Leak checking:
sudo privbind -u skaller env FLX_FINALISE=1 nohup valgrind --leak-check=full
--undef-value-
errors=no build/release/host/bin/flx_web --port=80 &
and my favourite thing ..
http://felix-lang.org/$/home/skaller/felix/nohup.out
The webserver reading its own log file and sending it t
Well felix webserver played up again. So now I thought, ha! I have log file
So I tried to look at it. And its full of shit:
Fetching line ...
Got the line length .. 0
Fetching line ...
Got the line length .. 0
Fetching line ...
Got the line length .. 0
Fetching line ...
Got the line length .
The webserver is now running again. Felix version bumped to 1.1.12.
On both my Mac and Ubuntu server I can run
FLX_MIN_MEM=0 FLX_FINALISE=1 make test
and it all works (except a couple of actual test failures). With the new system
this really hammers the GC.
However Mark reports his build still d
At present flx_web uses an escape like so:
http://localhost/$/usr/local
to get past "relative to webserver root". I did this because HTML URL's are
so very hard to work with I couldn't figure out how to refer to file / any
other way.
This works just fine on flx_web, however if you put
The good news is: Shayne and I have got Felix building so well
on Windows that the tools/webserver not only builds .. it actually
worked (first time after it compiled!!)
The bad news is: the OSX build works but the Linux webserver
will not show *.png files. I have no idea why: it's the same progra
On 12/04/2012, at 1:01 AM, ben wrote:
> hi
>
> The webserver in /usr/local/bin/webserver respond to static html
> requests. But when I try to call a fdoc it crashes with the following
> message:
>
Hmm .. that's weird. I just had the same problem on felix-lang.org.
Works fine on my Mac though
On 12/04/2012, at 8:54 AM, john skaller wrote:
>
> Also, which C++ compiler are you using?
> Fdocs display fine for me on a Macbook Pro running Snow Leopard, and
>
> i686-apple-darwin10-gcc-4.2.1 (GCC) 4.2.1 (Apple Inc. build 5666) (dot 3)
>
> It should work with clang 3.0 as well, though I ha
On 12/04/2012, at 1:01 AM, ben wrote:
> hi
>
> The webserver in /usr/local/bin/webserver respond to static html
> requests. But when I try to call a fdoc it crashes with the following
> message:
>
>
> $ /usr/local/bin/webserver
>
> [...]
> Spawned fthread running for socket 5
> got line fro
hi
The webserver in /usr/local/bin/webserver respond to static html
requests. But when I try to call a fdoc it crashes with the following
message:
$ /usr/local/bin/webserver
[...]
Spawned fthread running for socket 5
got line from socket 5: GET /web/document_index.fdoc HTTP/1.1
well formed ge
With help from koettermar...@gmx.de who responded to my
cry on libcurl mailing list, I think the webserver now works properly.
Should be no connection reset problems now. There's a 20 second
time limit for receiving acknowledgement from your browser
to shut down its transmit end of the connection,
in Felix code, require package "fred" is now hyperlinked to
config/fred.fpc. Also the Requires, flx_requires_driver
and includes fields are hyperlinked.
So you can now browse Felix to find Felix include files,
C/C++ header files #included anywhere in header or
body statements, package resource fi
On 01/11/2010, at 1:07 PM, Jeff Schultz wrote:
> On Mon, Nov 01, 2010 at 12:36:56AM +1100, john skaller wrote:
>>> And yes, I get Connection Interrupted reliably for
>>> http://184.106.135.43:1234/string. The data sent starts with PAGE NOT
>>> FOUND: string and is then interrupted.
>
>> Can yo
On 31/10/2010, at 10:04 PM, Jeff Schultz wrote:
> On Sun, Oct 31, 2010 at 04:15:35AM +1100, john skaller wrote:
>> The felix webserver is now up and running:
>>
>> http://184.106.135.43:1234/
>
> John, rat-on-speed testing found
lol !
>http://184.106.135.43:1234/windows.h
>
>NO FILE
On 31/10/2010, at 4:15 AM, john skaller wrote:
> The felix webserver is now up and running:
>
> http://184.106.135.43:1234/
>
This is now working well for me with a --close-delay=20.0, which means
my connection isn't shutdown for 20 seconds after the last write.
