Oh Great . . . What did they break this time?



  • I've used Firefox for a long time.  Althought I really, really dislike a lot of the UI changes that have been made over the past couple of years, I've never encountered any of the big problems that others have complained about (crashes, excessive memory use, etc.) and I don't recall ever running into any serious bugs that affected basic web browsing.  Just lucky, I guess.

    Then yesterday I tried to log on to my credit card's website.  I hadn't visited there in a long time, but everything always worked just fine in the past.  Login fails -- invalid username or password.  Make extra sure that I'm using the right password, still doesn't work.  After a couple of failed login attempts, I notice the problem -- my username containes underscores (for example, John_Q_Public) but when I enter it, the underscores turn into spaces (John Q Public) and so, of course, it doesn't match the name they have in their system.  Since I was using the horribly old and outdated Firefox v9 (from way, way back in December 2011), I updated to the latest version.  No good.  Same problem.

    So I fire up Chrome and Internet Explorer, and they both work perfecty fine.  After logging in with Chrome, I looked around in their customer service area and found that I could change my username.  Seems like a possible security WTF, but unfortunately necessary in this case.

    I suppose I should file a bug report, but at this point it hardly seems worth it.  I think Firefox is just fucking with me now.



  • Before you rant too severely about Firefox, I'll remind you that complex sites sniff browsers and often send different code to different browsers. It could be that the site has mangled the 'validation' scripts they are sending to firefox. You could try faking you user agent string and seing if it works, that is, if you are interested in debugging, instead of, like, living your life like a normal person.

    Just saying, it could be the bank. We all know how bad bank's web sites are!



  •  I'm pretty sure it's not Firefox's fault, as robbak said; I managed to log in to this account just fine.



  • @robbak said:

    Just saying, it could be the bank. We all know how bad bank's web sites are!
    You're right, I thought of that, it could be the bank.  They are a wonderful source of WTFs.  But it used to work fine in Firefox, so something has changed, and since Firefox almost certainly gets changed much more frequently that the bank's website, it seems pretty likely that the problem is with Firefox.

    Update:  Ran Chrome, spoofing useragent as Firefox 11,  underscores show up just fine, so clearly it's a Firefox problem and not a bank problem.



  • @I_Like_Underscores said:

    I'm pretty sure it's not Firefox's fault, as robbak said; I managed to log in to this account just fine.
     


    Rimshot!!
    <br>



  • @El_Heffe said:

    @I_Like_Underscores said:

    I'm pretty sure it's not Firefox's fault, as robbak said; I managed to log in to this account just fine.
     


    Rimshot!!
    <br>

    advice dog is always helpful



  • What! As a fine, upstanding TDWTF member, I would never do that!

    Though it might help if El_Heffe tried it, since sometimes the bottom few pixels in a textbox get cut off and aren't displayed, which makes underscores look like spaces; looking at the data actually going out over the wire would be one way to be sure that his username is being mangled like that client-side.

    I mean, it's not like there's anything else to do on the weekends...



  • @El_Heffe said:

    Update:  Ran Chrome, spoofing useragent as Firefox 11,  underscores show up just fine, so clearly it's a Firefox problem and not a bank problem.

    It's not clear at all. The bank's website may be using feature detection instead of useragent sniffing, in which case spoofing the useragent string has no effect.



  • @Ragnax said:

    It's not clear at all. The bank's website may be using feature detection instead of useragent sniffing, in which case spoofing the useragent string has no effect.
     

     Yup! Back when I worked for a company with a super legacy IE only website, during some downtime I tried to see what it would take to add cross-browser compatibility. I was initially stymied for about an hour, when it turned out that getting in with Firefox wasn't as simple as just changing my UA string. Through some debuggering on both client- and server-side, I discovered that an ancient programmer had gone above and beyond the call of duty when doing browser detection; even if the browser passed server-side UA sniffing, the client Javascript would still check for the existence of some bullshit IE-only Javascript feature that I have happily forgotten. If that wasn't present, the client would show you the standard "sorry, you can't use this website" error - but since it still got your present browser version from your UA string, it would paradoxically say "you're using IE 7, please use IE 6 or higher".



