On 18-Mar-05 00:24:39 Albert Wiersch wrote:
[...]
When I try it, I get a page back titled You are not authorized to view this
page. When I switch the server to accepting basic authentication, it works.
I bet that you are using an old release of ICS. The NTLM is present
in the release of this
Scrive Francois Piette [EMAIL PROTECTED]:
Anyone has complained that he cannot send an empty data
using the POST or PUT method?
Well, I have one of my apps that doesn't work in some wases: when data to
send is empty. And it is
perfectly OK for the app to have a post with empty data. And
Hello,
I noticed that the HttpCli component return the 404 status code not
only when the server return this code but even under different
situations. For example when it is unable to resolve the hostname or
when wsocket raise an error (please correct me if I'm wrong).
This is a problem because
Scrive Francois Piette [EMAIL PROTECTED]:
Probably a solution could be to expose the FRequestDoneError field.
The status code should be meaningful only if FRequestDoneError=0.
You mean reporting 10060 (and others) to the error field ?
If you mean FRequestDoneError yes.
Yes, probably
Scrive Francois PIETTE [EMAIL PROTECTED]:
Probably a solution could be to expose the FRequestDoneError field.
The status code should be meaningful only if FRequestDoneError=0.
I verifyed and actually the OnRequestDone error reflect the winsock error !
Who use sync operation will get an
Hello,
I discovered that the HttpCli authentication has a serious problem.
When both the host and the proxy requiring an authentication then an
endless loop will generated (a sort of ping pong between 407 and 401
status).
I'm working for a fix and will send it to Francois when ready.
Bye,
Hello,
while I'm working on the fix for the authentication I would like to remove one
think that I don't like. The component will overwrite the Username and
Password properties when an user and a password is specified in the url.
The problem is that when the user set another url without user
Scrive Tibor Csonka [EMAIL PROTECTED]:
Hello,
In my oppinion username and(or) passwords in the url are really useless to
implement. Think about Internet Explorer which inhibed this also.
It is already implemented so since long time, and I think this is not limited
to the http url.
On 28-Apr-05 07:56:34 Francois Piette wrote:
Maybe a problem with my bad english :-(
If your english is bad, what I must say of my english? ;-)
I read your message twice and don't understand. The changed behaviour you
describe is the curent behaviour.
The point is that actually the username
On 28-Apr-05 18:32:19 Corey Murtagh wrote:
Francois Piette wrote:
Maybe a problem with my bad english :-(
I read your message twice and don't understand. The changed behaviour you
describe is the curent behaviour.
What Maurizio is basically saying is that the component should get the
username
On 24-Apr-05 23:58:10 Maurizio Lotauro wrote:
Hello,
I discovered that the HttpCli authentication has a serious problem.
When both the host and the proxy requiring an authentication then an
endless loop will generated (a sort of ping pong between 407 and 401
status).
Update: the problem appear
Hello,
I made some authentication test with the THttpCli component. I use
Ethereal to see what the component send and receive. With my big
surprise, when the component made an authentication using NTLM,
Ethereal show me the credential as clear text!!!
At this point the question is: the NTLM is
On 06-May-05 08:54:32 Marcello Vezzelli wrote:
Maurizio Lotauro wrote:
Hello,
I made some authentication test with the THttpCli component. I use
Ethereal to see what the component send and receive. With my big
surprise, when the component made an authentication using NTLM,
Ethereal show me
Scrive Bob Herbst [EMAIL PROTECTED]:
Hello Francois,
I am doing some updating of my programs and I ask is there something that I
am doing wrong?
I tried http://132.163.4.102:13 and what I see in the browser is in the
header. This is why it seems that you don't receive anything.
Bye,
Hello,
someone has made some test with the fixed version of the HttpCli
component that Francois put in his web page?
Any feedback is needed (negative or positive) to evaluate if this
version is reliable or not.
TIA
Bye, Maurizio.
--
To unsubscribe or change your settings for TWSocket
Hello,
when a PC has more than one IP address (more than one NIC, one NIC
and an active connection with the modem, ...) is it possible to
detect through which local IP will be used to reach a specific IP
address?
