How is everyone else doing IE7 testing?

N

Nathan Sokalski

As we all know, IE7 is now available. However, it also cannot be installed
side by side with IE6. This makes it very hard to test pages on both
versions using the same machine. I am guessing that most businesses simply
have multiple machines, some with IE6 and some with IE7. However, being a
home developer who develops as both a personal hobby and potential career, I
need a way to test both versions from home with the same machine. The only
suggestion given to me from the microsoft.public newsgroups so far has been
to use VirtualPC, but that would be a lot of extra effort and not be very
efficient anyway. I am surprised that Microsoft has not created anything for
developers/designers that allows them to have IE6 and IE7 installed. If they
are willing to create utilities that prevent AutomaticUpdates from
installing it, I would think they would at least create a utility to run
them side by side if they want more people to start using IE7 (I know of
several universities that removed IE7 from the AutomaticUpdates at the
network level, so anybody not in a technology-related field might not even
know it exists for a while, and many businesses won't be installing it on
their machines for a while as well). I have been looking forward to IE7 and
following the blogs ever since I learned about them, but I'm not going to
give up having IE6 on my machine until people are no longer using it. What
are all of you doing as far as testing IE7 when limited to one machine?
Thanks.
 
R

Randy Webb

Nathan Sokalski said the following on 11/15/2006 12:25 AM:

What are all of you doing as far as testing IE7 when limited to one machine?

While I do have multiple machines to test on (both at home and work) and
keep my laptop with IE6 on it, check into a dual boot setup where you
can have multiple OS'es. I did that for a while with Win98 and WinXP at
home because I got tired of my wife fussing about me testing on her PC :)

But, as far as MS is concerned, IE6 will be old and dead shortly and the
track record for MS isn't very promising as far as supporting older
products that have an update out. Maybe it will be different for IE7 but
don't count on it.
 
A

Anthony Jones

Nathan Sokalski said:
As we all know, IE7 is now available. However, it also cannot be installed
side by side with IE6. This makes it very hard to test pages on both
versions using the same machine. I am guessing that most businesses simply
have multiple machines, some with IE6 and some with IE7. However, being a
home developer who develops as both a personal hobby and potential career, I
need a way to test both versions from home with the same machine. The only
suggestion given to me from the microsoft.public newsgroups so far has been
to use VirtualPC, but that would be a lot of extra effort and not be very
efficient anyway. I am surprised that Microsoft has not created anything for
developers/designers that allows them to have IE6 and IE7 installed. If they
are willing to create utilities that prevent AutomaticUpdates from
installing it, I would think they would at least create a utility to run
them side by side if they want more people to start using IE7 (I know of
several universities that removed IE7 from the AutomaticUpdates at the
network level, so anybody not in a technology-related field might not even
know it exists for a while, and many businesses won't be installing it on
their machines for a while as well). I have been looking forward to IE7 and
following the blogs ever since I learned about them, but I'm not going to
give up having IE6 on my machine until people are no longer using it. What
are all of you doing as far as testing IE7 when limited to one machine?
Thanks.

Frankly if you only have one machine then upgrade to IE7 and test on that.
It's more likely that things that work on IE6 will be broken on IE7 than the
other way round. Just make sure when developing you avoid using any new
features of IE7.
 
R

Randy Webb

Anthony Jones said the following on 11/15/2006 5:42 AM:

Frankly if you only have one machine then upgrade to IE7 and test on that.

And code to IE7, it's only a matter of time.
It's more likely that things that work on IE6 will be broken on IE7 than the
other way round.

Actually, its the other way around. Things that got fixed in IE7 will
still be broken in IE6.
Just make sure when developing you avoid using any new features of IE7.

Upgrade to IE7 and then ignore it?
 
J

Jesús López

I have experienced things working on IE7 and not working on IE6 and
viceversa, things woking on IE6 and not working on IE7.

Regards:

Jesús López
 
L

Laurent Bugnion

Hi,

Jesús López said:
> I have experienced things working on IE7 and not working on IE6 and
> viceversa, things woking on IE6 and not working on IE7.
>
> Regards:
>
> Jesús López

The IE team deliberately broke some things which worked on IE6 because
they were not compatible with the standards. This is especially true in
CSS. I think it was a clever and courageous decision. So as a rule of
thumb you can say that:

- What worked on IE6 and doesn't on IE7 was wrong to start with, and
should be corrected.
- What works on IE7 and doesn't on IE6 is just compliant with the
standards, which IE6 doesn't follow.