  • @Ragnax said:

    The bank's website may be using feature detection instead of useragent sniffing
    True. I've ruled out UA-sniffing, but it's much more difficult to figure out if they are feature-sniffing. And if they are feature-sniffing, why did things used to work but are now broken? That would seem to indicate that something in Firefox changed.



  • @El_Heffe said:

    And if they are feature-sniffing, why did things used to work but are now broken? That would seem to indicate that something in Firefox changed.

    Sideline: contractors designed our company website IE site and in doing so have the following representative line:

    <img src="some.jpg" onclick="onclick('somereference')">

    Yep, they created a handler called onclick to trap the onclick event. IE was fine with it, other browsers either threw errors or refused to work although I think the iPad and iPhone browsers (Safari?) coped okay.

    I recently upgraded Firefox from 3.some.random.no to 11 and it now copes okay. Meaning this functionality that shouldn't work.. does.

    TL;DR: things do change between versions and it could well be something internal to FF.



  • @Cassidy said:

    Yep, they created a handler called onclick to trap the onclick event.
    I fail to see how that is a problem.

    In fact, the handler will be called even if you don't click exactly on the image!



  • Okay, I just did a quick test on FF 12:

     

    <!DOCTYPE html>
    <html>
    <head>
    <script type="text/javascript">
    function onclick(id){
    console.log(id);
    }

    </script>
    </head>
    <body>

    <img src="file:///home/zecc/Pictures/wonderful_placeholder_pic.jpg" onclick="onclick('some_id')">

    </body>
    </html>

     

    All I got was a repeated "too much recursion" error and a hung Firefox. I don't know why.



  • @El_Heffe said:

    After a couple of failed login attempts, I notice the problem -- my username containes underscores (for example, John_Q_Public) but when I enter it, the underscores turn into spaces
     

    Unable to reproduce with FFX12 and my bank.

    Edit

    This quick bit of code confirms that under standard conditions, Firefox does absolutely nothing to underscores:

    <? echo $_POST["lolcare"]; ?>
    <form method="post" action="">
    lol: <input type="text" name="lolcare" />
    <input type="submit" class="submitbutton" name="" value="Submit" />
    </form>



  • @Zecc said:

    I fail to see how that is a problem.

    @Zecc said:

    All I got was a repeated "too much recursion" error and a hung Firefox. I don't know why.

    I rest my case.



  • I ran out of GDI objects last week – that was a new one on me for Firefox. The only other software that would do that, was GTK+ runtime for Windows.

    I keep my PC hibernated during the day, and since the 25th, Firefox is already holding over 5,000 GDI objects (29 tabs in two windows). (I just closed one window by moving its only tab to another window, and it's just lowered to 4,700 GDI objects.)

    I am at a loss to see how it could mangle underscores though – I imagine that's got to be the website doing something stupid. The primary issue that I have seen is whether space should be escape as + or %20, but the old escape() function doesn't like Unicode (any character in the local code page is encoded in 8-bit, otherwise Unicode, which is just silly). But underscores? Why would Firefox be messing with those?



  • @Zecc said:

    <img src="file:///home/zecc/Pictures/wonderful_placeholder_pic.jpg" onclick="onclick('some_id')">
     

    I'm sure I've mentioned this gem before:

    @no said:

    <a href="somewhere.html" id="somewhere_hyperlink" style="height:1px;width:1px;">Somewhere</a>

    <img src="somewhere.jpg" onclick="document.getElementById('somewhere_hyperlink').click();" />

    It gets better when the hyperlink also has a click handler that runs a javascript function whose sole purpose is "window.location=somwhere.html"



  • @Daniel Beardsmore said:

    I ran out of GDI objects last week – that was a new one on me for Firefox. The only other software that would do that, was GTK+ runtime for Windows.

    I keep my PC hibernated during the day, and since the 25th, Firefox is already holding over 5,000 GDI objects (29 tabs in two windows). (I just closed one window by moving its only tab to another window, and it's just lowered to 4,700 GDI objects.)

