ab -n 1 -c 1000 -k http://localhost/
If your website is slow, Rocket will only speed it up if you have lots
concurrent connections.
-tim
-Original Message-
From: "Michael Toomim"
Sent: Tuesday, March 16, 2010 5:32pm
To: "web2py-users"
Subject: [web2py] Re: no more cherrypy wsgiser
On Mar 13, 2010, at 8:40 AM, mdipierro wrote:
> I do not see an option to do that.
You could delete all the files, and leave just a README with a pointer to the
hg repository.
The history would still be there, but anyone who tried to fetch the head would
just get the README.
>
> Massimo
>
On Mar 13, 2010, at 7:21 AM, mdipierro wrote:
> Yarko, this is bad. Somebody should tell them. ca you do that? If not,
> can you point me to the installer and I will do so?
Perhaps you should turn off access, or at least public access, to the svn
repository. There's no good reason anyone should
On Mar 12, 2010, at 2:09 PM, Timothy Farrell wrote:
> OK, in testing mod_proxy I've hit a snag. I'd like for someone else to take
> a look. I have web2py running on port 8000. Here's the relevant section of
> my httpd.conf (this is Apache 2.2.14):
>
>
>SetEnv force-proxy-request-1.0 1
>
OK, in testing mod_proxy I've hit a snag. I'd like for someone else to
take a look. I have web2py running on port 8000. Here's the relevant
section of my httpd.conf (this is Apache 2.2.14):
SetEnv force-proxy-request-1.0 1
SetEnv proxy-nokeepalive 1
ProxyBadHeader StartBody
I'm not sure how you upgraded, but make sure you have a rocket.py in
your gluon folder.
-tim
On 3/12/2010 3:42 PM, Jose wrote:
On 11 mar, 16:08, mdipierro wrote:
We moved from cherrypy wsgiserver to Rocket, by Timothy Farrell.
I included an older version, need to include the latest one
On Mar 11, 2010, at 2:07 PM, mdipierro wrote:
> mod_proxy yes. The other servers options do not use it. They do not
> use wsgiserver now.
Right, thanks.
>
>
> On Mar 11, 4:02 pm, Jonathan Lundell wrote:
>> On Mar 11, 2010, at 9:50 AM, mdipierro wrote:
>>
>>> He explained it partially here:
>
On Mar 11, 2010, at 9:50 AM, mdipierro wrote:
> He explained it partially here:
>
>https://launchpad.net/rocket/+announcements
>
> We will stress-test it with different browsers anyway.
Not just browsers, right, but backending other servers and server variations?
(mod_wsgi, mod_proxy, ngin
well we shouln't worry a lot about IE 5.5, no? 5.5!!
let me link this:
http://farm4.static.flickr.com/3609/3629069606_3d1a1cd8fb_b.jpg
On Thu, Mar 11, 2010 at 10:39 PM, mdipierro wrote:
> That is good enough.
>
> Don't we all love IE? When we discover bugs in our won code we can
> think of IE
So I was testing with IE 5.5+. and I hit a bug uploading...but the bug
is in IE. It kept failing on uploading very large files and I couldn't
figure it out. Turns out, IE was sending this http header:
Content-Length: -556031510
Oops. This is in IE 5.5 all the way to IE 8. I suspect that an
What about testing this with Selenium[1]? You can use many browsers as
"plugins" (IE, Safari, Firefox etc.).
There is a Python wrapper[2].
There is a project made by brazilians (at Globo.com) that translates
"natural language" to Selenium commands, called pyccuracy[3] - it
could help too.
[1] http
That was FF 3.6 on Win7. I'm going to try some less well behaved
browsers (IE 5.5+ via IEtester) next.
On 3/11/2010 2:21 PM, mdipierro wrote:
Which browsers? The problem with cherrypy< 3.x was for example that
different browser treated in different ways the server delay and some
browser trunc
I tried larger files (2GB - 4.5GB) but Firefox wouldn't let me submit
the form. Something about DVD images I suppose ;-)
On 3/11/2010 2:18 PM, Timothy Farrell wrote:
Slight correction:
db.define_table('image',Field('upload', 'upload'))
I have successfully up- and downloaded files as larg
Slight correction:
db.define_table('image',Field('upload', 'upload'))
I have successfully up- and downloaded files as large as 480MB and apps
as large as 160MB (any larger apps crashed on unzipping). In all cases
I was testing over HTTPS.
-tim
On 3/11/2010 1:04 PM, mdipierro wrote:
Ro
On Mar 11, 2010, at 9:50 AM, mdipierro wrote:
> He explained it partially here:
>
>https://launchpad.net/rocket/+announcements
>
> We will stress-test it with different browsers anyway.
I'm personally not that concerned.
My point is that saying 0.x raises all sorts of red flags for manage
15 matches
Mail list logo
