Re: [fossil-users] artifacts received

[Ron Wilson]

On Sun, Jul 6, 2014 at 3:54 AM, Stephan Beal  wrote:

>
> Can't answer, but i see it only with 'pull', not with 'update'.
>
> [odroid@host:~/fossil/fossil]$ f pull
> Pull from http://step...@fossil-scm.org/index.html
> Round-trips: 1   Artifacts sent: 0  received: 74
> Pull finished with 453 bytes sent, 2644 bytes received
>
> [odroid@host:~/fossil/fossil]$ f up
> Autosync:  http://step...@fossil-scm.org/index.html
> Round-trips: 1   Artifacts sent: 0  received: 0
> Pull finished with 438 bytes sent, 612 bytes received
>

Maybe there was nothing left for update to pull?
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Automating a clone - any other way than the URL to pass in password?

[Ron Wilson]

On Sun, Jul 6, 2014 at 6:12 AM, Stephan Beal  wrote:

> Looking at the code (getpass() in user.c)... i'm not sure. It uses
> getc(stdin) to reach char by char, but doesn't seem to do anything unusual
> with the stream. Ah... that's the Windows/Android impl. On unix it uses
> getpass(3) (unistd.h), which might do something weird to prohibit piping in
> input.
>

As I recall, getpass(3) opens its own file descriptor to the controlling
"terminal", configures it with some reasonable, non-echo settings, then
reads the user's password.

Expect gets around this by using a pseudo terminal. the child side of the
fork assigns the "slave" side of the pseudo terminal to STDIN, STDOUT and
STDERR, then execs the command (or the shell to run the command). Then
expect monitors and controls the command via the control side of the pseudo
terminal.

FYI, xterm (and other Unix/Linux/etc command line windows) work the same
way.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Using fossil with Bitbake and the Yocto project

[Ron Wilson]

On Tue, Jul 1, 2014 at 5:00 PM, Ross Berteig  wrote:

>
> And I suspect that someday libfossil will provide the needed bits for that
> plugin. Just not today. :-)


My point was not whether Fossil (as Fossil or as libfossil) provided the
needed bits, but that writing a plug-in is harder than configuring a
command line interface. Just as an example, in my Slick Edit config, the
check in command is: fossil ci -M %c %f

As for Fossil providing the needed bits to write a plug-in to any given
tool, it is possible to write a plug-in that uses the command line Fossil.
Or, for that matter, a hybrid that uses libfossil for information and
fossil to perform commits, updates, etc.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Using fossil with Bitbake and the Yocto project

[Ron Wilson]

On Tue, Jul 1, 2014 at 2:39 PM, Ross Berteig  wrote:

>
> BitBake wants *both* MD5 and SHA256 of the tarball that it plans to
> download. Since the tarball contains a folder name supplied by URL
> parameters, fossil would have to construct the tarball with parameters and
> compute both checksums. I don't see any reason to do that work until that
> specific tarball is requested, and the path of least resistance is to just
> let BitBake compute them the first time. Unfortunately it does case BitBake
> to download the same tarball a second time, but on the grand scale of
> things that is just an annoyance.
>

I would guess this derives from the (increasingly less) common practice of
projects generating and publishing an "official" tarball and its hashes.
Curious it would still want MD5 as this has been deprecated for years in
favor of SHA256 or SHA512. Is MD5 actually a Yocto requirement?


>
> With Git, it can clone a repo and then checkout the requested version for
> building from its own clone of the repo. If I had more time and budget, I'd
> delve into that mechanism and figure out how to extend BitBake to have the
> similar abilities for a fossil repo.


If BitBake provides for configuring "generic" command line VCS clients,
this should be easy. However, I have noticed an increasing number of open
source projects "deprecating" general support for command line clients in
favor of "proper" plug-ins. In contrast, many commercial tools continue to
support command line clients either along side or instead of plug-ins.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Other tools

[Ron Wilson]

On Thu, Jun 19, 2014 at 11:32 PM, Noah Roberts 
wrote:

> Are there fossil plugins for things like Jenkins (or any other CI),
> eclipse mylyn, etc?
>
> I'm considering fossil as a basis for a new project and wondering if it
> will serve my needs.  I really like the idea of it, but I also want
> continuous integration.  I'd like mylyn integration but my main concern is
> with CI.
>

In my experience, it's usually the "other tool" that has plug-ins for
various VCSs. Some tools have straight forward ways to configure command
line based external tools. Unfortunately, many of the most popular, like
Eclipse, require coding actual plug-ins (or for someone to code a "command
line VCS" plug-in that could be configured as needed.

