Pier Fumagalli wrote: > "Remy Maucherat" <[EMAIL PROTECTED]> wrote: > >>Pier Fumagalli wrote: >> >> >>>As far as I would like to see WARP and its future development, it'll >>>probably end up following a different container architecture. The >>>extenization of the HTTP stack from the core of the container brings some >>>advantages to the engine, but as well this need to introduce a different >>>layering scheme for the components, such as the removal of <Host>. >> >>I'm not sure I quite understand why you say that. >> >>In that deployment scenario, you use: >>- Coyote HTTP/1.1 >>- Catalina 2.0 >>and this is not fundamentally different that the current Tomcat 4.1. The >>main difference is that the loader will probably be different, but >>you'll still be able to use the old one if you need to. > > > I am not inclined towards the usage of an HTTP protocol different from > Apache HTTPd, our flagship product. The servlet container should be only > that, a servlet container, without anything else (no HTTP, no CGI, no SSI, > no JSP, no TURBINE, no NOTHING)... Only in this scenario, with a clear > separation of what is what, we can suitably achieve a decent servlet > container, as JServ was back in the days.
Working as well as possible with Apache is my objective. Providing a great Java web server is also my objective. Also, you have to remember that in the long run VM based web servers may (huge speculation; no flames please) attract some interest eventually because of their inherent security advantages over native web servers. >>Actually, if you package the current o.a.c.connector classes with the >>webapp classes, you should be able (maybe after adding one or two new >>methods for the new Servlet API, but it shouldn't be too hard) to run >>Webapp 1.0 on TC 5. > > > As I said, I am not confident in Tomcat as of _NOW_ to run on _MY_ > production site. I don't want to think about a new upcoming and improved > release when IMO, there's still so much crap to do with the current code > base. I aim to change your mind eventually. I don't think you give the new code a fair trial, though. >>Rewriting WARP for Coyote wouldn't be that difficult, IMO. It just >>changes the way you communicate with your lower layer, and that's about >>it. I believe Costin wants to improve the "action" stuff used currently >>to be more flexible and powerful. > > > That would be a waste of time and CPU resources... You're looking in the > wrong direction... To get performance, and reliability, all you have to do > is simplify the code, removing layer after layer... Not adding them... Implementing both the connector AND the servlet API in the same classes is IMO the definition of messy code. So I am convinced that, to keep it clean, you have to have one additional layer. If everyone decides to think like you do, then I'll leave those persons in charge of the maintenance of the connectors. It is likely however that most people here are strongly against going back to the unified model. I recommend you review the Coyote HTTP/1.1 protocol handler, and tell me if you really think the extra layer adds any kind of performance penalty, or any added complexity. Personally, I think the Coyote HTTP/1.1 design makes me give up 1 or 2% of absolute maximum performance (ie, it's negligible), but in return the programming model is much simpler, and therefore I am able to get a lot closer from the theorical limit. >>In return, you get a GC friendly behavior of the connector, higher >>performance, and it should also be easier to adapt to future version of >>Tomcat and / or may be used by other containers. > > > FWIW, I don't care about other containers... And I won't care about future > revisions of Tomcat until this one "works for me"... I think this is reasonable :) I hope 4.1 will be that one, unless 4.0.4 + webapp does the job for you already. >>>This is where I want to end up to. Frankly at this point in time I don't >>>think that carrying on with the development of components such as the HTTP >>>connector, or other tomcat "features" (GZIP on-the-fly compression, CGI >>>support, JMX support), matters to me... >> >>I can understand. Usually, you lose interest in a project as soon as it >>does whatever you want it to do :) >>There are some exceptions, but I really can understand that. > > > Or you loose interest when you see that no matter what you do, noone even > listens to what you say... Personally, I think the feature completeness will happen first. But you obviously don't hink the same way I do, so .... Remy -- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
