Re: Re: HTTP::Cache vs HTTP::Download and HTTP::Header

Hello Michael, >> HTTP::Cache vs HTTP::Download and HTTP::Header >> >> If there is such thing as usability in PEAR, then I must admit, that >> above conflict directly relates to it. As an unexperienced developer >> I know which packages PEAR contain but I just don't have time to read >> each and all descriptions available on the site. MW> If one's not able to find and use information and transform MW> it into knowlegde, well, one can be called quite lost nowadays... MW> If you've got any suggestions to improve the situation, provide MW> them in a sensible way, we'll very much appreciate them. Well, I suggest to merge the functionality of them into one package, which could be called HTTP::Cache_Control. >> I can only presume, >> that HTTP::Download implements something like PEAR wget MW> HTTP_Download is no downloader. The name might *now* be confusing, MW> but when the project started it was a real tiny "send some data" class. MW> See the package description, cvs and the docs for more information. The name *is* confusing and I'm not interested in some package history as long as I can't do that I need. HTTP::Download - Quote from manual: ******** Intro Intro -- Introduction in usage of HTTP_Download Introduction With this package you can handle (hidden) downloads. It supports partial downloads, resuming and sending raw data ie. from database BLOBs. ATTENTION: You shouldn't use this package together with ob_gzhandler or zlib.output_compression enabled in your php.ini, especially if you want to send already gzipped data! ******** Where is something about Cache-Control here? What is "hidden downloads" BTW? I think it is downloader and descriptions confirm my thoughts. I can't imagine how can I use downloader to control cache on the users side? If Downloader has such Cache functionality I propose to detach this functionality into separate logical concept and establish a dependancy. That will be logicaly correct. If you want an argue - place a voting: "How should be called package implementing HTTP cache control mechanisms described in chapter > 13.1.3 of HTTP RFC http://www.faqs.org/rfcs/rfc2616.html ?" >> HTTP::Header is for some mysterious HTTP headers manipulations or for >> statistics or whatever. It was a big surprise to know them capable >> something else. MW> Well, then have another look. Many people are already using these MW> classes successfully. If you foind a bug while using them, please MW> report (if not needless to say). Many? Let's assume there is package download statistics and most of them using these packages to "handle (hidden) downloads. It supports partial downloads, resuming and sending raw data ie. from database BLOBs." How many of them use Cache-Control techniques from these packages? How many use different Cache-Control packages and why don't they use HTTP::Download and HTTP::Header for that purpose? >> I'd like to have some HTTP::Cache_Control, which specializes only on >> this part of HTTP protocol, even if it will bear some of the other >> packages functionality. But this one should implement the ultimate >> Caching mechanizm if it is possible. =) MW> I doubt that anybody will be against some UeberCache package... :) MW> As long as that package doesn't exist, my answer is quite simple: MW> You can use HTTP_Header_Cache for unpersonalized and HTTP_Download MW> for personalized pages, which should be cached on the client side. I want this package to exist and to be started at least. There are even people who want to write code for it. But seems, that other don't allow even to start with this UeberCache. What is Ueber, BTW? =) t --