Random String Search on Login - About xda-developers.com

Whenever I log into the site I end up on a search result page with a random string.
Generally looks like: "98ei10dj3j50dj55kal3l59zmn3z8832i", or similar. The page says "no result found", or equivalent. It's distinctly different than the Algolia search page or results.
Any idea? Seemed like the site was assigning me a random string and then searching by it.

artician said:
Whenever I log into the site I end up on a search result page with a random string. Generally looks like: "98ei10dj3j50dj55kal3l59zmn3z8832i", or similar. The page says "no result found", or equivalent. It's distinctly different than the Algolia search page or results. Any idea? Seemed like the site was assigning me a random string and then searching by it.
Click to expand...
Click to collapse
Maybe @MikeChannon can help you out with this or refer another Administrator who does.
I, personally, have never heard of such a thing purely by just logging in but, just because I've never heard of it doesn't mean that it doesn't happen.
Have you tried different browsers on the PC and/or the Android to see if it's consistent?
~~~~~~~~~~~~~~~
I DO NOT provide support via PM unless asked/requested by myself. PLEASE keep it in the threads where everyone can share.

It happens on Pale Moon (latest), in Win7x64 and Win8.1.
It's always a 32 character string.
The string changes between sessions.

Thank you!

Related

opera mini4 &5 keep switching to mobile sites instead of full

When I visit websites that also have sites optimized for mobile phones like nu.nl and cracked.com i am direcedt to those. But I want the original site. I can't find a way to change it. I turned mobile view off.
taarmen said:
When I visit websites that also have sites optimized for mobile phones like nu.nl and cracked.com i am direcedt to those. But I want the original site. I can't find a way to change it. I turned mobile view off.
Click to expand...
Click to collapse
You can't change it. It's not caused by Opera. It's the site that detects that you are using Opera Mini, and decides that you are to get a mobile site instead of the full site. Complain to the site.
Type opera:config in your address bar and poke around with the settings. It's in there somewhere.
EyeAmRubber said:
You can't change it. It's not caused by Opera. It's the site that detects that you are using Opera Mini, and decides that you are to get a mobile site instead of the full site.
Click to expand...
Click to collapse
And where do you think the site gets this information from? From Opera! - sending the wrong string as user agent in it's HTTP-requests.
EyeAmRubber said:
Complain to the site.
Click to expand...
Click to collapse
No. The site does deliver correctly what the browser requested. It's definitely Opera sending the wrong request.
SE-X1 said:
And where do you think the site gets this information from? From Opera! - sending the wrong string as user agent in it's HTTP-requests.
Click to expand...
Click to collapse
No, it's sending the right string. It's telling the site which browser it is. The wrong thing would be to lie about it.
No. The site does deliver correctly what the browser requested. It's definitely Opera sending the wrong request.
Click to expand...
Click to collapse
No, Opera is just telling the site that it's Opera. The site is choosing what it does when it encounters Opera.
If Opera changed its user agent string to become unrecognizable, browser statistics would be messed up, and sites relying on browsers to be truthful would break.
EyeAmRubber said:
No, it's sending the right string. It's telling the site which browser it is. The wrong thing would be to lie about it.
No, Opera is just telling the site that it's Opera. The site is choosing what it does when it encounters Opera.
If Opera changed its user agent string to become unrecognizable, browser statistics would be messed up, and sites relying on browsers to be truthful would break.
Click to expand...
Click to collapse
So this anoying problem can be fixed by just changing one string or telling 3476587653745344378 sites to change the way they handle requests.
I think it wouldn't even mess up statistics. There are more reliable methods to identify the browser than by using the user agent. The desktop version for example is capable of identifying itself as IE or Firefox and it does not mess up statistics.
I don't get mobile versions of pages shown in Opera mobile so this should be possible in Mini too.
SE-X1 said:
So this anoying problem can be fixed by just changing one string or telling 3476587653745344378 sites to change the way they handle requests.
Click to expand...