    I enabled the GDI objects column out of curiosity, and my audio card's control panel (which isn't even open) is using 5,550. Wow.

    My question is: why do GDI objects even have what appears to be an artificial ceiling? Are they statically allocated or something?



  •  They're allocated on their own heap, I think. In ye olde days, it was a 64kb heap, dunno if it still is.



  • I forget what Vista/7 changed, but for XP the limit was 10,000. Not 32,768¹ or 65,535 or any meaningful number, but 10,000². It's not the size of any handle, and surely Windows runtime didn't preallocate 10,000 objects in advance? Some limits are just weird.


    ¹How comes BBC BASIC used 32-bit integer variables on the 6502 CPU in 1981, but Visual Basic only allowed 16-bit variables on x86?

    ²At least it's over 9000.



  • @Daniel Beardsmore said:

    I forget what Vista/7 changed, but for XP the limit was 10,000. Not 32,768¹ or 65,535 or any meaningful number, but 10,000². It's not the size of any handle, and surely Windows runtime didn't preallocate 10,000 objects in advance? Some limits are just weird.

    Well considering it's one of those limits that you should never, ever, ever come even close to hitting, 10,000 is as good a number as 65,535. If you're halfway there, you've already fucked up.



  • Indeed. Firefox on this PC is already at 4,000 GDI objects since I logged in at 9 am (and it restored my session). What is it doing?

    Even so, I still dislike arbitrary limits.



  • What are you guys doing to your computers?  I've have mine running for over a month without a reboot and Firefox up the entire time.  Most GDI objects in use are 818 by Outlook.  Second is Firefox at 665.  Explorer is in third at 467.

    Heh, Windows Calculator comes in at a whopping 148, and that is with it running in "standard" mode.



  • Just out of curiosity I started switching between standard/scientific on a calculator I'd already got open (usually I have 20-30 open by the time I shutdown because I'm too lazy to do any mental arithmetic and my keyboard has a dedicated hotkey for it) and watching the readout. It was different every time. Sometimes more, sometimes less.



  • @pauly said:

    What are you guys doing to your computers?

    Making them toil and sweat and struggle in pain.



  • @Daniel Beardsmore said:

    Indeed. Firefox on this PC is already at 4,000 GDI objects since I logged in at 9 am (and it restored my session). What is it doing?

    Likely it is just not releasing them when it's supposed to. I call it a God Damn Irritating leak.



  • toH! It really is arbitrary.

    The session limit is 65,535, and the process limit is set to 10,000 by the following Registry value:

    HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows NT\CurrentVersion\Windows\GDIProcessHandleQuota

    lol, from that page:

    @HunterM said:

    windows 7? ....same bag of fit?

    honestly i brought 4GB RAM fr my laptop not knowing tht this bull *** is also there, im just a an effin normal customer tht wants to pay my money and buy a good system.... y the *** shud i hav to know about all this effin bottlenecks tht screw around with my operations? how do u increase GDI handle limit in windows 7, and does it make a difference????? im in the middle of my dissertation and pdf-xchange viewer keeps running out of resources to open files! die microsoft!$0 $0$0 $0 $0[Really, MSDN gonna let me edit this guys post??? I didn't even write this post... but I can not help myself, I think this post sucks and is not productive. I DO need to know these types of things... so users like chraniac here should not have too. I recommend using a different pdf viewer application... not Adobe's. His anger is real, his spelling is atrocious, but the direction of anger is misdirected IMO! I have said my peace... now lets see if I can "submit" this edit...]$0

    1 WTF point for a senseless rant against buggy $0[ software, and 2 bonus WTF points ]$0 $0 $0 $0 for a garbled MSDN comment entry $0. A final bonus WTF point for saying "Not Adobe's" when the commenter specifically noted that the bug lay, for a change, not with Adobe Reader but Tracker Software's PDF-Xchanger Viewer.

    At least for a change someone managed to differentiate Adobe as a company, from their products, though. Adobe Reader crashes, not Adobe. NVIDIA Corporation has been known to crash though:


    Edit: Firefox's newfound hatred for “http://” caused me to copy and paste the web address without it, creating an invalid URL. WTF Mozilla?



