kritton-devel Mailing List for Kritton

On Sat, 2005-09-10 at 14:02, Vassilis Pandis wrote:
> --- Nigel Horne <nj...@ba...> wrote:
> 
> > > > jason corcoran wrote:
> > > > > On Thu, Sep 08, 2005 at 09:41:14AM +0100, Nigel Horne wrote:
> > > > > 
> > > > >>Before the design stage, I would suggest that you decide on
> > your
> > > > "usp",
> > > > >>which is business jargon for "unique selling point". Once
> > that's
> > > > decided
> > > > >>upon, move to the design phase.
> > > > >>
> > > > >>-Nigel
> > > > > 
> > > > > 
> > > > > 
> > > > >>From Pandis last mail, our usp should be - ease of use and
> > security
> > > > 
> > > > > We'll need to fluff this out though :)
> > > > 
> > > > By ease of use, do you mean ease of configuration/installation?
> > > > By security, do you mean that you are going to ensure an
> > > > unknowledgable
> > > > sys-admin?
> > > > 
> > > > > Jason.
> > > > 
> > > > -Nigel
> > > > 
> > > 
> > > Well, we can't do that. Easy of use, easy of
> > > configuration/installation. Configuration can be simple, but not
> > > simplistic. 
> > 
> > You need to be sure of your target audience. It looks as though you
> > expect some level of knowledge, but not to guru status. Is that
> > right?
> > 
> 
> Well, administration cannot happen without some knowledge. Of course we
> could take matters to extremes and even provide a fancy wizard to
> handle the configuration. Or, we could do both. A wizard for the
> totally beginners and a more-advanced-but-still-easy-to-use tool (think
> of KPPP's wizard for a new connection and the manual way of setting up
> one).
> 
> 
> > > In other words, we can have a scalable and configurable
> > > server, but configuration shouldn't neccessarily happen by editing
> > > files. We could have a very cute graphical interface for that, to
> > make
> > > things much easier for administrators. 
> > 
> > Or you could consider writing a webmin plugin.
> 
> Indeed. Have you any experience with webmin ?

Only as a user ;-)

-Nigel

--- Nigel Horne <nj...@ba...> wrote:

> > > jason corcoran wrote:
> > > > On Thu, Sep 08, 2005 at 09:41:14AM +0100, Nigel Horne wrote:
> > > > 
> > > >>Before the design stage, I would suggest that you decide on
> your
> > > "usp",
> > > >>which is business jargon for "unique selling point". Once
> that's
> > > decided
> > > >>upon, move to the design phase.
> > > >>
> > > >>-Nigel
> > > > 
> > > > 
> > > > 
> > > >>From Pandis last mail, our usp should be - ease of use and
> security
> > > 
> > > > We'll need to fluff this out though :)
> > > 
> > > By ease of use, do you mean ease of configuration/installation?
> > > By security, do you mean that you are going to ensure an
> > > unknowledgable
> > > sys-admin?
> > > 
> > > > Jason.
> > > 
> > > -Nigel
> > > 
> > 
> > Well, we can't do that. Easy of use, easy of
> > configuration/installation. Configuration can be simple, but not
> > simplistic. 
> 
> You need to be sure of your target audience. It looks as though you
> expect some level of knowledge, but not to guru status. Is that
> right?
> 

Well, administration cannot happen without some knowledge. Of course we
could take matters to extremes and even provide a fancy wizard to
handle the configuration. Or, we could do both. A wizard for the
totally beginners and a more-advanced-but-still-easy-to-use tool (think
of KPPP's wizard for a new connection and the manual way of setting up
one).


> > In other words, we can have a scalable and configurable
> > server, but configuration shouldn't neccessarily happen by editing
> > files. We could have a very cute graphical interface for that, to
> make
> > things much easier for administrators. 
> 
> Or you could consider writing a webmin plugin.

Indeed. Have you any experience with webmin ?




