Related
I have an unusual problem that I can't wrap my head around and figure what is causing this and how to tackle it.
I've created a webpage that allows image uploads to the server! After the user uplaoded an image the host of the page receives an e-mail with the path to the uploaded image - really straight forward.
In the e-mail for instance a url like this is coming in.
https://www.something.com/uploads/FB_IMG_1525868856883.jpg
I can open that in Safari on Mac, I can open that in some cases in Chrome Incogntio, but it never renders in Internet Explorer or normal Chrome Window (in Incognito it does)
In the image you see something weird happening. In Incognito a different source is rendered than in a normal chrome window. (this is btw. not a caching issue)
you also see the page-title of the actual website resolved in the normal window, but not in incognito.
Why is that happening? What could cause that? is that some server setting? Some react setting I have to set?
I'm posting this in case someone else runs into it as well. I ran into this same issue where 500 kept happening for me and I knew it was not the extensions.
What fixed it for me:
Open Developer Tools in Chrome
Navigate to Applications Tab
Click "Clear Storage
Click Clear Site Data
As Nitish Phanse said, it is because the page is cached. Turning off browser caching will solve it.
I faced a similar problem, when any of the changes in CSS were not being reflected in the webpage, because the cached version was being used by chrome. However, the changes were visible in other browsers and devices. Turning off caching in chrome solved this problem for me.
I prefer to use incognito mode when developing website because when browsing normally the files are cached, thus, not reflecting the changes you made. When you use incognito mode, fresh cache and cookies are stored and the cache and cookies are disposed when you close the incognito window. Another plus point of using incognito mode for testing is that, it disables all the plugins, unless you explicitly enable them.
You might want to check out this question on Quora for further details on why to use incognito mode when developing websites.
Seems like you use quite amount of Chrome plugins. In Incognito-mode those are disabled by default. Try remove/disable AdBlock etc. stuff which blocks scripts and disallow them from running.
Makes quite a lot sense why they run in other browsers too.
I strongly recommend to NOT use plugins when you do web development. You might see different kind of behaviour with same code as user who do not use them. Only use this approach if you especially wan't to develop under certain conditions.
As most people suggest, it definitely sounds like a caching issue. Incognito doesn't load the page with cache. You could ctrl+f5 the page (reload without chache) and see if it still does it. If the caching itself proves to be an issue, you could add a ghost/phantom string (I don't know if that's the actual correct terminology). Basically you add,
<?php echo '?'.date('Y-m-d H:i:s'); ?>
at the end CSS or JavaScript import (whatever may cause the issue). I.e,
<script src="/js/regexps.js<?php echo '?'.date('Y-m-d H:i:s'); ?>" type="text/javascript"></script>
What this does, is that it will prevent people from caching your file, since the name of that file will change by the second. This way, you make sure that if you make changes to files that are often cached, that everyone gets the uncached changes as you update your CSS/JavaScript or whatever. This is very useful for files where you make JavaScript check for instance, in case you make changes to some regex etc.
So I know this might sound crazy, as it is technically a security concern which I understand. So I'm just trying to find out if there's any ideas on how to handle something like this.
Anyways, long story short, I was told to look into figuring out a possible way to scrape information from another browser window/tab. I have been asked to do this because, and I know this sounds crazy too, but the users of our website are incompetent enough to not be able to copy/paste and or type correctly something from a different website. I know it's tough for some to have to have several things in their workflow, but this is basically what they do: Go to their first website (after logging in) and bring up a record with information on it...including an identification number. Then, the user should take that number and go to the second website, our website (after logging in), and type it that number in a textbox (and eventually do some other stuff). But we have found that getting that identification number from the first website to ours is difficult for them. Some copy/paste correctly, some copy/paste too much text from the page, some write it down on paper then type it in our website, and some just seem to have trouble visually "copying" the number from site to site.
What I was thinking was that this could happen: the user would have already brought up the record on the first site, then they would come to ours. They could click a button, and that would run whatever I/we here come up with, that goes and finds the other browser window, finds the specific text needed, and puts it in our textbox. Sounds simple, right? HA.
