[Trinidad] added browser version support in skinning TRINIDAD-799

[Cristi Toth]

On Thu, Apr 17, 2008 at 9:58 PM, Blake Sullivan <blake.sullivan@...> wrote:

Andrew Robinson said the following On 4/17/2008 12:35 PM PT:

So do I read this correctly that for #3, 8 means 8.x so a max-version
of 8 means any browser agent with a major version of 8 or less an not
even look at the minor version?
 

I'm proposing that the version feature reflect the best floating point version number we can calculate for the browser, which will usually be a combination of the major and minor version, so the version for IE 5.5 will be the floating point number 5.5

Does anybody KNOW how the browsers express their  version exactly?
Because TrinidadAgent currently only has majorVersion and agentVersion (the latter being an unparsed string).
I will have to parse that agentVersion to also get out the minor version.
Does anybody know why this wasn't done initially? (any reason like too different between the browsers?)

 

8 is promoted to 8.0 and since max- means less-than-or-equal-to:max-version:8 means

version <= 8.0 == true

-- Blake Sullivan

--
Cristi Toth

-------------
Codebeat
www.codebeat.ro

[Blake Sullivan]

Cristi Toth said the following On 4/17/2008 2:00 PM PT:

On Thu, Apr 17, 2008 at 9:58 PM, Blake Sullivan <blake.sullivan@...> wrote:

Andrew Robinson said the following On 4/17/2008 12:35 PM PT:

So do I read this correctly that for #3, 8 means 8.x so a max-version
of 8 means any browser agent with a major version of 8 or less an not
even look at the minor version?
 

I'm proposing that the version feature reflect the best floating point version number we can calculate for the browser, which will usually be a combination of the major and minor version, so the version for IE 5.5 will be the floating point number 5.5

Does anybody KNOW how the browsers express their  version exactly?
Because TrinidadAgent currently only has majorVersion and agentVersion (the latter being an unparsed string).
I will have to parse that agentVersion to also get out the minor version.
Does anybody know why this wasn't done initially? (any reason like too different between the browsers?)

In the rich client agent code, we use different regular expressions for the different engines when extracting the floating point version  number.

-- Blake Sullivan

 

8 is promoted to 8.0 and since max- means less-than-or-equal-to:max-version:8 means

version <= 8.0 == true

-- Blake Sullivan

--
Cristi Toth

-------------
Codebeat
www.codebeat.ro

[Blake Sullivan]

Cristi Toth said the following On 4/17/2008 2:00 PM PT:

On Thu, Apr 17, 2008 at 9:58 PM, Blake Sullivan <blake.sullivan@...> wrote:

Andrew Robinson said the following On 4/17/2008 12:35 PM PT:

So do I read this correctly that for #3, 8 means 8.x so a max-version
of 8 means any browser agent with a major version of 8 or less an not
even look at the minor version?
 

I'm proposing that the version feature reflect the best floating point version number we can calculate for the browser, which will usually be a combination of the major and minor version, so the version for IE 5.5 will be the floating point number 5.5

Does anybody KNOW how the browsers express their  version exactly?
Because TrinidadAgent currently only has majorVersion and agentVersion (the latter being an unparsed string).
I will have to parse that agentVersion to also get out the minor version.
Does anybody know why this wasn't done initially? (any reason like too different between the browsers?)

I can't remember why I wrote the code the way I did 12 years ago, but assuming that I was lazy and didn't think that it probably mattered that much at the time is probably a good bet.

-- Blake Sullivan

 

8 is promoted to 8.0 and since max- means less-than-or-equal-to:max-version:8 means

version <= 8.0 == true

-- Blake Sullivan

--
Cristi Toth

-------------
Codebeat
www.codebeat.ro

[Cristi Toth]

On Thu, Apr 17, 2008 at 11:15 PM, Blake Sullivan <blake.sullivan@...> wrote:

Cristi Toth said the following On 4/17/2008 2:00 PM PT

Does anybody KNOW how the browsers express their  version exactly?
Because TrinidadAgent currently only has majorVersion and agentVersion (the latter being an unparsed string).
I will have to parse that agentVersion to also get out the minor version.
Does anybody know why this wasn't done initially? (any reason like too different between the browsers?)

I can't remember why I wrote the code the way I did 12 years ago, but assuming that I was lazy and didn't think that it probably mattered that much at the time is probably a good bet.

:)

I assume you can't "donate" prematurely those regex, right?
 