Of course IE7 is new, so it will have bugs.

My best memory from Mix06 is IE's product manager publicly apologizing
for what they did wrong in IE6 (5, 4...)

HTH,
Laurent
 
C

C A Upsdell

Laurent said:
- What worked on IE6 and doesn't on IE7 was wrong to start with, and
should be corrected.
- What works on IE7 and doesn't on IE6 is just compliant with the
standards, which IE6 doesn't follow.

My experience has been that, in some cases, different code is required
for IE6 and IE7, that 'fixing' the IE6 code in a way that makes it work
on IE7 may make it no longer work on IE6. So parallel testing for IE6
and IE7 is still necessary, and conditional comments continue to be a
vital tool.
Of course IE7 is new, so it will have bugs.

One problem is that Microsoft traditionally does not fix bugs, except
for security bugs, in minor updates. E.g., Microsoft may update IE7 to
fix security bugs, but not to fix (say) CSS bugs. So if IE7 misbehaves,
it will be necessary to code around it until IE8 (hopefully) comes out.
 
L

Laurent Bugnion

Hi,
My experience has been that, in some cases, different code is required
for IE6 and IE7, that 'fixing' the IE6 code in a way that makes it work
on IE7 may make it no longer work on IE6. So parallel testing for IE6
and IE7 is still necessary, and conditional comments continue to be a
vital tool.

Only if you wish to continue to support IE6. The need for that should be
evaluated on each case, I know that I won't continue to support IE6 for
my personal web applications, but I probably will have to support it for
the apps I develop in my firm. That said, I never had to resort to any
CSS tricks, I preferred to cancel (or modify) a requirement rather than
to use tricks. Of course I am lucky because the people in our marketing
department are reasonable, and accept the "not feasible in a standard
way" answer as a valid one.
One problem is that Microsoft traditionally does not fix bugs, except
for security bugs, in minor updates. E.g., Microsoft may update IE7 to
fix security bugs, but not to fix (say) CSS bugs. So if IE7 misbehaves,
it will be necessary to code around it until IE8 (hopefully) comes out.

I have a different experience. In one case at least (one case I know
of), Microsoft fixed a memory leak reported by us in an update
distributed through a security update this year. The advantage of fixing
bugs in security updates is that they are more likely to be installed ;-)

HTH,
Laurent
 
N

Nathan Sokalski

Do you know why Web Developers & Designers test their pages/sites on
multiple browsers?! They do it because there will never be a guarantee that
if it works in one browser it will work in another. Whether a page is more
likely to have problems in IE6 or IE7 is irrelevant here, because unless you
KNOW that a page works in both, you should test.
 
N

Nathan Sokalski

I could be wrong, but from what I understand VirtualPC is for multiple
operating systems. I do not want multiple operating systems on my machine, I
also don't want to install XP a second time and need to login twice in order
to use both browsers (not to mention that would mean each browser would be
running under a different username). Does anybody REALLY want to recreate
their whole system setup just to use another browser? Should we really need
to do that anyway?
 
M

Mark Rae

I could be wrong, but from what I understand VirtualPC is for multiple
operating systems.

That's right.
I do not want multiple operating systems on my machine,

So don't do it, then...Nobody's forcing you to do anything...
I also don't want to install XP a second time and need to login twice in
order to use both browsers

So decide whether you want IE6 or IE7 - you obviously can't have both
installed on the same machine...
(not to mention that would mean each browser would be running under a
different username).

??? No it wouldn't... Why would it mean that...?
Does anybody REALLY want to recreate their whole system setup just to use
another browser?

I have a virtual machine just for cross-browser testing - currently, it has
IE6, FireFox 2, Netscape 8.12, Opera 9 and Mozilla 1.7.3.

It takes about 10 seconds to launch and maybe another three seconds to log
in under exactly the same account as the host machine (or not...).