Click to collapse
Changing the string would break 3476587653745344378^10 other sites. And would remove it completely from browser stats. Which would get all the whiners going on about how no one is using Opera again.
I think it wouldn't even mess up statistics. There are more reliable methods to identify the browser than by using the user agent.
Click to expand...
Click to collapse
Such as?
The desktop version for example is capable of identifying itself as IE or Firefox and it does not mess up statistics.
Click to expand...
Click to collapse
It identifies as Opera by default. When it identifies as something else, it is not counted as Opera, but as a different browser.
I don't get mobile versions of pages shown in Opera mobile so this should be possible in Mini too.
Click to expand...
Click to collapse
Of course it's possible in Mini. This has got nothing to do with Opera Mini's capabilities. It's the site that chooses to send you to a mobile site.
You can argue as much as you want, it doesn't change anythink about the fact, that O-mini does not show the expected content although it could be very easily be fixed. That's ultra-anoying on any high-res-device.
There's even a checkbox in Opera's options to chose if you want to see mobile-versions if available or not, but it's just ignored.
I'm pretty sure the string can be altered in a way that does both, full-size-content AND right statistics.
Check showip.com with Desktop-Opera. It shows your user agent. It changes when you set Opera to claim it's IE or FF, but they all still include "Opera" +version in that string. Stats can still identify it correctly no matter what you select.
SE-X1 said:
You can argue as much as you want, it doesn't change anythink about the fact, that O-mini does not show the expected content although it could be very easily be fixed.
Click to expand...
Click to collapse
I'm not arguing. I'm pointing out the fact that Opera Mini is simply showing the content that's being sent to it. I'm pointing out the fact that it's the site that's doing this. You are arguing because you evidently don't understand the subject matter.
There's even a checkbox in Opera's options to chose if you want to see mobile-versions if available or not, but it's just ignored.
Click to expand...
Click to collapse
No, that checkbox actually controls what Opera Mini does. It changes the way Opera Mini handles a page. This is completely different from the problem you are referring to, which has to do with what the site does when it detects that you are using Opera Mini.
I'm pretty sure the string can be altered in a way that does both, full-size-content AND right statistics.
Click to expand...
Click to collapse
No, because the sites that are sending Opera Mini to mobile pages are doing so specifically for Opera Mini. If Opera Mini changes its UA string, those sites will simply adapt to the new string, and send that to a mobile page as well. You will have gotten nowhere, except making other people's life harder.
Check showip.com with Desktop-Opera. It shows your user agent. It changes when you set Opera to claim it's IE or FF, but they all still include "Opera" +version in that string. Stats can still identify it correctly no matter what you select.
Click to expand...
Click to collapse
Again, the sites that send Opera Mini to a mobile version are specifically looking for Opera Mini. If you change the string to include other browsers, the page will still specifically look for Opera Mini. It would solve nothing, as explained above.
Stats can identify it correctly if you don't remove Opera Mini from the string. But you seemed to suggest that they simply hide Opera Mini completely, in which case those stats sites would not be able to detect it.
Finally, what more reliable methods to identify the browser than the user agent string are there?
This is getting too much blah blah for a trivial thing. It could be fixed but it's not --> not good
It does work with Opera mobile, that's proof enough that it could work with mini too, no matter how it works.
You are right about the checkbox "mobile view", that's "Small screen rendering mode" which is something else... misleading label.
Im not the expert that can tell you how exactly the user Agent string has to look like, but it's obvious that it could work with the right one.
(Forget about "more reliable methods", seams like it does all work fine with user agents only)
SE-X1 said:
This is getting too much blah blah for a trivial thing. It could be fixed but it's not --> not good
Click to expand...
Click to collapse
It needs to be fixed by the site. It is not Opera Mini which decides which page it's being sent.
It does work with Opera mobile, that's proof enough that it could work with mini too, no matter how it works.
Click to expand...
Click to collapse
Opera Mini and Opera Mobile have different UA strings. Again, it's got nothing to do with "working with Opera Mini". This isn't caused by Opera Mini. It's caused by sites that detect that you are using Opera Mini.