The first website is not owned or managed by us in any way, otherwise this might be a little easier.
A little bit of background information: unfortunately, I'm technically targeting IE >= 10 through 9, so if there's a solution just for this (why I tagged vbscript), then that's great. If there's a broader solution (like with an applet or browser extensions... http://crossrider.com/ ), then that's even better, but not important. If it helps, we already have a hidden applet on the page that accesses the OS (yes, it has the mayscript attribute on the element so it is able to), so I thought that could be something to incorporate with. Also, the way I expect to know which window/tab to access is by URL and/or document title - either will be very specific.
We cannot install stuff on the users' computers, at least something outside of the browser (like extensions). I'm not sure how browser extensions work, so I'm wondering if they'd need to be "installed".
I know of HTML5's postMessage, but it only has partial support in IE (and none in IE <= 7)...and the partial support refers to not including exactly what I might need. It also requires that the other website be listening (which we don't have control over, but technically might be possible to include). So it doesn't count :)
The things I found with Java are to possibly find the list of processes currently running, but I don't know how to access/control one. Especially how to access the browser's Document.
And vbscript...I just don't know. I don't know if it's just me, but I can't seem to find good documentation on it, so I'm not sure what can be done with it.
Even if I could get control of the other browser window, I don't know how I would get information from it (like the DOM).
I'm not looking for code, just ideas...I'll do the research. And although it may sound impossible, don't just brush it off because Javascript can't do it - I haven't.
UPDATE:
I ended up developing a browser extension with http://www.crossrider.com/ which wasn't ideal, but works.
You could use a bookmarklet for this ... the user would have to drag the bookmarklet into their bookmarks bar on their browser, but if doing that wasn't beyond your user's abilities/the technical restrictions you've mentioned, then you'd definitely be able to send the information you need back to your site that way.
You'd just need to give your users instructions to:
i) drag the bookmarklet into their bookmarks bar on their browser
ii) go to the website in question and click the bookmarklet
you could code the bookmarklet so that it would grab the info you need, and redirect the browser to your website. All done in one click.
I think you may be thinking about it in the wrong way when you talk about posting from one 'window' to another. You could write the bookmarklet so that it would do a http post of whatever information you wanted into your site from the other site, and it could also redirect the window that they were looking at when they clicked it (the other site) to your site. Or if, for some reason, you didn't want to redirect the the window that they had the 'other' site in to your site, then you could add a listener to your site so that once the bookmarklet had posted the info you require then the window with your displaying could automatically update. The first option would make more sense and be easier though.
Maybe to open the other site from button/link resided in your site using window.open() method?
How do I take a deeper look at what's going on in the network level of my web browser than the developer tools network panel will provide? For example, I'm trying to debug an issue on a page right now in Firefox (18.0.1) where the loading "spinner" keeps going continuously long after the page has finished loading.
The firebug net panel shows all resources loaded ("200 OK") but the loading icon on the tab is still spinning and the text on the bottom of the window still shows "Transferring data..". How would I take a closer look and see what is supposedly still transferring? Note that this only seems to happen with Firefox and only on a fresh load (not when resources are cached).
After checking your link with Wireshark and with HTTPScoup (a different tool), my conclusion from both was that all the data arrives successfully and no connection remains open with your server's IP. The problem in FF I think is related to something else
Google reveals that in some cases, it is possible to confuse FF into this state. It doesn't have to be related to open network connections.
Take a look at this questions for example:
FireFox sits on "Transferring data from..." or "Read...."
The issue there was unclosed tags in the HTML body.. so try to find problems like this one too. You can try fixing some validation errors by running a validator such as this one. Since you have about 200 validation errors, I suggest you start by deleting half your HTML code until the problem goes away and this way zero-in on the problem location. Then go in and fix the tag mess.. looking a little in the HTML source gave me a headache :)
I would use a tool like Fiddler or Wireshark to do this. Both of these tools provide more low level packet information.