The delay preceds a shutdown(
The felix webserver is now up and running:
http://184.106.135.43:1234/
This is still very rudimentary (both the server and the website).
A few things don't work. At my end, I sometimes get
firefox telling me the connection was reset whilst it was loading
a page .. please try it out and let me k
On 28 October 2010 12:42, john skaller wrote:
>
> On 28/10/2010, at 9:18 PM, Rhythmic Fistman wrote:
>
>>> Date: Thu, 28 Oct 2010 09:12:24 +1100
>>> From: john skaller
>>
>>
>>> I guess I use open() then dup2() to create fd's for 0,1,2 so
>>> the webserver can still do I/O.. not sure what happens
On 28 October 2010 12:43, john skaller wrote:
>
>>
>> Why are you daemonising? What's wrong with
>> nohup tools/webserver&
>
> No idea, didn't think of it :)
> That's why I asked for help :)
Keep an eye on stdout and stderr. I'm not sure what happens there.
--
>
> Why are you daemonising? What's wrong with
> nohup tools/webserver&
No idea, didn't think of it :)
That's why I asked for help :)
--
john skaller
skal...@users.sourceforge.net
--
Nokia and AT&T present the 2010
On 28/10/2010, at 9:18 PM, Rhythmic Fistman wrote:
>> Date: Thu, 28 Oct 2010 09:12:24 +1100
>> From: john skaller
>
>
>> I guess I use open() then dup2() to create fd's for 0,1,2 so
>> the webserver can still do I/O.. not sure what happens
>> to C++ streams cin, cout, cerr or C streams stdin,
> Date: Thu, 28 Oct 2010 09:29:54 +1100
> From: john skaller
> I will have to put the daemonisation into flx_(a)run
> and make it happen *before* the demux thread is open.
>
> The problem is .. how do I *tell* it?
>
> Hmm .. it would be better if there was just a "daemon" command,
> so I could do
> Date: Thu, 28 Oct 2010 09:12:24 +1100
> From: john skaller
> I guess I use open() then dup2() to create fd's for 0,1,2 so
> the webserver can still do I/O.. not sure what happens
> to C++ streams cin, cout, cerr or C streams stdin, stdout. stderr
who even uses that stuff?
---
On 27/10/2010, at 12:53 AM, john skaller wrote:
>>
>>
>> OK, I have checked the remote when the Content-Length field is set,
>> and that works. So it seems to be necessary. Grrr.. rewrite of webserver
>> coming up ;(
>
>
> Ok, it is done and checked in, and now everything works fine on local
>
>
> OK, I have checked the remote when the Content-Length field is set,
> and that works. So it seems to be necessary. Grrr.. rewrite of webserver
> coming up ;(
Ok, it is done and checked in, and now everything works fine on localhost,
however with the remote server, the old problem is back
On 25/10/2010, at 11:23 PM, john skaller wrote:
>
> On 25/10/2010, at 10:41 PM, john skaller wrote:
>
>>
>> HTML 0.9 (what the server runs) is a frameless protocol.. why do these
>> clients
>> think that the page wasn't finished?
>
>
> BTW: this problem doesn't occur on my localhost. Howeve
On 25/10/2010, at 10:41 PM, john skaller wrote:
>
> HTML 0.9 (what the server runs) is a frameless protocol.. why do these clients
> think that the page wasn't finished?
BTW: this problem doesn't occur on my localhost. However whilst curl gets all
the
data across the internet .. Firefox doesn
On linux I get this problem with the webserver:
curl
return o + ')';
}
}
}
curl: (56) Failure when receiving data from the peer
What am I doing wrong here? That is in fact the whole page. Firefox says
"connection reset whilst the page was loading". In fact it displays the p
if you launch the webserver by
flx tools/webserver --root=INSTALL_DIR
then you can now browse the library "lib" using hyperlinked include
statements. A good link is:
http://localhost:1234/std/__init__.flx
as it includes most of the standard library.