		
___________________________________________________________ 
To help you stay safe and secure online, we've developed the all new Yahoo! Security Centre. http://uk.security.yahoo.com

> > jason corcoran wrote:
> > > On Thu, Sep 08, 2005 at 09:41:14AM +0100, Nigel Horne wrote:
> > > 
> > >>Before the design stage, I would suggest that you decide on your
> > "usp",
> > >>which is business jargon for "unique selling point". Once that's
> > decided
> > >>upon, move to the design phase.
> > >>
> > >>-Nigel
> > > 
> > > 
> > > 
> > >>From Pandis last mail, our usp should be - ease of use and security
> > 
> > > We'll need to fluff this out though :)
> > 
> > By ease of use, do you mean ease of configuration/installation?
> > By security, do you mean that you are going to ensure an
> > unknowledgable
> > sys-admin?
> > 
> > > Jason.
> > 
> > -Nigel
> > 
> 
> Well, we can't do that. Easy of use, easy of
> configuration/installation. Configuration can be simple, but not
> simplistic. 

You need to be sure of your target audience. It looks as though you
expect some level of knowledge, but not to guru status. Is that right?

> In other words, we can have a scalable and configurable
> server, but configuration shouldn't neccessarily happen by editing
> files. We could have a very cute graphical interface for that, to make
> things much easier for administrators. 

Or you could consider writing a webmin plugin.

> 
> Don't know, this is just a suggestion. But don't you feel that we
> should not aim for yet-another-unix-webserver? If you don't like this
> idea, or have a better one, say so - everything is under discussion! 

Personally the project needs to do all it can to avoid that. So that is
why I am prompting discussion.

> Pandis

-Nigel 

--- Nigel Horne <nj...@ba...> wrote:

> jason corcoran wrote:
> > On Thu, Sep 08, 2005 at 09:41:14AM +0100, Nigel Horne wrote:
> > 
> >>Before the design stage, I would suggest that you decide on your
> "usp",
> >>which is business jargon for "unique selling point". Once that's
> decided
> >>upon, move to the design phase.
> >>
> >>-Nigel
> > 
> > 
> > 
> >>From Pandis last mail, our usp should be - ease of use and security
> 
> > We'll need to fluff this out though :)
> 
> By ease of use, do you mean ease of configuration/installation?
> By security, do you mean that you are going to ensure an
> unknowledgable
> sys-admin?
> 
> > Jason.
> 
> -Nigel
> 

Well, we can't do that. Easy of use, easy of
configuration/installation. Configuration can be simple, but not
simplistic. In other words, we can have a scalable and configurable
server, but configuration shouldn't neccessarily happen by editing
files. We could have a very cute graphical interface for that, to make
things much easier for administrators. 

Don't know, this is just a suggestion. But don't you feel that we
should not aim for yet-another-unix-webserver? If you don't like this
idea, or have a better one, say so - everything is under discussion! 

Pandis


	
	
		
___________________________________________________________ 
Yahoo! Messenger - NEW crystal clear PC to PC calling worldwide with voicemail http://uk.messenger.yahoo.com

jason corcoran wrote:
> On Thu, Sep 08, 2005 at 09:41:14AM +0100, Nigel Horne wrote:
> 
>>Before the design stage, I would suggest that you decide on your "usp",
>>which is business jargon for "unique selling point". Once that's decided
>>upon, move to the design phase.
>>
>>-Nigel
> 
> 
> 
>>From Pandis last mail, our usp should be - ease of use and security 
> We'll need to fluff this out though :)

By ease of use, do you mean ease of configuration/installation?
By security, do you mean that you are going to ensure an unknowledgable
sys-admin?

> Jason.

-Nigel

-- 
Nigel Horne. Arranger, Adjudicator, Band Trainer, Composer, Typesetter.
NJH Music, Barnsley, UK.  ICQ#20252325
nj...@de... http://www.bandsman.co.uk

On Thu, Sep 08, 2005 at 09:41:14AM +0100, Nigel Horne wrote:
> Before the design stage, I would suggest that you decide on your "usp",
> which is business jargon for "unique selling point". Once that's decided
> upon, move to the design phase.
> 
> -Nigel


From Pandis last mail, our usp should be - ease of use and security 
We'll need to fluff this out though :)

Jason.
-- 
Jason.