I would start with Fiddler. Fiddler is meant for http traffic analysis whereas Wireshark is a full-fledged packet sniffing tool for all layers of network traffic.
Another advantage to both of these tools is that you can use them with any browser. One disadvantage to fiddler is I believe it is only available for Windows.
I would recommend a very nice Firefox extension called HttpFox which is a bit easier to use than wireshark and friends
It lets you see the HTTP calls made by the page in much greater detail, and also the flow between pages (fast redirects and such). Here is what it looks like:
I use Firefox to develop a web site and at the same time to browse the web, read my gmail, etc.
The problem is every now and then I need to delete the cache and or remove the cookies of the web app, but I want to stayed logged in in the other web pages I am visiting.
Do you know a Firefox plugin (or Firefox trick) that can help with this issue?
Preferences → Privacy → remove individual cookies
They are grouped by domain, and you can eliminate all cookies from selected domain with one click.
Update:
Option → Privacy → remove individual cookies
There are two add-ons for FF that every web developer needs and they are Web Developer Toolbar and Firebug, the Firebug extension YSlow comes in very handy too.
Web Developer Toolbar has great cache and cookie control down to individual cookies. Firebug lets you mess with the DOM and CSS directly for a page and YSlow is good for page weight and response times.
Be careful having Firebug enabled for JavaScript heavy sites as they really slow down. GMail will warn you about this and tell you to turn it off. I have had a problem trying to disbable sites with the menus when the tool is embedded in the browser at the bottom but opening it in its own windows the sites menu works fine.
I personnaly also like IE Tab which means I can quickly view a page in IE without leaving FF and I can also get it to load anything that only works in IE with it.
Cookie Monster might work?. Also, have you considered just using two profiles in Firefox? You could setup two profiles with two icons and use one profile for the website and the other for the non-development. The only problem would be you can only have one open at once.
Also, two browsers (Chrome, IE, Firefox, etc) might work well too.
WebDeveloper extension allows to delete cookies for domain. It also can clear cache, although for me Ctrl-F5 is enough usually.
I had the same problem and found a firefox addon to remove the cookies for the current site that you are displaying:
https://addons.mozilla.org/en-US/firefox/addon/remove-cookies-for-site/
Using that I don't have to navigate the firefox menus to find my site in the list (which was too time consuming). This does not cover the cache-issue. For that there is the Ctrl+F5 other people have mentioned. It is covered in good detail here:
What requests do browsers' "F5" and "Ctrl + F5" refreshes generate?
Best solution:
1) Just Click CTRL+H and search the site.
2) right click on it and choose Forget About this site
(source: https://superuser.com/a/733154/249349 )
If you would like to use firefox, clear cache and cookies for one domain and stay logged in (and retain cache) at other sites, you could run a second firefox profile concurrently using the '-no-remote' parameter. I use this all the time to view sites I am building from an anonymous and administrative perspective at the same time.
One way to do this (in windows xp) is: make two profiles, then startmenu->run type 'firefox -profilemanager', untick the 'don't ask next time' box and load profile 1, then startmenu->run type 'firefox -no-remote' and load profile 2. You should now have two seperate firefox profiles running.
Clearing cookies for a single domain is easy with any one of a number of tools. Clearing the cache for a single domain or at the individual page level is sometimes useful, but hasn't turned out so easy for me thus far.
One specific situation is changing a meta redirect, where you can't just ctrl-F5 (you're being redirected, after all). Using a different profile isn't an acceptable general solution. Web Developer toolbar can disable the cache to test that the change is working but it's only a temporary fix, and not one that can readily be suggested to a client in order for them to view the change.
I use SQLite Manager for firefox to remove single sites from cache. All you have to do is open places.sqlite in your user profile folder using this plugin and remove the site. It's not exactly an easy solution but it does work. Make sure you backup your profile 1st!
Ross, I have the perfect answer for you. Get on the current tab that you want cookies cleared. Click tools > Page Info. Or install Page Info Forms & links for a shortcut. Then click the Security Tab. Then View Cookies button. It will bring up all the cookies for the relevant domain. Click the ones you want to delete, and click remove.
