From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.2 (2018-09-13) on dcvr.yhbt.net X-Spam-Level: X-Spam-ASN: X-Spam-Status: No, score=-4.0 required=3.0 tests=ALL_TRUSTED,AWL,BAYES_00 shortcircuit=no autolearn=ham autolearn_force=no version=3.4.2 Received: from localhost (dcvr.yhbt.net [127.0.0.1]) by dcvr.yhbt.net (Postfix) with ESMTP id 55CEF1F45A for ; Fri, 17 Apr 2020 09:11:03 +0000 (UTC) From: Eric Wong To: meta@public-inbox.org Subject: [PUSHED 0/2] relnotes and some notes about the future Date: Fri, 17 Apr 2020 09:11:01 +0000 Message-Id: <20200417091103.22488-1-e@yhbt.net> MIME-Version: 1.0 Content-Transfer-Encoding: 8bit List-Id: So I've got some vague ideas of what 1.5 will entail... I'm still disatisfied with memory consumption and performance; so more on that. Maybe some more optional Inline::C, but ease-of-installation still must be taken into account. I started writing Documentation/technical/data_structures.txt a some time ago and realized we could reduce, eliminate and reuse some data structures. That would allow us to document less things, reduce cognitive load on hackers, AND save some RAM :> Another place that bothers me is we build up big, per-message memory buffers for HTML/Atom feeds. Since most HTTP requests are compressed anyways, I've been slowing moving towards buffering directly to gzip, allowing our internal memory representation to be compressed. Though it may not help a lot because deflate contexts take several dozen kilobytes of RAM. Still, that allows us to avoid Plack::Middleware::Deflater and the memory-wasting closures it imposes. And one less dependency in the INSTALL file, too. Eric Wong (2): public-inbox 1.4.0 doc: update 1.4.0 relnotes with date, start 1.5.0 Documentation/RelNotes/v1.4.0.eml | 29 ++++++++++++++++++++++------- Documentation/RelNotes/v1.5.0.eml | 13 +++++++++++++ MANIFEST | 1 + Makefile.PL | 4 ++-- 4 files changed, 38 insertions(+), 9 deletions(-) create mode 100644 Documentation/RelNotes/v1.5.0.eml