Граф коммитов

23 Коммитов

Автор SHA1 Сообщение Дата
bryner%uiuc.edu 8b80c7c363 Bug 70595 - making nsIPrompt accessible to PSM. r=darin, sr=mscott. 2001-03-06 05:10:33 +00:00
dougt%netscape.com 128f95aa9b Relanding Necko Changes.
Revising nsIChannel to allow for overlapped i/o. This consists of three parts:

1. Factoring nsIChannel into a protocol specific part, the nsIChannel, and a socket specific, the nsITransport.
2. Derive the nsIChannel from a nsIRequest.
2. Changes the notification system from necko and the URILoader to pass the nsIRequest interface instead of nsIChannel interface.

This goal stems from wanting to be able to have active AsyncRead and AsyncWrite operations on nsSocketTransport.
This is desired because it would greatly simplify the task of maintaining persistent/reusable socket connections
for FTP, HTTP, and Imap (and potentially other protocols). The problem with the existing nsIChannel interface is
that it does not allow one to selectively suspend just one of the read or write operations while keeping the other active.

r=darin@netscape.com
sr=rpotts@netscape.com
2001-02-21 20:38:08 +00:00
disttsc%bart.nl 3d2d80d536 Back out dougt's channel changes 2001-02-12 03:14:23 +00:00
dougt%netscape.com 1b9ca82439 Revising nsIChannel to allow for overlapped i/o. This consists of three parts:
1. Factoring nsIChannel into a protocol specific part, the nsIChannel, and a socket specific, the nsITransport.
2. Derive the nsIChannel from a nsIRequest.
2. Changes the notification system from necko and the URILoader to pass the nsIRequest interface instead of nsIChannel interface.

This goal stems from wanting to be able to have active AsyncRead and AsyncWrite operations on nsSocketTransport.
This is desired because it would greatly simplify the task of maintaining persistent/reusable socket connections
for FTP, HTTP, and Imap (and potentially other protocols).  The problem with the existing nsIChannel interface is
that it does not allow one to selectively suspend just one of the read or write operations while keeping the other active.

The full details of the change on written up in the netlib newsgroup.

r=darin@netscape.com
sr=rpotts@netscape.com
2001-02-10 00:16:26 +00:00
darin%netscape.com 6a7567992b Bug 62566. nsIChannel::AsyncWrite interface revision and related cleanup.
r=dougt@netcape.com, sr=mscott@netscape.com.
2001-01-23 22:10:34 +00:00
rpotts%netscape.com 749abe34f1 bug #48546. The cache channel should delegate the handling of load attributes to the cache channel (ie. disk/memory channel)... 2000-08-24 22:54:45 +00:00
warren%netscape.com e05eef45e9 Bug 46777. Redesigned stream interfaces to allows stream observers to be decoupled from pipe implementation. Needed for embedding to fix 'spin' problem. r=rpotts,valeski,ruslan 2000-08-22 07:03:33 +00:00
neeti%netscape.com 70e2ba8ffa fix for bug 40084, [CRASH] Crash in disk cache code, backed out changes made on 7/27, put new fix
fix for bug 46014, No keywords loaded on bugzilla pages, r/a=gagan
2000-07-29 02:32:14 +00:00
warren%netscape.com 512c8bf433 Renaming nsIAllocator to nsIMemory (and nsAllocator to nsMemory). API cleanup/freeze. Bug #18433 2000-06-03 09:46:12 +00:00
warren%netscape.com 727f312b32 Necko API changes: primarily nsIChannel, changing initialization parameters to accessors. Got javascript: evaluation to happen at the right time (when AsyncRead is called) as well as on the right thread. 2000-03-29 03:58:50 +00:00
gagan%netscape.com c6349ecc8b Fix for 30385. We were leaking channel due to the no longer needed "proxy channel" by the cache transport (mem cache) Also removed nsIHttpEventSink.h since its now an IDL. Also fixed 31667, 31222, 31657 (the assertion failures for thread safety) 2000-03-14 03:24:56 +00:00
warren%netscape.com c0497e31be Bug 21556: Making linux be thread-safe. Making tons of classes implement threadsafe AddRef/Release. a=jar 2000-03-05 21:26:01 +00:00
rpotts%netscape.com 04d8935869 bug #24711 (r=gagan). Changed cache channels to behave like transports rather than protocol channels... nsCacheEntryChannel.cpp 2000-02-29 04:45:14 +00:00
gagan%netscape.com 2113f0d630 We had a nsCOMPtr on a class instead on a interface (along with the problem that its default constructor was protected) This was preventing the cache from being built on AIX. Should work now. r=self(& scc) See bug 24876 2000-01-25 03:11:16 +00:00
mscott%netscape.com 4ab4a29bdf Bug #24796 --> the fix for the cache manager friday night broke quoting msgs
for imap (which uses the memory cache). You don't need a load group to run a url
through the cache....so don't return an error if the load group is null
r=bienvenu
2000-01-24 00:32:14 +00:00
valeski%netscape.com 09d0af3430 24250. r=gordon,mscott,warren. We now ensure that a cache entry adds itself to the loadgroup *before* the load group goes away. This allows the editor to 'Edit Page's again. 2000-01-21 23:58:29 +00:00
rpotts%netscape.com 672c0938ff Removed over aggressive assert when the loadgroup listener factory does not exist 2000-01-04 05:46:47 +00:00
dougt%netscape.com 43d538707a Including nsIAllocator.h directly. 2000-01-03 23:23:33 +00:00
fur%geocast.com a516109cda Added assertion and trivial recovery code to catch the case of a load group
listener factory that reports success, but nevertheless returns a NULL stream
listener.
1999-12-18 14:59:59 +00:00
fur%netscape.com e422da8a9b It turns out that PR_Now() has only one second resolution on the Mac, hence it's unsuitable for computing download rates for the cache. Thanks to sdagley for pointing this out. 1999-12-08 05:46:31 +00:00
fur%netscape.com 771d577a8e Fixed incorrect calculation of content download time and profitability of caching a given URL. r:none 1999-12-07 08:51:03 +00:00
sfraser%netscape.com d7bec0c5bc Some casts and a .get() so that it compiles on Mac. r=fur 1999-12-07 06:31:14 +00:00
fur%netscape.com a3e4a5596f Landing of CacheIntegration_BRANCH in support of bug #8305 (Implement cache) 1999-12-07 02:07:31 +00:00