User talk:Sheldon Rampton/archive 2

question re: page
Hi Sheldon; HNY to thee, and keep this great project rolling. Question: it looks as if Daniel Pipes himself rewrote the page pertaining to him, and wondered if this should be reverted... Maybe some of the materials he has provided could be used, but i object to the (1) uncritical info that now appears on this objectionable character and (2) the links to the organizations he is affiliated with now point to these organizations instead of the internal D* ones... Keep up the good work Kind rgds PaulR

robots.txt problems again?
Hi Sheldon, can i please draw your attention to User:Bonzai/Google-index and Bonzai's comments on my user talk page: User talk:Neoconned. It looks like there may be a problem with SourceWatch's robots.txt file again... Cheers, --Neoconned 20:59, 13 Dec 2004 (EST)


 * Hi Sheldon, I noticed you put a robots.txt file. Thanks for the quick action. However it only contains 1 line and I think it would be better to change it to (without the space at the start of the line):

User-agent: * Disallow:
 * Just in case. Bonzai 04:29, 14 Dec 2004 (EST)


 * I noticed you also added the second line. Not sure if that extra line helped or not, but the good news is that the snippet for the homepage is showing again in Google's search results. Bonzai 18:53, 15 Dec 2004 (EST)

temporary communication
Hi Sheldon: I have no clue on how to revert a text to the previous one. it is just that these changes are wrong: (diff) (hist). . National Endowment for Democracy; 13:38. . 205.188.116.131 (Talk) (Founding of the NED) (diff) (hist). . National Endowment for Democracy; 13:37. . 205.188.116.131 (Talk) (Founding of the NED) NB: the name of these organizations has changed over time, and they even have gone back and forth with their acronyms. In the NED website it lists the names as previously shown.... Kind rgds PaulR

Upgrade
Hello Sheldon,

I just noticed the upgrade to the new wiki-software and now look of the site has been done. Some fine-tuning might still be needed but it looks good. Congratulations.

One point that might need some extra attention is the new URLs that are now used.

For example, the URL for Internet activism used to be:

http://www.SourceWatch.org/wiki.phtml?title=Internet_activism

but that is now:

http://www.SourceWatch.org/wiki.phtml/Internet_activism

The old-style URLs are still working, so links from search engines and other sites to pages on SourceWatch will result in the correct article being displayed. Having a URL without the '?' character is better to get pages in search engines after all, so in the long run that is an improvement too. At wikipedia even ".phtml" is no longer part of the URL, but that's not that important. What might be more of a problem is that pages that have inbound links (links from other sites to SourceWatch) will end up double in the search engines. For Google that means that the PageRank is divided over 2 identical pages with different URLs so SourceWatch-pages will be displayed lower in the search results. To avoid this, I will contact you later. By the way, as can be seen on User:Bonzai/Google-index, the number of SourceWatch-pages in Google's index is stabilizing (most likely at the old level) one month after fixing the robots.txt file problem. I don't have access to the site statistics, but according to Alexa (see 'Daily Traffic Trend Rank' at []) visitors to the site are more-or-less back to old levels.

Another point is that links to previously visited pages are now hard to see. A link to another page on SourceWatch is usually shown as blue text (unless it's an article that doesn't exist which is shown as red text). Visited articles are dark green (?) and look rather similar to the black text. Best Regards, Bonzai 17:51, 27 Nov 2004 (EST)

Encoding
Hello Sheldon,

Currently (most likely since the upgrade), the header of the pages contain the line:



This code informs the browser that the received page is in UTF-8 (Unicode) format.

Some older pages in Google's cache have the line:



So in the recent past the the encoding was changed from 'iso-8859-1' to 'utf-8'.

It looks however, that the pages served are still in 'iso-8859-1' format. If you go to the books-section of Michel Salomon, you will see the original (French) titles of his books are shown incorrectly.

In case you use Microsoft Internet Explorer and go to ALT-v (View) -- Encode and select Western European (ISO) to temporarily change the encoding for that page, you will see the French titles are then displayed correctly.