-- Blake Sullivan

 

8 is promoted to 8.0 and since max- means less-than-or-equal-to:max-version:8 means

version <= 8.0 == true

-- Blake Sullivan

--
Cristi Toth

-------------
Codebeat
www.codebeat.ro

[Jeanne Waldman]

I think if I saw max-version: 5, I'd think that all minor versions of 5
would work, too.
Otherwise, I would have written max-version: 5.0.
- Jeanne


Blake Sullivan wrote, On 4/17/2008 12:58 PM PT:

> Andrew Robinson said the following On 4/17/2008 12:35 PM PT:
>> So do I read this correctly that for #3, 8 means 8.x so a max-version
>> of 8 means any browser agent with a major version of 8 or less an not
>> even look at the minor version?
>>  
> I'm proposing that the version feature reflect the best floating point
> version number we can calculate for the browser, which will usually be
> a combination of the major and minor version, so the version for IE
> 5.5 will be the floating point number 5.5
>
>
> 8 is promoted to 8.0 and since max- means
> less-than-or-equal-to:max-version:8 means
>
> version <= 8.0 == true
>
> -- Blake Sullivan
>
>> If so, I like 3 as well.
>>
>> -Andrew
>>
>> On Thu, Apr 17, 2008 at 1:31 PM, Blake Sullivan
>> <blake.sullivan@...> wrote:
>>  
>>> If we agree that we like the we like the media query syntax and that
>>> the
>>> only issue is how to handle less than (as opposed the <=) for the
>>> max-version, then we can just collect up the proposals and pick one:
>>>
>>>  1) The verbose and explicit  (max-version-less-than:8).
>>>  2) Define that for the version feature, max-version means < not <=.
>>> Inconsistent with other uses of max (max-version:8)
>>>  3) Let the skinning author provide enough precision to avoid the
>>> need to
>>> distinguish between < 8 and <= a number that apporaches 8
>>> (max-version:7.99)
>>>  4) Add an operator suffix (max-version-lt:8)
>>>
>>>  1) is gross
>>>  2) is potentially confusing due to inconsistency
>>>  3) might not be immediately obvious and could theoretically have
>>> precision
>>> problems
>>>  4) is not immediately obvious either but incredibly flexible
>>>
>>>  I vote for 3) since it gets the job done and doesn't preclude doing
>>> more
>>> later.
>>>
>>>  -- Blake Sullivan
>>>
>>>
>>>
>>>
>>>  Andrew Robinson said the following On 4/17/2008 11:53 AM PT:
>>>
>>>
>>>
>>>    
>>>> http://www.w3.org/TR/REC-CSS2/media.html
>>>>
>>>> @import url("loudvoice.css") aural;
>>>>
>>>> so here are multiple groups of characters that show that spaces are
>>>> acceptable (import url and aural keywords in one "bunch")
>>>>
>>>> url("loudvoice.css")
>>>> shows that parenthesis with at least one argument is acceptable
>>>>
>>>> @media screen, print {
>>>> Shown that a comma separated list, unlike normal CSS selectors applies
>>>> to the whole @ (meaning that it wasn't "@meda screen, @media print")
>>>>
>>>> From css3 (http://www.w3.org/TR/css3-reader/):
>>>> @import "my-print-style.css" print;
>>>> here, a quoted string is permissible (goes with the url values in CSS
>>>>      
>>> rules)
>>>    
>>>> <?xml-stylesheet href="style1.css" type="text/css"
>>>>  media="screen and (color) and (max-width: 400px"?>
>>>> <?xml-stylesheet href="style2.css" type="text/css"
>>>>  media="reader and (max-device-ratio: 1/1)"?>
>>>> Hmmm.... interesting, but do we want to reuse something that relates
>>>> to CSS but is not in a CSS file?
>>>>
>>>> @media reader and (grid: 0)
>>>> Ah, now we are talking. This looks like what Blake was referring to
>>>>
>>>> From http://www.css3.info/preview/media-queries/:
>>>> @media all and (min-width: 640px) {
>>>> Even better, showing an "all" keyword and having "normal CSS
>>>> properties" in parens.
>>>>
>>>> http://www.css3.info/preview/attribute-selectors/:
>>>> Do we dare take RegExp like syntax from attr. selectors and apply them
>>>> to @agent rules?
>>>>
>>>>
>>>> So I can see Blake's suggestion being backed by these, but IMO
>>>> "max-version-less-than:8" is too long to remember.
>>>>
>>>> Perhaps just:
>>>> IE 5.5 or greater:
>>>> @agent ie and (min-version: 5.5)
>>>>
>>>> IE 5.0 or greater:
>>>> @agent ie and (min-version: 5)
>>>>
>>>> IE >= 5.0 and < 6.0:
>>>> @agent ie and (version: 5)
>>>> or (I like this one less):
>>>> @agent ie and (major-version: 5)
>>>>
>>>> IE <= 6.0:
>>>> @agent ie and (max-version: 6)
>>>>
>>>> IE < 6:
>>>> @agent ie and (max-version: 5.9)
>>>>
>>>> IE >= 6.0 and < 8.0:
>>>> @agent ie and (min-version: 6) and (max-version: 7.9)
>>>> same as:
>>>> @agent ie and (min-version: 6) and (max-version: 7)
>>>>
>>>> IE >= 6.0 and <= 8.0:
>>>> @agent ie and (min-version: 6) and (max-version: 8.0)
>>>>
>>>> IE >= 6.0 and <= 8.x:
>>>> @agent ie and (min-version: 6) and (max-version: 8)
>>>>
>>>> So x.y (ie 5.5) means precisely that, no vagueness and x (ie 6) means
>>>> major version x regardless of minor version. If it is too hard to
>>>> parse the decimal and remember it, "max-major-version",
>>>> "min-major-version" and "major-version" could be used for integer only
>>>> comparison with the major version and "max-version", "min-version" and
>>>> "version" could be used for full major.minor comparison.
>>>>
>>>> I think using something like 7.9 or  7.99 could theoretically be used
>>>> for less than but not equal to. I think the fewer number of keywords
>>>> the clearer it will be to use. Just my opinion.
>>>>
>>>> Just adding some thoughts to chew on since concrete ideas were
>>>> asked for.
>>>>
>>>> -Andrew
>>>>
>>>>
>>>> On Thu, Apr 17, 2008 at 12:26 PM, Cristi Toth <cristi.toth@...>
>>>>      
>>> wrote:
>>>    
>>>>      
>>>>> Hi guys,
>>>>>
>>>>> You're right, I should have discussed the format before committing
>>>>> it.
>>>>> I started fixing the issue using the format that was specified
>>>>> there...
>>>>> (there weren't to many comments on that issue btw...)
>>>>>  During I was fixing it, I noticed that XSS suppported multiple
>>>>>        
>>> versions,
>>>    
>>>>> so I adapted what was suggested on the issue to support that too.
>>>>>
>>>>> Anyway, lets get this subject out in a new thread
>>>>> and stick here to discussing the format.
>>>>>
>>>>> Guys, those of you that suggested some general guidelines, they all
>>>>>        
>>> sound
>>>    
>>>>> good,
>>>>> but please try to think of some concrete format that comply with
>>>>> those
>>>>> guidelines.
>>>>>
>>>>> If we decide a final format and implement it until its get released,
>>>>>        
>>> then no
>>>    
>>>>> big harm done.
>>>>>  So please be constructive ;)
>>>>>
>>>>> Thanks for any feedback!
>>>>>
>>>>> cheers,
>>>>> --
>>>>>
>>>>> Cristi Toth
>>>>>
>>>>> -------------
>>>>> Codebeat
>>>>> www.codebeat.ro
>>>>>
>>>>>
>>>>>        
>>>    
>
>