Im not the expert that can tell you how exactly the user Agent string has to look like, but it's obvious that it could work with the right one.
Click to expand...
Click to collapse
I'm telling you that it couldn't. These sites that send Opera Mini mobile content do so because they chose to specifically do so for whatever reason. Changing the UA string would just piss these people off, and update their scripts to send mobile pages to the new UA string as well.
It's getting boring. Have fun with ugly unreadable pages while I surf the web how it's supposed to look.
You admit that you were mistaken then, I presume?
EyeAmRubber said:
You admit that you were mistaken then, I presume?
Click to expand...
Click to collapse
No, you're simply wrong, I just gave up convining you.
So I am "simply wrong" even though you made all the false assertions, such as "more reliable methods" (showing that you have no idea what you are talking about).
Telling me that I'm wrong when all your arguments so far have been devastated is not exactly convincing.
EyeAmRubber said:
So I am "simply wrong" even though you made all the false assertions, such as "more reliable methods" (showing that you have no idea what you are talking about).
Telling me that I'm wrong when all your arguments so far have been devastated is not exactly convincing.
Click to expand...
Click to collapse
I was wrong in a detail like I already said above. But that doesn't change anything about the rest. The problem can definitely be fixed in the Software, that's a provable fact I'm not going to discuss anymore. Saying the software is ok, just the rest of the world has to change to become compatible is just nonsense because it will not fix the problem because it will never happen.
Besides the alternative view does make sense on low end devices so it has to be Opera to decide witch view to load depending on the resulution of the device or better by the settings the user made. And don't tell me again it's not possible, that's just nonsense because it's not and does already work with other browsers.
The "problem" can't be "fixed" by Opera without breaking it for everyone else.
The rest of the world doesn't need to change. You are the one who wants sites who specifically send Opera Mini users to a mobile site to change.
As I already explained, this is not about "Opera views", this is about what kind of content a site sends to Opera. It "works" in other browsers because the sites are not sending them to a mobile site.
Again: Changing the useragen string would break ****loads of websites, and Opera would disappear from the stats. That, or the change will achieve exactly nothing, because the sites that are sending Opera Mini users to a mobile site are specifically looking for Opera Mini in the first place.
You need to stop making assertions when you are clearly severely lacking in knowledge on the subject. It gets worse when you blame Opera for something the site is doing, and claim that it's in Opera's hands.
Sure it's the Site that sends different content, nobody stated anything else. But it depends on the request made by the browser! And when I say it works in other browsers I'm NOT talking about desktop browsers. I use Opera mobile and it does what i want although I got the same results as with Mini with default settings.
In Opera mobile you just go to the setting and set "Spoof UserAgent ID" to "2" to mask as Mozilla. That does still leave "Opera 9.7" mentioned in the User agent and is still identified in browser statistics correctly, but this way you allways get fullsize content without breaking anything. Sure thing the same can be done to any other browser easily.
Maybe you just try it yourself and then we talk again who's got a lack of knowledge.
It's so funny reading all the time what's impossible while everyone can download other mobile browsers and see with own eyes that it's not.
SE-X1 said:
Sure it's the Site that sends different content, nobody stated anything else. But it depends on the request made by the browser!
Click to expand...
Click to collapse
In what way, specifically? What kind of request?
And when I say it works in other browsers I'm NOT talking about desktop browsers. I use Opera mobile and it does what i want although I got the same results as with Mini with default settings.
Click to expand...
Click to collapse
Again, that is because other browsers have a different UA string, and the sites haven't started specifically sending those to a mobile page.
In Opera mobile you just go to the setting and set "Spoof UserAgent ID" to "2" to mask as Mozilla. That does still leave "Opera 9.7" mentioned in the User agent and is still identified in browser statistics correctly, but this way you allways get fullsize content without breaking anything.
Click to expand...
Click to collapse
No, masking as Mozilla completely removes "Opera". "Identify as" doesn't.
How do you know that it's identified in browser statistics?
BTW, please give me an example of a site which sends mobile content to Opera Mobile 7 before changing this setting.