The IDE I currently use at work, Slick Edit (a "pay ware" product) does
support command line VCSs. For Fossil, I was able to configure sensible
command templates for most of the commands Slick Edit supports (Add,
Checkin, Difference, Get, History (the rest I ignored)), then handle
sync/push/pull, etc, from the command line.

Depending on what you want, a plug-in to run the command line Fossil might
not be harder than a plug-in that would use libfossil. It could be as
simple as launching an xterm or command window to run Fossil. Your results
would be displayed in the window and you'd be able to answer any questions
Fossil asks. Not pretty, but functional. Interfacing to a CI tool, I
assume, would log Fossil messages and return Fossil's exit status to the
tool.

(Our process requires :"change packages" to be committed to branches,
reviewed, then merged (one at at time) and tested by the "release engineer"
before being committed to the trunk (or release branch). We have discussed
using a CI tool, but have not worked out how we would fit it in to our
process.)
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Playing with ticket in console

[Ron Wilson]

On Fri, Jun 20, 2014 at 12:39 PM, Andreas Kupries 
wrote:

> On Fri, Jun 20, 2014 at 4:55 AM, Stephan Beal 
> wrote:
> > Hi! The tickets subsystem is www-only. There are neither cli nor json
> > interfaces for it.
>
> So what is documented by
> fossil help ticket
> ?
>
> Yes, the ticket system does have some cli commands for working it.
> While nothing for reconfiguration of the system in itself, access to
> the tickets is present.


I had forgotten about this. We set up a script for our Project Managers to
export tickets to spreadsheets for reports. fortunately, we haven't had to
import from spreadsheets, so haven't a script for that.

Unfortunately, this doesn't seem to have any means for accessing
attachments, not even a field for listing artifact IDs of attachments.

Otherwise, all fields seem to be readable, and most fields seem to be
set-able/append-able through this.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] fossil CLI tricks: negative pattern matching

[Ron Wilson]

On Thu, Jun 19, 2014 at 12:09 PM, Alysson Gonçalves de Azevedo <
agalys...@gmail.com> wrote:

> Hi all. Just like Beal shared with us the # trick, today i'll share how i
> can commit all except one file using negative patterns
> ,
> something like !(file).
>
> /tmp/fossil $ fossil commit -m message folder/!(ipsum)
> New_Version: 35556f15c25ebe4c5cef11cda3b7e7d59d89a4e2
>

Which shell? bash  gives me the error:  !: event not found
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] How to deal with incomplete code?

[Ron Wilson]

On Sat, Jun 14, 2014 at 4:02 PM,  wrote:

> 2. Can a branch contain only the single incomplete file I'm working on,
> but somehow to 'load' this branch together with the 'trunk'?  I.e., I want
> to have all the files of the current trunk plus the single (or few) file(s)
> of my special WIP branch.  Is this possible?  If yes, how?


Merging only merges in to your working space, so if you fetch a commit from
your branch, then merge from Trunk, your working space will contain what
you are requesting. Usually, after resoling any conflicts, the result would
be committed to the branch, but that is not actually required. If you
don't, then any fetch from your branch will result in a working space
containing out of date files. That is, those files you haven't made changes
to will be as they were when the branch was created. Usually this is not
desirable as you would then always have to merge in  from Trunk. Where as
if you committed the result of your merge, you only have to merge when
Trunk actually changed.

There is only a very tiny cost to committing merged-in files not changed by
a branch. The resulting delta will effectively be a pointer to a commit in
Trunk (or other branch). Also, any record keeping advantages to not
committing the merge is significantly outweighed by the cost of forgetting
to merge in changes when checking out your branch.

With an up to date branch, the -n option for a merge from your branch will
give you a list of the changed files. Also, in the time line, it is easy to
distinguish changes from merge-commits, so you can still easily see what
files you've been working on.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] autocrlf like in Git?

[Ron Wilson]

On Sat, Jun 7, 2014 at 8:37 PM, Nico Williams  wrote:

> The same is true for git, and Mercurial, and...  It doesn't mean it
> can't be done, just that the VCS has to know how to canonicalize the
> file's contents, and that is awful.
>
> But sometimes someone has no choice.
>
> Depends on how you define "no choice".

I can't speak to git, et al, not turns out one of my coworkers was a
student of the creator of RCS, Walter Tichy, and did work on RCS at that
time.

(Disclaimer: I am NOT advocating building the auto conversion into Fossil,
only relaying a story of how one port of RCS acquired the auto conversion.
(I really don't think it should be in Fossil or other VCS.))

When he graduated, his first job was at a company that was, at the time,
using terminal connected to a VAX running BSD Unix. Then company started
getting PCs, which ran MS-DOS back then. At first, they continued to
develop on Unix, only pulling down source files when they needed to compile
on MS-DOS. Of course, this required runing the files through unix2dos, but
this was part of the script that pulled the files from the Unix system.