[Blake Sullivan]

Cristi Toth said the following On 4/17/2008 2:25 PM PT:

On Thu, Apr 17, 2008 at 11:15 PM, Blake Sullivan <blake.sullivan@...> wrote:

Cristi Toth said the following On 4/17/2008 2:00 PM PT

Does anybody KNOW how the browsers express their  version exactly?
Because TrinidadAgent currently only has majorVersion and agentVersion (the latter being an unparsed string).
I will have to parse that agentVersion to also get out the minor version.
Does anybody know why this wasn't done initially? (any reason like too different between the browsers?)

I can't remember why I wrote the code the way I did 12 years ago, but assuming that I was lazy and didn't think that it probably mattered that much at the time is probably a good bet.

:)

I assume you can't "donate" prematurely those regex, right?

I'm sure we can.

Matthias, can you give them to Christi.  They are in AdfAgent.getAgent().

-- Blake

 

-- Blake Sullivan

 

8 is promoted to 8.0 and since max- means less-than-or-equal-to:max-version:8 means

version <= 8.0 == true

-- Blake Sullivan

--
Cristi Toth

-------------
Codebeat
www.codebeat.ro

[Jeanne Waldman]

Actually, I take that back. I think if I saw max-version: 5, I'd think 5.0.
The reason is people talk about versions, like, 2.0.0.10 and 2.0.0.4,
etc, so if I saw 2.0, I'd think 2.0.0.0

Otherwise I like #3.