It's so funny reading all the time what's impossible while everyone can download other mobile browsers and see with own eyes that it's not.
Click to expand...
Click to collapse
Evidently, those other mobile browsers are not being singled out by the site and sent mobile content. Once again you show that you completely fail to understand the fundamentals of browser sniffing.
in my view i dont think mini5 is esstially much speedy & good than mini4.2 so i am still use 4.2 now

Little "non-useful" tips/hacks

Hi all.
This is a little collection of things that i have been noticing while testing hacking issues on the phone.
Remember that those are "non-useful" (not to jailbreak) the phone, and just curiousity as topic.
Easy Hidden Menu Call
Do you need a search on the net to remember the hidden menu code? No more!. Test this phone-number string instead:
(Edited now): ##PROGRAMNITT
Max size for an app name/web favorite
Seems to be no max per se, but after doing some test, where i created title as: "chunk1chunk2chunk......chunkN" i was able to load a 1691124 characters title. Further than that, the browser seems to crash.
That's about a 3MB text string, just for the title. Would work well, when testing if several of them pinned reduce our 8gb storage(use storage) or doesn't (uses other).
Btw, you can pip up to 67 apps, (51 new) so... that's a max anyway,
Application Menu "About:blank" hack
Test this in the browser bar as direction: "about:blank". Kin IE will yell that it's not a supported protocol. Yeah, that's right. Let's dev a page on a local webserver with:
PHP:
<html>
<head>
<title>Mad redirection!</title>
</head>
<body>
<h3>Mad redirection tool!</h3>
<p> Testing: <div id="testTab"></div></p>
<p> Errors:
<div id="errorsTab"></div>
</p>
<script type="text/javascript">
var urlToTest = "about:blank";
try {
var test = document.getElementById("testTab");
test.innerHTML = urlToTest;
window.location=urlToTest;
}
catch (error) {
var err = document.getElementById("errorsTab");
err.innerHTML = "Error going to " +urlToTest+"<br/>"+error.message;
}
</script>
</body>
</html>
Browse it with the kin and you will land in the about:blank page, with the ability to be pinned on the application menu. Of course it will work, having the App link on the App menu, with a non working link (Kin still yells if you use it from menu).
Useless, but weird...
I do know that this is pure thread necromancy and that those are old news but:
a) if you are able to do the trick (using the sample html i posted) you can see that indeed it comes to about:blank and is shown as that on the title: "ABOUT:BLANK".
b) if you are so smart to change it to "about:lame" it goes there but shows a "Action canceled" webpage, where it suggest you to press the "refresh" button or use menu opcion "File -> work offline".
Like if you could.. rofl.
That means:
1) "about:" protocol is supported (at least about:blank) to be navigated BUT is nerfed from the direction bar. So other protocols could work. For example, smtp and ftp does trigger a popup from the IE, but res:// file:// and rtsp:// do not (even if they crash later, and rtsp opening zune for streaming).
2) This is a pure IE (with file menu,hopefully )
3) some other things can be tested, and every person can!
I upgraded the posted code, so it outputs an error when the redirection doesnt work (almost allways).
If you try it, remember not to end your url with \ (backslash) as it interferes with the doublequotes.
I've just completed testing a couple of things.
First, I successfully tested the "about:blank".
I also tried "about:", "about:about", "about:cache", and "about:home". These each resulted in the action canceled page described above.
I also tried the "file://" protocol, with the address "file://localhost/c:/" and received the following:
Errors:
Error going to file://localhost/c:/
Could not complete the operation due to error 80070005.
[edit] It seems that error 80070005 is given when you do not permission. The solution? Log on with administrator privileges... (see link)
Upon further testing:
about:desktopitemnavigationfailure works and displays "navigation cancelled" page.
about:navigationcanceled works and displays "navigation cancelled" page.
about:navigationfailure works and displays "navigation cancelled" page.
about:noadd-ons displays "navigation cancelled" page.
about: offlineinformation works and informs the user that the current page can not be viewed off line.
about: postnotcached works and informs the user that to refresh the current page, information entered in a form will have to be re-posted.
about:securityrisk displays "navigation cancelled" page.
about:tabs (unsuprisingly) displays "navigation cancelled" page.