Then another graduate was hired and he brought with him a very early MS-DOS
client for Unix file servers. Very soon, management asked if there was a
way to run RCS on MS-DOS. My (now) coworker volunteered to port RCS to
MS-DOS. Because they were also versioning WordStar and other types of
non-text files, his port of RCS did not use MS-DOS's "text mode" to
read/write files, so unix2dos and dos2unix were still required. At first,
that continued to be hidden in the makefiles, but then management decided
the team should start doing their development on MS-DOS and only push files
to the Unix system for compiling.

At first, my friend just reworked the makefiles to convert and commit any
changed files as part of the build process, thus leaving the source files
in CRLF format on the PCs. And created and update script that would
automatically update a developer's working spacing from RCS. But soon,
people started using RCS directly, like they had on Unix. This created a
big mess. Management's response was to move the conversion to the Unix
side, leaving the sources in RCS in CRLF format. This worked until older
versions was needed. Although the makefiles still had the conversion in
them, editing the files became an issue and people was constantly
forgetting to use unix2dos.

Then management decided the conversion had to be done in RCS. At first, my
friend created a .bat file around RCS for the developers, but this had
problems, so he finally had to add the conversion to RCS. Actually, it was
just the logic to decide when to convert.

He doesn't if it was his port to MS-DOS actually made it into the master
RCS repo, but whichever one did, it's likely it did arise from a situation
like this one, where it was the expedient thing to do.

While I suspect Linus and company may not have included the auto conversion
in git, later maintainers likely decided it was expedient to do so, whether
from their own experience or just bowing to a plentitude of requests from
the user community.

(Since SVN was intended to replace CVS and CVS evolved out of RCS, I am
guessing that SVN "inherited" auto conversion from RCS.)
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] autosync from GUI

[Ron Wilson]

On Sat, Jun 7, 2014 at 4:37 PM, Scott Robison 
wrote:

> On Jun 7, 2014 1:27 PM, "Ron Wilson"  wrote:
> >
> > On Sat, Jun 7, 2014 at 2:23 PM, Stephan Beal 
> wrote:
> >>
> >> For the local UI case, sure, i can see it being useful, but people
> would also expect it to work remotely, and it often wouldn't.
> >
> > When running the local UI, that is seen as part of the local client, but
> when accessing a remote server the perception is different. I think it
> reasonable that the "server Fossil" not try to autosync to another server
> Fossil.
>
> Please forgive what may be a stupid observation, but if there are users
> that need ticket or wiki access (workflows that are typically ui only),
> might not the best approach be to give them appropriate access to the
> master via http (as someone already says they do indirectly by making that
> the master copy)?
>
Not stupid. I suspect many Fossil installations do that. The Fossil and
SQLite main sites do that.

Sometimes, thought, that isn't practical. When those people aren't in the
office, its not too hard for them to understand that the syncs don't
happen. In the office, however, it is natural for them to expect the syncs
are done automatically.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Cannot determine user when using -A

[Ron Wilson]

On Sun, Jun 8, 2014 at 3:25 PM, B Harder  wrote:

> Fossil certainly will spawn external programs if you:
>
> 1) start a commit w/o a command-line commit message
>
> 2) run a did with an external diff-command configured.
>
I had been thinking about hooks and discussion of multi platform support
issues.

I had forgotten about invoking an external editor or diff tool.

(I had not forgotten about Fossil forking copies of itself. I just wasn't
counting those as external tool invocations.)
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] autocrlf like in Git?

[Ron Wilson]

On Sat, Jun 7, 2014 at 2:32 PM, Richard Hipp  wrote:

>
>
> Really?  What tool are you using where the line ending makes a difference?
>

At least in my software team, no, but we do share test scripts with the
test engineering team and their tools do care about line endings. Our work
around, since those scripts have extensions that do not conflict with our
source files, is to configure our editors to convert line endings on file
write.  (We don't like this, but it keeps the test team from whiling at
their manager (and, therefore, the director of engineering from harassing
us.)
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] autosync from GUI

[Ron Wilson]

On Sat, Jun 7, 2014 at 2:23 PM, Stephan Beal  wrote:

> For the local UI case, sure, i can see it being useful, but people would
> also expect it to work remotely, and it often wouldn't.
>

When running the local UI, that is seen as part of the local client, but
when accessing a remote server the perception is different. I think it
reasonable that the "server Fossil" not try to autosync to another server
Fossil.

Granted, in a large enough project to have a hierarchy of  core developers,
lack of server-to-server autosync might be inconvenient, but as far as I
know, large projects using git worry about this.