It also means that I can e.g. run a web app in debug mode on my host
machine, browse to it from one of the browsers on the virtual machine and
actually step through the code - same scenario with the various browsers
installed on another virtual machine running SuSE Linux, and the various
browsers installed on my Mac Mini...
Should we really need to do that anyway?

You don't *NEED* to do anything at all. You asked how people are currently
doing cross-browser testing, and the above is how I'm currently doing it.
 
A

Anthony Jones

Nathan Sokalski said:
Do you know why Web Developers & Designers test their pages/sites on
multiple browsers?! They do it because there will never be a guarantee that
if it works in one browser it will work in another. Whether a page is more
likely to have problems in IE6 or IE7 is irrelevant here, because unless you
KNOW that a page works in both, you should test.

Of course and if your concerned that it works in IE6 as well as IE7 then you
will probably want to test in Firefox (1.5 and 2), Netscape, Opera etc as
well. The established professionals take this as read.

However you stated that you only had one machine and setting up (possibly
having to purchase) a virtual machine was too much of a faf. You also
stated that you were going to stick with IE6 for a while. IMO that would be
a mistake go to IE7.

If you want to do things professionally buy a test machine, some extra
copies of appropriate OSes and and some ghost software. Now you can test
your various target environments properly.
 
M

Mark Rae

Of course and if your concerned that it works in IE6 as well as IE7 then
you
will probably want to test in Firefox (1.5 and 2), Netscape, Opera etc as
well. The established professionals take this as read.
Obviously.

However you stated that you only had one machine and setting up (possibly
having to purchase) a virtual machine was too much of a faf. You also
stated that you were going to stick with IE6 for a while. IMO that would
be
a mistake go to IE7.

I agree, but simply can't understand the problem with Virtual PC. It's free
to download, takes about a minute to install, and then you have an unlimited
amount of virtual machines (subject obviously to hard disk space and
licensing *) which you can create and destroy as your business needs
dictate...

If you require nothing more than cross-browser testing on Windows XP, this
would be a one-off exercise which would take less than an hour...
If you want to do things professionally buy a test machine, some extra
copies of appropriate OSes and and some ghost software. Now you can test
your various target environments properly.

* Or, alternatively, get a basic MSDN subscription, which will allow you to
deploy as many copies of the various flavours of Windows on as many machines
(real or virtual) as you like, so long as this deployment is for development
and/or testing purposes...
 
R

Randy Webb

Anthony Jones posted the following on 11/16/2006 4:46 AM:
Of course and if your concerned that it works in IE6 as well as IE7 then you
will probably want to test in Firefox (1.5 and 2), Netscape, Opera etc as
well. The established professionals take this as read.

And most "established professionals" know that Netscape 8 uses the
Mozilla engine and the IE engine to render/process with so testing in FF
and then NS8 is moot.
However you stated that you only had one machine and setting up (possibly
having to purchase) a virtual machine was too much of a faf. You also
stated that you were going to stick with IE6 for a while. IMO that would be
a mistake go to IE7.

Everybody said the same thing about IE6. To date, my personal experience
has been that when the browser is first released to the public, stay
away from it as the majority of users will still be using the previous
version. That was true when going from IE4 to 5.0, from 5.0 to 5.5, from
5.5 to 6.0 and now from 6.0 to 7.0.

The converse is true with Opera and FF though where the users tend to
update them pretty frequent as they chose to use that browser, most IE
users don't have a choice (either don't have one or don't know they have
one).
If you want to do things professionally buy a test machine, some extra
copies of appropriate OSes and and some ghost software. Now you can test
your various target environments properly.

Really? So to do things "professionally" I need a second machine, some
extra copies of OSes and ghost software? I do my work professionally and
I do it from the same PC I "play" on. You can post a URL to

comp.lang.javascript,
comp.info.www.authoring.html and comp.infosystems.www.authoring.stylesheets

And ask for a critique or how it looks in other browsers and the
regulars there are more than glad to test it for you. Some are more
critical than others.

So no, you don't have to spend a ton of money to do things
"professionally", you just have to know how to get it done. I have been
writing - professionally - cross-browser/cross-OS scripts for a little
over 6 years now and I have yet to own a mac.
 
A

Anthony Jones

Randy Webb said:
Anthony Jones posted the following on 11/16/2006 4:46 AM:

And most "established professionals" know that Netscape 8 uses the
Mozilla engine and the IE engine to render/process with so testing in FF
and then NS8 is moot.

