Re: Simple Proposal for setting HTTP headers

Hello Dirk, Ms2ger, others,

On 2013/07/23 5:26, Dirk Pranke wrote:
> On Mon, Jul 22, 2013 at 12:41 AM, Ms2ger<ms2ger@gmail.com>  wrote:
>
>> On 07/22/2013 06:29 AM, "Martin J. Dürst" wrote:
>>
>>> On 2013/07/21 17:07, Tobie Langel wrote:

>>>> Existing code (.htaccess) has a dependency we don't want to have:
>>>> Apache. So either way we'll be writing code.
>>>
>>> I got as far as understanding that some people don't want to depend on
>>> Apache. I haven't yet understood why. Pointer or explanation appreciated.
>>>
>>
>> Mozilla has the requirement that tests can be run locally by our
>> developers (volunteers as well as paid staff), and I think it's unlikely
>> that an Apache dependency will be considered acceptable.

I agree that it's good to be able to run tests locally. Actually, that 
was one of the motivations to starting this discussion; I want to be 
able to "test" the tests we are working on here locally before uploading 
them. But locally for me doesn't mean just accessing files in the file 
system, it means using HTTP locally. Otherwise, I don't see the point of 
discussing HTTP headers. And for using HTTP locally, one needs some kind 
of server.

Do the Mozilla tests run locally from the file system, or over HTTP? If 
the later, what kind of server is used? If the former, how would HTTP 
headers be relevant?

> For the record, Chromium/Blink's testing infrastructure currently requires
> either Apache or Lighttpd. We're not necessarily overjoyed to depend on
> Apache, but it's not upsetting either (i.e., it's quite acceptable).
> WebKit's requires Apache. Both are fully capable of running the tests
> locally, and I'm a bit puzzled by why that would be difficult. (It is
> admittedly harder on Win32).

I think "quite acceptable" is pretty close to my own experience. 
Actually, I have never had any kind of issues for myself or any of my 
students setting up Apache locally on Windows. I haven't done it 
recently, but it comes out of the box. Looking at 
http://httpd.apache.org/download.cgi, I see that Windows binaries are 
not available for 2.4.x, but they are available for 2.2.x, and I don't 
think we need the newest and greatest functionality for our tests.

Compared to installing Apache, compiling Mozilla Firefox from source was 
a seriously more difficult undertaking for those of my students who 
tried it.

> That said, coming up with something lighter weight wouldn't be a bad thing.

Agreed.

Regards,   Martin.

> However, we have a lot of proprietary tests that also require Perl or PHP,
> and so even if the W3C tests switched to all using Python, we might still
> require Perl/PHP (and hence Apache/Lighttpd) for some indefinite period of
> time.
>
> -- Dirk
>

Received on Tuesday, 23 July 2013 08:32:18 UTC