So, I think it would be worth it to have autosync for "fossil ui" without
also having it for the various "real" server modes.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] autosync from GUI

[Ron Wilson]

On Sat, Jun 7, 2014 at 1:26 PM, Eric Rubin-Smith  wrote:

> Two of Fossil's self-proclaimed major benefits are: (a) its GUI and
> (b) its autosync feature.  That the GUI does not attempt to implement
> autosync at all *is* surprising.
>
> Note that in non-tiny teams, there is often a "project manager"
> type whose job includes defining ticket reports, categorizing and
> prioritizing tickets, editing the wiki, and so on.  This type of person
> might exclusively use the GUI.  Forcing them to go use the CLI (even
> after they use the GUI Admin page to choose the 'autosync' setting)
> feels stranger still, since they would have no reason to use the CLI
> otherwise.
>

Because my team is using Fossil without support from IT, we made the
Project Manager's PC (which is under his desk) our main server.

But if not for this, we would have to setup his PC to run a scheduled sync
many times per day.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Cannot determine user when using -A

[Ron Wilson]

On Sat, Jun 7, 2014 at 11:52 AM, Stephan Beal  wrote:

>
> On Sat, Jun 7, 2014 at 5:18 PM, Will Parsons 
> wrote:
>
>> Or you should be able to simply set the environment variable:
>>
>> $ USER=philip_bennefall
>>
>> and then you don't need to worry about the -A or --user options.
>>
>
> That would work, but note that many programs use $USER to create strings
> (paths) and such, so reassigning it is generally not a good idea. However...
>
> export FOSSIL_USER=philip_bennefall
>
> should do the trick without endangering any other apps.
>

Also:

env USER= philip_bennefall fossil command options etc

Will run fossil will that value of USER but not affect other things (unless
fossil invokes them, but fossil doesn't do that)
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Outside contribution feature

[Ron Wilson]

On Thu, Jun 5, 2014 at 6:38 PM, B Harder  wrote:

>
> On Jun 5, 2014 10:55 AM, "Richard Hipp"  wrote:
> >
> > If you included all the artifacts that you need, that would immediately
> make the subrepo larger than you might expect.  Certainly the subrepo would
> be smaller if you only included artifacts that actually changed, and even
> then only included deltas and not the complete text.
> >
> > But maybe there should be an open to the "fossil submit" or "fossil
> subrepo" command (whatever it ends up being called)
>
> A fossil "partial"?
>
Thinking about this, I am wondering what would happen if:
1. clone the main repo
2. Mark the root of the clone "private", making the whole clone private
3. Start a branch for the local changes and mark that branch not private
4. Commit changes to the new branch
5. Clone the clone

Would the second clone be just the new branch?
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] DRH's PGCon 2014 Keynote (with Fossil sighting!)

[Ron Wilson]

On Mon, Jun 2, 2014 at 12:48 PM, Stephan Beal  wrote:

>   On Mon, Jun 2, 2014 at 6:41 PM, Scott Robison 
> wrote:
>
>> DOH! Ah well, usually it was S-Q-L. :)
>>
>
> To be fair, though, he only did it (once) in the context of the MS product
> named SQL Server, and if i'm not mistaken that how MS says it.
>

FWIW, the IT people at my office call MS SQL Server "sequel sever" and the
Oracle server "Oracle SQL" (or just Oracle). (As I understand it, "sequel
server" is the backend for the Sharepoint server and "Oracle SQL" is the
backend for the "SAP Enterprise Document Management System".)
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] fossil clone on localhost

[Ron Wilson]

On Fri, May 30, 2014 at 8:30 AM, Stephan Beal  wrote:

>
> On Fri, May 30, 2014 at 2:28 PM, Michai Ramakers 
> wrote:
>
>> Could anyone give an example of when it makes sense to clone a repo on
>> localhost at all, assuming it ever makes sense..? (I am currently
>>
>
> i can only think of one (small) reason: to avoid having to know the path
> to the repo file. Functionally, it will be equivalent to cloning from a
> local copy of the repo.
>

Related to that, git and Mercurial make their repositories an integral part
of a workspace, so users coming from git and Mercurial (and maybe others)
might be more comfortable cloning repo+workspace rather than having
multiple workspaces for a single repo.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Autosync retry?

[Ron Wilson]

On Fri, May 30, 2014 at 5:55 AM, Martin Gagnon  wrote:


> Le 29 mai 2014 16:10, "Stephan Beal"  a écrit :
>
> >
> > Wasn't even aware of pull-only until earlier today.
>
>   snip
>
> Same for me, I always use autosync=1 together with the dont-push=1 setting
> for that. Look like an option got added by someone that didn't know about
> the other.
>
As I recall, the 2 options have slightly different affects. "pull-only"
only affects auto sync, while "dont-push" affects both manual and auto
sync. A manual push will, of course, still push.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Autosync retry?

