f7a410eca9 2008-10-18       drh: From: "Ralf S. Engelschall"
f7a410eca9 2008-10-18       drh: Date: October 18, 2008 1:40:53 PM EDT
f7a410eca9 2008-10-18       drh: To: drh@hwaci.com
f7a410eca9 2008-10-18       drh: Subject: Some Fossil Feedback
f7a410eca9 2008-10-18       drh: 
f7a410eca9 2008-10-18       drh: I've today looked at your Fossil project (latest version from your
f7a410eca9 2008-10-18       drh: repository). That's a _really_ interesting and very promising DVCS.
f7a410eca9 2008-10-18       drh: While I looked around and tested it, I stumbled over some issues I want
f7a410eca9 2008-10-18       drh: to share with you:
f7a410eca9 2008-10-18       drh: 
f7a410eca9 2008-10-18       drh: o No tarball
f7a410eca9 2008-10-18       drh: 
9a482dd701 2008-10-24       drh:   You currently still do not provide any tarballs of the Fossil sources.
9a482dd701 2008-10-24       drh:   Sure, you are still hacking wild on the code and one can on-the-fly
9a482dd701 2008-10-24       drh:   fetch a ZIP archive, but that's a manual process. For packagers (like
9a482dd701 2008-10-24       drh:   me in the OpenPKG project) this is very nasty and just means that
9a482dd701 2008-10-24       drh:   Fossil will usually still not be packaged. As a result it will be not
9a482dd701 2008-10-24       drh:   spreaded as much as possible and this way you still do not get as much
9a482dd701 2008-10-24       drh:   as possible feedback. Hence, I recommend that you let daily snapshot
9a482dd701 2008-10-24       drh:   tarballs (or ZIP files) be rolled. This would be great.
f7a410eca9 2008-10-18       drh: 
f7a410eca9 2008-10-18       drh: o UUID
f7a410eca9 2008-10-18       drh: 
9a482dd701 2008-10-24       drh:   Under http://www.fossil-scm.org/index.html/doc/tip/www/concepts.wiki
9a482dd701 2008-10-24       drh:   you describe the concepts and you clearly name the artifact ID a
9a482dd701 2008-10-24       drh:   "UUID" and even say that this is a "Universally Unique Identifier".
9a482dd701 2008-10-24       drh:   Unfortunately, a UUID is a 128-bit entity standardized by the ISO
9a482dd701 2008-10-24       drh:   (under ISO/IEC 11578:1996) and IETF (under RFC-4122) and hence it is
9a482dd701 2008-10-24       drh:   *VERY MUCH* confusing and unclean that your 160-bit SHA1-based ID is
9a482dd701 2008-10-24       drh:   called "UUID" in Fossil.
9a482dd701 2008-10-24       drh: 
9a482dd701 2008-10-24       drh:   I *STRONGLY* recommend you to either use real UUIDs (128-bit
9a482dd701 2008-10-24       drh:   entities, where a UUID of version 5 even is SHA1-based!) or you name
9a482dd701 2008-10-24       drh:   your 160-bit SHA1 entities differently (perhaps AID for "Artifact
9a482dd701 2008-10-24       drh:   Identifier"?).
f7a410eca9 2008-10-18       drh: 
f7a410eca9 2008-10-18       drh: o "fossil cgi <script>"
f7a410eca9 2008-10-18       drh: 
9a482dd701 2008-10-24       drh:   Currently we have only "fossil cgi <script>" where <script> contains
9a482dd701 2008-10-24       drh:   "repository: <file>". This is a neat idea and should be kept. But
9a482dd701 2008-10-24       drh:   this way one cannot easily host multiple repositories over this
9a482dd701 2008-10-24       drh:   CGI interface without creating such a script for every individual
9a482dd701 2008-10-24       drh:   repository.
