Introduction

The purpose of this analysis is to estimate the usefulness of the ACK/NACK system as used in OpenVPN project's development. This study extends back two years, and the data is obtained from Git logs:

$ git log --since="2 years ago"

This data is cross-referenced with the emails stored in the ​openvpn-devel mailing list archive. As reviewing all patches and ACKs is not possible, a random sample is taken.

The hypothesis is that the ACK/NACK process is most suitable for reviewing individual patches or small patchsets sent to the openvpn-devel mailing list by developers not intimately familiar with OpenVPN. A less bureaucratic process might be more suitable for core developers, especially when less security-sensitive parts are being modified.

Current ACK/NACK code review system

Scope

Most patches have required an ACK before going into Git "master". There are a only few exceptions:

• Patches from James Yonan's ​Subversion repository are merged without review
• Anything else?

Evolution

The original ACK system required an ACK from one developer for the code to enter the Git repository. This process was later (add date?) split into two logical parts:

• Feature-ACK: this should be the primary consideration when reviewing patches. Merging useless patches makes no sense whatsoever.
• Code-ACK: if the feature makes sense, the next step would be to review that the code itself makes sense (e.g. has no obvious errors and follows established conventions)

Still later (add date?) a move towards ​lazy consensus was taken. So, as it stands now (Mar 2012), code can go to the Git repository without an ACK, but with a considerable delay.

Pros

The formal ACK/NACK system was imposed to increase OpenVPN code quality by decreasing the chances of bad code getting into the tree. The mandatory ACK forces peer review. In software engineering jargon, it fulfills the roles of

• ​Rationale management (feature-ACKs)
• ​Static testing (code-ACKs)

Limitations

There are limits to what the ACK process can do. For example:

• Runtime errors may be very difficult to notice
• Seeing the "big picture" may be difficult, especially with big patchsets
• Whether a particular problems is noticed depends highly on the reviewer's skills

Cons

The ACK/NACK process has it's cons:

• Requiring an ACK slows down flow of code from contributors to the "master" development tree. This problem gets worse as the patchset size increases.
• Due to the above, it takes longer to get the code into wide circulation. This means bugs are reported later than if the code had been merged as soon as it's ready.

Raw data

This raw data is based on Git clone done on 5th Apr 2012:

Generic statistics

• Commit count: 635
• ACK number: 458
• Average ACK count: .72

ACKs by person

• Adriaan de Jong 20
• Alon Bar-Lev 3
• Davide Guerri 1
• David Sommerseth 155
• Eric F Crist 1
• Gert Doering 107
• Gilles Espinasse 1
• Heiko Hund 3
• James Yonan 116
• Jan Just Keijser 2
• Kazuyoshi Aizawa 2
• krzee 3

Generic statistics:

• Commit count: 622
• ACK number: 439
• Average ACK count: 0.70

ACKs by person:

• Adriaan de Jong: 18
• Alon Bar-Lev: 3
• Davide Guerri: 1
• David Sommerseth: 148
• Eric F Crist: 1
• Gert Doering: 101
• Gilles Espinasse: 1
• Heiko Hund: 3
• James Yonan: 113
• Jan Just Keijser: 2
• Kazuyoshi Aizawa: 2
• krzee: 3
• Peter Stuge: 11
• Samuli Seppänen: 32

• Peter Stuge 11
• Samuli Seppänen 33

Effects of the ACK system on patches

The following Git command was used to generate a list of all patches in Trac format:

git log --date=short --since="2 years ago" --pretty=format:"||%ad||%h||%s||-||%an||-||||"

A random sample of 64 patches (out of 634) was then taken analyzed in detail.