In other words, which gateway will be used for a remote IP?
Bye, Maurizio.
--
Scrive zayin [EMAIL PROTECTED]:
Hi,
If you are sending the command you can bind to a NIC.
If you are not binding, then the address of the NIC can be a clue as to
which one will be used.
What I need is to know which connection will be used for a specific ip address
to read the associated
On 28-May-05 23:04:13 Bob Herbst wrote:
[...]
The ICS HttpGet with or without the http:// (new updated version 26 May
2005), still does not show the browser output: Received 0 bytes.
Do you saw my post? The result is in the header and not in the
received stream.
Bye, Maurizio.
--
To
Scrive Francois PIETTE [EMAIL PROTECTED]:
I'm only read the changes in the code. IIRC you sayd in the past that
you don't like to have code based on external dll.
Yes, still the same. But this is the first step. The next step is to use
delphi code instead of the DLL. There are zlib
On 11-Jun-05 14:13:41 Francois PIETTE wrote:
Is there someone willing to implement NTLM
authentication into HTTP server component ?
If no other has offered yourself I can help you in this but only if
you have a documentation how to it should work. Otherwise I have no
idea how to start it.
Bye,
On 09-Jun-05 07:32:15 Francois PIETTE wrote:
[...]
If you want I can try to add a logic similar to the one that I made for
authentication (I'm not speaking of the fix).
You mean a class to encapsulate the compression feature ?
Well, why not ?
I looked into the changes but they are some steps
On 21-Jun-05 08:13:21 Francois Piette wrote:
Yes, and now I partially know how the client works internally. But I
never worked on the source of the server, so I should first learn how
it works internally.
It's quite simple. Each client has his own TWSocket instance. This reduce the
problem to
On 23-Jun-05 07:40:28 Francois Piette wrote:
As you can see the server must at least generate the Challenge when
speak with an ICS client, and keep it to elaborate the Message3.
While from the client side we are free to set flags and fields with
specific values, the server should be able to
Hello,
I finally get some time to check the changes proposed to handle the
gzip content encoding.
First I think that it should be better not to add specific gzip
handling but a generic class to handle the content encoding, using a
registration machanism for each encoding, like what happen with
On 11-Jul-05 19:23:20 Francois PIETTE wrote:
Welcome back Francois :-)
a) where is the best place to decode the received stream? Xavier do
this in the GetBodyLineNext when the end of document is reached.
I think decoding should be done on the fly.
What do you mean? Data should be decoded
Before start, I get your answer from the archive of this ML because
it seems that my ISP has lost a whole day of emails :(
On 12-Jul-05 02:41:45 Francois PIETTE wrote:
a) where is the best place to decode the received stream? Xavier do
this in the GetBodyLineNext when the end of document is
Scrive Francois PIETTE [EMAIL PROTECTED]:
class function THttpContentCoding.GetCoding: String;
const
BASE_CLASS_NAME = 'THttpContentCoding';
begin
if Pos(BASE_CLASS_NAME, ClassName) = 1 then
Result := Copy(ClassName, Length(BASE_CLASS_NAME), MAXINT)
else
Result := 'Unknown';
On 23-Jul-05 14:36:18 GSE Sas wrote:
Yes is last version.
Actually I don't use Proxy username and password because I don't where I can
find this data!!
Try user and password used to login into windows.
Bye, Maurizio.
--
To unsubscribe or change your settings for TWSocket mailing list
please
On 23-Jul-05 17:04:48 Merijn Terheggen wrote:
I noticed some behavior of HttpCli that caused me a LOT of headaches
before I discovered what was going on and made a work-around.
[...]
Are you using the original HttpCli or an inherited version made by
you?
Bye, Maurizio.
--
To unsubscribe or
On 24-Jul-05 12:01:55 Xavier Le Bris wrote:
Hello,
A contribution to compression with gzip and some answers to Maurizio's
questions :
It is a good idea to put the whole stuff in another unit.
a/ The site 'http://www.pipeboost.com/' works today for me in T_HttpCliGz.