[Ron Wilson]

On Thu, May 29, 2014 at 11:38 AM, Andy Bradford 
wrote:

> I agree that for network related failures, retry won't help. Others have
> reported non-network related failures (primarily due to locking or other
> similar problems).


Intermittent network failures can be a problem.So, when I'm not at the
office, I usually turn off auto sync and run a sync script that tries the
sync 5 times, wating a minute between tries. This usually works, but is
rather slow.

(Besides the office, my work involves "hands on" time in various labs and
manufacturing facilities that are electrically noisy. (The machinery is
connected to an ethernet that has no connection to even the company
intranet, let alone the internet))
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] encrypted repos?

[Ron Wilson]

On Thu, May 29, 2014 at 2:52 PM, Michai Ramakers 
wrote:

>
> I was thinking the resulting encrypted repo would change a lot, when
> only certain blocks in the unencrypted repo change. Would this not be
> so?
>

When encrypting the file, you should be doing the encryption in CBC mode.
In that case, any change will affect all the subsequent cipher text. In
fact, each time you encrypt, you should use a different Initialization
Vector (IV), which will result in the entire cipher text being different.

It is possible to encrypting using ECB mode, which will only affect the
cipher text around each change, but is extremely insecure. Counter Mode is
another option, but is still very insecure.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Server static files to load JSON APP

[Ron Wilson]

On Wed, May 28, 2014 at 2:23 PM, Ron Wilson  wrote:

> On Tue, May 27, 2014 at 5:00 PM, Petrica Clement Chiriac (Tica2) <
> petrica_chir...@fluxinternet.ro> wrote:
>
>> Hi fossil-users,
>>
>> I have simple JS app (build with GWT) and
>> these static files are in one folder   ./jsout
>>
>> You can get the "raw" content of any file stored in Fossil.
>

Also see the following for an additional way:

   http://fossil-scm.org/index.html/doc/tip/www/embeddeddoc.wik
i
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Server static files to load JSON APP

[Ron Wilson]

On Tue, May 27, 2014 at 5:00 PM, Petrica Clement Chiriac (Tica2) <
petrica_chir...@fluxinternet.ro> wrote:

> Hi fossil-users,
>
> I have simple JS app (build with GWT) and
> these static files are in one folder   ./jsout
>
> You can get the "raw" content of any file stored in Fossil. For example:

http://fossil-scm.org/index.html/raw/src/export.c?name=tip

In your case, the following should work:

   http://localhost:8080/index.html/raw/jsout/index.html?name=tip
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Running script on push

[Ron Wilson]

On Tue, May 27, 2014 at 8:34 AM, Abilio Marques wrote:

> I was looking for an easier path. My set of tests can be run from command
> line, and I really like shell scripting, so I was wondering, does fossil
> has a trigger mechanism (on commit / push) run this command?
>
> Do you see any utility on such an idea, or is just me?
>

Launching "hook" processes from Fossil has been discussed several times
over the last few years. The problem with doing so has to do with the
different ways this is done on the various platforms.

For the most part, the alternatives Fossil does support, RSS feed and
"pinging" a server via an URL, have served well enough that no one has
taken up the task of launching processes from Fossil.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Location of ".fossil" Database

[Ron Wilson]

On Wed, May 21, 2014 at 1:17 PM, Stephan Beal  wrote:

> On Wed, May 21, 2014 at 7:04 PM, Igor de Oliveira Couto <
> i...@semperuna.com> wrote:
>
>> That structure is *absolutely guaranteed* to exist, as it is necessary
>> for the way the System works. The way Apple structures their hard drive and
>> sub-divides the directories has changed very little since the first release
>> of OS X.
>>
>
> LOL! And $HOME has not changed since... 1973? ;)
>

In Microsoft land, %HOME% became %HOMEPATH% at some point, and the under
lying home directory location has changed a few times as well. And how
%APPDATA% is used.

I wish I didn't have deal with Microsoft, but at work, we have a mixed
environment with most people using Microsoft.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Tracking System Configuration Files - Best Practices

[Ron Wilson]

They told me they use the open source version of Tripwire (
http://sourceforge.net/projects/tripwire/).

They won't let me even see the scripts, but said they started with scripts
they found on the web. I tried looking, but only found general stuff.