The following random sample of patches was analyzed:

Commit date	Identifier	Patch name	Patchset	Author	Modified	Delay
2010-03-02	d04b858	Several updates to openvpn.8 (man page updates)	-	Karl O. Pinc	-
2011-06-16	0f2bc0d	Do some file/directory tests before really starting openvpn	-	David Sommerseth	-
2011-06-23	b01cb9e	Refactored crypto initialisation functions	-	Adriaan de Jong	-
2012-02-29	fbae7d2	build: plugins: properly use CC, CFLAGS and LDFLAGS	-	Alon Bar-Lev	-
2011-03-15	58704ea	Updated INSTALL-win32.txt	-	Samuli Seppänen	-
2011-04-27	b70d99f	Fix compile issues when using --enable-small and --disable-ssl/--disable-crypto	-	Gustavo Zacarias	-
2011-06-23	4a5a603	Refactored NTLM DES key generation	-	Adriaan de Jong	-
2011-03-25	dc2ccc8	Clarify --tmp-dir option	-	chantra	-
2011-10-11	359adbf	Raised D_PID_DEBUG_LOW from level 3 to 4 to reduce replay error verbosity at level 3.	-	James Yonan	-
2010-07-27	75dfe3d	Added "net stop dnscache" and "net start dnscache" in front of existing --register-dns commands.	-	James Yonan	-
2009-09-24	e478770	* make possible to x-compile openvpn/win32 in Linux	-	JuanJo? Ciarlante	-
2010-04-27	7d5e26c	Fix certificate serial number export	-	Davide Brini	-
2010-03-10	1e02046	On TARGET_LINUX define _GNU_SOURCE if not defined	-	David Sommerseth	-
2011-07-25	576dc96	Merge remote branch SVN 2.1 into the git tree	-	David Sommerseth	-
2011-06-30	3e44ea5	Refactored tls-verify script code	-	Adriaan de Jong	-
2010-02-20	a698464	* undo mroute.c changes related to ipv6 payload, nothing to do w/ipv6 transport afterall.	-	JuanJo? Ciarlante	-
2012-02-14	8e5613c	Migrated x509_get_sha1_hash to use the garbage collector	-	Adriaan de Jong	-
2010-10-30	f0eac1a	Make "topology subnet" work on Solaris (ifconfig + route metric changes by Kazuyoshi Aizawa, adding of local "connected subnet" route by me)	-	Gert Doering	-
2011-10-31	54628d1	Moved prng_uninit out of crypto_uninit_lib	-	Adriaan de Jong	-
2011-07-01	8c96419	Fixed a missing include in ssl_backend.h	-	Adriaan de Jong	-
2011-10-16	eaacf8d	Moved to PolarSSL 1.0.0:	-	Adriaan de Jong	-
2011-03-21	1df945e	Version 2.1.3n	-	James Yonan	-
2011-06-29	d22a379	Made domake-win builds to use easy-rsa/2.0/openssl-1.0.0.cnf	-	Samuli Seppänen	-
2011-09-16	8ca19c0	Platform cleanup for NetBSD	-	Gert Doering	-
2011-04-12	e51935d	For Mac OSX, when DARWIN_USE_IPCONFIG is defined, retry ipconfig command on failure once every second for up to 15 seconds. This is necessary to work around an issue observed on OSX 10.5 where the ipconfig command sometimes fails if executed immediately after the tun device open.	-	James Yonan	-
2011-07-28	c94eff3	Added back checks for ks->authenticated in verify_user_pass	-	Adriaan de Jong	-
2011-02-11	a7f0fc3	Added configure.h and version.m4 variable parsing to win/config.py	-	Samuli Seppänen	-
2010-11-13	22178d0	Merge branch 'svn-BETA21' into bugfix2.1	-	David Sommerseth	-
2011-11-03	8407991	Fixed client issues with DHCP Router option extraction/deletion when using layer 2 with DHCP proxy:	-	James Yonan	-
2010-11-15	7581c8f	Removed functions not being used anywhere	-	David Sommerseth	-
2011-10-31	1d90851	Moved from strsep to strtok, for Windows compatibility	-	Adriaan de Jong	-
2010-07-16	4c91fc8	Fixes openssl-1.0.0 compilation warning	-	chantra	-