If not, check if zlib.dll
On 26-Jul-05 20:37:54 Xavier Le Bris wrote:
Hello,
Hello,
[...]
FGzTime is necessary, not for me, but for anyone who wants to choose the
best compression level/time ratio.
I think that this is the job for a test application, not for the
component. You can use OnBeforeHeaderSend and OnDocEnd.
On 26-Jul-05 17:47:41 Francois PIETTE wrote:
[...]
We already talked about RcvdCount.
Yes, but we haven't still get a conclusion :-)
I think I said that RcvdCount that a choice has to be made and Ithat I have
no defitive answer. The idea is to break as less as possible existing code.
RcvdCount
Scrive Merijn Terheggen [EMAIL PROTECTED]:
I'm trying to write a spider that collects web pages.
In the LocationChange event, I check the properties of the new
location. For instance:
Ok, nothig strange that can interact with the message queue.
Francois, I reproduced the problem using
Scrive Francois Piette [EMAIL PROTECTED]:
[...]
I doubt that RcvdCount could be used to allocate storage. The body
data will be put into RcvdStream that is a stream, and normally a
stream is able to allocate the storage itself.
RcvdStream is not the only way to get data from the
Scrive Francois PIETTE [EMAIL PROTECTED]:
RcvdStream is not the only way to get data from the component. Some (many
?)
applications sue
OnDocData event to get data on the fly.
In that case the decompression should be made completly by the
application, or
should we add some
On 27-Jul-05 19:05:28 Francois PIETTE wrote:
Well, something went wrong with the attached document.
I've uploaded it at: http://www.overbyte.be.arch/OverByteHome.kmz
(This is a kind of hyperlink for Google Earth).
This doesn't work for me.
Bye, Maurizio.
--
To unsubscribe or change your
On 30-Jul-05 01:54:44 Maurizio Lotauro wrote:
On 28-Jul-05 08:34:58 Francois Piette wrote:
Do you want that the data passed to OnDocData is decompressed?
Yes I do.
Are you really sure? Ok Ok, I'll try to do it :-)
Done.
Now I have some points that I would like to discuss.
a) Exception
On 08-Aug-05 19:08:36 Francois PIETTE wrote:
Another interesting link is this one:
http://codecentral.borland.com/Item.aspx?id=16213
It contains code to validate usercode/password on a Windows domain.
It is a Delphi translation of the code in the article:
I repost some points to discuss:
a) Exception in THttpContentCoding.GetCoding method
This method is called indirectly during the initialization. It seems
that this is not the best moment to raise an exception.
When run from Delphi, if the Stop on Delphi exception is not
enabled, the developer see
On 13-Aug-05 08:46:47 Francois PIETTE wrote:
Thank Francois for your answer. I hope other people will partecipate
to this discussion.
a) Isn't it possible to defer call to THttpContentCoding.GetCoding until it
is really needed so that exceptions are raise at that time ?
I made some changes, now
On 15-Aug-05 10:52:10 Francois PIETTE wrote:
[...]
b.1) There is already a property Options which is a set of properties.
It
is better to extent this set.
First, I mention record but I mean object.
Usually I don't use anymore a set for a property mainly for two
reasons. The first is that set
On 16-Aug-05 14:21:00 Angus Robertson - Magenta Systems Ltd wrote:
If you don't understand what we are talking about, please ask some
questions to clarify. There are no stupid question, only people afraid
to ask.
I've not taken much notice of this because:
1 - the authentication change to
On 16-Aug-05 16:30:00 Angus Robertson - Magenta Systems Ltd wrote:
1. Your choice. You have now the opportunity to be sure the new
version doesn't break existing code, specially yours.
No, it's the job of the person introducing the changes to make sure the
component is backward compatible,
On 17-Aug-05 10:32:00 Angus Robertson - Magenta Systems Ltd wrote:
How do you think that all we heavily tested for weeks if nobody,
except the developer(s), made any test?