Jeanne Waldman wrote, On 4/17/2008 2:43 PM PT:

> I think if I saw max-version: 5, I'd think that all minor versions of
> 5 would work, too.
> Otherwise, I would have written max-version: 5.0.
> - Jeanne
>
>
> Blake Sullivan wrote, On 4/17/2008 12:58 PM PT:
>> Andrew Robinson said the following On 4/17/2008 12:35 PM PT:
>>> So do I read this correctly that for #3, 8 means 8.x so a max-version
>>> of 8 means any browser agent with a major version of 8 or less an not
>>> even look at the minor version?
>>>  
>> I'm proposing that the version feature reflect the best floating
>> point version number we can calculate for the browser, which will
>> usually be a combination of the major and minor version, so the
>> version for IE 5.5 will be the floating point number 5.5
>>
>>
>> 8 is promoted to 8.0 and since max- means
>> less-than-or-equal-to:max-version:8 means
>>
>> version <= 8.0 == true
>>
>> -- Blake Sullivan
>>
>>> If so, I like 3 as well.
>>>
>>> -Andrew
>>>
>>> On Thu, Apr 17, 2008 at 1:31 PM, Blake Sullivan
>>> <blake.sullivan@...> wrote:
>>>  
>>>> If we agree that we like the we like the media query syntax and
>>>> that the
>>>> only issue is how to handle less than (as opposed the <=) for the
>>>> max-version, then we can just collect up the proposals and pick one:
>>>>
>>>>  1) The verbose and explicit  (max-version-less-than:8).
>>>>  2) Define that for the version feature, max-version means < not <=.
>>>> Inconsistent with other uses of max (max-version:8)
>>>>  3) Let the skinning author provide enough precision to avoid the
>>>> need to
>>>> distinguish between < 8 and <= a number that apporaches 8
>>>> (max-version:7.99)
>>>>  4) Add an operator suffix (max-version-lt:8)
>>>>
>>>>  1) is gross
>>>>  2) is potentially confusing due to inconsistency
>>>>  3) might not be immediately obvious and could theoretically have
>>>> precision
>>>> problems
>>>>  4) is not immediately obvious either but incredibly flexible
>>>>
>>>>  I vote for 3) since it gets the job done and doesn't preclude
>>>> doing more
>>>> later.
>>>>
>>>>  -- Blake Sullivan
>>>>
>>>>
>>>>
>>>>
>>>>  Andrew Robinson said the following On 4/17/2008 11:53 AM PT:
>>>>
>>>>
>>>>
>>>>  
>>>>> http://www.w3.org/TR/REC-CSS2/media.html
>>>>>
>>>>> @import url("loudvoice.css") aural;
>>>>>
>>>>> so here are multiple groups of characters that show that spaces are
>>>>> acceptable (import url and aural keywords in one "bunch")
>>>>>
>>>>> url("loudvoice.css")
>>>>> shows that parenthesis with at least one argument is acceptable
>>>>>
>>>>> @media screen, print {
>>>>> Shown that a comma separated list, unlike normal CSS selectors
>>>>> applies
>>>>> to the whole @ (meaning that it wasn't "@meda screen, @media print")
>>>>>
>>>>> From css3 (http://www.w3.org/TR/css3-reader/):
>>>>> @import "my-print-style.css" print;
>>>>> here, a quoted string is permissible (goes with the url values in CSS
>>>>>      
>>>> rules)
>>>>  
>>>>> <?xml-stylesheet href="style1.css" type="text/css"
>>>>>  media="screen and (color) and (max-width: 400px"?>
>>>>> <?xml-stylesheet href="style2.css" type="text/css"
>>>>>  media="reader and (max-device-ratio: 1/1)"?>
>>>>> Hmmm.... interesting, but do we want to reuse something that relates
>>>>> to CSS but is not in a CSS file?
>>>>>
>>>>> @media reader and (grid: 0)
>>>>> Ah, now we are talking. This looks like what Blake was referring to
>>>>>
>>>>> From http://www.css3.info/preview/media-queries/:
>>>>> @media all and (min-width: 640px) {
>>>>> Even better, showing an "all" keyword and having "normal CSS
>>>>> properties" in parens.
>>>>>
>>>>> http://www.css3.info/preview/attribute-selectors/:
>>>>> Do we dare take RegExp like syntax from attr. selectors and apply
>>>>> them
>>>>> to @agent rules?
>>>>>
>>>>>
>>>>> So I can see Blake's suggestion being backed by these, but IMO
>>>>> "max-version-less-than:8" is too long to remember.
>>>>>
>>>>> Perhaps just:
>>>>> IE 5.5 or greater:
>>>>> @agent ie and (min-version: 5.5)
>>>>>
>>>>> IE 5.0 or greater:
>>>>> @agent ie and (min-version: 5)
>>>>>
>>>>> IE >= 5.0 and < 6.0:
>>>>> @agent ie and (version: 5)
>>>>> or (I like this one less):
>>>>> @agent ie and (major-version: 5)
>>>>>
>>>>> IE <= 6.0:
>>>>> @agent ie and (max-version: 6)
>>>>>
>>>>> IE < 6:
>>>>> @agent ie and (max-version: 5.9)
>>>>>
>>>>> IE >= 6.0 and < 8.0:
>>>>> @agent ie and (min-version: 6) and (max-version: 7.9)
>>>>> same as:
>>>>> @agent ie and (min-version: 6) and (max-version: 7)
>>>>>
>>>>> IE >= 6.0 and <= 8.0:
>>>>> @agent ie and (min-version: 6) and (max-version: 8.0)
>>>>>
>>>>> IE >= 6.0 and <= 8.x:
>>>>> @agent ie and (min-version: 6) and (max-version: 8)
>>>>>
>>>>> So x.y (ie 5.5) means precisely that, no vagueness and x (ie 6) means
>>>>> major version x regardless of minor version. If it is too hard to
>>>>> parse the decimal and remember it, "max-major-version",
>>>>> "min-major-version" and "major-version" could be used for integer
>>>>> only
>>>>> comparison with the major version and "max-version", "min-version"
>>>>> and
>>>>> "version" could be used for full major.minor comparison.
>>>>>
>>>>> I think using something like 7.9 or  7.99 could theoretically be used
>>>>> for less than but not equal to. I think the fewer number of keywords
>>>>> the clearer it will be to use. Just my opinion.
>>>>>
>>>>> Just adding some thoughts to chew on since concrete ideas were
>>>>> asked for.
>>>>>
>>>>> -Andrew
>>>>>
>>>>>
>>>>> On Thu, Apr 17, 2008 at 12:26 PM, Cristi Toth <cristi.toth@...>
>>>>>      
>>>> wrote:
>>>>  
>>>>>    
>>>>>> Hi guys,
>>>>>>
>>>>>> You're right, I should have discussed the format before
>>>>>> committing it.
>>>>>> I started fixing the issue using the format that was specified
>>>>>> there...
>>>>>> (there weren't to many comments on that issue btw...)
>>>>>>  During I was fixing it, I noticed that XSS suppported multiple
>>>>>>        
>>>> versions,
>>>>  
>>>>>> so I adapted what was suggested on the issue to support that too.
>>>>>>
>>>>>> Anyway, lets get this subject out in a new thread
>>>>>> and stick here to discussing the format.
>>>>>>
>>>>>> Guys, those of you that suggested some general guidelines, they all
>>>>>>        
>>>> sound
>>>>  
>>>>>> good,
>>>>>> but please try to think of some concrete format that comply with
>>>>>> those
>>>>>> guidelines.
>>>>>>
>>>>>> If we decide a final format and implement it until its get released,
>>>>>>        
>>>> then no
>>>>  
>>>>>> big harm done.
>>>>>>  So please be constructive ;)
>>>>>>
>>>>>> Thanks for any feedback!
>>>>>>
>>>>>> cheers,
>>>>>> --
>>>>>>
>>>>>> Cristi Toth
>>>>>>
>>>>>> -------------
>>>>>> Codebeat
>>>>>> www.codebeat.ro
>>>>>>
>>>>>>
>>>>>>        
>>>>    
>>
>>
>

