NetTalk Central

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - Stu

Pages: 1 2 3 [4] 5 6 ... 33
46
Yes!

Fantastic, keep thinking :) This would be very very cool.

47
Thanks Bruce.

So, just because I'm thick-headed and maybe not reading this right - There's no way to prepend or append a piece of text to a cookie-name on the server level, so we could have server-specific cookies?

48
Hi Jim,

I've run into this a bit, and it can be frustrating.

Basically, if the service needs to access printer drivers, or the nettalk->pdftools->pdf in browser (can't remember the reason, but it's set in my mind), then I add another windows account, log into it, make sure there's a default printer (doesn't have to exist physically, but the driver needs to be there, and defaulted). Once that's squared away, log back into the main account and set the service to run with the new windows account credentials (username/password).

However. I'm curious.

What do you mean by "doesn't respond to any requests" - Do you mean it's not actually serving when you use a browser to go to the url?

If so, a couple of things.

1. Make sure you have it serving fine in normal exe mode (which sounds like you have), visit it on a browser on the server itself to check locally.
2. Run as service, check both the url and also, if you're not already, try http://127.0.0.1:<port>, and http://localhost:<port>.
3. Make sure you have DebugView++ (or some debugview) running. Turn on as much logging as possible in the SelfService AND the Nettalk templates.
4. Recompile and run.
5. Watch the debug log for the starting of the service. Check any warnings / errors.
6. Then try visiting the url with the browser (locally). See what is being logged, check it, check it again :)

Hope that helps.

49
Hey Bruce,

I know this is a couple of years back, the post, but have been testing a new website that while in testing is on the same sub-domain as our intranet, and unless I use different browsers / incognito (which is fine, am happy to do this), I get logged out of the intranet.

Back then you were talking about the possibility of cookies being different per sub domain.

However, what's happening at the moment is that the servers are on the same sub-domain, it's just that the PORT numbers are different.

So.

QUESTION 1: Would the "Bind Session to IP Address" option in the Webserver template also take the port number into consideration? (easy enough for me to test, which I'm doing now)

[Edit] Didn't seem to recognise the port in testing, so probably not a question that needs asking.

QUESTION 2: If not, is there any way we could have the option to put a custom prefix/suffix on cookie names in the webserver template, to help out differentiate between two servers on the same domain/subdomain but different ports?

Cheers!

50
Hey Bruce, you get that I know it's hard right?

I'm not posting here for fun, or because I haven't worked on this problem for a while now.

I'm here because I'm at the end, I can't work it out, and I can't seem to work out a good way to even get started on debugging this.

That's why this post is here, to see if someone can help me.

So please - Come on. I know it's hard to debug like this.

-

I'll have a look at the stage on the form. From memory I did this, but it's foggy, so I'll give it a go, see if something falls out.

51
Have just had an instance where it happened to me.

In firefox, with firebug going.

Popup form from browse. No errors from firebug console.

Click on Cancel button.

Code: [Select]
<?xml version="1.0" encoding="utf-8"?>
<ajax-response></ajax-response>

That's the response.

But the window won't close on clicking Cancel.

I'm guessing Save won't work, but I can't click the save button, cause it's live data.

The only way out is to refresh via the address bar or click on the X in the popup window or to press ESC.

On pressing ESC, the firebug console shows the response is exactly the same as above:

Code: [Select]
<?xml version="1.0" encoding="utf-8"?>
<ajax-response></ajax-response>

Does this mean that the server is registering what it thinks is the close on the browse side, but the popup itself is not closing?

From memory this issue does happen with not-popup forms as well, but I guess I'm going to have to go back and check it and try and get it to break.

52
Thanks Mike, will check that out.

The issue isn't 100% of the time, and doesn't happen with everyone/location (ip/network). It's a frustrating one to track down. From my experience, if there's a javascript issue, then it will always trigger.

The only thing I can think of is that depending on the content on the page, some of which is custom built, it might be stuffing up for that record. But then after a few goes the save button will work, with the same record.

But yeah, I'll do a check of the javascript etc.

Stu

53
Can confirm that the issue still occurs with jGrowl removed.

54
Hey Bruce,

Definitely possible to remove the jGrowl, yeah. Will do that and see.

But definitely NOT possible to remove the custom autocomplete template stuff unfortunately. It's way too prevalent across the system. Would need to really spend a lot of time on it - or if you want to upgrade the nettalk autocomplete .. eh eh? Heh heh.