2011-04-24	d549726	Added 'dir' flag to "crl-verify" (see man page for info).	-	James Yonan	-
2012-01-22	62c613d	Platform cleanup for FreeBSD	-	Gert Doering	-
2011-07-05	557624e	Hardening: periodically reset the PRNG's nonce value	-	Adriaan de Jong	-
2010-02-18	058f3d0	[PATCH] Change verify-cn so cn is no longer hardcoded in openvpn's config file	-	Karl O. Pinc	-
2011-03-17	d02a86d	Renamed branch to reflect that it is no longer beta.	-	James Yonan	-
2011-08-31	d90428d	add --mark option to set SO_MARK sockopt	-	Heiko Hund	-
2011-05-26	21fc2ed	Don't define ENABLE_PUSH_PEER_INFO if SSL is not available	-	David Sommerseth	-
2010-03-07	94d50a1	when deleting a route on win32, also add gateway address (otherwise netsh.exe will succeed, but silently ignore request)	-	Gert Doering	-
2011-05-26	739fa98	Fix a Visual Studio 2008 build error in options.c	-	Samuli Seppanen	-
2011-06-20	54c739e	Revert "Add new openssl.cnf to easy-rsa/Windows"	-	David Sommerseth	-
2010-04-22	892e64b	Fix missing /bin/bash -> /bin/sh	-	Davide Brini	-
2011-08-16	f0257ab	For all accesses to "struct route_list * rl", check first that rl is non-NULL	-	Gert Doering	-
2009-10-05	e293510	* TODO.ipv6 update	-	JuanJo? Ciarlante	-
2010-09-30	58f8d94	Add HTTP/1.1 Host header	-	Lars Hupel	-
2011-08-19	2627335	CC_PRINT character class now allows any 8-bit character value >= 32. This is done to allow UTF-8 and restrict the use of control characters in usernames, passwords, common names, etc.	-	James Yonan	-
2011-07-05	a9bf901	Added an extra define to allow building without PKCS#11	-	Adriaan de Jong	-
2012-02-18	4ebc587	define access mode flag X_OK as 0 on Windows	-	Heiko Hund	-
2011-02-11	4e4aa65	Several modifications to win/make_dist.py to allow building the NSI installer	-	Samuli Seppänen	-
2010-02-19	adfe37f	verb 5 logging wrongly reports received bytes	-	David Sommerseth	-
2010-11-15	33ee747	Fixed potential misinterpretation of boolean logic	-	David Sommerseth	-
2011-08-17	14a382a	remove wrapper code for Windows CryptoAPI function	-	Heiko Hund	-
2011-06-27	eab0cf2	Refactored TLS_PRF to new hmac and md primitives	-	Adriaan de Jong	-
2012-02-29	51bd56f	build: autotools: first pass of trivial autotools changes	-	Alon Bar-Lev	-
2012-02-29	553d95d	cleanup: memcmp.c: remove unused source	-	Alon Bar-Lev	-
2010-10-21	59afc4a	Fix problem with special case route targets ('remote_host')	-	Gert Doering	-
2010-04-16	add7fe0	Updated the man page to reflect the behavioural change of create_temp_file()	-	David Sommerseth	-
2011-10-31	7ac7170	Reordered functions to ensure warning-free Windows build	-	Adriaan de Jong	-
2011-04-03	9ed122e	Fixed bug in port-share that could cause port share process to crash with output like this:	-	James Yonan	-
2012-03-28	3144411	Enable pedantic in windows compilation	-	Alon Bar-Lev	-
2011-04-12	2a12831	For Mac OSX, when DARWIN_USE_IPCONFIG is defined, retry ipconfig command on failure once every second for up to 15 seconds. This is necessary to work around an issue observed on OSX 10.5 where the ipconfig command sometimes fails if executed immediately after the tun device open.	-	James Yonan	-
2012-03-24	405f338	build: windows: set vendor to openvpn project + cleanups	-	Alon Bar-Lev	-
2011-02-28	a75c7dd	Fixed typo in plugin.h	-	Stefan Hellermann	-

Analysis