9a482dd701 2008-10-24       drh: 
9a482dd701 2008-10-24       drh:   Perhaps a "fossil cgi --repository <file>" would help here, as this
9a482dd701 2008-10-24       drh:   way one can use a generic CGI script which first figures out the
9a482dd701 2008-10-24       drh:   name <name> of the individual project and then runs "fossil cgi
9a482dd701 2008-10-24       drh:   --repository /path/to/repositories/<name>.db". But perhaps I'm just
9a482dd701 2008-10-24       drh:   overlooking something and this is already possible...
f7a410eca9 2008-10-18       drh: 
f7a410eca9 2008-10-18       drh: o "fossil db <operation>"
f7a410eca9 2008-10-18       drh: 
9a482dd701 2008-10-24       drh:   In Monotone DVCS we have a "mtn db" command for the low-level SQLite
9a482dd701 2008-10-24       drh:   database manipulations. For instance a "mtn --db=<monotone-repository>
9a482dd701 2008-10-24       drh:   db dump" is more or less equal to a "sqlite3 <monotone-repository>
9a482dd701 2008-10-24       drh:   .dump". A "mtn --db=<monotone-repository> db exec '<SQL>'" is equal
9a482dd701 2008-10-24       drh:   "echo '<SQL>' | sqlite3 <monotone-repository>", etc. The point is
9a482dd701 2008-10-24       drh:   that the DVCS user usually has no SQLite CLI at hand but as the DVCS
9a482dd701 2008-10-24       drh:   already contains more or less the complete SQLite it is very useful to
9a482dd701 2008-10-24       drh:   also include the SQLite CLI functionality and wire it into the DVCS
9a482dd701 2008-10-24       drh:   CLI via a "db" command.
f7a410eca9 2008-10-18       drh: 
f7a410eca9 2008-10-18       drh: o "fossil version"
f7a410eca9 2008-10-18       drh: 
9a482dd701 2008-10-24       drh:   Mostly all VCS I know if either support a command "version" or a
9a482dd701 2008-10-24       drh:   command-line option "--version" or most of the time even both. Please
9a482dd701 2008-10-24       drh:   output the "This is fossil version [9e80dc66cf] 2008-10-18 13:03:36"
9a482dd701 2008-10-24       drh:   there (at least additionally) instead of at the end of "fossil help".
f7a410eca9 2008-10-18       drh: 
f7a410eca9 2008-10-18       drh: o "--port" vs. "-port"
f7a410eca9 2008-10-18       drh: 
9a482dd701 2008-10-24       drh:   In the "fossil help server" there is "--port" while under
9a482dd701 2008-10-24       drh:   http://www.fossil-scm.org/index.html/doc/tip/www/quickstart.wiki there
9a482dd701 2008-10-24       drh:   is "-port". I personally like GNU-long-style options "--xxxx" more
9a482dd701 2008-10-24       drh:   than the CC-style options "-xxx". But anyway, independent which one is
9a482dd701 2008-10-24       drh:   correct (well, perhaps both even work) only one should be documented
9a482dd701 2008-10-24       drh:   to avoid confusion.
f7a410eca9 2008-10-18       drh: 
f7a410eca9 2008-10-18       drh: o User creation on the CLI
f7a410eca9 2008-10-18       drh: 
9a482dd701 2008-10-24       drh:   There is "fossil user new ?USERNAME?" which interactively prompts
9a482dd701 2008-10-24       drh:   for the capabilities and the password -- even from the TTY and not
9a482dd701 2008-10-24       drh:   from STDIN. One needs "fossil user new ?USERNAME? ?CAPABILITIES?
9a482dd701 2008-10-24       drh:   ?PASSWORD?" to be able to create a repository in batch without having
9a482dd701 2008-10-24       drh:   to hack the user into the "user" table via the SQLite CLI. Similar:
9a482dd701 2008-10-24       drh:   the "fossil user password USERNAME" should be actually "fossil user
9a482dd701 2008-10-24       drh:   password USERNAME ?PASSWORD?", please.
f7a410eca9 2008-10-18       drh: 
f7a410eca9 2008-10-18       drh: o "-R <repository"
f7a410eca9 2008-10-18       drh: 
9a482dd701 2008-10-24       drh:   There is the "-R" option for specifiying the repository. But it is
9a482dd701 2008-10-24       drh:   a sub-command option and not a global option. And on "fossil ui" or
9a482dd701 2008-10-24       drh:   "fossil server" it even is an argument and not an option. This is
9a482dd701 2008-10-24       drh:   confusing. Every time one has to figure out how to set the repository
9a482dd701 2008-10-24       drh:   on the CLI. Monotone simply uses a *global* option "--db" and that's
9a482dd701 2008-10-24       drh:   it. So, I recommend that Fossil also uses a single global option
9a482dd701 2008-10-24       drh:   "--repository"/"-R" instead of a command-specific option. Sure, not
9a482dd701 2008-10-24       drh:   all commands might require a repository, but IMHO it is better to
9a482dd701 2008-10-24       drh:   ignore the global option there than having to figure out every time
9a482dd701 2008-10-24       drh:   how the repository has to be specified.