  • @Daniel Beardsmore said:

    Edit: Firefox's newfound hatred for “http://” caused me to copy and paste the web address without it, creating an invalid URL. WTF Mozilla?
     

    about:config
    browser.urlbar.trimURLs
    Set to "False"

     [/BecauseHavingUsersAlterConfigFilesToRecoverCommonExpectedSecurityFeaturesIsAGoodIdeaDotDotDotNOT!]



  • @Ragnax said:

    @El_Heffe said:
    Update:  Ran Chrome, spoofing useragent as Firefox 11,  underscores show up just fine, so clearly it's a Firefox problem and not a bank problem.

    It's not clear at all. The bank's website may be using feature detection instead of useragent sniffing, in which case spoofing the useragent string has no effect.

    I'd be flabbergasted to see a bank website smart enough to use feature detection, yet still dumb enough to use it to send a broken alternate version of a script instead of having a non-broken script in the first place.



  • @Lorne Kates said:

    @Daniel Beardsmore said:

    Edit: Firefox's newfound hatred for “http://” caused me to copy and paste the web address without it, creating an invalid URL. WTF Mozilla?
     

    about:config
    browser.urlbar.trimURLs
    Set to "False"

     [/BecauseHavingUsersAlterConfigFilesToRecoverCommonExpectedSecurityFeaturesIsAGoodIdeaDotDotDotNOT!]

    I'm the first person to bitch about Firefox, but the https:// is still there on https URLs. Also, Chrome does the same thing, removing http:// from http sites but leaving https:// alone. However, Chrome does have a much nicer pull-down which gives useful HTTPS info. Also, it actually warns you when a site includes insecure JS or CSS, which Firefox does not seem to do at all.


  • Discourse touched me in a no-no place

    @Daniel Beardsmore said:

    Edit: Firefox's newfound hatred for “http://” caused me to copy and paste the web address without it, creating an invalid URL. WTF Mozilla?
    Really? It usually prepends it if the first character in the address bar is part of what's being copied. (At least in 11.0/Linux)



  • @Lorne Kates said:

    browser.urlbar.trimURLs

    Set to "False"

    Ting! Sorted.

    Thanks, Kates!



  • @PJH said:

    Really? It usually prepends it if the first character in the address bar is part of what's being copied. (At least in 11.0/Linux)

    Yes it usually does, but not always. It doesn't work if you want to copy only part of the address, and in this instance I have no idea what went wrong. I did edit the address to remove the parenthetical part that is allegedly what causes MSDN links to break all the time. (Which, if that's true, would mean that Microsoft could write a pretty trivial mod_rewrite instruction to deal with the fallout. Or rather, IIS's equivalent facility. Or simply have an ASP script that handles the redirect directly. There's just no excuse to have such widespread link breakage.)

    I did know about the about:config workaround, but I didn't care whether "http://" was visible or not. It's just that appearance isn't the only issue it causes.



  • @pauly said:

    What are you guys doing to your computers?  I've have mine running for over a month without a reboot and Firefox up the entire time.  Most GDI objects in use are 818 by Outlook.  Second is Firefox at 665.  Explorer is in third at 467.
    Quick glance at the GDI column shows that around 9000 handles are in use, of which 1435 are used by ... Wordpad?! GIMP was way down there with 200 handles.



  • @Daniel Beardsmore said:

    Firefox's newfound hatred for “http://” caused me to copy and paste the web address without it
     

    clicking the address bar selects the entire url which invisibly includes http.

    Manually drag-selecting from the end of the url to the front also includes http.

    What demented sequence of actions are you performing to copy a url from the address bar?



  • @ender said:

    Quick glance at the GDI column shows that around 9000 handles are in use, of which 1435 are used by ... Wordpad?! GIMP was way down there with 200 handles.

    Maybe it renders each character separately. Or something. Maybe in GIMP, being cross-platform, a single GDI handle is a wrapper around a group of quirky somethings.

    These numbers are mystifying.



  • @dhromed said:

    What demented sequence of actions are you performing to copy a url from the address bar?