This problem will occur for any page with special characters, sometimes even in the title of the article like:
 * C.J.F. B&ouml;ttcher
 * Daniel A. D'Aniello
 * Gerhard St&ouml;hrer
 * Here's the rest of the list
 * International Women's Democrat Union
 * Guant&aacute;namo
 * Gonzalo S&aacute;nchez de Lozada
 * etc.

I hope that even if you don't understand much about encoding, you have enough information to solve this problem. If not, please let me know. Bonzai 19:46, 30 Nov 2004 (EST)

Stub template
Thanks for the advice on the skin problem. My SourceWatch has now been returned to modernity. I have taken the liberty of creating a stub template: Template:stub. Would it be possible for someone to lock this page? Laurence

Footer
I was informed on my talk-page that the footer has a typo ("tog et" should be "to get"). I have no idea how to correct this (maybe I don't even have the needed permissions anyway), so I can only forward the request to correct this small mistake. Bonzai 04:03, 2 Dec 2004 (EST)

Editing
Since the update of the software, I have a problem with the edit box. No matter what value I enter at Preferences -- Editing -- Columns the lines are to long to fit in my screen. Changing the Rows however, works as expected. I will do some more trial & error to find a workaround, but I thought you might need to be informed. Bonzai 04:03, 2 Dec 2004 (EST)


 * I have been having the same problem. Thought changing the column nos had fixed it but was wrong. Now other users have had the same problem. I'm using IE Version 6 - what about you Bonzai? --144.131.118.231 04:18, 2 Dec 2004 (EST)


 * On the PC I usually use with IE version 6/Windows 2000-prof., there is a problem. On an older PC with IE version 5/Windows98 I had no problem (both as user:Bonzai). So somehow it depends on the browser, OS, and/or cookie. Bonzai 19:00, 2 Dec 2004 (EST)


 * I logged out and editing (without saving) went correctly on IE 6/Windows 2000 prof. Now I try again (logged in) and problem looks solved. Not sure if some system setting has been changed by Sheldon or so in the last few minutes, but my editing problem seems to be over now. Bonzai 19:12, 2 Dec 2004 (EST)


 * I did change a setting but didn't have time to write a note about it. If that's fixed the problem for now, great. FYI, my experience dealing with this and some other problems have convinced me that IE 6 for Windows is a bug-ridden piece of typical Microsoft junk. Before launching the upgrade publicly, I checked it out on half a dozen different web browsers including Safari, Firefox, Mozilla, Opera and IE 5/Mac OS X. It displayed fine on all of them, but there were several problems with IE 6. Maybe the day will come at some point in the future when Bill Gates manages to sell a single piece of software that isn't a disgrace, but that day hasn't arrived yet. (NOW EXITING RANT MODE) --Sheldon Rampton 00:04, 3 Dec 2004 (EST)

Blocking Users
Would you have an idea of what Maynard would consider a positive contribution?

Let's start with how his contribution to my Talk page is of benefit? Weide 17:41, 4 Dec 2004 (CST)

New Format Messing Up Google Results
The new format is causing Google to regard the SourceWatch boilerplate text at the top of the pages as content. As a result, Google usually no longer shows the first lines of the article in the description part of a Google search. It's important to have a meaningful bit of text show, since that is often how people decide on whether or not to click through to the article. Since Google heavily weights the initial text in its ranking algorithm, this is also adversely affecting SourceWatch's Google page ranks. For instance, the article on Richard Perle got bumped from first to third on the Google search for "perle". Do that search and note how meaningful text has almost disappeared from the description, and been replaced by uninformative (about Perle, anyway) boilerplate. --Paul


 * Perhaps Google, and other bots, could be served a customized "skin", or no skin. This suggestion could appropriately go all the way to MediaWiki it seems, if necessary. --Maynard


