Old browsers and RT3rc3

Hi All,

Just wondering if RT is going to continue being usable by all browsers.

I was just showing someone RT3 who uses Netscape 4 (most of our clients
are still using Netscape 4) and found it was not working so well in
Netscape 4.

I haev so far discovered the following when following the link to a
ticket from the create email (/Ticket/Display.html?id#).

  • Login box was hard left.
  • after login I got an error because for some reason the URL was reduced
    to “/Ticket”
  • Once I manually typed in the url, the top left hand preferences/logout
    and Logged in as… was missing.

2 of them are breaking the application for Netscape 4 users, one is only
looks (login box).

Sorry to sound like a winger here, but I thought one of the better things
about RT was that it worked with older browsers.

I will try and see if I can’t offer a patch for this, but in the mean time
I figured I should atleast mention it.

Cheers,

Stewart

Hi All,

Just wondering if RT is going to continue being usable by all browsers.

I was just showing someone RT3 who uses Netscape 4 (most of our clients
are still using Netscape 4) and found it was not working so well in
Netscape 4.

I haev so far discovered the following when following the link to a
ticket from the create email (/Ticket/Display.html?id#).

  • after login I got an error because for some reason the URL was reduced
    to “/Ticket”

I’d bet quite strongly that that’s a configuration error, not a bug in RT.

Sorry to sound like a winger here, but I thought one of the better things
about RT was that it worked with older browsers.

Older no, widely used, yes.

Netscape 4 is one of the most horribly broken pieces of crap out there,
with regard to Cascading Style Sheets. It would be one thing if it just
didn’t support them, but no. It supports them wrong. When building
RT3, we had a choice between having our UI work well in a wide variety
of current browsers (including Mozilla, Netscape 6, IE5 and 6, lynx and w3m)
or having it work well in Netscape 4. If you need support for netscape
4, Best Practical can of course build you a custom frontend designed for
Netscape 4.

Request Tracker... So much more than a help desk — Best Practical Solutions – Trouble Ticketing. Free.

  • after login I got an error because for some reason the URL was reduced
    to “/Ticket”
    I’d bet quite strongly that that’s a configuration error, not a bug in RT.

Looks like you could be right. Now i have had a chance to look deeper into
this, my installation is not Returning anything from the code for the FORM
(). It seems to fail at the UNIVERSAL::can($r, ‘uri’)
portion (I ttested by doing (UNIVERSAL::can($r, ‘uri’) || “HELLO”), which
of course means the second part is not functioning and the web page is
getting a ACTION=“” which seems to be occuring for mozilla aswell,
obviously moz is alot saner in how it handles ACTION=“” than netscape.

Netscape 4 is one of the most horribly broken pieces of crap out there,
with regard to Cascading Style Sheets. It would be one thing if it just
didn’t support them, but no. It supports them wrong. When building
RT3, we had a choice between having our UI work well in a wide variety
of current browsers (including Mozilla, Netscape 6, IE5 and 6, lynx and w3m)
or having it work well in Netscape 4. If you need support for netscape
4, Best Practical can of course build you a custom frontend designed for
Netscape 4.

I have to agree with you on this. If it was not in such wide use I would
be happy to ignore the problems myself, The prefs/logout one is a stupid
problem (for some bizarre reason Netscape 4 wants a second closing SPAN),
and I do not care about the Login box being hard left. Now if I can just
figure out why UNIVERSAL::can is failing things should be OK.

I have no intention of making the system look as pretty in NS4 as it does
in any decent browser, but I will have to be sure it functionally works.

IF anyone can tip me off as to why UNIVERSAL::can is failing it would
really be appreciated. I’m running debian woody with perl 5.6.1 (Package
Version: 5.6.1-8.2).

Cheers,

Stewart