    The rule appears to be that, so long as … hmm. Just copied the address of this page minus the last character, and: http://forums.thedailywtf.com/forums/AddPost.aspx?ReplyToPostID=289062&Quote=Fals

    OK, so "http://" is added (in v11 at least – I was using v12 earlier) so long as the address bar selection contains at least the first character. I thought that it was only added so long as you copied the entire address, but not any substring, start, middle or end. I've already forgotten what mental notes I made as to when it does and doesn't work (which I think were for v11). Or that about:config fixes it :)

    I believe I just selected the address, ctrl+C, ctrl+V. Nothing unusual – Firefox simply forgot to include http:// that time.



  • @Daniel Beardsmore said:

    @PJH said:
    Really? It usually prepends it if the first character in the address bar is part of what's being copied. (At least in 11.0/Linux)
    Yes it usually does, but not always. It doesn't work if you want to copy only part of the address, and in this instance I have no idea what went wrong. I did edit the address to remove the parenthetical part that is allegedly what causes MSDN links to break all the time. (Which, if that's true, would mean that Microsoft could write a pretty trivial mod_rewrite instruction to deal with the fallout. Or rather, IIS's equivalent facility. Or simply have an ASP script that handles the redirect directly. There's just no excuse to have such widespread link breakage.)

    I did know about the about:config workaround, but I didn't care whether "http://" was visible or not. It's just that appearance isn't the only issue it causes.

    There also is a really retarded bug in Community Server's handling of urls.  For example, type  msdn.microsoft.com  into any web browser and it will work just fine.  But, if I use that url to create a link in this post, like this, msdn.microsoft.com , Community Server turns the link into  http://forums.thedailywtf.com/forums/msdn.microsoft.com

    Nice!!

     



  • @El_Heffe said:

    There also is a really retarded bug in Community Server's handling of urls.  For example, type  msdn.microsoft.com  into any web browser and it will work just fine.  But, if I use that url to create a link in this post, like this, msdn.microsoft.com , Community Server turns the link into  http://forums.thedailywtf.com/forums/msdn.microsoft.com

    Nice!!


    That doesn't seem [i]that[/i] retarded to me; it's just assuming you've put in a relative reference and converting it to an absolute one. Probably a poor assumption in general, but I can understand the logic. I've seen the same thing on other sites too.

    On the GDI object count: winner on my system at the moment is TOAD with 4682, closely followed by IE on 4038 (I suspect this is either the instance of IE that's running Oracle Fusion Applications, or the one that's logged into Oracle Support). Informatica Designer is next with 1,442 and Explorer rounds out the >1k list with 1,285. Firefox comes way down the list with 96.



  • @El_Heffe said:

    There also is a really retarded bug in Community Server's handling of urls.  For example, type  msdn.microsoft.com  into any web browser and it will work just fine.  But, if I use that url to create a link in this post, like this, msdn.microsoft.com , Community Server turns the link into  http://forums.thedailywtf.com/forums/msdn.microsoft.com
     

    Hey. Erm, that's not a bug in CS.

    That is how link tags work, you know.

    Since ever.

    Am I correct in assuming you've never written HTML before?



  • @dhromed said:

    That is how link tags work, you know.

    Well, it's a little more than that. CS is actually replacing a relative link with an absolute link; it's generating

    <a href="http://forums.thedailywtf.com/forums/msdn.microsoft.com">msdn.microsoft.com</a>

    rather than

    <a href="msdn.microsoft.com">msdn.microsoft.com</a>

    though of course both would go to the same place, as you pointed out.