[Andrew Robinson-5]

I'll be happy either way, but I think I now bend to the below
explanation of 5 == 5.0 from Jeanne's reasoning

In the future we could add:

@agent ie and (major-version: 5)

where only the major-version is compared.

-Andrew

On Thu, Apr 17, 2008 at 6:07 PM, Jeanne Waldman
<jeanne.waldman@...> wrote:

> Actually, I take that back. I think if I saw max-version: 5, I'd think 5.0.
>  The reason is people talk about versions, like, 2.0.0.10 and 2.0.0.4, etc,
> so if I saw 2.0, I'd think 2.0.0.0
>
>  Otherwise I like #3.
>
>  Jeanne Waldman wrote, On 4/17/2008 2:43 PM PT:
>
>
>
> > I think if I saw max-version: 5, I'd think that all minor versions of 5
> would work, too.
> > Otherwise, I would have written max-version: 5.0.
> > - Jeanne
> >
> >
> > Blake Sullivan wrote, On 4/17/2008 12:58 PM PT:
> >
> > > Andrew Robinson said the following On 4/17/2008 12:35 PM PT:
> > >
> > > > So do I read this correctly that for #3, 8 means 8.x so a max-version
> > > > of 8 means any browser agent with a major version of 8 or less an not
> > > > even look at the minor version?
> > > >
> > > >
> > > I'm proposing that the version feature reflect the best floating point
> version number we can calculate for the browser, which will usually be a
> combination of the major and minor version, so the version for IE 5.5 will
> be the floating point number 5.5
> > >
> > >
> > > 8 is promoted to 8.0 and since max- means
> less-than-or-equal-to:max-version:8 means
> > >
> > > version <= 8.0 == true
> > >
> > > -- Blake Sullivan
> > >
> > >
> > > > If so, I like 3 as well.
> > > >
> > > > -Andrew
> > > >
> > > > On Thu, Apr 17, 2008 at 1:31 PM, Blake Sullivan
> > > > <blake.sullivan@...> wrote:
> > > >
> > > >
> > > > > If we agree that we like the we like the media query syntax and that
> the
> > > > > only issue is how to handle less than (as opposed the <=) for the
> > > > > max-version, then we can just collect up the proposals and pick one:
> > > > >
> > > > >  1) The verbose and explicit  (max-version-less-than:8).
> > > > >  2) Define that for the version feature, max-version means < not <=.
> > > > > Inconsistent with other uses of max (max-version:8)
> > > > >  3) Let the skinning author provide enough precision to avoid the
> need to
> > > > > distinguish between < 8 and <= a number that apporaches 8
> (max-version:7.99)
> > > > >  4) Add an operator suffix (max-version-lt:8)
> > > > >
> > > > >  1) is gross
> > > > >  2) is potentially confusing due to inconsistency
> > > > >  3) might not be immediately obvious and could theoretically have
> precision
> > > > > problems
> > > > >  4) is not immediately obvious either but incredibly flexible
> > > > >
> > > > >  I vote for 3) since it gets the job done and doesn't preclude doing
> more
> > > > > later.
> > > > >
> > > > >  -- Blake Sullivan
> > > > >
> > > > >
> > > > >
> > > > >
> > > > >  Andrew Robinson said the following On 4/17/2008 11:53 AM PT:
> > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > > http://www.w3.org/TR/REC-CSS2/media.html
> > > > > >
> > > > > > @import url("loudvoice.css") aural;
> > > > > >
> > > > > > so here are multiple groups of characters that show that spaces
> are
> > > > > > acceptable (import url and aural keywords in one "bunch")
> > > > > >
> > > > > > url("loudvoice.css")
> > > > > > shows that parenthesis with at least one argument is acceptable
> > > > > >
> > > > > > @media screen, print {
> > > > > > Shown that a comma separated list, unlike normal CSS selectors
> applies
> > > > > > to the whole @ (meaning that it wasn't "@meda screen, @media
> print")
> > > > > >
> > > > > > From css3 (http://www.w3.org/TR/css3-reader/):
> > > > > > @import "my-print-style.css" print;
> > > > > > here, a quoted string is permissible (goes with the url values in
> CSS
> > > > > >
> > > > > >
> > > > > rules)
> > > > >
> > > > >
> > > > > > <?xml-stylesheet href="style1.css" type="text/css"
> > > > > >  media="screen and (color) and (max-width: 400px"?>
> > > > > > <?xml-stylesheet href="style2.css" type="text/css"
> > > > > >  media="reader and (max-device-ratio: 1/1)"?>
> > > > > > Hmmm.... interesting, but do we want to reuse something that
> relates
> > > > > > to CSS but is not in a CSS file?
> > > > > >