On Thu, Apr 24, 2014 at 1:28 PM, Chad Perrin  wrote:

> On Thu, Apr 24, 2014 at 12:23:29AM -0400, Ron Wilson wrote:
> > On Wed, Apr 23, 2014 at 9:38 AM, Will Parsons  >wrote:
> >
> > > This kind of stuff isn't a "project", and you don't
> > > need the extra stuff that Fossil (or Git, Mercurial, Bazaar,
> > > Subversion, or CVS) provide.  I've tracked system files for over a
> > > decade with RCS (and before that with SCCS) and see no reason to
> > > change.
> > >
> >
> > I disagree.  Very often system changes have to be coordinate across
> several
> > config files. Most distributions have admin tools to take care of this,
> but
> > don't track the history of the changes. For most uses, this is fine.
> >
> > Where I work, the IT people already use a tool like Tripwire to monitor
> the
> > status (including ownership and permissions) of system critical
> directories
> > and files. Another part of this tool is used to reset the permissions,
> > ownership, etc of these files when changes are made. Because of this,
> they
> > can - and do - use Fossil to track system configuration changes.
> >
> > By using this combination of distribution provided tools, a few custom
> > tools, the Tripwire like tool and Fossil, they actually have more and
> > better control of configuration. And they save the company several $10k
> per
> > year in licensing fees for commercial system management suites.
>
> What are the chances you could produce a description of how this setup
> is used such that it would provide a howto for setting up similar
> systems in other people's networks, or convince someone else to write
> such a thing?  This could make for an excellent article somewhere.
>
> --
> Chad Perrin [ original content licensed OWL: http://owl.apotheon.org ]
> ___
> fossil-users mailing list
> fossil-users@lists.fossil-scm.org
> http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
>
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Tracking System Configuration Files - Best Practices

[Ron Wilson]

On Wed, Apr 23, 2014 at 9:38 AM, Will Parsons wrote:

> This kind of stuff isn't a "project", and you don't
> need the extra stuff that Fossil (or Git, Mercurial, Bazaar,
> Subversion, or CVS) provide.  I've tracked system files for over a
> decade with RCS (and before that with SCCS) and see no reason to
> change.
>

I disagree.  Very often system changes have to be coordinate across several
config files. Most distributions have admin tools to take care of this, but
don't track the history of the changes. For most uses, this is fine.

Where I work, the IT people already use a tool like Tripwire to monitor the
status (including ownership and permissions) of system critical directories
and files. Another part of this tool is used to reset the permissions,
ownership, etc of these files when changes are made. Because of this, they
can - and do - use Fossil to track system configuration changes.

By using this combination of distribution provided tools, a few custom
tools, the Tripwire like tool and Fossil, they actually have more and
better control of configuration. And they save the company several $10k per
year in licensing fees for commercial system management suites.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Copy User & Category Configuration

[Ron Wilson]

On Tue, Apr 22, 2014 at 6:57 AM, Paulus Tuerah  wrote:

> Hi,
>
> I have many repositories, and I want all the repositories to have same
> user & category capabilities.
> So I set the users & categories capabilities in one repository, and then
> export it to file, then import it to other repositories.
>
> Users and categories are imported. But there are 2 problems:
> 1. Sometimes not all users & categories are exported, sometimes category
> "Reader" is not exported, sometimes all categories are not exported (only
> normal users are exported). Is this bugs? I use fossil configuration export
> user filename --repository "path-to-repo.fossil"
>

This seems like a bug to me.


> 2. In other repositories, I can see the user, but I can not login with
> that user because of wrong password.
> I change the password in the web ui to be the same, and I can login
> normally.
> I try to export user in that other repository, and I see the password is
> different.
> The original repository: pw 'c0644e265cbb14d2c97921e7485293ff00abaa77'
> The imported configuration repository: pw
> 'a7592474d2116164ebfe0f6513c1824cd6e8511c'
> Even though the password is the same. I guess different repository has
> different password key.
> Is there a way to copy the password as well to other repository?
>
> Fossil has a way to designate one repository as the master for
authenticating user login. You still have to export user information to
each new repository and configure each to use the master for password
authentication, but once setup, all authentication for any of the
repositories is handled via the one.

(Unfortunately, I can't find the documentation for this feature.)
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] How to unset global settings in local settings without the global taking precedence?

[Ron Wilson]

On Fri, Apr 18, 2014 at 11:51 AM, Andy Goth  wrote:

> which is a no-op ("does nothing, successfully"):
>

Being in the world of "bare silicon",  this gave me a chuckle. "does
nothing, successfully" has a side effect, therefore actually does
something. Specifically, it declares success. For us, no-op is just "does
nothing", leaving the state the same as before. (Ok, technically, time has
passed and the instruction pointer register has incremented, so nothing
actually "does nothing".).
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] "default user" confusion

[Ron Wilson]

On Fri, Apr 11, 2014 at 4:12 PM, Stephan Beal  wrote:

>
> Just to have said it: sharing a repo over a network mount is fundamentally
> a bad idea, and you will get very little sympathy when things go wrong with
> it vis-a-vis a fossil db file.
>

One of the big advantages of Fossil is that it makes it somwhat easy to
avoid sharing a repo through a network share. I say "somewhat easy"
because, at least in my experience, the main reason for sharing a repos
through a network share is not being able to setup a server to host the
main repo. Fossil allows easy peer-to-peer sync between the team members
currently connected to the team's network (such as the office network in
the case of a business). Any member offline has to rely on other team
members to get caught up.