I only have a single development environment, and I need to be able to
produce bug fix versions of several applications at
On 17-Aug-05 15:17:26 Guillaume MAISON wrote:
[...]
Hi Maurizio (and others),
Hi Guillaume,
i'm sorry but i've crashed my old mail reader.
:-(
Could it be possible to send back the overall design pattern used for
the content encoding extension ?
It is very easy to understand if you know how
On 17-Aug-05 14:54:56 Francois Piette wrote:
[...]
The opinion I want to hear is about the design decisions that need to be done
about the HTTP client component supporting compression. See Maurizio messages
and answer to the technical questions he ask.
Don't forget to answer to Set properties
About recent discussion of HttpCli changes, I made a reply (on topic) to three
people: Angus, Guillaume and Francois.
I received my posts and see they in the archive of elists, but I never get a
reply and none are present in the archive.
Is it a problem of the ML or all want drop the discussion?
On 23-Aug-05 19:54:50 Francois PIETTE wrote:
I haven't noticed any problem with the list.
I have noticed that it is very difficult to have people actively
participating in development :-((
Unfortunately, most people here are interested in using ICS but not
developping it. Sometimes it is easier
On 23-Aug-05 19:39:00 Angus Robertson - Magenta Systems Ltd wrote:
About recent discussion of HttpCli changes, I made a reply (on topic)
to three people: Angus, Guillaume and Francois.
I received my posts and see they in the archive of elists, but I never
get a reply and none are present in
Scrive Francois Piette [EMAIL PROTECTED]:
Anyway, I would like to close this enhancement. What I absolute need
is to decide about how add the new parameters.
I spoke about the problemd with sets and VFI, i.e. Visual Form
Inheritance. Should I explain what kind of problem is?
Do you know
Scrive Angus Robertson - Magenta Systems Ltd [EMAIL PROTECTED]:
So you get it. Can you keep it and read when you have better time?
I've already said I will try to test your HTTP component when I have
the time, but that I will not use anything with a DLL, which is what
the zip I
Scrive Guillaume MAISON [EMAIL PROTECTED]:
Francois PIETTE a écrit :
[...]
I want that the two new properties be added to the Options property.
Names: httpoEnableContentCoding and httpoUseQuality
i'm wondering about the utility of such properties. Why not use the
ContentCoding and
On 24-Aug-05 22:13:53 Gunnar wrote:
The problem is that a set is practically a single value.
That's not a problem!
Yes it is when you use the VFI.
Take for example the
TFont.Style property. If you set the bold in an inherited form then in the
dfm you will see Font.Style = [fsBold]. Now if
On 24-Aug-05 19:28:11 Francois PIETTE wrote:
[...]
I think this behaviour is the normal expected behaviour of the way
inheritance work.
Yes and no. The feedback using the OI is that you change a single
property value, but effectly you change the value of a single bit.
[..]
In either case
On 26-Aug-05 09:05:35 Guillaume MAISON wrote:
Maurizio Lotauro a écrit :
i'm wondering about the utility of such properties. Why not use the
ContentCoding and Quality within the ContentCoding component and let the
HttpCli use it only when its ContentCoding property is assigned.
ContentCoding
?
There are a snipped of messages from Francois in this mailing list:
8
Subject: [twsocket] HTTP client component and document encoding
Date: 21 Jul 2005 17:15:19 GMT
Maurizio Lotauro is working on the HTTP client component to extend the work
already done by Xavier Le Bris (gzip support
On 29-Aug-05 09:11:37 Guillaume MAISON wrote:
Maurizio Lotauro a écrit :
I'm not sure that I understood what you mean. Can you elaborate? :-)
Well to make it simple :
[...]
or, as for what happens with TBitmap, TJpeg, etc... this can be
registered within initialization of the unit
On 22-Sep-05 14:12:39 Francois Piette wrote:
The best is that you try the latest version, both the latest released version
and the beta version Maurizio has done (both available from ICS page at my
website).
I doubt that the beta solve this problem because the changes are in a
different part of
Scrive Arno Garrels [EMAIL PROTECTED]:
By default UseBandwidthControl is defined in the project options.
