean by "learning a language". Is it sufficient to memorize fifty words each in eight different langauges, translating from English and/or guessing missing word in context and spelling it correctly? If so, my existing 1970-1995-2002 technology can do that for anyone willing to spend 20 minutes per day for three weeks. But if by "learning a language" you mean memorizing ten thousand words in each language, that would take a bit longer to "learn 8 languages" at the rate of 25 new words per day, appx. nine years.

More errata:
http://www.voanews.com/english/2009-09-30-voa7.cfm
"Some [neighborhood] have lost electricity and telecommunications."
(should be plural)

> Go to Menu -> Settings -> Security.

OK, next:

> Enter your own four-digit lock code.

OK, I've done that, now I see these options:
- Lock phone
- Change lock code
- Emergency numbers
- Erase contacts

> Choose "Selective Call Blocking"

There's no way to choose that. It's not one of the available options.

<<<                                       Information about the current document
                 Lynx 2.8.6rel.5 (09 May 2007) (latest release)
File that you are currently viewing
   Linkname: HTML Link Tags Examples
        URL: http://www.mountaindragon.com/html/links.htm#local
Link that you currently have selected
   Linkname: Your harddrive.
        URL: file://localhost/C:/
That link doesn't work here (on FreeBSD Unix).

CSIXConnect#15575
Understanding the [clients] environment
That should be possessive (with apostrophe) rather than, or in addition to, plural.
.
phJobSource#27143
Must be proficient in Excel and Publisher software to work on [a] upcoming project.
Should read "an".
.
phJobSource#27141
Our client is a fab-less semiconductor startup located in Fremont. [The] are developing next generation wireless chipsets ...
Should read "they".
.
imdb.com
Frederick Carson: Makes me wanna go home and shoot my [Chinee] cook.
.
dpsinfo.com
A song writer named Pee Wee King [wrote wrote] The Tennesee Waltz died in early 2000, ... (double word, first of the two should probably read "who") A similarly named person who'd claimed to have written "Itsy Bitsy Yello Polka Dot Bikini" died in September 2006 ... (Correct song title is "Itsy Bitsy Teeny Weenie Yellow Polka Dot Bikini")

I posted a messages to a Yahoo! Group, and right after it confirms that the message was sent it asks me:
Would you like to share all your activities (messages, photos, etc.) in this Yahoo Group via Yahoo! Updates? What's this?
I click on that link, and I see a FAQ. The first item is: What is Yahoo! Updates? so I click on that, taking me to a page where the first thing it tells me is:
You have reached the maximum number of saved articles. Your oldest saved article will be replaced with the new one.
WHAT THE FUCK??? I have never asked it to save an article, and I have no idea what that even means, like where would an article be saved and how would I get access to it. And I have no idea why, when I ask for information about what is this feature it's trying to get me to use, that the FAQ would try to save an article for me and complain about there being a problem with it.

w3schools: Each code line in PHP must end with a semicolon. The semicolon is a separator and is used to distinguish one set of instructions from another.
If semicolon were a separator, as it is in Algol 60, it would be required only *between* statements, i.e. statement;statement, not needed after the last statement, and if there *is* a semicolon after the block then it'd cause the compiler to generate a null statement after the semicolon. So this tutorial tells an untruth here.

I've manually edited four of my own tweets and one of HDNR's tweets that are linked by reply and/or explicit atsignUser/updateNumber, to show how a two-directional thread-tracking service (value-added on top of Twitter), i.e. a Directed Acyclic Graph (DAG), might look. This took me most of three days, which was really too much effort, but once I got started it was hard to find a good place to stop until I had finished editing to show:

• What looks like @CalRobert/5774571844 in the tweet is actually a link to http://twitter.com/CalRobert/status/5774571844
• Above the showing of the tweet is an option to expand to quotes of each of the referenced tweets, i.e. the tweet replied-to and any tweet explicitly linked via @CalRobert/5774571844 syntax.
• The list of referred-to tweets above the main tweet is in the same enhanced format with links.
• That list is numbered, with numbered refereces to them in the main tweet.
• Below the showing of the tweet is an option to conduct a search for later tweets that reference this tweet by either reply or explicit link. First click there just gives totals of each type.
• List of later tweets referring to this has link to fully show them all, again in same format, but split in two groups, replies and explicit links.
• Each such later tweet displayed has brackets around the link back to the main tweet.
• If you already have expanded above or below, you can then expand the other, to see both together. (I got tired and didn't do these additional combinations except for the one tweet that starts this whole demo.)

wordpress.com You should now see a row returned for each Customer and the total number [or] orders placed in the year 1996--including the customers that didn't place an order.
.
mysqltutorial.org:
After SELECT keyword is a list of column [name] in which you want to retrieve the data.
If the columns' name in those tables are the same, you have to [explicit] declare [the] as the form table_name.column_name, ...
Next you will list the main table and then a list of [table] you want to join.

Link from Yahoo startup in InterNet Explorer: article/107602:
One of the items it says is obsolete is "Long-Distance Call", on the claim that there is no longer any distinction between local and long distance calls. In fact they are totally wrong. I am totally unable to make any phone calls from my landline to any point outside my local service area, and even within that service area it costs money to call anywhere beyond the Palo Alto or San Jose West/North areas. Even San Jose South is a toll call from here. If I want to call outside the local service area, I have to call from my cell-phone, which costs 18 cents per minute, much more expensive than calls from my land-line to places within the local service area.

php-mysql-tutorial.com errata:
If either of [this condition] is not met then the one accessing this page [haven't] [login] yet.
In both [application] we use the session variable ...