Cool we can take one of those out of the list then.
Everybody said the same thing about IE6. To date, my personal experience
has been that when the browser is first released to the public, stay
away from it as the majority of users will still be using the previous
version. That was true when going from IE4 to 5.0, from 5.0 to 5.5, from
5.5 to 6.0 and now from 6.0 to 7.0.

The difference is that in many cases it's just going to arrive via windows
update and much has been made of IE6's past security difficencies. IMO IE7
will have a much faster uptake than the IE5/5.5 to IE6 transition. If you
only have one machine put IE7 on it.
The converse is true with Opera and FF though where the users tend to
update them pretty frequent as they chose to use that browser, most IE
users don't have a choice (either don't have one or don't know they have
one).


Really? So to do things "professionally" I need a second machine, some
extra copies of OSes and ghost software? I do my work professionally and
I do it from the same PC I "play" on. You can post a URL to

comp.lang.javascript,
comp.info.www.authoring.html and comp.infosystems.www.authoring.stylesheets

And ask for a critique or how it looks in other browsers and the
regulars there are more than glad to test it for you. Some are more
critical than others.

Yeah "professional" is probably too broad a term and it really does depend
on who your customers are and what they are going to do with the site you
are developing and... the list goes on. If my customers asked me how I
tested that my application works on multiple browsers, I'm fairly sure they
would be dismayed by 'Oh I made it publically available for all and sundry
to see, posted a link to it in some newsgroups and asked people the to come
and hack around with it'. I'm certain that they won't consider that to be
"professional".
So no, you don't have to spend a ton of money to do things
"professionally", you just have to know how to get it done. I have been
writing - professionally - cross-browser/cross-OS scripts for a little
over 6 years now and I have yet to own a mac.

Well done.
--
Randy
Chance Favors The Prepared Mind
comp.lang.javascript FAQ - http://jibbering.com/faq
Javascript Best Practices -
http://www.JavascriptToolbox.com/bestpractices/
 
M

Mark Rae

If my customers asked me how I
tested that my application works on multiple browsers, I'm fairly sure
they
would be dismayed by 'Oh I made it publically available for all and sundry
to see, posted a link to it in some newsgroups and asked people the to
come
and hack around with it'. I'm certain that they won't consider that to be
"professional".

I couldn't agree more! Anyone who believes that the above constitutes
testing really isn't worth bothering with...
 
D

Dave Anderson

Laurent said:
- What worked on IE6 and doesn't on IE7 was wrong to start with, and
should be corrected.
- What works on IE7 and doesn't on IE6 is just compliant with the
standards, which IE6 doesn't follow.

Of course IE7 is new, so it will have bugs.

It looks like I got a new version of jscript.dll when I upgraded to IE7, but
the Number.toFixed() bug still has not been fixed. Pathetic. I otherwise
like IE7 so far.
 
D

Dave Anderson

Randy said:
To date, my personal experience has been that when the browser
is first released to the public, stay away from it as the
majority of users will still be using the previous version.
That was true when going from IE4 to 5.0, from 5.0 to 5.5,
from 5.5 to 6.0 and now from 6.0 to 7.0.

What is different this time is that Microsoft plans to distribute the
browser as a high-priority update:

"To help our customers become more secure and up-to-date,
Microsoft will distribute Internet Explorer 7 as a high-
priority update via Automatic Updates soon after the final
version of the browser is released for Windows XP later
this year.[1]"

Furthermore, Microsoft advised companies to deploy the Blocker Toolkit
before November 2006, so this update could be imminent. Lastly, Windows
Vista users will not have IE6. I expect IE7 adoption will be much faster
than IE6. But that's just my opinion.
 
D

d d

This blocker toolkit you speak of. That stops you being
forced to get the IE7 update? I have Virtual Machines
setup with IE4, 5, 5.5, 6 and I don't want any of them to
receive IE7. They range from 98 to XPsp2.

Do you have a link? I realize it might be earlier in this
thread but yours is the only message I see in this thread.

~dd
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Members online

Forum statistics

Threads
473,744
Messages
2,569,483
Members
44,902
Latest member
Elena68X5

Latest Threads

Top