5 - XXX: Upgrades don't pull updated tags, breaking git describe --tags!
6 Fix this for the future, and figure out how to make everyone else happy!
7 - XXX: Quota parsing code is buggy on some lockers, for example /afs/athena.mit.edu/org/p/physics
8 - XXX: Some installs are throwing spurious errors; investigate
9 - XXX: Some installs are locked; we should automatically break locks if they're old
10 - XXX: Prolly would be nice to have some information about how many installs actually succeeded
11 - XXX: parallel-find.pl incorrectly reports Wizard-style autoinstalls twice, because
12 the find succeeds on both .scripts-version and .scripts.
13 - If you try to do an install on scripts w/o sql, it will sign you up but fail to write
14 the sql.cnf file. This sucks.
16 - wizard install wordpress should ask for password. One problem with this is that
17 Wordpress will still send mail with the wrong username and password, so Wordpress
18 will need to be patched to not do that. Alternatively we can initally set the admin
19 email to a null address and then fix it manually.
20 - --raw parameter for install which means an arbitrary commit can be installed
21 - Parse output HTML for class="error" and give those errors back to the user (done),
22 then boot them back into configure so they can enter in something different
23 - Get rid of our custom sizing code and use dialog's built-in sizing (i.e. width=0, height=0).
24 Maybe our sizing code is superior, maybe not.
26 - Replace gaierror with a more descriptive name (this is a DNS error)
28 - Pre-emptively check if daemon/scripts-security-upd
29 is not on scripts-security-upd list (/mit/moira/bin/blanche)
31 - Web application for installing autoinstalls has a hard problem
32 with credentials (as well as installations that are not conducted
33 on an Athena machine.) We have some crazy ideas involving a signed
34 Java applet that uses jsch to SSH into athena.dialup and perform
38 - Tidy up common code in callAsUser and drop_priviledges in shell
39 - Summary script should be more machine friendly, and should not
40 output summary charts when I increase specificity
41 - Summary script should do something intelligent when distinguishing
42 between old-style and new-style installs
43 - Report code in wizard/command/__init__.py is ugly as sin. Also,
44 the Report object should operate at a higher level of abstraction
45 so we don't have to manually increment fails. (in fact, that should
46 probably be called something different). The by-percent errors should
48 - Move resolutions in mediawiki.py to a text file? (the parsing overhead
50 - If a process is C-ced, it can result in a upgrade that has
51 an updated filesystem but not updated database. Make this more
53 - PHP end of file allows omitted semicolon, can result in parse error
54 if merge resolutions aren't careful. `php -l` can be a quick stopgap
57 - Figure out why Sphinx sometimes fails to crossref :func: but wil
58 crossref :meth:, even though the dest is very clearly a function.
59 Example: :func:`wizard.app.php.re_var`
60 - The TODO extension for Sphinx doesn't properly force a full-rebuild
62 - Make single user mass-migrate work when not logged in as root. The
63 primary difficulty is making the parallel-find information easily
64 accessible to individual users: perhaps we can do a single-user
65 parallel-find on the fly.
66 - Don't use the scripts heuristics unless we're on scripts with the
67 AFS patch. Check with `fs sysname`
68 - Make 'wizard summary' generate nice pretty graphs of installs by date
69 (more histograms, will need to check actual .scripts-version files.)
70 - It should be able to handle installs like Django where there's a component
71 that gets installed in web_scripts and another directory that gets installed
73 - ACLs is a starting point for sending mail to users, but it has
74 several failure modes:
75 - Old maintainers who don't care who are still on the ACL
76 - Private AFS groups that aren't mailing lists and that we
78 A question is whether or not sending mail actually helps us:
79 many users will probably have to come back to us for help; many
80 other users won't care.
82 PULLING OUT CONFIGURATION FILES IN AN AUTOMATED MANNER
84 advancedpoll: Template file to fill out
85 django: Noodles of template files
86 gallery2: Multistage install process
88 mediawiki: One-step install process
89 phpbb: Multistage install process
90 phpical: Template file
93 wordpress: Multistage install process
95 COMMIT MESSAGE FIELDS:
97 Installed-by: username@hostname
98 Pre-commit-by: Real Name <username@mit.edu>
99 Upgraded-by: Real Name <username@mit.edu>
100 Migrated-by: Real Name <username@mit.edu>
101 Wizard-revision: abcdef1234567890
102 Wizard-args: /wizard/bin/wizard foo bar baz
106 Committer: Real Name <username@mit.edu>
107 Author: lockername locker <lockername@scripts.mit.edu>
111 - It is not required nor expected for update scripts to exist for all
112 intervening versions that were present pre-migration; only for it
113 to work on the most recent migration.
115 - Currently all repositories are initialized with --shared, which
116 means they have basically ~no space footprint. However, it
117 also means that /mit/scripts/wizard/srv MUST NOT lose revs after
122 * Some parts of the infrastructure will not be touched, although I plan
123 on documenting them. Specifically, we will be keeping:
125 - parallel-find.pl, and the resulting
126 /mit/scripts/.htaccess/scripts/sec-tools/store/scriptslist
128 * The new procedure for generating an update is as follows:
129 (check out the mass-migration instructions for something in this spirit,
130 although uglier in some ways; A indicates the step /should/ be automated)
132 0. ssh into not-backward, temporarily give the daemon.scripts-security-upd
133 bits by blanching it on system:scripts-security-upd, and run parallel-find.pl
135 1. [ see doc/upgrade.rst ]
137 [ENTER HERE FROM CREATING A NEW REPO]
139 9. Push all of your changes in a public place, and encourage others
140 to test, using --srv-path and a full path.
142 [ XXX: doc/deploy.rst ]
143 GET APPROVAL BEFORE PROCEEDING ANY FURTHER;
144 THIS IS PUSHING THE CHANGES TO THE PUBLIC
146 NOTE: The following commands are to be run on not-backward.mit.edu.
147 You'll need to add daemon.scripts-security-upd to
148 scripts-security-upd to get bits to do this. Make sure you remove
149 these bits when you're done.
151 10. Run `wizard research appname`
152 which uses Git commands to check how many
153 working copies apply the change cleanly, and writes out a logfile
154 with the working copies that don't apply cleanly. It also tells
155 us about "corrupt" working copies, i.e. working copies that
156 have over a certain threshold of changes.
158 11. Run `wizard mass-upgrade appname`, which applies the update to all working
161 12. Run parallel-find.pl to update our inventory
163 [ XXX: doc/upgrade.rst ]
164 * For mass importing into the repository, there are a few extra things:
166 * When mass producing updates, if the patch has changed you will have to
167 do a special procedure for your merge:
169 git checkout pristine
170 # NOTE: Now, the tricky part (this is different from a real update)
171 git symbolic-ref HEAD refs/heads/master
172 # NOTE: Now, we think we're on the master branch, but we have
173 # pristine copy checked out
174 # NOTE: -p0 might need to be twiddled
175 patch -p0 < ../app-1.2.3/app-1.2.3.patch
177 # reconstitute .scripts directory
178 git checkout v1.2.2-scripts -- .scripts
180 # NOTE: Fake the merge
181 git rev-parse pristine > .git/MERGE_HEAD
183 You could also just try your luck with a manual merge using the patch
186 [ XXX: doc/layout.rst ]
187 * The repository for a given application will contain the following files:
189 - The actual application's files, as from the official tarball
191 - A .scripts directory, with the intent of holding Scripts specific files
192 if they become necessary.
194 - .scripts/dsn, overriding database source name