> > > > > > @media reader and (grid: 0)
> > > > > > Ah, now we are talking. This looks like what Blake was referring
> to
> > > > > >
> > > > > > From http://www.css3.info/preview/media-queries/:
> > > > > > @media all and (min-width: 640px) {
> > > > > > Even better, showing an "all" keyword and having "normal CSS
> > > > > > properties" in parens.
> > > > > >
> > > > > > http://www.css3.info/preview/attribute-selectors/:
> > > > > > Do we dare take RegExp like syntax from attr. selectors and apply
> them
> > > > > > to @agent rules?
> > > > > >
> > > > > >
> > > > > > So I can see Blake's suggestion being backed by these, but IMO
> > > > > > "max-version-less-than:8" is too long to remember.
> > > > > >
> > > > > > Perhaps just:
> > > > > > IE 5.5 or greater:
> > > > > > @agent ie and (min-version: 5.5)
> > > > > >
> > > > > > IE 5.0 or greater:
> > > > > > @agent ie and (min-version: 5)
> > > > > >
> > > > > > IE >= 5.0 and < 6.0:
> > > > > > @agent ie and (version: 5)
> > > > > > or (I like this one less):
> > > > > > @agent ie and (major-version: 5)
> > > > > >
> > > > > > IE <= 6.0:
> > > > > > @agent ie and (max-version: 6)
> > > > > >
> > > > > > IE < 6:
> > > > > > @agent ie and (max-version: 5.9)
> > > > > >
> > > > > > IE >= 6.0 and < 8.0:
> > > > > > @agent ie and (min-version: 6) and (max-version: 7.9)
> > > > > > same as:
> > > > > > @agent ie and (min-version: 6) and (max-version: 7)
> > > > > >
> > > > > > IE >= 6.0 and <= 8.0:
> > > > > > @agent ie and (min-version: 6) and (max-version: 8.0)
> > > > > >
> > > > > > IE >= 6.0 and <= 8.x:
> > > > > > @agent ie and (min-version: 6) and (max-version: 8)
> > > > > >
> > > > > > So x.y (ie 5.5) means precisely that, no vagueness and x (ie 6)
> means
> > > > > > major version x regardless of minor version. If it is too hard to
> > > > > > parse the decimal and remember it, "max-major-version",
> > > > > > "min-major-version" and "major-version" could be used for integer
> only
> > > > > > comparison with the major version and "max-version", "min-version"
> and
> > > > > > "version" could be used for full major.minor comparison.
> > > > > >
> > > > > > I think using something like 7.9 or  7.99 could theoretically be
> used
> > > > > > for less than but not equal to. I think the fewer number of
> keywords
> > > > > > the clearer it will be to use. Just my opinion.
> > > > > >
> > > > > > Just adding some thoughts to chew on since concrete ideas were
> asked for.
> > > > > >
> > > > > > -Andrew
> > > > > >
> > > > > >
> > > > > > On Thu, Apr 17, 2008 at 12:26 PM, Cristi Toth
> <cristi.toth@...>
> > > > > >
> > > > > >
> > > > > wrote:
> > > > >
> > > > >
> > > > > >
> > > > > >
> > > > > > > Hi guys,
> > > > > > >
> > > > > > > You're right, I should have discussed the format before
> committing it.
> > > > > > > I started fixing the issue using the format that was specified
> there...
> > > > > > > (there weren't to many comments on that issue btw...)
> > > > > > >  During I was fixing it, I noticed that XSS suppported multiple
> > > > > > >
> > > > > > >
> > > > > >
> > > > > versions,
> > > > >
> > > > >
> > > > > >
> > > > > > > so I adapted what was suggested on the issue to support that
> too.
> > > > > > >
> > > > > > > Anyway, lets get this subject out in a new thread
> > > > > > > and stick here to discussing the format.
> > > > > > >
> > > > > > > Guys, those of you that suggested some general guidelines, they
> all
> > > > > > >
> > > > > > >
> > > > > >
> > > > > sound
> > > > >
> > > > >
> > > > > >
> > > > > > > good,
> > > > > > > but please try to think of some concrete format that comply with
> those
> > > > > > > guidelines.
> > > > > > >
> > > > > > > If we decide a final format and implement it until its get
> released,
> > > > > > >
> > > > > > >
> > > > > >
> > > > > then no
> > > > >
> > > > >
> > > > > >
> > > > > > > big harm done.
> > > > > > >  So please be constructive ;)
> > > > > > >
> > > > > > > Thanks for any feedback!
> > > > > > >
> > > > > > > cheers,
> > > > > > > --
> > > > > > >
> > > > > > > Cristi Toth
> > > > > > >
> > > > > > > -------------
> > > > > > > Codebeat
> > > > > > > www.codebeat.ro
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > > >
> > > >
> > >
> > >
> > >
> >
> >
>

