"SfR Fresh" - the SfR Freeware/Shareware Archive 
Member "pan-0.133/NEWS" of archive pan-0.133.tar.gz:
As a special service "SfR Fresh" has tried to format the requested source page into HTML format using source code syntax highlighting with prefixed line numbers.
Alternatively you can here view or download the uninterpreted source code file.
That can be also achieved for any archive member file by clicking within an archive contents listing on the first character of the file(path) respectively on the according byte size field.
1 Pan 0.14.90 is the first beta of a ground-up rewrite of Pan in C++.
2
3 Some of the best user-visible improvements:
4
5 * The memory footprint for large newsgroups has been cut by 2/3. (A month's
6 worth of headers in a sample binaries group dropped from 283M to 94M.)
7
8 * The time to download headers has been cut by 2/3 by parallelizing
9 the download across multiple conections to the server. (In the same
10 sample group, downloading a months' worth of headers dropped from
11 17 minutes to 5 minutes.)
12
13 * The time to load large newsgroups from disk has been cut by 90%.
14 (In the same sample newsgroup, the time to load from disk
15 dropped from 14 seconds to 1.3 seconds.)
16
17 * Multiserver support. Pan can now download files in parallel not just
18 with multiple connections to one server, but also to other servers.
19
20 * NZB support. Pan can now import NZB files. It can also be started in an
21 NZB client mode that doesn't create the newsreader window at all.
22
23 * All of the user interface windows and dialogs have been redesigned.
24
25
26 Some of the best backend improvements:
27
28 * The source code's line count has dropped from 44,382 to 23,093 --
29 and is far cleaner.
30
31 * Fewer library prerequisites: Pan no longer requires GNet or libxml2.
32
33 * Better decoupling.
34 _ Unit tests that were impossible before, such as tasks running and
35 talking to a mock news server, can now be done.
36 _ Pan can now run 'headless' to process queued tasks.
37 _ The task and application layers are separated from the backend,
38 so a database backend could be added in the future if desired.