    But since it is bothering to generate an absolute link, it's quite sensible to suggest that in this use case (i.e. links entered in forum comments without http://) it would be more appropriate to generate an absolute link as

    <a href="http://{whatever the user entered}">{whatever the user entered}</a>


  • Talking about anchor tags, this came through in an e-mail from Adobe late last year – I don't know whether Adobe broke this, or the person who forwarded it to me, or Exchange, or my Outlook, or what, but it's pretty impressive:

    Enter the site at <a href="%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20https:/licensing.adobe.com/sap/bc/bsp/sap/zliclogin/login.htm%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20"><font color="#0066cc"><span style="color:#0066CC;text-decoration:none">https://licensing.adobe.com/sap/bc/bsp/sap/zliclogin/login.htm</span></font></a>


  • @Scarlet Manuka said:

    But since it is bothering to generate an absolute link, it's quite sensible to suggest that in this use case (i.e. links entered in forum comments without http://) it would be more appropriate to generate an absolute link as

    <a href="http://{whatever the user entered}">{whatever the user entered}</a>

     

    That is entirely true.

     



  • @dhromed said:

    @Scarlet Manuka said:
    But since it is bothering to generate an absolute link, it's quite sensible to suggest that in this use case (i.e. links entered in forum comments without http://) it would be more appropriate to generate an absolute link as

    <a href="http://{whatever the user entered}">{whatever the user entered}</a>

    That is entirely true.

    Which was exactly my point in my previous comment, although I may not have stated it clearly.  Slapping a "http://" onto the front of whatever link a person enters makes perfect sense -- "that's how link tags work".   Inserting "http://forums.thedailywtf.com/forums/" makes no sense at all.  It may not be a bug but it's a really bad design decision.

     



  • @El_Heffe said:

    Inserting "http://forums.thedailywtf.com/forums/" makes no sense at all.  It may not be a bug but it's a really bad design decision.
     

    I think it's not CS's fault, or Tiny's fault, but more a crazy way of how ContentEditable operates by default. As you type and the view updates, it applies user-changes to the DOM, then reads them back and in between the browser absolufied the links, just as it would under normal circumstances with any link that does not explicitly contain http://.

    Still, I think it's fixable by either CS or Tiny, so...



  • @dhromed said:

    @El_Heffe said:

    Inserting "http://forums.thedailywtf.com/forums/" makes no sense at all.  It may not be a bug but it's a really bad design decision.
     

    I think it's not CS's fault, or Tiny's fault, but more a crazy way of how ContentEditable operates by default. As you type and the view updates, it applies user-changes to the DOM, then reads them back and in between the browser absolufied the links, just as it would under normal circumstances with any link that does not explicitly contain http://.

    Still, I think it's fixable by either CS or Tiny, so...

    It's not ContentEditable, it's just a plain 'ol RTF field. It's been a few years, but I believe you are correct about how RTF engines function; I recall having to write JS to convert absolute links to relative links before submitting the contents of the RTF field. The real joy was that the site could have any number of hostnames it appeared under (SEO WTF) so I had to load a big list of hostnames from the server.



  • @morbiuswilters said:

    It's not ContentEditable, it's just a plain 'ol RTF field.
     

    Oh, I just heard the word once and figured that was the standard entry point for doing edity things in wepsites. I've never actually looked at an implementation directly.

    That sounds kind of like a WTF for someone in my line of work, and I suppose it is, but I've never been directly in need of understanding it, so it kept slipping aside.



  • @dhromed said:

    @morbiuswilters said:

    It's not ContentEditable, it's just a plain 'ol RTF field.
     

    Oh, I just heard the word once and figured that was the standard entry point for doing edity things in wepsites. I've never actually looked at an implementation directly.

    That sounds kind of like a WTF for someone in my line of work, and I suppose it is, but I've never been directly in need of understanding it, so it kept slipping aside.

    Fair enough. Most of the RTF stuff is a huge, platform-dependent WTF, so I'd be more-than-happy to not ever have used it. It's possible TinyMCE works differently now, but the version CS is running is pretty close to the version I was using 7 years ago.



  • @dhromed said:

    Oh, I just heard the word once and figured that was the standard entry point for doing edity things in wepsites.

    You should really be using wpasites.



  • @boomzilla said:

    @dhromed said:
    Oh, I just heard the word once and figured that was the standard entry point for doing edity things in wepsites.

    You should really be using wpasites.

    Actually wpa2-using-ccmp-and-with-wps-disabledsites.


Log in to reply
 

Looks like your connection to What the Daily WTF? was lost, please wait while we try to reconnect.