I read that about:mozilla works in older versions of IE. However, it displayed the "navigation cancelled" page. You can also supposedly access the about:mozilla page using the following URL: res://mshtml.dll/about.moz
However, while this "res" protocol appears to be supported, I received the same permissions error as referenced in the above post.
I tested the mms protocol on a couple of working mms streams, but received the notification that the protocol is not supported.
I tried view-source://(random web address) and unsuprisingly was told that the protocol isn't supported. While this protocol works with some browsers, it doesn't seem to work on internet explorer even on a regular computer.
I tried the javascript protocol and it seems to work, but is different than about:, http:, etc. Mainly, it processes the javascript without leaving the script "address" in the address bar like we see with about: and http:
I was a little disappointed in this one, hoping to bookmark a javascript to test the videohamster flash video viewer for ipods, or itransmogrify for other flash files.
very nice work here. I like what you have done with this.
I'm glad that other than about:blank works (apart of the "action cancelled").
I took my time to install a wm6.5 emulator and test where do this "Action cancelled" come from in the pocket IE url bar.
They are from " res://.....navcancl.dll ".
Maybe there's a way to bypass the restrictions (the permission error) by calling some parameter in the "about:XXXX", but i can't bet on it.
Edit:
about:version seems to work (it auto-says "cannot find server", although my python custom-made-for-exploits server says that it delivered my html). But it keeps loading after the javascript redirection happens.... lol, so random .
One thought I had, that I have not had time to experiment with yet, is how deep the permissions restrictions go. For example, at times I have been logged on to a windows-based computer and have access to certain user-specific files but not to system files or to files or folders closer to the root. So for instance, we may be able to access the WinCE equivalent of "C:\Documents and Settings\<UserName>" using the file:// or res:// protocols even though we don't have permission to access "C:\".
Here's another potential avenue for information related to the "res" protocol. Apparently, it can be used to enumerate the software on a machine by identifying certain executables or dlls. (see here).
Unfortunately, the example cited in the article is not available so I can't view the code on how it was done. However, the results can be viewed here, where incidentally you can see the software installed on the computer that crawled this webpage.
Luckily, a manual or how-to paper is available here. I will try to check it out and see if I can figure out something useful.
i checked, it doesnt yell at you if you use a res:// but either if using ftp:// so the big problem is that you must pre-know the res:// uri before testing.
And in the best case, you will just get an image shown, ad js cannot give you the binary data.
anyway, i'm interested in this things....
Here's a couple other likely non-useful tidbits.
The browser will attempt to open the following filetypes with the Zune player:
.avi
.3gp
.mov
.fli
.mp4
.wmv
.wmx
When you open a VBScript in the browser, the script isn't executed, but it is displayed.
The mailto: protocol works from the browser and opens up the email dialog.
The following script causes the browser to hang (and deleting temporary files does not resolve the problem--but restarting the Kin does):
HTML:
<html><body onLoad=Demo()><script>
// MoBB Demonstration
function Demo() {
var a = new ActiveXObject("Internet.HHCtrl.1");
var b = unescape("XXXX");
while (b.length < 256) b += b;
for (var i=0; i<4096; i++) {
a['Image'] = b + "";
}
}
</script>
</body></html>
I haven't played around with the logs at all, but would this provide an error that gives some useful log output?
After some further testing, I discovered the Kin does not yell about the following protocols as being unsupported (in other words, they seem to be supported):
gopher://
nntp://
telnet://
news://
snews://
windowsmail.url.mailto://
windowsmail.url.news://
windowsmail.url.nntp://
windowsmail.url.snews://
johnkussack said:
Maybe there's a way to bypass the restrictions (the permission error) by calling some parameter in the "about:XXXX", but i can't bet on it.
Click to expand...