IMDB.Com:
"In 2005, Hugh attempted to create [and] Indian version of Playboy"

Twitter user AwOLonline (Name: Philippe Delmar / Location New York City Area) posted: status/4776219208 which shows as:
@EllenAMuth This only shows that you have to ask a lot of questions before meeting strangers and ... in your case, always bring a chaperone! 8:20 PM Oct 10th from web in reply to EllenAMuth
AwOLonline
Philippe Delmar
I tried to click on that "in reply to EllenAMuth" link, but got an error: 404 Not Found.
I was curious what terrible dating event Ellen had posted about, so I sent AwOLonline a reply, which shows as:
@AwOLonline You replied to Ellen's tweet that doesn't exist. Did she delete it after you replied? Did you save a copy of her tweet? about 20 hours ago from API in reply to AwOLonline
CalRobert
Robert Maas
All that AwOLonline had to do was to click on the "in reply to AwOLonline" link to see what I was referring to, but apparently AwOLonline didn't do that. Instead AwOLonline replied as follows:
@CalRobert Which of Ellen's tweets are you referring to? about 20 hours ago from web in reply to CalRobert AwOLonline
Philippe Delmar
then I replied:
@AwOLonline Click on the link in AwOLonline/status/4776219208 > "in reply to EllenAMuth" > EllenAMuth/status/4708100730 > 404 Not Found about 14 hours ago from API in reply to AwOLonline
CalRobert
Robert Maas
then AwOLonline replied:
@CalRobert Can't find the link. Do you recall the content of the tweet? about 5 hours ago from web in reply to CalRobert
AwOLonline
Philippe Delmar
then I copied the whole conversation here (above) and will shortly post a link to it.

momento24.com typo:
The most destructive eruption came in February [1841] ...
All other sources say it was 1814. I tried to post a comment about the typo, but it caused a 500 internal server error, because lynx on VT100 term on Unix don't have JavaScript.

Trying to post to Twitter gives:

                                          503 Service Temporarily Unavailable
                        Service Temporarily Unavailable 
   
   The server is temporarily unable to service your request due to
   maintenance downtime or capacity problems. Please try again later.
   
   Additionally, a 403 Forbidden error was encountered while trying to use
   an ErrorDocument to handle the request.

America's biggest economic problem? We're all broke. Literally. ; New Deal 2.0
Typo: which in turn could further [weaker] house prices
(should read "weaken")
.
I keyed-in this comment:
Let each foreclosure go to auction, under condition that anyone winning the auction *must* rent back to original mortgage holders. USA government always make a reasonable bid, and in most cases the USA government will in fact make the highest bid and win the aution, so USA will hire property managers to handle actual rent paymets, but USA will be formal owner/landlord. but when I clicked the HTML-form-submit button:

Link that you currently have selected
   Linkname: Submit Comment
     Method: POST
    Enctype: application/x-www-form-urlencoded
     Action: http://www.newdeal20.org/wp-comments-post.php

http://www.huffingtonpost.com/marshall-auerback/debtors-revolt_b_256548.html
[Unitl] greedy and corrupt Banks are put in their place

.
http://www.quizjungle.com/?act_id=3&gid=279&rg_id=1001&pid=90&cid=1478&aid=5392&sid=35491-72120313
"Vent [abut] your problems."
(typo)
.
http://whyslopes.com/etc/solvingLinearEquations/BeyondStickDiagrams03.html
"And after you written a solution and checked it, you may see a simpler or quicker way."
(missing word "have")

The city of Sunnyvale generates bad HTML:
sunnyvale.ca.gov: 13 Errors, 9 warning(s)
bit.ly/dBnrfW -> Solid+Waste+and+Recycling: 124 Errors, 22 warning(s) bit.ly/b3KXsa -> Earth+Day+Event.htm: 75 Errors, 15 warning(s)

cutestkidever said:
The easy link is going to the spammer's profile and clicking "report for spam."
then I replied:
I'm looking at http://twitter.com/RyksAffiliates which is the spammer's timeline. How to get to spammer's profile?
then cutestkidever replied:
Timeline = Profile for my purposes.
then I replied:
I'm looking at http://twitter.com/RyksAffiliates where neither "report" nor "spam" appears anywhere on the page.
then cutestkidever replied:
In the right column, it's under "Actions" and above "Following."
Here's the HTML being served when I fetch http://twitter.com/RyksAffiliates, and here's a save (text dump) of exactly what I see on-screen.
I really don't see what you're referring to, anywhere on that page.

Problems with the computers in lab (2009.Sep.16):
- Neither middle nor right computer has Mozilla Firefox anywhere on hard disk. (Left computer is currently busy so I can't check what's there.)
- Not only is clock three years wrong (year 2006) on right computer, but there's no way to set the date+time to fix the problem, so there's no way to log into Web sites that require correct certificate, and any file edited/created would have wrong date so basically it's not possible to do any real work there.
.
Complaint (2009.Sep.24): On these public computers, whenever I type a password to a Web site, it interrupts me to suggest that I authorize it to remember the password and thereby reveal it to anyone else who uses the computer after me the same day.