 * Thanks for pointing this out. It will take probably take me awhile to figure out the best way to deal with this. Once we've finished running the current survey, that text will disappear from the top of the page, and that alone will help somewhat. --Sheldon Rampton 17:35, 5 Dec 2004 (EST)


 * I agree with Paul that the text at the top of the page is important to search engines like Google. Serving special pages for search engines is against Google's Webmaster Guidelines:
 * &bull; Make pages for users, not for search engines. Don't deceive your users, or present different content to search engines than you display to users.
 * So that's not the solution. Once the current survey text is gone, I will look at it to see if more needs to be done. Bonzai 00:48, 6 Dec 2004 (EST)

Disambigous Pages
Hi Sheldon. I don't understand know how to do this and rather than mess things up, I wanted to ask you how I should handle the following: Control Risks Group's CFO is Michael Martin, a name that already exists as a SourceWatch article. They are not the same person, as far as I can see, and I can only find little to no real info on CRG's Michael Martin, so I can't really create much of a page for him. but anyone who who clicks on his name when visiting CRG will get the wrong info for their CFO. Any suggestions? Thank you. User talk:Spacegrit

Thanks for the idea. I seperated the two by title, moving the speaker to Hon. Michael Martin.

Error on 'Maintenance - Disambiguation pages'
Hello Sheldon, just for your information.

After reading the previous topic, I added 'Michael Martin' to the Links to disambiguating pages and tried the Maintenance-'Disambiguation pages' to check it. Unfortunately I got an error:
 * "wfSpecialDisambiguation is broken. Link tables have changed... "

The function used to work nicely for quickly getting all the links to Disambiguation pages (usually those links should go to one of the articles mentioned on a disambiguation page). It is not urgent and it seems this is a known problem when upgrading from 1.2.x to 1.3.x and the problem should be solved in version 1.4 so nothing really to worry about. Just to keep you informed. Bonzai 20:04, 6 Dec 2004 (EST)
 * FYI, I am just getting a from-scratch 1.3.9 installation up and running (G5mac.us) and I also appear to be having this issue. Does this make sense, or is it more likely that I've broken something in getting this system up and running? -Matt McCabe [01/03/2005]

Fair Use?
Hi Sheldon. Sorry to bother you again. Regarding pay sites like Hoovers and intellectual property rights and copyright laws, can I place information here, in my own words, that is available and obtained from those sites, information that is not available elsewhere? Does this fall under fair use laws, or do they actually own the right to display that information. After reading their copyright page and the fair use page here, I'm not too clear what the parameters are. Any clarifications would be helpful. User talk: Spacegrit

Gracias........ That answers a bit.

robots.txt missing after name change to sourcewatch
Hi Sheldon,

I like the new name! (sourcewatch) BUT robots.txt is now missing again.


 * Ie. http://www.sourcewatch.org/robots.txt is missing
 * and http://www.disinfopedia.org/robots.txt is also missing

Sourcewatch's google rankings will crash again, if this is not sorted out ASAP...

--Neoconned 11:11, 22 Jan 2005 (EST)


 * Oh my god, you're right. Of all the details to forget...OK, I fixed it. Thanks for the catch. --Sheldon Rampton 12:27, 22 Jan 2005 (EST)


 * No problem. I really do like the new name... --Neoconned 06:05, 23 Jan 2005 (EST)

Diacritics mangled by database-wide replacement of Disinfopedia -> Sourcewatch?
Hi Sheldon, I think the name change may possibly have created another gremlin... :(( I notice that PaulR has been sorting out single quotes that have become question marks. Eg:
 * see the mention of Shannon Oâ€™Connell in NDI Board of Directors & Senior Advisory Committee (2004)
 * Looking in the history for this page, I see that O'Connell appears as Shannon O?Connell for ALL previous revisions until PaulR fixed it today.
 * Yet, looking in Google's cached copy of the page, I see that as at 16 Jan 2005, O'Connell appeared correctly as Shannon O'Connell in the article.
 * Therefore, I have a horrible feeling that there is a possibility that something in the name change operation messed up at least some of the diacritics in the database, on a db-wide basis.