So yeah, I can remove one, but not the other.

55
Hi Johan,

Clarion 10.0.11975 (have not upgraded the last couple of updates, heard some stuff about SQL breaking, got scared off).
Nettalk 9.12.

Hi Bruce,

No, not using vtip. I did try once to get it working, but it created lots of havoc with .. With things not working similarly to this (but in a more wide-spread manner and 100% of the time).

That's the thing, if it's something like vtip (say, I use jGrowl to popup little messages, the messages being wrapped in a p_web._jsok() ) .. If it's something like that, why wouldn't it happen 100% of the time?

And the places this is happening, there is no usage of jGrowl (which is the only custom jquery thing I can think of. No. Wait.

I do use jQuery Autocomplete, my own custom templated version of it.

But again .. If it's not happening 100% of the time for the same information (ie, the exact same record with the same html appearing), then could it still be an xhtml issue?

More questions.

Thanks Bruce.

56
Hi Folks,

I'm trying to track down a particularly nasty issue, that could be multiple issues, that is occurring live and as of yet I haven't been able to solve it.

Basically, it's this.

1. User is in the webserver interface, adding information via netwebforms, saving information. Some of the forms are complex, some are simple.
2. Multiple offices, in different physical locations, accessing the webserver.
3. The user will be, for example, adding/updating Person information.
4. The user will add a few Person records, and update a couple.
5. Then, for the unknown reason, the user will have a Person record open (either Insert or Change), and the Save button won't work. The Cancel button won't work.
6. After clicking on the X button and refreshing the browser, the user will try again with the same information / Person record.
7. Sometimes they will try once and it will work okay, sometimes a few times.
8. And sometimes, they let me know, and I log into the webserver and have a look.
9. On occasion I've been able to replicate the issue, once, and then try again and it does work for me. Other times it's worked for me straight away.

Thoughts:
* My initial thought was that this might have something to do with an XHTML issue, where there is some custom html somewhere on the form, causing problems. But if that's the case, then I'm not smart enough to understand why with the same entity/information, it will sometimes work and sometimes not.

Does anyone have any idea what this might be, or various troubleshooting actions I could take to try and track down what might be happening.

And Bruce, if you're here .. Be gentle. I'm still sore after the shelacking our 7s rugby team got from some mob over your way.

57
Hi Johan,

Sometimes it pays to go in and click on the "Refresh Classes" button in the Classes tab of the extension template.

58
Hey Mike,

Is there any clue from say Firebug/Console as to what is happening?

Sounds like it might be that pesky xhtml "i don't work properly for heaps of different reasons" thing.

59
Hi Jim,

Does the user need to preview the pdf report?

If not, you could just generate it on the Send button click.

OR better still, call a webservice from the send button click which then goes away and does the stuff (generate pdf, send email), and allows the user to carry on.

OR a little more excitement, you could call the webservice but instead of generating the pdf just include a special custom link in the email (tailored for the person the email is sent to) that allows them to click and download the pdf report from their device.

Lots of good stuff comes out of that (tracking, understanding, streamlining process).

But if they need to see the report on the netwebform, then those things won't solve that problem.

It's a bit of a hack, but you could:
1. On the memory form remove the cancel button and change the save to a "Close".
2. Add a "Send" button to the form, hide it based on a session value (say sendreport:Okay).
3. On opening form, set to zero, p_web.SSV('sendreport:Okay',0).
4. When the report is generated, set to 1, p_web.SSV('sendreport:Okay',1).
5. Make sure you are refreshing the custom "Send" button after report is generated.
6. Send button should then be readonly/hidden (whichever you choose in the template) until the pdf is generated.

Hope that helps.

60
Hi Ashley,

One thing to be aware of is that there is no hard link between a menu and it's items and a browse/form entity in the way you are thinking.

The menu is just that. It's a bunch of links.

Setting a menu item to only be seen by a particular user filter doesn't affect the form itself, it only affects that menu item.

So it's in the form itself that you muck about with the security for that form.

If someone is able to access a form when they shouldn't be, then we probably need to go into more specific details about the issue.

* A good way to test this is to, on login, set a made up session variable to one, like .. p_web.SSV('oktoaccess:ContactForm',1), and then on the form, set the special security to check for that particular session variable being 1 (remove the level specific check).

That'll help you understand a little more what might be happening.

Pages: 1 2 3 [4] 5 6 ... 33