On the downside:
BUGGER the old problem of
On 17/10/2010, at 11:53 PM, john skaller wrote:
>>
>> For GNU cp, the "-a" ("archive") option might help with that (or
>> "--preserve=all") by preserving timestamps when copying.
>>
>> -- James
>>
>> cp -a
>
>
> Actually cp -ra since I'm copying a whole tree... trying it..
> Thanks!
Actuall
On 17/10/2010, at 7:57 PM, James Dennett wrote:
> On Sun, Oct 17, 2010 at 1:18 AM, john skaller
> wrote:
>> The latest version of the tools/webserver.flx automagically colours Felix
>> files,
>> keywords, strings, comments etc.
>>
>> Stay tuned for the next version, which will hyperlink includ
On Sun, Oct 17, 2010 at 1:18 AM, john skaller
wrote:
> The latest version of the tools/webserver.flx automagically colours Felix
> files,
> keywords, strings, comments etc.
>
> Stay tuned for the next version, which will hyperlink include files.
>
> BTW: I think there is a bug in the install. Whe
The latest version of the tools/webserver.flx automagically colours Felix files,
keywords, strings, comments etc.
Stay tuned for the next version, which will hyperlink include files.
BTW: I think there is a bug in the install. When I compile using the
installed Felix, it takes ages, whenI use the
On Sat, Oct 16, 2010 at 2:28 AM, john skaller
wrote:
>
> Really? I thought it didn't do that, since migration is expensive in terms
> of cache coherency .. that being the idea behind affinity.. did this
> change?
I've seen some articles claim that the schedulers in linux 2.5 and 2.6
do a pretty g
On 16/10/2010, at 5:16 AM, James Dennett wrote:
> On Fri, Oct 15, 2010 at 9:29 AM, john skaller
> wrote:
>>
>> On 16/10/2010, at 2:34 AM, Rhythmic Fistman wrote:
>>
>> Of course .. this is what actual os threads do on a multi-core processor.
>> Although AFAIK in Linux a thread always runs on t
On Fri, Oct 15, 2010 at 9:29 AM, john skaller
wrote:
>
> On 16/10/2010, at 2:34 AM, Rhythmic Fistman wrote:
>
> Of course .. this is what actual os threads do on a multi-core processor.
> Although AFAIK in Linux a thread always runs on the same CPU.
> [Called "affinity" ..]
By default Linux migra
On 16/10/2010, at 2:34 AM, Rhythmic Fistman wrote:
>
> I also had some unclearness when I modified variables "above" the
> fthread's closure (?).
> Apple's "block" implementation makes outer variables read only, but fthreads
> are
> often synchronous (that's why I use them), so modification can
On 15 October 2010 16:57, john skaller wrote:
>
> On 15/10/2010, at 2:16 AM, Rhythmic Fistman wrote:
>
>>> From: john skaller
>>
>>> I have finally got the webserver to work. Unsurprisingly there were NO bugs
>>> in any
>>> of the C++ library code!
>>
>> Woohoo!
>>
>>> RF also ran into this prob
On 15/10/2010, at 2:16 AM, Rhythmic Fistman wrote:
>> From: john skaller
>
>> I have finally got the webserver to work. Unsurprisingly there were NO bugs
>> in any
>> of the C++ library code!
>
> Woohoo!
>
>> RF also ran into this problem with closures.
>
> Yes, I often had separate fthread
> From: john skaller
> I have finally got the webserver to work. Unsurprisingly there were NO bugs
> in any
> of the C++ library code!
Woohoo!
> RF also ran into this problem with closures.
Yes, I often had separate fthreads magically sharing local variables.
I'll add node.js to Erick's mong
Ok, the Felix webserver (in tools directory) is not behaving itself.
There seems to be an attempt to write on a closed connection,
after which the server just exits for unknown reason.
I note that webserver does NOT read the whole of the client's request,
only the first line (the GET line).
If t
I've been trying to get the tools/webserver working, because I'm thinking
of modifying it to act as a documentation server (by formatting stuff in
our special undecided way).
I have noticed broken pipes and many other bugs and I have seen this
ages ago on linux too. So whilst RF's demux code may b
40 matches
Mail list logo