[Andy Schwartz]

On Thu, Apr 17, 2008 at 9:21 PM, Andrew Robinson
<andrew.rw.robinson@...> wrote:
> I'll be happy either way, but I think I now bend to the below
>  explanation of 5 == 5.0 from Jeanne's reasoning

FWIW, I actually agree with Jeanne's first opinion. :-)

That is, I think of "5" as "5.*".  "5.0" as "5.0.*", etc.

Regarding the use of floating points to represent versions...  I was
wondering whether we should avoid this since it would prevents us from
supporting "major.minor.reallyminor" version strings.  I don't know
that we will ever need to go further than major.minor, though the
Gecko versions use the third digit, so perhaps we should pick a
solution that doesn't preclude us from supporting this?

(BTW, sorry all about my little digression earlier on the thread...)

Andy

[Andrew Robinson-5]

Well that is difficult isn't it? Perhaps what I suggested a long time
back in my bug is best and allow regexp:

@agent blah and (matches-version: /someRegExp/)

example to match 6.x through 7.x:
@agent ie and (matches-version: /[67](\.\d)*/)

yeah it is harder to write, but then we can write it once and it
handles pretty much all use cases.

The other twist is to give the code entire user agent string:

@agent matches(/MSIE\s+[67]/)

This really gives the user all the control they need and it is pretty
easy to parse without having to code many syntax improvements over
time as new requirements come up.