Click to collapse
I tried playing around with about:____, such as with the following types of addresses:
about:<input%20type=file>
about:<a%20href=C:\windows\>Click-Here</a>
but without luck.
I also tried the shell handler "Shell:" which seems to be another supported protocol, but again without luck. I tried the following Shell commands:
Shellrofile
ShellrogramFiles
Shell:System
Shell:ControlPanelFolder
Shell:Windows
Shell:::{21EC2020 shell:::{21EC2020-3AEA 3AEA-1069 1069-A2DD A2DD-08002B30309D}
Here are a couple more that I found other people sometimes try that I haven't tried (at least not yet):
shell:ControlPanelFolder
shell:::{35786D3C-B075-49b9-88DD-029876E11C01}
shell:::{208D2C60-3AEA-1069-A2D7-08002B30309D}
shell:::{7007ACC7-3202-11D1-AAD2-00805FC1270E}
shell:::{20D04FE0-3AEA-1069-A2D8-08002B30309D}
shell:::{450D8FBA-AD25-11D0-98A8-0800361B1103}
shell:::{E17D4FC0-5564-11D1-83F2-00A0C90DC849}
Ok, so this will be my last post in this thread tonight . For some unknown reason, you can access your emotes when in camera mode.... It doesn't do anything if you try to use one though.
great to hear about the shell::XXXX thing.
Does it trigger something? like about:blank or the other trigger a blank or a "cannot go" page.
btw, a real path on the phone (granted by the logs) is:
\Windows\eri.bin
That's assured , with the start backslash ("\\" if used on js code)
these hacks arent nonuseful
you should have called these hacks something other than non useful because we can use these little tips and tricks in combination with others to actually create an in browser jailbreak using the unrestricted protocols.
shell commands
try the net user admin <username> <password> console command in the shell protocol and see if you an bypass restrictions. theres no reason why console commands shouldnt work even though i havent tried this myself.
X-15D9W8491 said:
try the net user admin <username> <password> console command in the shell protocol and see if you an bypass restrictions. theres no reason why console commands shouldnt work even though i havent tried this myself.
Click to expand...
Click to collapse
Sorry, I'm not sure where you mean to do this. Unless I completely missed the revelation, so far, nobody has been able to get any type of shell/console access (as it doesn't really exist on a Windows Mobile OS anyway).
i called them as is, cause in first place, they were non useful, lol.
Although now, it could be a good try to get some "jailbreak" procedure.
as we dunno what windows mobile i6 can do, i guess we should/must try into a real mobile device (maybe my old pda too), or a win mobile 6.5 emulator, to test procedures (less restrictions), and then repeat on the kin (restricted).
I always though that the browser was the weakest part anyway
if you do tel: in the browser, and write anything after that it opens it up in a bubble....it lets you call letters, although it gives an error in the phone app
When using the TRACERT (Trace Route) in the programnitt menu I found a quirk.
Using 127.0.0.1 to Trace replies: WindowsCE
...that's obvious but interesting.
Using 127.0.0.0 to Trace replies: * 87 (30 times, hits limit and stops)
I have no idea why it would reply with the voicemail number....

Searching for phrases on XDA

Hello
Every time I try to search more than a single word it looks for any of the individual words - even when I am putting the words in quotes "" or as word1+word2+word3.
What am I missing here? Is there no way to search for phrases on here?
I found this :
Basic Search Syntax
Search multiple words use Boolean operators such as:
- AND (ex. ajax AND instant)
- OR (ex. ajax OR instant)
- NOT (ex. ajax NOT instant) or use the minus sign (-) (ex. ajax -instant)
Search for an exact phrase using quotes:
- (ex. "instant email")
Search for wild cards using asterisks (*)
- (ex. *mail)
black3yed said:
I found this :
Basic Search Syntax
Search multiple words use Boolean operators such as:
- AND (ex. ajax AND instant)
Search for an exact phrase using quotes:
- (ex. "instant email")
Click to expand...
Click to collapse
Yes, but as I said neither of these seems to work on XDA .