f7a410eca9 2008-10-18       drh: 
f7a410eca9 2008-10-18       drh: o Setup pages
f7a410eca9 2008-10-18       drh: 
9a482dd701 2008-10-24       drh:   When hitting "Apply changes" on any setup page, one keeps staying on
9a482dd701 2008-10-24       drh:   this page. Sure, that's what an "apply" button usually is about. But
9a482dd701 2008-10-24       drh:   I at least would have liked to see a button (either instead of the
9a482dd701 2008-10-24       drh:   "apply changes" or at least in addition) which applies the changes and
9a482dd701 2008-10-24       drh:   goes back to the setup main page (from where one usually come).
f7a410eca9 2008-10-18       drh: 
f7a410eca9 2008-10-18       drh: o _FOSSIL_
f7a410eca9 2008-10-18       drh: 
9a482dd701 2008-10-24       drh:     Very nice idea to use another SQLite database for the _FOSSIL_
9a482dd701 2008-10-24       drh:     control file. BUT: Why "_FOSSIL_"? CVS's "CVS" directory was ugly for
9a482dd701 2008-10-24       drh:     decades. Today we have ".svn", ".git", ".hg" and "_MTN"/".mtn" to get
9a482dd701 2008-10-24       drh:     rid of those ugly control files or directories of the DVCS! Sure,
9a482dd701 2008-10-24       drh:     dot-files are disliked by Windows people. But that's no problem, one
9a482dd701 2008-10-24       drh:     can accept "_XXX" in addition to ".XXX" easily, of course.
9a482dd701 2008-10-24       drh: 
9a482dd701 2008-10-24       drh:     So, I really would like to see the file "_FOSSIL_" to be renamed
9a482dd701 2008-10-24       drh:     to ".fossil" under Unix and "_fossil" under Windows or (if the
9a482dd701 2008-10-24       drh:     upper-case is important to you) ".FOSSIL" and "_FOSSIL". But to see
9a482dd701 2008-10-24       drh:     an ugly "_FOSSIL_" at the top of every source tree is really rather
9a482dd701 2008-10-24       drh:     boring for Unix people!
f7a410eca9 2008-10-18       drh: 
f7a410eca9 2008-10-18       drh: o "fossil open", "fossil checkout", "fossil update".
f7a410eca9 2008-10-18       drh: 
9a482dd701 2008-10-24       drh:   I'm personally confused that "fossil open" is what "checkout" does in
9a482dd701 2008-10-24       drh:   other DVCS and that "checkout" is just a variant of "update". Hmmm...
9a482dd701 2008-10-24       drh:   for me it looks at least one should be eleminated to reduce confusion.
9a482dd701 2008-10-24       drh:   The difference between checkout and update could become an option of
9a482dd701 2008-10-24       drh:   a single option. And the remaining to perhaps better should be either
9a482dd701 2008-10-24       drh:   "open" + "update" or "checkout" + "update". Well, perhaps I'm still
9a482dd701 2008-10-24       drh:   not understanding Fossil enough. But I can only tell that I got rather
9a482dd701 2008-10-24       drh:   confused.
f7a410eca9 2008-10-18       drh: 
f7a410eca9 2008-10-18       drh: o "fossil commit"
f7a410eca9 2008-10-18       drh: 
9a482dd701 2008-10-24       drh:   There is "fossil commit" but no short-hand "fossil ci". But there
9a482dd701 2008-10-24       drh:   is "fossil status" and even just "fossil st" which clearly shows
9a482dd701 2008-10-24       drh:   that Fossil allows abbreviations or at least prefix-matching on the
9a482dd701 2008-10-24       drh:   commands. Sure, "ci" is not a prefix of "commit" but mostly all VCS
9a482dd701 2008-10-24       drh:   support the good old RCS/CVS "ci" and similar aliases. So I recommend
9a482dd701 2008-10-24       drh:   that Fossil does not only alias matching, but also provides aliases:
9a482dd701 2008-10-24       drh:   "ci" for "commit", "co" for "checkout", "log" for "timeline", etc.
