5 - If you try to do an install on scripts w/o sql, it will sign you up but fail to write
6 the sql.cnf file. This sucks.
7 - Make an awesome wizard subcommand that you can point at a file that is DOS
8 when the source wants it to be UNIX, or the other way around, and it will
9 redo the merge on that file correctly. Alternatively, just make wizard
10 sufficiently awesome to detect line-ending mismatch and fix it w/o any
12 - Use diff3 conflictstyle
14 - wizard install wordpress should ask for password. One problem with this is that
15 Wordpress will still send mail with the wrong username and password, so Wordpress
16 will need to be patched to not do that. Alternatively we can initally set the admin
17 email to a null address and then fix it manually.
18 - --raw parameter for install which means an arbitrary commit can be installed
19 - Parse output HTML for class="error" and give those errors back to the user (done),
20 then boot them back into configure so they can enter in something different
21 - Get rid of our custom sizing code and use dialog's built-in sizing (i.e. width=0, height=0).
22 Maybe our sizing code is superior, maybe not.
24 - Replace gaierror with a more descriptive name (this is a DNS error)
26 - Pre-emptively check if daemon/scripts-security-upd
27 is not on scripts-security-upd list (/mit/moira/bin/blanche)
29 - Web application for installing autoinstalls has a hard problem
30 with credentials (as well as installations that are not conducted
31 on an Athena machine.) We have some crazy ideas involving a signed
32 Java applet that uses jsch to SSH into athena.dialup and perform
36 - Tidy up common code in callAsUser and drop_priviledges in shell
37 - Summary script should be more machine friendly, and should not
38 output summary charts when I increase specificity
39 - Summary script should do something intelligent when distinguishing
40 between old-style and new-style installs
41 - Report code in wizard/command/__init__.py is ugly as sin. Also,
42 the Report object should operate at a higher level of abstraction
43 so we don't have to manually increment fails. (in fact, that should
44 probably be called something different). The by-percent errors should
46 - Move resolutions in mediawiki.py to a text file? (the parsing overhead
48 - If a process is C-ced, it can result in a upgrade that has
49 an updated filesystem but not updated database. Make this more
51 - PHP end of file allows omitted semicolon, can result in parse error
52 if merge resolutions aren't careful. `php -l` can be a quick stopgap
55 - Figure out why Sphinx sometimes fails to crossref :func: but wil
56 crossref :meth:, even though the dest is very clearly a function.
57 Example: :func:`wizard.app.php.re_var`
58 - The TODO extension for Sphinx doesn't properly force a full-rebuild
60 - Make single user mass-migrate work when not logged in as root. The
61 primary difficulty is making the parallel-find information easily
62 accessible to individual users: perhaps we can do a single-user
63 parallel-find on the fly.
64 - Don't use the scripts heuristics unless we're on scripts with the
65 AFS patch. Check with `fs sysname`
66 - Make 'wizard summary' generate nice pretty graphs of installs by date
67 (more histograms, will need to check actual .scripts-version files.)
68 - It should be able to handle installs like Django where there's a component
69 that gets installed in web_scripts and another directory that gets installed
71 - ACLs is a starting point for sending mail to users, but it has
72 several failure modes:
73 - Old maintainers who don't care who are still on the ACL
74 - Private AFS groups that aren't mailing lists and that we
76 A question is whether or not sending mail actually helps us:
77 many users will probably have to come back to us for help; many
78 other users won't care.
80 PULLING OUT CONFIGURATION FILES IN AN AUTOMATED MANNER
82 advancedpoll: Template file to fill out
83 django: Noodles of template files
84 gallery2: Multistage install process
86 mediawiki: One-step install process
87 phpbb: Multistage install process
88 phpical: Template file
91 wordpress: Multistage install process
93 COMMIT MESSAGE FIELDS:
95 Installed-by: username@hostname
96 Pre-commit-by: Real Name <username@mit.edu>
97 Upgraded-by: Real Name <username@mit.edu>
98 Migrated-by: Real Name <username@mit.edu>
99 Wizard-revision: abcdef1234567890
100 Wizard-args: /wizard/bin/wizard foo bar baz
104 Committer: Real Name <username@mit.edu>
105 Author: lockername locker <lockername@scripts.mit.edu>
109 - It is not required nor expected for update scripts to exist for all
110 intervening versions that were present pre-migration; only for it
111 to work on the most recent migration.
113 - Currently all repositories are initialized with --shared, which
114 means they have basically ~no space footprint. However, it
115 also means that /mit/scripts/wizard/srv MUST NOT lose revs after
120 * Some parts of the infrastructure will not be touched, although I plan
121 on documenting them. Specifically, we will be keeping:
123 - parallel-find.pl, and the resulting
124 /mit/scripts/.htaccess/scripts/sec-tools/store/scriptslist
126 * The new procedure for generating an update is as follows:
127 (check out the mass-migration instructions for something in this spirit,
128 although uglier in some ways; A indicates the step /should/ be automated)
130 0. ssh into not-backward, temporarily give the daemon.scripts-security-upd
131 bits by blanching it on system:scripts-security-upd, and run parallel-find.pl
133 1. [ see doc/upgrade.rst ]
135 [ENTER HERE FROM CREATING A NEW REPO]
137 9. Push all of your changes in a public place, and encourage others
138 to test, using --srv-path and a full path.
140 [ XXX: doc/deploy.rst ]
141 GET APPROVAL BEFORE PROCEEDING ANY FURTHER;
142 THIS IS PUSHING THE CHANGES TO THE PUBLIC
144 NOTE: The following commands are to be run on not-backward.mit.edu.
145 You'll need to add daemon.scripts-security-upd to
146 scripts-security-upd to get bits to do this. Make sure you remove
147 these bits when you're done.
149 10. Run `wizard research appname`
150 which uses Git commands to check how many
151 working copies apply the change cleanly, and writes out a logfile
152 with the working copies that don't apply cleanly. It also tells
153 us about "corrupt" working copies, i.e. working copies that
154 have over a certain threshold of changes.
156 11. Run `wizard mass-upgrade appname`, which applies the update to all working
159 12. Run parallel-find.pl to update our inventory
161 [ XXX: doc/upgrade.rst ]
162 * For mass importing into the repository, there are a few extra things:
164 * When mass producing updates, if the patch has changed you will have to
165 do a special procedure for your merge:
167 git checkout pristine
168 # NOTE: Now, the tricky part (this is different from a real update)
169 git symbolic-ref HEAD refs/heads/master
170 # NOTE: Now, we think we're on the master branch, but we have
171 # pristine copy checked out
172 # NOTE: -p0 might need to be twiddled
173 patch -p0 < ../app-1.2.3/app-1.2.3.patch
175 # reconstitute .scripts directory
176 git checkout v1.2.2-scripts -- .scripts
178 # NOTE: Fake the merge
179 git rev-parse pristine > .git/MERGE_HEAD
181 You could also just try your luck with a manual merge using the patch
184 [ XXX: doc/layout.rst ]
185 * The repository for a given application will contain the following files:
187 - The actual application's files, as from the official tarball
189 - A .scripts directory, with the intent of holding Scripts specific files
190 if they become necessary.
192 - .scripts/dsn, overriding database source name