Seems we have cookies covered here.
As for cache, the trick is to make your development server force a refresh.
In httpd, you probably have some some cache directives to swap out for these:
Header set Cache-Control "no-cache, must-revalidate"
ExpiresActive On
ExpiresDefault "now"
In header:
<meta http-equiv="cache-control" content="no-cache, no store"/>
The benefit of doing it the httpd way is that your scripts/images/css get refreshed.
In that way you are not restricting yourself to just browser with web developer toolbar, you can use other browsers such as chrome that does not do the no-cache thing so easily.
Another extension that handles the cookie part of the question with a little more ease than the rest is Close'n forget
https://addons.mozilla.org/en-US/firefox/addon/closen-forget
You can use it by toolbar button/context menu/keyboard shortcut in order to close the current page, discarding any cookies it held.
Optionally, the site can be removed from browser history and (again optionally) all the pages you visited in the current page can be removed from browser history too.
Simple solution is simple - use two different browsers.
Say chrome for your email and Firefox for dev.
Also, you have Cookie Swap. but it doesn't really work quite as smoothly as you'd like.
What Firefox add-ons do you use that are useful for programmers?
I guess it's silly to mention Firebug -- doubt any of us could live without it. Other than that I use the following (only listing dev-related):
Console2: next-generation error console
DOM inspector: as the title might indicate, allows you to browse the DOM
Edit Cookies: change cookies on the fly
Execute JS: ad-hoc Javascript execution
IE Tab: render a page in IE
Inspect This: brings the selected object into the DOM inspector
JSView: display linked javascript and CSS
LORI (Life of Request Info): shows how long it takes to render a page
Measure IT: a popup ruler.
URL Params: shows GET and POST variables
Web Developer: a myriad of tools for the web developer
Here are mine (developer centric):
FireBug - a myriad of productivity enhancing tools, includes javascript debugger, DOM inspector, allows you to edit the CSS/HTML on the fly which is highly valuable for troubleshooing layout and display problems.
Web Developer - again another great developer productivity tool. I mostly use it for quickly validating pages, disabling javascript (yes I disable javascript sometimes, don't you?), viewing cookies, etc.
Tamper Data - lets you tamper with http headers, form values, cookies, etc. prior to posting back to a page, or getting a page. Incredibly valuable for poking and prodding your pages, and seeing how your web app responds when used with slightly malicious intent.
JavaScript Debugger - has a few more features than javascript debugger provided by firebug. Although I must admit, I sparingly use this one since firebug has largely won me over.
Live HTTP Headers - invaluable for troubleshooting, use it frequently. Lets you spy on all HTTP headers communicated back and forth between client and server. It has helped me track down nefarious problems, especially when debugging issues when deploying your web app between environments.
Header Spy - nice addon for the geeky types, shows you the web server and platform a web site runs on in the status bar.
MeasureIt - I don't use this all too frequently, but I've still found it valuable from time to time.
ColorZilla - again, not something I use all that frequently, but when I need it, I need it. Valuable when you want to know a color and you don't want to dig through a CSS file, or open up a graphics editing app to get a color embedded in some image.
Add N Edit Cookies - this has been a great debugging tool in web farms where the load balancer writes a cookie, and uses the cookie value to keep your session "sticky". It allowed me to switch at will between servers to track down problems on specific machine. Also a good tool if you want to try to mess with a site that uses cookies to track your login status/account, and you want to see how your code responds to malformed or hacked info.
Yellowpipe Lynx Viewer Tool - yeah I know what your thinking, lynx, who needs it, its so 1994. But if you are developing a site that needs to take web accessibility into account (meaning accessible to users with visual impairments who use screen readers), or if you need to get a sense of how a web spider/indexer "sees" your site, this tool is invaluable. Granted, you could always just go out and grab Lynx for yourselfhere's the windows xp port that I use.