Arno, the BandWidth properties are actually only in the SSL version.
Bye, Maurizio.
This mail has been sent using Alpikom webmail
Scrive Arno Garrels [EMAIL PROTECTED]:
Hello,
It's the first time I use the HttpCli component (I'm disappointed).
Basic authorization after basic proxy-authorization does not work (infinite
loop).
NTLM authorization in combination with basic proxy-authorization does not
work.
NTLM
On 13-Oct-05 15:33:48 Arno Garrels wrote:
Francois Piette wrote:
[...]
You can apply it to the version in ICS-SSL V4.
Those not using ICS-SSL may use the beta already available from my
website.
That's a good idea, I'll test it hard ;-)
Maurizio, can you please send me a copy, so I can add
On 13-Oct-05 15:33:39 Arno Garrels wrote:
Maurizio Lotauro wrote:
[...]
It is a know issue. I already fix it and a corrected version of httpcli
is on the ics site as beta.
Good to know, I was going to waste my time fixing it myself (not easy).
Don't remaind me, it needed two weeks of work
Scrive Arno Garrels [EMAIL PROTECTED]:
Maurizio Lotauro wrote:
[...]
Just tested you code, so far it is working smoothly, good job!
Thank you :-)
However, I changed the string-type of new properties ServerAuth as well
as ProxyAuth to a new custom type:
THttpAuthType = (httpAuthNone
On 16-Oct-05 07:29:49 Arno Garrels wrote:
Maurizio Lotauro wrote:
Can you explain what happen?
Because you get the loop problem I imagine that your situation is:
- the client request an URI
- the proxy answer with a 407 code
- the client authenticate to the proxy and start the SSL
On 17-Nov-05 20:00:16 Francois PIETTE wrote:
The question is: Should the HTTP client component implement this relative
path removal algorithm ?
If it is formally valid yes. But it could be tricky to do because
sometimes the relative path could be a little strange :-)
And the problem that Angus
On 19-Nov-05 09:59:28 Francois PIETTE wrote:
The question is: Should the HTTP client component implement this relative
path removal algorithm ?
If it is formally valid yes. But it could be tricky to do because
sometimes the relative path could be a little strange :-)
Can you give and example
On 19-Nov-05 19:48:43 Francois PIETTE wrote:
Am I right in thinking the HTTP compression stuff has not
yet made it into a released ICS?
Indeed. It was out of my head.
I should re-read the last messages published with that subject...
If you need any help to refresh your memory I'm there :-)
On 20-Nov-05 11:04:30 Francois PIETTE wrote:
The question is: Should the HTTP client component implement this
relative
path removal algorithm ?
If it is formally valid yes. But it could be tricky to do because
sometimes the relative path could be a little strange :-)
Can you give and example
On 20-Nov-05 12:10:00 Angus Robertson - Magenta Systems Ltd wrote:
[...]
I'm vaguely planning on adding ZLIB compression to the FTP client and
server (which FileZilla supports) and that will need a proper ZLIB
components using C OBJ files, once that is done HttpProt can use the
same stuff.
I
On 20-Nov-05 12:25:37 Francois PIETTE wrote:
[...]
As the compression is encapsulated in the HttpProt component, one could use
anything it likes for actual compression library with minimal effort. It is
not mandatory to use any DLL. My first idea was to validate the code using
any compressions
On 20-Nov-05 10:55:56 Francois PIETTE wrote:
[...]
I wonder if we shouldn't simply push the compression code into that beta.
This is probably the only way to have people really testing it (well beside
pushing it into the release).
Anyone having any opinion about that ?
I fully agree with you
On 20-Nov-05 14:59:00 Angus Robertson - Magenta Systems Ltd wrote:
[...]
The Borland VCL uses OBJ files for it's TCustomZlibStream class in
Delphi 7 and later (and for JPEG images), so if it's good enough for
them, it should be good enough for us. That's what I'll try and use,
although I might
Scrive Fastream Technologies [EMAIL PROTECTED]:
- Original Message -
From: Francois Piette [EMAIL PROTECTED]
To: ICS support mailing twsocket@elists.org
Sent: Monday, November 21, 2005 9:42 AM
Subject: Re: [twsocket] HTTP compression
I don't know the FTP component, maybe what
Scrive Guillaume MAISON [EMAIL PROTECTED]:
[...]
i do agree to say that using zlib or any other compression method has
more its place within a demo (why not write a special demo for using
compression with http based upon any of the actual demo, but named
specifically HttpCompression - for
On 21-Nov-05 12:15:00 Angus Robertson - Magenta Systems Ltd wrote:
[...]
If the new component needs external code adding to actually decompress a
page, we are effectively no further forward than several years ago, and
I'd just ignore any new attempt to add compression and use my old code,
Scrive Fastream Technologies [EMAIL PROTECTED]:
[...]
The server then starts compression (deflate works by packet-by-packet way)
I see that some ftp program can download a .gz version of a file. This is a
different think, right?
Bye, Maurizio.
On 24-Nov-05 21:05:00 Angus Robertson - Magenta Systems Ltd wrote:
I've merged your HttpProt changes into the latest version, and added
the LocationChangeMaxCount feature to prevent endless relocation looping.
Do you mean latest release, ICS beta or ICS-SSL?
The component is backward
On 25-Nov-05 10:03:07 Dan wrote:
[...]
Has Francois considered using CVS or something similar for ICS? Could come
in very handy, you could maintain the normal and SSL branches and could
merge fixes across both.
The standard ICS and SSL version share the same code. The specific
SSL parts are in
On 27-Nov-05 10:55:08 Paul wrote:
Hi Maurizio,
Anohter question. Is there involved an authentication?
Yes, it is and the user is requested to enter the proxy username and
password.
The proxyname and port are auto-detect.
Are you using HTTP 1.0 or 1.1? If the client doesn't keep alive the
On 28-Nov-05 09:21:48 Arno Garrels wrote:
Hello,
I'm actually writing a simple user manager component that I would like
to contribute to ICS. The component makes use of classes
TCollection/TCollectionItem to realize a tree-structure. A user group
contains accounts, accounts may have parameters.
On 30-Nov-05 15:21:24 Fastream Technologies wrote:
Hello,
I chose the second method (THttpCli) but the below code fails in a thread:
[...]
It gives no exception but throws ESocketError just after DNSLookup which is
successful. I tried with www.intel.com and the IE works just fine from here.
On 30-Nov-05 08:19:53 Arno Garrels wrote:
Maurizio Lotauro wrote:
[...]
Will they contains passwords?
Yes, with an option to save the MD5-hash only.
What is the purpose of the component?
The target is server only or client too?
[...]
Are there alternatives?
Why don't use the streaming
On 02-Dec-05 11:16:33 Arno Garrels wrote:
Maurizio Lotauro wrote:
On 30-Nov-05 08:19:53 Arno Garrels wrote:
Maurizio Lotauro wrote:
[...]
Will they contains passwords?
Yes, with an option to save the MD5-hash only.
What is the purpose of the component?
The target is server only
On 03-Dec-05 12:56:55 Fastream Technologies wrote:
Hello,
Does THttpCli support chunked encoding? If so, how do we indicate we can
accept it from server?
The history say that it is introduced on December 28, 2003.
Looking on the code it seems that there is no way to automatically
add the
On 10-Dec-05 19:33:07 Paul wrote:
No, this fine component from Maurizio detects proxy info from IE using
WinInet
Yes, but there is one point that is not clear to me, i.e. if the
WinInet is installed as part of OS and IE use it, or if it is part of
IE installation. It could be even possible that
On 23-Dec-05 18:57:05 Arno Garrels wrote:
Maurizio Lotauro wrote:
[...]
But that error is on a dialog box instead on the message windows. In this
case TForm1 is used in the dfm so probaly the compiler doesn't like the
define around cleass declaration.
I'm still using D5.
I tried
Scrive Francois PIETTE [EMAIL PROTECTED]:
I renamed ICS-V6 files to have ICS files clearly identified. The consequence
is that all source code using the component must be updated to change file
names in their uses clause.
This will not a big problem when it generate an error at compile
Scrive Wilfried Mestdagh [EMAIL PROTECTED]:
Hello,
I wants to wish TeamICS, Francois, and the whole community a happy new
year, good business in 2006 and most of all interesting projects :)
Thanks Wilfried, you too and to all others people in this ML.
Bye, Maurizio.
Scrive Francois PIETTE [EMAIL PROTECTED]:
A user reported having problem with a server handling badly NTLM. Using the
option httpoNoNTLMAuth doesn't really disable NTLM. The user has derived a
new class from the component and has added in the constructor
FServerAuth = httpAuthBasic;
Hello,
I just discovered a small bug in the HttpContCod unit. in the
THttpContentCoding.GetCoding the line
Result := Copy(ClassName, Length(BASE_CLASS_NAME), MAXINT)
must be
Result := Copy(ClassName, Length(BASE_CLASS_NAME) + 1, MAXINT)
Actually this bug is harmless, so no need to
Scrive Arno Garrels [EMAIL PROTECTED]:
Maurizio Lotauro wrote:
[...]
I just discovered that when Arno changed FServerAuth and FProxyAuth from
string to an ordinal type some code was modified in the wrong way.
Shall I change my job? It looks like that :(
I don't think so. Usually more
Scrive Francois PIETTE [EMAIL PROTECTED]:
A user reported having problem with a server handling badly NTLM. Using
the option httpoNoNTLMAuth doesn't really disable NTLM. The user has
derived a new class from the component and has added in the constructor
FServerAuth = httpAuthBasic;
Scrive Tibor Csonka [EMAIL PROTECTED]:
Hello list,
I've just ran into a small bug in NTLM authentication from HttpCli.
There is a possibility that the user enters the username as
domain\username.
In this case HttpCli will send the username and the domain as the
username, wich is not
Scrive Tibor Csonka [EMAIL PROTECTED]:
[...]
You have a similar situation with IIS? I mean, if you don't include the
domain
you will be authenticated?
It seemes to me that you have the same problem as me. Try to separe the
user and the domain. If it works let me know so I can patch
Scrive Paul [EMAIL PROTECTED]:
You can't know what the proxy wants,
so you have to test all possible situations and
pick the one that works
As Maurizio said, some users have to add the domain
and some not (on the same proxy!)
Is it possible for you to test with proxy that need [EMAIL
Scrive Francois PIETTE [EMAIL PROTECTED]:
That is what I ment also. Now the question is that should or should not
ICS allow other separators than backslash?
Should ICS be responsible for domain/user parsing out of a single property
or should a new property with domain added ?
I suggest
Scrive Tibor Csonka [EMAIL PROTECTED]:
I would like to correct the bug report I posted before:
[...]
A new property for the domain can be added for proxy auth and
separatelly for web auth or a property in which users can set the domain
name separator this can be one for each of the
Scrive Arno Garrels [EMAIL PROTECTED]:
[...]
I think they arn't so smart, but just calling LogOnUser() note
that [EMAIL PROTECTED] valid, not only for FTP-proxies.
I don't think that the server use LogOnUser to authenticate the request
because with NTLM it will not receive the password in
Scrive DZ-Jay [EMAIL PROTECTED]:
Fastream Technologies wrote:
Hello,
Thank you both for your replies. I found the problem myself: IE6 has a bug
that makes it expect a comma before Realm=
That's really weird. Does adding the comma break it on Firefox or
Opera? The RFC does
Scrive DZ-Jay [EMAIL PROTECTED]:
Maurizio Lotauro wrote:
Scrive Fastream Technologies [EMAIL PROTECTED]:
Firefox and Opera tolerates the comma well. You can try here:
http://www.fastream.net/beta/NetFileServer.exe (8.0.8B2). Just add a user
from HTTP-Users and then assign the user
1 - 100 of 222 matches
Mail list logo