--Neoconned 19:50, 23 Jan 2005 (EST)


 * That looks like the same problem as mentioned above in the section 'Encoding'. Some setting for 'iso-8859-1' to 'utf-8' did the trick last time. Bonzai 06:26, 24 Jan 2005 (EST)


 * OK, I think I've got it fixed again. The problem arose because in the transition from Disinfopedia to SourceWatch I also upgraded the MediaWiki software from version 1.3.7 to 1.3.9, and some settings got lost in the process. --Sheldon Rampton 00:50, 25 Jan 2005 (EST)

Negative -- the "?" appear all over the place, and some of the html accented letter have been mangled also... PaulR 05:05, 25 Jan 2005 (EST)


 * Can you give some examples of pages where this is happening? --Sheldon Rampton 11:46, 25 Jan 2005 (EST)


 * Hi Sheldon, yes I'm afraid there are still problems:
 * As pointed out by User:Spacegrit, Private Military Corporations has problems in its first para (see any recent previous version as Spacegrit has fixed it now).
 * In the Bill Moyes article, the pound sterling symbol (£) has been replaced by a capital-A-circumflex-plus-pound-symbol (Â£).
 * Cheers, --Neoconned 18:31, 25 Jan 2005 (EST)


 * That pound sterling symbol was added in the Bill Moyes article on Jan. 23, 2005 (11:18). At that time the browser of the user who was editing a page got from the header that the data was in Unicode (UTF-8) while it wasn't. So special characters edited after switching to version 1.3.9 and before restoring the setting by Sheldon Rampton might now be incorrect. Bonzai 18:47, 25 Jan 2005 (EST)


 * Yes, that user was me. And the same explanation applies to the Private Military Corporations example. So, problem sorted. Thanks for the explanation, --Neoconned 21:03, 27 Jan 2005 (EST)

sourcewatch might bring random visits
it isn't helpful really, but expect an occasional random visit under this domain name from persons familiar with and/or looking for open-source software information. I like the name, but if i'd seen it first out of context, i would have thought it to be related to coding. There are plusses and minuses with an added geekiness factor that may be inadvertantly derived from this.

--Hugh Manatee 01:37, 25 Jan 2005 (EST)

redirect issue
Sheldon, do you mind determining why a search on MEMRI doesnt land one in the destination page? NB: i recently did a MOVE PAGE, and it seems that the original MEMRI redirect didnt accompany.

In advance, thx PaulR


 * Hi, Paul. Sorry for not getting back to you sooner. I did searches for both "memri" and "Middle East Media Research Institute," and both search returned the expected results. --Sheldon Rampton 01:46, 28 Jan 2005 (EST)

Hi Sheldon; This is what i get: Search "MEMRI" >> MEMRI: Middle East Media Research Institute (Redirected from MEMRI)

REDIRECT Middle East Media Research Institute << Kind rgds


 * OK, I get it now. The problem is that the article named "MEMRI" was a redirect to an article named "MEMRI: Middle East Media Research Institute," which in turn was a redirect to the article named "Middle East Media Research Institute." The MediaWiki software doesn't do multiple redirects (which is a good thing, because then there would be a potential infinite loop problem if some created a redirect pages that point back to each other). It therefore redirected from the first article to the second one in the chain, but didn't go the rest of the way to the article you actually wanted. I fixed it by editing "MEMRI" so that it redirects straight to the intended target article. --Sheldon Rampton 11:07, 31 Jan 2005 (EST)

Hi Sheldon; How were you able to reach that page... that was what was the sticking point. Kind rgds PaulR


 * When you arrive at a page through a redirect, you'll see a line underneath the page title that says, "Redirected through (something)" -- in this case, "Redirected through MEMRI." You can click on that, and it will take you to the page that does the redirection, which you can then edit. --Sheldon Rampton 15:31, 31 Jan 2005 (EST)