Currently, at my office, I have a second PC that is server as the de facto
main repo server for my team. But I only have that PC because I volunteered
to be the first (in the office) try Win7, 3 years ago, so IT let me keep
the XP PC I had. By the time all the tools and applications were updated
for Win7, the old PC was considered obsolete, so IT has no use for it. I
have since installed Debian Testing on it.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Partial checkouts, a possible methodology ...

[Ron Wilson]

On Fri, Apr 11, 2014 at 1:40 PM, Stephan Beal  wrote:

> Just coincidentally, i stumbled across this topical entry in the git
> random doc generator:
>
> NAME
> git-pause-subtreepause all unstaged subtrees from all logged remotes
>
Actually sounds useful.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Fossil CGI trick: extensionless CGIs under Apache

[Ron Wilson]

On Fri, Apr 11, 2014 at 1:30 PM, Eduardo Morras  wrote:

> Check this addon for firefox
> https://addons.mozilla.org/en-US/firefox/addon/user-agent-switcher/
>
> It allows change the user-agent and identify firefox as a IE.


Thanks. I tried that and several others. Didn't help. As I said, these are
"aggressive" about enforcing IE only. The stated reason being "even a
'standards compliant ' browser isn't compliant and we don't even have time
to test on all versions of IE the company demands we support."

Though, there might be hope. In the last several months, I've seen the CEO
and other C*Os (and other high level management) using iPads for work
related use.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Partial checkouts, a possible methodology ...

[Ron Wilson]

On Fri, Apr 11, 2014 at 12:11 PM, Matt Welland  wrote:

> Partial checkouts is something I find myself desiring to do surprising
> often.
>

The only times I used partial check outs in any VCS were for clients whose
teams just kept adding new projects in to the same repository. Partially
this was because they were used to files based systems, so cherry-picked
check outs were the norm. Also, adding a new repository required involving
the IT people.


>
> Would a feature like this be of use to others? Implementing this in the
> existing fossil code base might be both disruptive and difficult and is
> certainly beyond my personal C skills. In spite of this I thought the idea
> worthy of discussion and I'm curious to hear what others think. The
> alternative, nested fossils, is a hassle and unless fossil can propagate
> commands to sub fossils it is too weakly coupled to work for my scenarios.
>

I'm sure that for teams who inherited monster repos, this feature would be
a huge blessing. While I currently am not in that situation, I would not be
surprised to land in it, again, in the future.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Partial checkouts, a possible methodology ...

[Ron Wilson]

On Fri, Apr 11, 2014 at 12:26 PM, Stephan Beal wrote:

> On Fri, Apr 11, 2014 at 6:19 PM, Stephan Beal wrote:
>
>> i was thinking that nested repos ("external repos" in SVN jargon) would
>> serve the same niche and possibly be less intrusive to existing code. But
>> that's just my spontaneous - the effort might not be as invasive as i
>> imagine it being.
>>
>
> What happens if...
>
> - dir xyz is a subrepo
> - fossil checkin xyz/def/* abc/*
>
> we're doing a partial commit against a subrepo which includes files
> outside that subrepo.
>
> What's the correct behaviour?
>

First perform the operation on the contents of the "host" check out,
queuing sub check outs as they are encountered. Then perform the operation
on each sub check out in turn. This would basically be a "width first"
traversal of the nested check outs.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Fossil CGI trick: extensionless CGIs under Apache

[Ron Wilson]

On Fri, Apr 11, 2014 at 12:44 PM, Stephan Beal wrote:

> In hindsight, i should not have expected it to fail (and not sure why i
> did. Browser cache? Dunno.), because /index.php/ is also valid.
>

Still fails under IE, but I only tried it as an additional data point.
(Normally I only use IE to access certain company internal websites that
"aggressively" refuse to work with anything other than IE (any version of
IE).)
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Fossil CGI trick: extensionless CGIs under Apache

[Ron Wilson]

On Fri, Apr 11, 2014 at 9:22 AM, Richard Hipp  wrote:

> On Fri, Apr 11, 2014 at 9:13 AM, Stephan Beal wrote:
>
>>
>> This works:
>> http://fossil.wanderinghorse.net/repos/test/blah
>>
>> but this does not:
>> http://fossil.wanderinghorse.net/repos/test/blah/
>>
>> (notice the trailing slash)
>>
>>
>
> Both links work for me, using FF and Chrome.  Did you change something on
> the server since this email was posted?
>

Which link did you try first? when I tried FF, I accidently tried the
working link first. Though even after I removed
fossil.wanderinghorse.netfrom history, cleared cache and restarted FF,
the "bad" link still worked.