9a482dd701 2008-10-24       drh:   Sorry, having to type the long "fossil commit" every time is too long
9a482dd701 2008-10-24       drh:   for us Unix hackers ;-)
f7a410eca9 2008-10-18       drh: 
9a482dd701 2008-10-24       drh:   Additionally, Fossil seems to use GnuPG when installed and --nosign is
9a482dd701 2008-10-24       drh:   not specified. Hmm... two questions arise here for me: 1. should the
9a482dd701 2008-10-24       drh:   use of a cryptographically strong signature really be just _optional_
9a482dd701 2008-10-24       drh:   (Monotone for instance RSA signs every commit) and 2. is GnuPG here
9a482dd701 2008-10-24       drh:   really the right tool (Monotone does a plain RSA signing and is even
9a482dd701 2008-10-24       drh:   able to leverage a running SSH agent for the operation -- something
9a482dd701 2008-10-24       drh:   which is very cool both under Unix with "ssh-agent" and under Windows
9a482dd701 2008-10-24       drh:   with "pagent"). OTOH, GnuPG 2.x supports a gpg-agent, so this might be
9a482dd701 2008-10-24       drh:   no longer a big issue. But Fossil should document this a little bit
9a482dd701 2008-10-24       drh:   futher: how to create the necessary GnuPG key, how to setup gpg-agent,
9a482dd701 2008-10-24       drh:   etc.
f7a410eca9 2008-10-18       drh: 
f7a410eca9 2008-10-18       drh: o "fossil diff"
f7a410eca9 2008-10-18       drh: 
9a482dd701 2008-10-24       drh:   There is "Usage: fossil diff|gdiff ?-i? ?-r REVISION? FILE...". Two
9a482dd701 2008-10-24       drh:   questions arise: Why do I have to specify a "FILE"? I would expect
9a482dd701 2008-10-24       drh:   that a simple "fossil diff" recursively shows a "diff" of all changed
9a482dd701 2008-10-24       drh:   files in the source tree. Additionally, how does one do a diff between
9a482dd701 2008-10-24       drh:   two particular revisions (one seems to be not able to specifiy "-r"
9a482dd701 2008-10-24       drh:   twice).
f7a410eca9 2008-10-18       drh: 
f7a410eca9 2008-10-18       drh: o "manifest" / "manifest.uuid"
f7a410eca9 2008-10-18       drh: 
9a482dd701 2008-10-24       drh:   Above I was already bothered by the _FOSSIL_ file, but now that I
9a482dd701 2008-10-24       drh:   played with Fossil I see that "manifest" and "manifest.uuid" files
9a482dd701 2008-10-24       drh:   showed up in my source tree. Why is this not implicitly in the
9a482dd701 2008-10-24       drh:   database and instead externally represented in the source tree.
9a482dd701 2008-10-24       drh:   Hmmm... then I recommend that you use a .fossil *DIRECTORY*, place
9a482dd701 2008-10-24       drh:   the control file into .fossil/control (or whatever) and the manifest
9a482dd701 2008-10-24       drh:   stuff into .fossil/manifest and .fossil/manifest.uuid. But please do
9a482dd701 2008-10-24       drh:   not clutter the top-level of the source three with control data of the
9a482dd701 2008-10-24       drh:   DVCS.
f7a410eca9 2008-10-18       drh: 
f7a410eca9 2008-10-18       drh: o "fossil mv"
f7a410eca9 2008-10-18       drh: 
9a482dd701 2008-10-24       drh:   There is "fossil add" and "fossil rm", but no "fossil mv". Hopefully
9a482dd701 2008-10-24       drh:   this doesn't mean a file move is an "add" and a "remove" bundle, as
9a482dd701 2008-10-24       drh:   this way Fossil would have trouble to know that the file was renamed
9a482dd701 2008-10-24       drh:   and hence the full history tracking of a file seems to be broken.
f7a410eca9 2008-10-18       drh: 
f7a410eca9 2008-10-18       drh: o "fossil changes" vs. "fossil status"
f7a410eca9 2008-10-18       drh: 
9a482dd701 2008-10-24       drh:   Finally, I got confused by the "fossil changes" and "fossil status"
9a482dd701 2008-10-24       drh:   commands. "fossil status" seems to be a super-set of "fossil changes".
9a482dd701 2008-10-24       drh:   Looks a little bit less orthogonal than necessary. I would remove
9a482dd701 2008-10-24       drh:   "fossil changes" or at least do not show the file details on "fossil
9a482dd701 2008-10-24       drh:   status".
f7a410eca9 2008-10-18       drh: 
f7a410eca9 2008-10-18       drh: Yours,
9a482dd701 2008-10-24       drh:                                                                             Ralf S. Engelschall
f7a410eca9 2008-10-18       drh: