Beta Config (Sep 19 2010)

[sidki3003]

(Sep. 26, 2010 10:37 PM)turtle Wrote:  

(Sep. 26, 2010 08:48 PM)sidki3003 Wrote:  Maybe they fix it before IE9 is going gold.


You could try to do a little troubleshooting yourself:
- Is the URL (i.e. web address) which shows up in the address bar for "not found" pages actually the one you wanted to visit (e.g. http://www.google.com/)
- You can temporarily turn off almost all advanced mode filters by entering "l.4" in Proxomitron's filter list windows (at the bottom, w/o quotes), unchecking the filters you see. A good start is the header filters window. You're probably a professional troubleshooter by then ( ) and are possibly able to narrow it down to a single offending "l.4" filter.

I tried to have a look at it, but came across too many obstacles. The IE9 beta doesn't install in a WinXP image. I have only 64bit Win7 images, which don't run in my VMWare, which happens to be installed on 32bit Win2K. The only 32bit Vista image i have doesn't seem to be genuine enough to allow for Vista SP2, which in turn is a prerequisite for IE9. ... ... ...

Thanks, I'll give it a try. Glad it works in 1.4 mode!

Found it -- "JS "location.replace" to Link 09.06.04 (fail) [pr sd] (d.1 l.4)"

Found a 32bit Win7 image, installed IE9 beta in a VM.

If i navigate for instance to http://prxbx.com/forums/ in advanced mode, the address bar (upper left text box) shows:

Code:

http://prxbx.com/forums/function%20(l)%20%7B%20%20%20%20if%20(prxO.oiLoaded%20%3C%203)%20%7B%20prxO.inCatchLocURL%20=%20l;%20return%20false;%20%7D%20%20%20%20window.location.replace%20=%20PrxRLreplace;%20%20%20%20return%20location.replace(l);%20%20%7D

This URL doesn't exist and results in a "not found" (404) error.

The respective filter is just activating a JS function (after testing for IE/Opera), which never worked in Gecko browsers (e.g. Firefox) and results in a redirection loop in Konqueror (default browser in some Linux distros).

Question is what to do with that function. Fix it or remove it?
Since - as a Firefox user - i can't verify its correct functioning in everyday browsing, i'd say remove it in the next release config.


Another thing is that most pages with an XHTML doctype, such as these forum pages or my site, trigger "IE9 document mode". in this mode a couple of things are broken. For instance, the stylesheet for the Proxomitron menu isn't applied.
Trying to adjust the config makes no sense here, IMO, since IE9 is just a beta, and such things get hopefully fixed by MS later.

Only work-around i found thus far is to press the ALT key, select "Tools", select "Compatibility View Settings", and check "Display all websites in Compatibility View". This forces IE7 standards. Actually, IE8 standards would work too (just not IE9 XHTML).

If someone knows a way to force "IE8 document mode", let me know!
The latter can be manually selected from within Developer Tools (F12).

[whenever]

(Oct. 02, 2010 03:30 PM)sidki3003 Wrote:  If someone knows a way to force "IE8 document mode", let me know!

Does this help? http://msdn.microsoft.com/en-us/library/...px#SetMode

[wammie]

(Oct. 02, 2010 10:57 AM)sidki3003 Wrote:  People, read the documents! They were written to answer common questions, give background info, and to explain not-so-easy to understand aspects of the config.

With all due respect, I have "RTFM" and consider myself an intermediate user. So again, I suggest to consider the following points:
1) The default configuration provided by Scott Lemmon is this:

Code:

UseSSLeay = FALSE

2) The Sidki beta configuration is this:

Code:

UseSSLeay = TRUE

3) Sidki's own Readme file states this:

Code:

*******************************************************************************
Installation
*******************************************************************************
...
- Load sidki_xxxx-xx-xx.ptron by using Proxomitron's main window or the tray
  menu, and you're ready to go!  But wait...


- Open Proxomitron's Main Window and push the "Headers" button.
  There are five config modes, starting with "! |||||||||||| 1.1 Minimal Mode". These modes are described in "Config_Control.txt".

  In short:
  If you are new to Proxomitron, you may want to uncheck "Standard Mode" and check "Light Mode" or "Minimal Mode".

...and this:

Code:

*************************** Below Steps Are Optional **************************
...
- If you want Proxomitron to filter secure pages and/or want to activate the
  "Use Half-SSL" option in the "Header Filters" window, make sure that the SSL
  DLLs - SSLeay32.dll and Libeay32.dll - are installed on your system.

4) I/we understand the SSLEAY part of the beta config is a program option. However, the config file as a whole is altered from the original, so this line must also be considered as part of the beta software offering.

5) The process being observed is this:
a. A new user may follow the Readme file, simply extracting the beta config into Proxomitron's directory and using the light mode.
b. The "optional" (as per the Sidki Readme) step of installing the required SSL dll files, by virtue of being an option, may not be executed.
c. Error messages concerning the SSL configuration may be generated if the user had not previously installed the correct SSL files in the correct places.
d. The beta tester reporting this observation, after having used Sidki's configuration sets for years, and providing feedback multiple times in an effort to help improve the software, is being told to RTFM.

There was no "general flaw" in concept assumed, merely a reporting of an apparent flaw in a new beta config. If there is this much resistance to reporting troubles with beta configs, I'll have to reconsider devoting my time, thought, and participation as a beta tester.

[JJoe]

(Oct. 02, 2010 04:07 PM)wammie Wrote:  c. Error messages concerning the SSL configuration may be generated if the user had not previously installed the correct SSL files in the correct places.

Wammie, have you set the browser to use the Proxomitron for HTTPS?
Do you see the errors when the browser is not set to use the Proxomitron for HTTPS?

Please read http://prxbx.com/forums/showthread.php?tid=1670

[sidki3003]

(Oct. 02, 2010 03:59 PM)whenever Wrote:  

(Oct. 02, 2010 03:30 PM)sidki3003 Wrote:  If someone knows a way to force "IE8 document mode", let me know!

Does this help? http://msdn.microsoft.com/en-us/library/...px#SetMode

Probably not.
If we would inject an IE8 meta tag (or header) into every HTML document, we wouldn't get those Quirks-Mode pages right that would require a lower (IE7) mode.
If i understood that correctly, "Compatibility View Settings" allows for setting a "highest" mode, which unfortunately is IE7, not IE8.

[sidki3003]

(Oct. 02, 2010 04:07 PM)wammie Wrote:  d. The beta tester reporting this observation, after having used Sidki's configuration sets for years, and providing feedback multiple times in an effort to help improve the software, is being told to RTFM.

Again, there is no difference in SSL handling between this beta and previous sidki configs:
- "UseSSLeay = TRUE" has been set since my first public config (2003-11-19).
- If you read Abbreviations.txt, which is part of the documents that come with each config, you'll understand that the SSL filters you posted as being active are actually inactive.
- On the contrary, this beta comes with a new list "Bypass-SSL.ptxt" which allows you to add secure sites, where Proxomitron should go into SSL-bypass mode right from the beginning.

[wammie]

(Oct. 02, 2010 04:33 PM)JJoe Wrote:  Wammie, have you set the browser to use the Proxomitron for HTTPS?
Do you see the errors when the browser is not set to use the Proxomitron for HTTPS?

No, my browser was not set to use the Proxomitron for HTTPS and was not set that way in testing.

(Oct. 02, 2010 04:33 PM)JJoe Wrote:  Please read http://prxbx.com/forums/showthread.php?tid=1670

Good stuff. What happened is that I had incompatible DLL files in the Proxomitron directory. However, in my previous configuration, UseSSLeay was set to false. Loading up the new beta config activated the UseSSLeay switch and caused the errors.

[JJoe]

(Oct. 02, 2010 05:59 PM)wammie Wrote:  What happened is that I had incompatible DLL files in the Proxomitron directory. However, in my previous configuration, UseSSLeay was set to false. Loading up the new beta config activated the UseSSLeay switch and caused the errors.

Glad you figured it out.
Now that you have been given an education and some

What is your opinion? "UseSSLeay = TRUE" or "UseSSLeay = FALSE" as set default?

Consider:
The user must decide or choose to filter https by sending it to the Proxomitron.
The user must add the SSL DLLs to the Proxomitron.
While there may be some https filtering without the SSL DLLs, it's probably not very useful in sidki's sets and those who know about it will probably know to reset 'UseSSLeay'.

[wammie]

(Oct. 02, 2010 07:15 PM)JJoe Wrote:  What is your opinion? "UseSSLeay = TRUE" or "UseSSLeay = FALSE" as set default?

JJoe, it is not clear to whom this question was directed. In my case the errors persist even after the incorrect DLL files are removed from the Proxomitron directory, meaning I probably represent a typical user who has not intentionally installed the specific DLL files as needed by Proxomitron.

Unless I'm still missing something, I perceive the choice boils down to these scenarios:

1) With the TRUE switch, many typical users, new or otherwise, wanting to just install the Sidki configuration file and be "ready to go!" as indicated in the Sidki Readme, not having already installed the proper SSL DLLs and choosing SSL filtering in the browser (an optional procedure), or wanting SSL filtering in the first place, will encounter these annoying DLL errors and have to discover for themselves how to rectify the trouble. These "first use" errors, in my opinion, are a turn off for the general user and may lower the user base.

2) With the FALSE switch, the typical user will not be bothered with these errors. Those wanting to explore the optional SSL option will seek out what needs to be done and carry through.

As it is, with the default 2010-09-19 switch being TRUE, users are forced at the onset to deal with a supposedly optional feature.

[sidki3003]

(Oct. 02, 2010 07:15 PM)JJoe Wrote:  What is your opinion? "UseSSLeay = TRUE" or "UseSSLeay = FALSE" as set default?

Taking this discussion out of the beta thread, since it isn't beta related at all, turning it into a public poll, would have been the better decision, IMO, but okay. I'm setting UseSSLeay to FALSE with the next release.

After the next release things will change. I'll place a notification in the "About..." thread, in time.

[JJoe]

(Oct. 02, 2010 10:27 PM)wammie Wrote:  

(Oct. 02, 2010 07:15 PM)JJoe Wrote:  What is your opinion? "UseSSLeay = TRUE" or "UseSSLeay = FALSE" as set default?

JJoe, it is not clear to whom this question was directed.

For you and only you, "Glad you..", "Now that you", "What is your". I've been trying to lead you down the path to "why I think "UseSSLeay = TRUE" is the proper setting". This path has been traveled before.

But your reports do worry me. I want to know why you are seeing what you are seeing. I may have to reevaluate my position.

Should I change my mind, I won't be asking for a blankety-blank poll.

(Oct. 02, 2010 10:27 PM)wammie Wrote:  In my case the errors persist even after the incorrect DLL files are removed from the Proxomitron directory, meaning I probably represent a typical user who has not intentionally installed the specific DLL files as needed by Proxomitron.

Unless I'm still missing something,

Did you restart the Proxomitron after you removed the DLLs?

I have: exited the Proxomitron, removed the DLLs, set the browser to not use the Proxomitron for https, restarted the Proxomitron (with "UseSSLeay = TRUE"), attempted to load https://mail.yahoo.com/, and did not see the errors.

Would you, wammie, try the same and post your results.

The Proxomitron's log window should show no https traffic from the browser during this test since the browser won't be using the Proxomitron for https.

(Oct. 02, 2010 10:27 PM)wammie Wrote:  1) With the TRUE switch, many typical users, new or otherwise, wanting to just install the Sidki configuration file and be "ready to go!" as indicated in the Sidki Readme, not having already installed the proper SSL DLLs and choosing SSL filtering in the browser (an optional procedure), or wanting SSL filtering in the first place, will encounter these annoying DLL errors and have to discover for themselves how to rectify the trouble. These "first use" errors, in my opinion, are a turn off for the general user and may lower the user base.

I believe that when the browser is not set to send https to the Proxomitron and Half-SSL is not used, the errors are not generated.

Users who want to filter https will correct errors.
Users who mistakenly set the browser to send https to the Proxomitron may be alerted to their mistake by the errors.

(Oct. 02, 2010 10:27 PM)wammie Wrote:  2) With the FALSE switch, the typical user will not be bothered with these errors. Those wanting to explore the optional SSL option will seek out what needs to be done and carry through.

Users who mistakenly set the browser to send https to the Proxomitron will not be alerted to their mistake by the errors.

[sidki3003]

The "Whois" Prox menu entry is pointing to http://samspade.org/whois , which is defunct since at least several weeks.

Whois alternatives i've found:
http://whois.domaintools.com/foo.com
http://www.dnsstuff.com/tools/whois/?ip=foo.com
http://network-tools.com/default.asp?pro...st=foo.com

I've opted for the first one. The respective entry for proxjs-x-menu.js is:

Code:

x[x.length] = beSelect; x[x.length] = "prxO.oInt.inMyPop('http://whois.domaintools.com/' + escape(q));\">Whois &#187; </a></div>";

If there is a better choice, please let me know.

[sidki3003]

Fixing a regression:

Code:

[Patterns]
Name = "<!>: Fix Comments     10.09.26 (multi) [sd] (d.2 l.3)"
Active = TRUE
Multi = TRUE
URL = "$TST(hCT=*html)(^$TST(keyword=*.(a_code|a_comm_f|i_level:[12]).*))"
Limit = 1024
Match = "<!--(^ -- >| <s(cript|tyle)|$TST(script=*))"
        "("
        "[ ]+-(---+ >|((-| --(^ >))+\1(-[ -]+?++{0,1} >)\0&$TST(\0=*--(? >| >))))"
        "|"
        "(([ ]+[^\r\n-]*)\1(-[ -]+?++{0,1} >)\0&$TST(\0=*--? >))"
        ")"
Replace = "<!--\1-->"

[apacallyps]

(Oct. 01, 2010 08:00 PM)Kye-U Wrote:  <steppingin>
Sidki is a valuable asset to the Proxomitron community... We all love him. For the future, I'd like to ask everyone to fully appreciate his work.

Well, I LOVE Sidki and I haven't even met him personally! I think what he's doing for the Proxomitron community is GREAT and as far as loving him, well, I REALLY do, that's absolutely true, but that's because I'm a born-again, Bible-believing, God-fearing, Christian; (See John 3:3). I believe about 6,000 years ago God created the heavens and the earth. 4,400 years ago there was a word-wide flood that destroyed the world and dropped everybody’s property value to zero. 2,000 years ago Jesus came, died on the cross, and here we are today. Here we are today waiting for the Lord to come back in about 5 minutes. The Bible timeline not only explains how the dinosaurs died, but how so many fossils were preserved. A gigantic flood with tons of dirt and water covering you instantly would result in a fossil.

"Whereby the world then was, being overflowed with water perished." 2 Peter 3:3-6

So, when I say I love Sidki and everyone else in this forum, I really do. I pray that all of you trust Christ as your Savior and get saved because this universe had a Designer and you will answer to Him real soon. Be ready for that. Thank you so much.

PS, I'm not trying to get off topic or start a debate here, I'm just responding to Kye-U's great post and stating my position on this. Kye-U is right and a great friend to Sidki. God bless you all, and have a great rest of the day. :-)

[wammie]

(Oct. 01, 2010 08:00 PM)Kye-U Wrote:  <steppingin>
Sidki has done an excellent job with his config, not only in the sense of it being effective at cleaning up the web, but also making it easy to configure. The nature of Proxomitron and Sidki's config (or any config) allows you to personalize it and make it your own; his especially, where you simply need to tick/untick Header filters.

Sidki is a valuable asset to the Proxomitron community; he's doing all of this work for free (side-note: Sidki, when are you going to set up a PayPal account?! ). We all love him. For the future, I'd like to ask everyone to fully appreciate his work. I'm not saying that users shouldn't point out glitches/bugs or make suggestions, but dictating what should/should not be done goes against showing appreciation.
</steppingin>

OK, use of the word "ought" in my previous post wasn't the best choice. The intent wasn't at all to give a directive, as I've already stated. The intent was to say "it would be prudent to see if something in this new beta config is causing new errors," with an implied belief that the software author would be the most qualified person to know. From a dictionary definition for "ought": used to express propriety, appropriateness, etc.

Logically, an equivalent statement "here's an apparent issue with beta software, the author's duty is to consider if this is the case," is only an affirmation of the whole concept of beta testing.

I have great respect and admiration for Sidki's work, as evidenced in many of my previous posts here! The suggestion of anything otherwise was misplaced. It was entirely unexpected that this request for an inquiry into a source cause would be understood as anything other than a helpful suggestion to correct a possible flaw.

I'm usually very careful selecting the words I use, as I'm a professional technical writer, and thus my keen attention to documentation. Please excuse this language "slip" as an honest lapse in word choice, and accept my apologies for any apparent inconsideration or change in tone from the usual respectful discourse in this forum.

So moving on, let's continue to focus on issues, filters, and config set improvements, rather than trying to characterize people as disrespectful or assuming certain levels of involvement or expertise. Negative accusations just set the stage for people to take a defensive posture.

Thanks once again, Sidki, for your efforts!