Fortune :
SIMM crosstalk.

Before the design stage, I would suggest that you decide on your "usp",
which is business jargon for "unique selling point". Once that's decided
upon, move to the design phase.

-Nigel

-- 
Nigel Horne. Arranger, Adjudicator, Band Trainer, Composer, Typesetter.
NJH Music, Barnsley, UK.  ICQ#20252325
nj...@de... http://www.bandsman.co.uk

Ok, I full agree on this with Jason. We need to set a goal, make a list
of things to be done, do them, and release. The main problem with the
project is that we began creating a webserver, without having a precise
idea of what KIND of a webserver it should be. I lately had the idea of
creating a really easy to use webserver. We could have configuration
and everything handled by KDE/Gnome tools and make a webserver that
will never reach the breadth of features that Apache offers, but will
be much easier to setup, smaller and more secure. And I believe that
there are newbies that would like to get a webserver running without
the command line. So, we have to make a graphical frontend to Kritton,
enhance Kritton and make sure it's stable. We can also try to implement
remote-administration tools like this :
http://www.zeus.com/products/zws/management/img/zws_admin.gif although
I doubt their usefulness.

What do you guys think? Is this an interesting target? What else do you
have to propose?

Pandis



--- jason corcoran <ja...@jc...> wrote:

> 
> I have been off line for the last month or so as my wife has given
> birth to a lovely baby girl, Naoise, unfortunately we were expecting
> twins and her sister, Leah, did not survive the birth.
> 
> So what do we do next, project wise. I definitely think we need to
> think of the aims of the project and translate that into some sort of
> requirements document. Just bullet points.
> 
> Next we should get a design doc, i.e. how we want to fulfill these
> requirements. This will allow us to agree upon a method to get the
> work done.
> 
> Any comments?
> 
> Jason.
> -- 
> Jason.
> 
> Fortune :
> 1909: Plastic Invented
> Result: Baywatch
> 
> 
> -------------------------------------------------------
> SF.Net email is Sponsored by the Better Software Conference & EXPO
> September 19-22, 2005 * San Francisco, CA * Development Lifecycle
> Practices
> Agile & Plan-Driven Development * Managing Projects & Teams * Testing
> & QA
> Security * Process Improvement & Measurement *
> http://www.sqe.com/bsce5sf
> _______________________________________________
> Kritton-devel mailing list
> Kri...@li...
> https://lists.sourceforge.net/lists/listinfo/kritton-devel
> 



		
___________________________________________________________ 
How much free photo storage do you get? Store your holiday 
snaps for FREE with Yahoo! Photos http://uk.photos.yahoo.com

I have been off line for the last month or so as my wife has given birth to a lovely baby girl, Naoise, unfortunately we were expecting twins and her sister, Leah, did not survive the birth.

So what do we do next, project wise. I definitely think we need to think of the aims of the project and translate that into some sort of requirements document. Just bullet points.

Next we should get a design doc, i.e. how we want to fulfill these requirements. This will allow us to agree upon a method to get the work done.

Any comments?

Jason.
-- 
Jason.

Fortune :
1909: Plastic Invented
Result: Baywatch

Hi,
just sending to say that I'm here. Rather busy at the moment, will
check the changes tomorrow.

See ya
Pandis

--- Nigel Horne <nj...@ba...> wrote:

> I've fixed a few minor issues tonight, committed to CVS.
> 
> They are mostly to do with the handling of CGI scripts.
> 
> -Nigel
> 
> 
> 
> -------------------------------------------------------
> SF.Net email is Sponsored by the Better Software Conference & EXPO
> September 19-22, 2005 * San Francisco, CA * Development Lifecycle
> Practices
> Agile & Plan-Driven Development * Managing Projects & Teams * Testing
> & QA
> Security * Process Improvement & Measurement *
> http://www.sqe.com/bsce5sf
> _______________________________________________
> Kritton-devel mailing list
> Kri...@li...
> https://lists.sourceforge.net/lists/listinfo/kritton-devel
> 



		
___________________________________________________________ 
How much free photo storage do you get? Store your holiday 
snaps for FREE with Yahoo! Photos http://uk.photos.yahoo.com

I've fixed a few minor issues tonight, committed to CVS.

