pan-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Pan-devel] ANN: SSL Support


From: SciFi
Subject: Re: [Pan-devel] ANN: SSL Support
Date: Wed, 26 Oct 2011 08:39:43 +0000 (UTC)
User-agent: Pan/0.135 (Tomorrow I'll Wake Up and Scald Myself with Tea; GIT 045ef68 (github.com/judgefudge/pan2/master); x86_64-apple-darwin10.8.0; gcc-4.2.1 (build 5666 (dot 3)); 32-bit mode)


Hi,

I got your git master 045ef68 built and running here,
including --with-openssl their cvs master at around September 2011
(most latest patches there are to the FIPS stuff which I don't build).

I had to make sure of the new settings for this Pan using SSL,
stuff like going thru the news-server settings
pointing to a 'real' addy for GN & AW & Gmane
that supplies their SSL connections
(I gleaned from my stunnel.conf)
and making sure the new option is set for 'Secure (TLS) SSL' down there.

After making sure all connections continued to work,
even after knocking-down all stunnel tasks,
I saw an incredible high i/o rate being computed & shown by this Pan
at several places
(I'll include little PNG screengrabs as attachments,
 here we are fetching new headers from GN/primary and AW/fallback).

I say 'incredible' because those i/o rates are quite impossible
as the ISP here allows these maximum rates:
>   Connection Speed:                                                           
>                                                                               
>    
>   o  Incoming: 6016 kbps
>   o  Outgoing: 768 kbps
(directly shown by the modem's internal webpages).

Note the ISP uses small-'b' for 'bitrate'
while Pan uses Big-'B' for 'byterate'.
Before your SSL build, we've never seen any rate above ~600 KiB/s.
So, somehow, your code is now showing us a _doubled_ i/o rate
or something near.
;)

I've been running this build for a couple days.
Everything seems to be working fine
vis-à-vis stunnel.

I then saw an update or two for XQuartz,
catching-up on the X.org entities
as can be seen at the usual place
<http://xquartz.macosforge.org/trac/wiki/X112.7.0>
but they also elected to turn-on "fontconfig autohint"
which makes things look quite "improved".
(I thought "autohint" was provided by freetype… oh well….)
(btw you'll see how Pan now looks in the aforementioned attachments)

I _have_ noticed a difference in one font I've been using for many months.
Below I'm snipping pieces of 'preferences.xml' from my Pan settings.
After updating to XQuartz-2.7.0_rc1
with its fontconfig enabling the "autohint" feature,
suddenly this font is no-longer showing its "fixed-point"ness:
<string name='body-pane-font' value='ProFontWindows Medium 10'/>
e.g. 'i' is now a smaller width than other letters etc.
while this font does look "fixed-point" as expected:
<string name='monospace-font' value='Monospace 10'/>
and others such as "Monaco" etc. do behave as expected.
btw I chose ProFontWindows because it is /suppose/ to let us see the
fancier 'nfo' text (ascii shading & boxes etc)
otherwise it is okay that I not use it
but rather use a 'real' fixed-point font.  ;)

As it is, your code seems to be working fine.
But I can never figure-out if we're running SSL "for reals"
(yes the servers did reject/not-respond-to the "plain-text" setting
 but I don't think that is enough proof ;) ).
And the doubled i/o rate calculations are something that needs
further study, please
(this is during header-fetch _and_ downloading-binaries).

Thank you for all your work.







reply via email to

[Prev in Thread] Current Thread [Next in Thread]