I've got a handful of other addons that I've used from time to time that I'll just quickly mention: FireFTP (one I installed wasn't stable and I've not tried a newer release), Html Validator (also found this one unstable, least back when I installed like a year ago), IE Tab (I usually just have both IE and FireFox open concurrently, but that is just me, I know many others that find this addon useful).
I'd also recommend the Web Developer extension by Chris Pederick.
As far as web development, especially for javascript, I find Firebug to be invaluable. Web developer toolbar is also very useful.
The ones I have are...
Y-SLow
Live Headers
Firebug
Dom Inspector
One that wasn't mentioned yet is this HTML Validator extension that I found very useful.
#Flávio Amieiro
MeasureIt is an unnecessary extension to have if you install the Web Developer Toolbar. Web Developer Toolbar includes a ruler as one of its features. Under the "Miscellaneous" category for Web Developer click the option "Display Ruler" to use a ruler identical to the MeasureIt one.
That will allow you reduce the number of extensions needed by at least one.
Firefox addons:
FireBug:helps web developers and designers test and inspect front-end code. It provides us with many useful features such as a console panel for logging information, a DOM inspector, detailed information about page elements, and much, much more.
Web Developer-gives you the power disable CSS, edit CSS on the fly, measure certain areas of a page and much more.
ColorZilla
Just click on the icon, hover over the area you'd like to know the hex color for, and click.
Window Resizer
to make sure the layout is displayed properly in the standard resolutions of today.
Total Validator
validating websites much easier by checking HTML, links, CSS and doing a lot more.
Web Developer for web development. Scribefire if you're a blogger-progammer
For web developing I use the Web Developer Toolbar, CSS Viewer and MeasureIt.
But I'm really not one of those who has a thousand of extensions to do everything. I like to keep things simple.
EDIT: Thanks to Dan's answer I don't need MeasureIt anymore. Can't believe I've never seen that! I guess I'll just have to pay more atention to this WebDeveloper toolbar.
Adding to everyones lists, Tamper Data is quite useful, lets you intercept requests and change the data in them.
It can be used to bypass javascript validation and check whether the server side is doing its thing.
I use Web Developer, it's a real time saver.
+1 for LORI ("life-of-request-info"). It's a very convenient alternative for rough measurements of the load time of a particular web page -- the kind of thing that you might otherwise use an external stopwatch for.
New Tab Homepage. Combined with a "speed dial"-type homepage (a personal, fast-loading page of links that you use frequently), helps you get where you're going faster when you open a new browser tab.
LastTab. Changes the behavior of Ctrl+Tab to let you navigate back and forth between your most-recently-used tabs with repeated presses of Ctrl+Tab, the same way that Alt+Tab works in Windows. Also provides a nice view of all open tabs while Ctrl is still being held down for easy navigation. (The resultant behavior is very similar to the Ctrl+Tab behavior in recent releases of Visual Studio.)
FireFTP is good for grabbing/uploading any necessary files.
I find Hackbar to be quite useful. Very useful if you want to edit the querystring part of the url, to test for vulnerabilities, or just general other types of testing where you might end up with complicated query string values.
I was learning DOM inspector, but I've switched to Firebug.
Some of which has been missed above are here
Load Time Analyzer – View detailed graphs of the loading time of web pages in firefox. The graphs display events like page requests, image loading times etc.
Poster – A must have tool for web developers enabling them to interact with web services and other web resources.
Aardvark – A cool extension for web developers and designers, allows them to view CSS attributes, id, class by highlighting page element individually.
Fiddler is a really great debugging proxy. Think of it as a more powerful version of the "Net" panel in Firebug or the Live HTTP headers.
It used to be an IE-only extension, now it also has hooks into Firefox.
Groundspeed, is useful for testing server side code. It was created for input validation tests during pentest, but can be useful for any test that require manipulating input (similar to TamperData).
It lets you control the form elements in the page, you can change their type and other attributes (size, lenght, javascript event handlers, etc). So for example you can change a hidden field or a select to a textbox and then enter any value to test the server response and stuff like that.