They are mostly to do with the handling of CGI scripts.

-Nigel

Hi all,

The list of what has been done can be easily found in the CVS
Changelog. I feel that there have been plenty of changes, certainly
enough to justify a release. I agree with Nigel however, that we have
to determine the goals that we want to achieve. 
My own view of the subject is that Kritton should focus on providing an
efficient, stable webserver. Primary focuses should be on stability. In
terms of features I also believe that we should try to incorporate as
many as possible, enough at least to satisfy most peoples' needs. We
can't provide the funtionality of Apache, but we can outweigh the
feature disadvantage by avoiding complex installations. Apache is under
the Apache license with renders it non-free, yet open source and
permissive. It is a very long shot, but we can aim to provide an
alternative to Apache, which is though under the GNU GPL.
Thinking of it in the short term, I think that development should aim
on the following:

- Making the server compatible with the HTTP 1.1 protocol. The basic
aspects of the protocol are already working, but it definitely needs
more work.
- Support for PHP and CGI. CGI is already there, but not completely.
- Improve logging.
- Support for server side scripting.
- ASP perhaps?

I don't know what else to suggest right now. Add to the above list, and
then we can discuss the roadmap, saying what will be implemented, and
at what version. We should also try to set very rough release dates. We
should try to aim to have all the features implemented by 0.9, so that
the 0.9.x releases are all bug fix releases for 1.0 which ought to be
very stable, and secure.

Do you feel that 0.5, as it is now, deserves promotion? We do not want
to promote before it is too early, but we cannot attain perfect
stability before promoting because we won't have users to report bugs.
Do you have any ideas as to how promotion can succeed?

Pandis


		
___________________________________________________________ 
How much free photo storage do you get? Store your holiday 
snaps for FREE with Yahoo! Photos http://uk.photos.yahoo.com

On Tue, Jul 19, 2005 at 05:00:18PM +0100, Nigel Horne wrote:
> For what it's worth, I think we need to think aims/design of the project
> before worrying about developing.
> 
> -Nigel


TEST, I have been sending mails to the list, but not receiving them. Anyway if this one makes it, I agree. We also need some sort of Documentation stating what the design ( top level type Product requirements for the next phase) so we can get a definitive road plan for V0.6

J.
-- 
Jason.

Fortune :
If you must choose between two evils, pick the one you've never tried before!

> Fleuh and Jason,
> I am of the opinion that we should determine a release date to release
> 0.5.0 and begin to get more people involved in Kritton. In essence, I
> believe that Kritton promotion should be a principal goal. Of course, I
> am open to all suggestions. How do you think Kritton should be promoted
> ? If none of the other developers I think that we should use the
> 'Project Help' feature of Sourceforge to find one or two more
> developers to help out. Do you agree ?

For what it's worth, I think we need to think aims/design of the project
before worrying about developing.

-Nigel

On Tue, Jul 19, 2005 at 04:47:00PM +0100, Vassilis Pandis wrote:
> 
> --- jason corcoran <ja...@jc...> wrote:
> Jason,
> Sorry, misunderstanding. I didn't overlook the message, it probably
> never reached me (I just browsed through my inbox). Nevermind.  I was
> discussing with Fleuh who is also still with us that we need to set
> some targets and define exactly what we are trying to do. 
> 
> Fleuh and Jason,
> I am of the opinion that we should determine a release date to release
> 0.5.0 and begin to get more people involved in Kritton. In essence, I
> believe that Kritton promotion should be a principal goal. Of course, I
> am open to all suggestions. How do you think Kritton should be promoted
> ? If none of the other developers I think that we should use the
> 'Project Help' feature of Sourceforge to find one or two more
> developers to help out. Do you agree ?
> 
> Pandis
> 


Pandis,

I think thats a good idea. We should list out some functionality that we want to add, a kind of road map, so that when new developers come alone they can get there teeth into something straight away.