-Andrew

On Fri, Apr 18, 2008 at 12:10 PM, Andy Schwartz
<andy.g.schwartz@...> wrote:

> On Thu, Apr 17, 2008 at 9:21 PM, Andrew Robinson
>
> <andrew.rw.robinson@...> wrote:
>
> > I'll be happy either way, but I think I now bend to the below
>  >  explanation of 5 == 5.0 from Jeanne's reasoning
>
>  FWIW, I actually agree with Jeanne's first opinion. :-)
>
>  That is, I think of "5" as "5.*".  "5.0" as "5.0.*", etc.
>
>  Regarding the use of floating points to represent versions...  I was
>  wondering whether we should avoid this since it would prevents us from
>  supporting "major.minor.reallyminor" version strings.  I don't know
>  that we will ever need to go further than major.minor, though the
>  Gecko versions use the third digit, so perhaps we should pick a
>  solution that doesn't preclude us from supporting this?
>
>  (BTW, sorry all about my little digression earlier on the thread...)
>
>  Andy
>

[Matt Cooper]

Hi all,

The regex would be powerful though I'm afraid that it would not as
obvious or easy to use for non-technical designers/skinners.

I think something like this would be clearer:

@agent ie and (min-major-version: 6) and (min-minor-version: 1) and
(max-major-version: 6) {
  /* styles for IE agent version 6.1 through 6.x (inclusive) */
}

Regards,
Matt

On Fri, Apr 18, 2008 at 12:30 PM, Andrew Robinson
<andrew.rw.robinson@...> wrote:

> Well that is difficult isn't it? Perhaps what I suggested a long time
>  back in my bug is best and allow regexp:
>
>  @agent blah and (matches-version: /someRegExp/)
>
>  example to match 6.x through 7.x:
>  @agent ie and (matches-version: /[67](\.\d)*/)
>
>  yeah it is harder to write, but then we can write it once and it
>  handles pretty much all use cases.
>
>  The other twist is to give the code entire user agent string:
>
>  @agent matches(/MSIE\s+[67]/)
>
>  This really gives the user all the control they need and it is pretty
>  easy to parse without having to code many syntax improvements over
>  time as new requirements come up.
>
>  -Andrew
>
>  On Fri, Apr 18, 2008 at 12:10 PM, Andy Schwartz
>
> <andy.g.schwartz@...> wrote:
>
>
> > On Thu, Apr 17, 2008 at 9:21 PM, Andrew Robinson
>  >
>  > <andrew.rw.robinson@...> wrote:
>  >
>  > > I'll be happy either way, but I think I now bend to the below
>  >  >  explanation of 5 == 5.0 from Jeanne's reasoning
>  >
>  >  FWIW, I actually agree with Jeanne's first opinion. :-)
>  >
>  >  That is, I think of "5" as "5.*".  "5.0" as "5.0.*", etc.
>  >
>  >  Regarding the use of floating points to represent versions...  I was
>  >  wondering whether we should avoid this since it would prevents us from
>  >  supporting "major.minor.reallyminor" version strings.  I don't know
>  >  that we will ever need to go further than major.minor, though the
>  >  Gecko versions use the third digit, so perhaps we should pick a
>  >  solution that doesn't preclude us from supporting this?
>  >
>  >  (BTW, sorry all about my little digression earlier on the thread...)
>  >
>  >  Andy
>  >
>

[Andrew Robinson-5]

The problem already brought up is that minor version may not be
enough. In a 3 part version (ie 2.0.10) the minor of 0 isn't helpful
if you want to do something based on the 10 value

-Andrew

On Fri, Apr 18, 2008 at 12:40 PM, Matt Cooper <matt.faces@...> wrote:

> Hi all,
>
>  The regex would be powerful though I'm afraid that it would not as
>  obvious or easy to use for non-technical designers/skinners.
>
>  I think something like this would be clearer:
>
>  @agent ie and (min-major-version: 6) and (min-minor-version: 1) and
>  (max-major-version: 6) {
>   /* styles for IE agent version 6.1 through 6.x (inclusive) */
>  }
>
>  Regards,
>  Matt
>
>  On Fri, Apr 18, 2008 at 12:30 PM, Andrew Robinson
>
>
> <andrew.rw.robinson@...> wrote:
>  > Well that is difficult isn't it? Perhaps what I suggested a long time
>  >  back in my bug is best and allow regexp:
>  >
>  >  @agent blah and (matches-version: /someRegExp/)
>  >
>  >  example to match 6.x through 7.x:
>  >  @agent ie and (matches-version: /[67](\.\d)*/)
>  >
>  >  yeah it is harder to write, but then we can write it once and it
>  >  handles pretty much all use cases.
>  >
>  >  The other twist is to give the code entire user agent string:
>  >
>  >  @agent matches(/MSIE\s+[67]/)
>  >
>  >  This really gives the user all the control they need and it is pretty
>  >  easy to parse without having to code many syntax improvements over
>  >  time as new requirements come up.
>  >
>  >  -Andrew
>  >
>  >  On Fri, Apr 18, 2008 at 12:10 PM, Andy Schwartz
>  >
>  > <