Unfortunately, vBulletin's search function lacks recognition of Boolean operators. It makes searching for exact phrases difficult.
Try to use the keywords from your phrase. For example, instead of 'root for captivate', try 'samung captivate root guide' or something similar. The more specific you make it, the higher the chance of your top hits coming back with all your keywords.
It's a known issue, and there's been a lot of discussion on what can be done to get around it. Unfortunately with it being a 'feature' of vBulletin, it may take some time for this to occur
Good luck! If I can help with your search at all, feel free to send me a PM
Hate to resurrect this thread from the dead but, this is especially frustrating for searching within threads... Trying to search by phrase or just two words gives an impossible result
For example I'm try to search for "change color" unfortunately I get just the individual words results, and 125 pages of results lol :crying: Seems like I run into this wall on a weekly basis, frustrating. Usually I can just do a google search ie. site:forum.xda-developers.com "my phrase" but this doesn't let me search within specific threads.
There is still no work around for this right?
quick_ said:
Hate to resurrect this thread from the dead but, this is especially frustrating for searching within threads... Trying to search by phrase or just two words gives an impossible result
For example I'm try to search for "change color" unfortunately I get just the individual words results, and 125 pages of results lol :crying: Seems like I run into this wall on a weekly basis, frustrating. Usually I can just do a google search ie. site:forum.xda-developers.com "my phrase" but this doesn't let me search within specific threads.
There is still no work around for this right?
Click to expand...
Click to collapse
Use double quote marks round your query?
"like this"
pulser_g2 said:
Use double quote marks round your query?
"like this"
Click to expand...
Click to collapse
Unfortunately that doesn't work anymore, just gives results of the individual words.
that's so incredible that this simple thing cannot be solved on the biggest forum for smartphone, pfffff
in applications forums all the message are in a database and that is VERY easy to make a query for exact phrase with php.
I think xda dont want, for unknow reason ...
ANd i am afraid to read several same bad answers here ( and one by an "Developer Admin" ! OMG ! ) : ==> use quote '"like this"
====>
READ OR TRY , BECAUSE THAT DONT WORK !
anyway i give you my contribution , use Google ,
with the "trick" : site:forum.xda-developers.com
example :
"smallest rom" site:forum.xda-developers.com
unfornunately that dont work for specific smartphone forum then : give in more some indications to Google :
"smallest rom" I9300 s3 site:forum.xda-developers.com
jimi27 said:
that's so incredible that this simple thing cannot be solved on the biggest forum for smartphone, pfffff
in applications forums all the message are in a database and that is VERY easy to make a query for exact phrase with php.
I think xda dont want, for unknow reason ...
Click to expand...
Click to collapse
lol. xda DB is huge. Querying the database directly is not a good idea since it can heavily impact the performance.
Thus the need of an external service for indexing and searching, such as Apache Solr..

Outside links on xda are broken

There's something wrong with the links on xda. They redirect through api.viglink.com.
This is bad for a couple of reasons:
- Manipulating links so that they point to a different address than displayed in the status bar is generally considered bad behaviour equal to malware.
- It breaks context menu functions like "open in another browser," "send shortcut to desktop," "translate," "search," etc.
Can you repair the links on xda so that what you see in the status bar is what you get? You don't want to modify links in a way that breaks browser functions, do you? I understand that you need to make money somehow, but please try to do so in a non-destructive way.
rogier666 said:
There's something wrong with the links on xda. They redirect through api.viglink.com.
This is bad for a couple of reasons:
- Manipulating links so that they point to a different address than displayed in the status bar is generally considered bad behaviour equal to malware.
- It breaks context menu functions like "open in another browser," "send shortcut to desktop," "translate," "search," etc.
Can you repair the links on xda so that what you see in the status bar is what you get? You don't want to modify links in a way that breaks browser functions, do you? I understand that you need to make money somehow, but please try to do so in a non-destructive way.
Click to expand...
Click to collapse
Sounds like you may have some malware installed on your PC or your web browser screwing with your links, you might want to check that. All of the outbound links in the XDA forums that I checked just now link directly to the external URL, standard "A" tag links.