We should look at getting a list of functionality that has been added to 0.5.0 and see if there is enough added to warrant a major release.

J.
-- 
Jason.

Fortune :
It was the most I ever threw up, and it changed my life forever.

		-- Homer Simpson
		   Homer Goes To College

--- jason corcoran <ja...@jc...> wrote:
Jason,
Sorry, misunderstanding. I didn't overlook the message, it probably
never reached me (I just browsed through my inbox). Nevermind.  I was
discussing with Fleuh who is also still with us that we need to set
some targets and define exactly what we are trying to do. 

Fleuh and Jason,
I am of the opinion that we should determine a release date to release
0.5.0 and begin to get more people involved in Kritton. In essence, I
believe that Kritton promotion should be a principal goal. Of course, I
am open to all suggestions. How do you think Kritton should be promoted
? If none of the other developers I think that we should use the
'Project Help' feature of Sourceforge to find one or two more
developers to help out. Do you agree ?

Pandis

> 
> Pandis,
> 
> I sent a mail to you last week, didn't receive a reply?
> 
> J.
> -- 
> Jason.
> 
> Fortune :
> Make sure all variables are initialised before use.
>             - The Elements of Programming Style (Kernighan &
> Plaugher)
> 
> 
> 




	
	
		
___________________________________________________________ 
Yahoo! Messenger - NEW crystal clear PC to PC calling worldwide with voicemail http://uk.messenger.yahoo.com

Pandis,

I sent a mail to you last week, didn't receive a reply?

J.
-- 
Jason.

Fortune :
Make sure all variables are initialised before use.
            - The Elements of Programming Style (Kernighan & Plaugher)

Hello guys,
as you will have (hopefully) noticed, Kritton development has been
totally zero for the past couple of months or so. I had absolutely zero
spare time to devote to Kritton but now I can begin coding again.If you
are still interested in continuing Kritton development, please e-mail
me. If you do not e-mail me within two weeks from today, it probably
means that you don't even read your mail anymore and you will be
deleted from the development list. It would be nice before moving on
with development to set a firm target for a release and whatever else
may come up in discussion and form an actual target for Kritton.

Thanks a lot
Pandis


	
	
		
___________________________________________________________ 
Yahoo! Messenger - NEW crystal clear PC to PC calling worldwide with voicemail http://uk.messenger.yahoo.com

Ok, that's all there is to it. I updated netio.c (fixed the missing
header problem) and now raise_error uses the zero copy mechanism. Zero
copy currently is supposed to work under Linux, Solaris and FreeBSD.
The only other OS which I am aware of with zero copy support is
HP-UX/AIX which uses the function send_file. This means that there is a
name conflict with the function send_file defined in netio.c so we'll
have to change lots of things to accomodate for AIX (will do later).

Pandis


	
	
		
___________________________________________________________ 
Yahoo! Messenger - want a free and easy way to contact your friends online? http://uk.messenger.yahoo.com

(Finally, thought I was alone :)
Sorry, was me. netio.c is a newfile with all of the send and error
handling functions. With the Zero copy support main.c was getting
rather big and so netio.c seemed a good idea. If someone has access to
a Solaris or a FreeBSD box, I'd like them to check the Zero copy code
because I haven't tried it, just used the man pages of those OS
(sendfile() is not POSIX standard). I'm currently working on
raise_error, so that is uses the sendfile() too, where available. I'll
leave the emergency functions as they are, since they should probably
never be called so optimizing would be pointless. To summarise the past
days, CGI 1.1 support is close to complete, and I finally got a vague
understanding of how POST is supposed to work although there are a
number of questions in my head (what to do with a POST request for a
non-CGI file? how the parse the arguments of the POST request, and
when?)