My university is gonna like that. It's happening on their computers.
That'll teach them for forcing internet explorer on the entire department
(But it only happens with outside links on xda, not on other sites. Is there some xda-specific malware running wild?)
Nah, it's not malware or anything. It's something the owners decided to implement at some point. Sorry don't have any specific information though. Just letting you know it's not anything on your end.
This happens on almost every link I click on on my nexus 5 from any browser.

[Q] Contacts not recognized when first name contains Umlaut (ä, ö, ü) or ß

Hi,
if you want to phone somebody - let's call him Jochen Altmann - you can dial J-O-C... (5-6-2...) and the contacts get checked for matches. The more letters you type, the less possible hits are displayed. This also works when you start by typing the last name A-L-T...
Strangely, this does not work if the first name contains an Umlaut (ä, ö, ü) or a German ß. If that is the case, the contact won't be displayed. E.g. if I wanted to call 'Jürgen Altmann', neither typing J-Ü-R... (5-8-7...) nor A-L-T... makes the contact show up. The contact is totally ignored, as if it didn't exist.
This problem does NOT appear, if only the LAST name contains ä, ö, ü or ß.
Also, the problem does NOT appear, if the first name contains any other special letter like é.
The problem seems to affect a number of people, not only my phone. And it seems to be a Fairphone issue, not an Android issue.
Any solution available?
Best,
smartn
Hi,
smartn said:
Hi,
if you want to phone somebody - let's call him Jochen Altmann - you can dial J-O-C... (5-6-2...) and the contacts get checked for matches. The more letters you type, the less possible hits are displayed. This also works when you start by typing the last name A-L-T...
Strangely, this does not work if the first name contains an Umlaut (ä, ö, ü) or a German ß. If that is the case, the contact won't be displayed. E.g. if I wanted to call 'Jürgen Altmann', neither typing J-Ü-R... (5-8-7...) nor A-L-T... makes the contact show up. The contact is totally ignored, as if it didn't exist.
This problem does NOT appear, if only the LAST name contains ä, ö, ü or ß.
Also, the problem does NOT appear, if the first name contains any other special letter like é.
The problem seems to affect a number of people, not only my phone. And it seems to be a Fairphone issue, not an Android issue.
Any solution available?
Best,
smartn
Click to expand...
Click to collapse
This is really an Android problem. Also because the search only matches the begin of a word you can not search for "rgen" for example.
_keesj said:
Hi,
This is really an Android problem. Also because the search only matches the begin of a word you can not search for "rgen" for example.
Click to expand...
Click to collapse
1. No, as stated before, this is NOT an Android problem - it did not occur on other Android 4.2.2 mobiles that I've tried.
2. You're right, the search term has to match the beginning of a word. But this is the way it works, no matter what specific letters the search term contains, so it's not a bug but the way it's meant to be.
Hello
smartn said:
1. No, as stated before, this is NOT an Android problem - it did not occur on other Android 4.2.2 mobiles that I've tried.
2. You're right, the search term has to match the beginning of a word. But this is the way it works, no matter what specific letters the search term contains, so it's not a bug but the way it's meant to be.
Click to expand...
Click to collapse
I did not look at the code I tried entering a few name containing special chars and once I was able to reproduce the problem searched the web for an answer:
I found the following thread from 2009!
https://code.google.com/p/android/issues/detail?id=3732
Thanks for your effort, _keesje, but after flying over a good portion of the 235 entries on the website which you linked, it seems to me as if they were dealing with another bug in the search function: the problem that a contact search is only limited to the names, so you can't search by company etc.
Surely a problem worth mentioning, but not the least helpful in the matter I'm dealing with - or did I miss anything? If so, please provide the number of the entry.
smartn
In the meantime, Fairphone has replied to my direct request:
"Thank you for your email. Our development team is aware of this issue, and we hope to include a fix the future update.
Cheers,
Rick"
I hope so, too.
But meanwhile, I'm still thankful for any other approach...

Categories

Resources