Pandis

--- Jason Corcoran <ja...@jc...> wrote:
> On Tuesday 03 May 2005 13:16, Vassilis Pandis wrote:
> > There has been some inactivity lately on the list. Is there a
> specific
> > reason for this or just boredom/lack of time ? In the meantime,
> > regarding the POST method, I will implement it in the days to come
> as
> >
> 
> sorry about the inactivity, have been out celebrating and visiting
> family 
> since finding out about the twins [http://www.jcorcoran.net/twins]
> 
> I performed a cvs update this morning and it seems there is a file
> [netio.h] 
> missing from the project.
> 
> Could whoever checked in the netio.h into cvs add the c file.
> 
> Thanks 
> Jason
> 
> -- 
> ja...@jc...
> 
> Fortune:
> That government is best which governs least.
>   -- Henry David Thoreau, "Civil Disobedience"
> 
> 
> -------------------------------------------------------
> This SF.Net email is sponsored by: NEC IT Guy Games.
> Get your fingers limbered up and give it your best shot. 4 great
> events, 4
> opportunities to win big! Highest score wins.NEC IT Guy Games. Play
> to
> win an NEC 61 plasma display. Visit http://www.necitguy.com/?r=20
> _______________________________________________
> Kritton-devel mailing list
> Kri...@li...
> https://lists.sourceforge.net/lists/listinfo/kritton-devel
> 

Send instant messages to your online friends http://uk.messenger.yahoo.com 

On Tuesday 03 May 2005 13:16, Vassilis Pandis wrote:
> There has been some inactivity lately on the list. Is there a specific
> reason for this or just boredom/lack of time ? In the meantime,
> regarding the POST method, I will implement it in the days to come as
>

sorry about the inactivity, have been out celebrating and visiting family 
since finding out about the twins [http://www.jcorcoran.net/twins]

I performed a cvs update this morning and it seems there is a file [netio.h] 
missing from the project.

Could whoever checked in the netio.h into cvs add the c file.

Thanks 
Jason

-- 
ja...@jc...

Fortune:
That government is best which governs least.
  -- Henry David Thoreau, "Civil Disobedience"

The Linux one works for sure, the FreeBSD I don't know (don't know even
if it compiles, no access to a FreeBSD box but I may manage to use the
compile farms). A detailed account of zero copy can be found at
http://www.cs.duke.edu/ari/trapeze/freenix/node6.html 

http://www.linuxjournal.com/article/6345

TCP_CORK was not implemented because it is not always faster than
normal TCP and in the case of a webserver would probably hinder
performance instead of improving it.


	
	
		
___________________________________________________________ 
Yahoo! Messenger - want a free and easy way to contact your friends online? http://uk.messenger.yahoo.com

Added more CGI environmental variables, just a few left now. Also, the
server now responds with the appropriate HTTP version. ie. If the
client requests with HTTP/1.0, we say HTTP/1.0 OK.

Pandis

Send instant messages to your online friends http://uk.messenger.yahoo.com 

Work towards CGI specification conformance. Added some env variables,
fixed SERVER_NAME variable. I'll work more on this to achieve full CGI
conformance.

Pandis

Send instant messages to your online friends http://uk.messenger.yahoo.com 

There has been some inactivity lately on the list. Is there a specific
reason for this or just boredom/lack of time ? In the meantime,
regarding the POST method, I will implement it in the days to come as
it turned out to be significantly simpler than what I and Florent
thought. After we implement it, I feel that we should use the 0.5
series to produce Kritton 0.5.0-0.5.3 or so to allow us to make sure we
have a sturdy package, release quality. I also submitted the Kritton
homepage to Yahoo! and Google directories but as of now, it has not
been added to either. How do you think Kritton development should
proceed? What should be done, on what should we emphasize? For the time
being, I think that bug hunting and fixing is quite essential.

I also think that after we have managed to produce a quality web
server, have written documentation and released 0.6, we could request a
review from magazines like Linux Format or Linux Magazine, and online
sites. Of course, this should not be rushed. We should strive to
promote Kritton to establish a user database which will provide
feedback and finally see Kritton compete other webservers.

Give me your opinions on the matter.
Pandis

Send instant messages to your online friends http://uk.messenger.yahoo.com