I'm at work, so tried it with IE. It returned an empty page.

When I get home, I will try it with FF (as a different user) and wget.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] a bit of healthy fossil/git rivalry...

[Ron Wilson]

Quite funny. Thanks for sharing.


On Fri, Apr 11, 2014 at 5:02 AM, Stephan Beal  wrote:

> Maybe fossil needs such a tool...
>
> http://git-man-page-generator.lokaltog.net/
>
> Be sure to tap F5 a few times and re-read the page each time.
>
> --
> - stephan beal
> http://wanderinghorse.net/home/stephan/
> http://gplus.to/sgbeal
> "Freedom is sloppy. But since tyranny's the only guaranteed byproduct of
> those who insist on a perfect world, freedom will have to do." -- Bigby Wolf
>
> ___
> fossil-users mailing list
> fossil-users@lists.fossil-scm.org
> http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
>
>
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] fossil sync/pull/push --once broken as of [64aa75260f]

[Ron Wilson]

On Thu, Apr 10, 2014 at 10:19 AM, Andy Bradford wrote:

>
> I think I may have brought it up once or twice with respect to SSH. When
> using it with SSH and a shared account, for example, it is almost always
> wrong to  choose the  remote user  as the local  fossil username.  It is
> almost always correct to choose the $USER.
> ...
> Yes, I realize it's a messy situation now not one with a simple solution
> that I can  see, that's why I  asked on the list. In  addition to adding
> two users  on cloning, we  would also  have to mark  one of them  as the
> ``defaultuser'' so Fossil  automatically picks that one.  Again, I'm not
> suggesting this is the correct way to deal with it, I'm just looking for
> ideas.
>

I have ideas, but also questions. (I haven't had to deal with your
situation,
but I can see where I likely will in the future.)

For a shared SSH account, I assume, on the server, Fossil is set up to use
REMOTE_USER, so "divorcing" the Fossil user ID from the log in ID
should not be an issue.

Q1. When cloning from a server accessed by a shared SSH account, what
is the most reasonable way to tell the local Fossil to ignore the user name
in
the URL?

Q2. When ignoring the name in the URL, what user ID should Fossil use?

Q3. What happens if my local Fossil user ID is not the same as my Fossil
user ID on the server? (Example, my personal user ID (and local Fossil user
ID) is "rxw", but my remote Fossil user ID is "ronw" (in REMOTE_USER).)

I will probably have more questions, but those are a start.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Very short UUID abbreviations

[Ron Wilson]

On Mon, Apr 7, 2014 at 11:43 PM, Andy Goth  wrote:

> On 4/7/2014 8:59 PM, Andy Bradford wrote:
>
>> To be fair, there is still one type of UUID that will work at a
>> shorter length than 4; events currently do not go through collision
>> detection, so it's possible to view an event with a length shorter
>> than 4:
>>
>> http://fossil.bradfords.org:8080/info/a
>>
>> Should I also add event collision handling to this branch?
>>
>
> I do prefer consistency, but we're going to need to think more about
> whether we want to break something which used to work just for the sake
> of consistency.


Bypassing collision handling is not good. It results in information being
hidden. As such, adding it would not be "just for the sake of consistency."
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Authentication problems

[Ron Wilson]

On Sun, Apr 6, 2014 at 8:47 AM, Kai Lauterbach  wrote:

>
>
> Am 06.04.2014 13:12, schrieb Stephan Beal:
> > On Sun, Apr 6, 2014 at 1:09 PM, Kai Lauterbach  wrote:
> >
> >> fossil clone "http://localhost:8080"; repo2.fossil
>
> I have done this on my local windows machine.
>
> > That _might_ be the problem: IIRC server mode operates slightly
> differently
> > vis-a-vis permissions when listening on localhost (i might be wrong - a
> > long time ago i had some confusion related to which IP i accessed a local
> > server instance over). Try using the local IP (or a host alias which
> > resolves to it) of your pi instead.
> >
>
> The hostname "pi" is always resolved to 192.168.128.37 and "localhost"
> on my windows machine resolves to 192.168.128.29, so at this point there
> is no naming confusion.
>

Unless you manually overrode your configuration, "localhost" always
resolves to 127.0.0.1, which is always self referential. that is, if you
"fossil clone http://127.0.0.1:8080 repo.fossil" on your desk PC, it will
always try to connect to a Fossil server running on your desk PC. If you do
it on your Pi, it will always try to connect to a Fossil server running on
your Pi.

Anyway, Fossil does grant full privileges to a client connecting from
127.0.0.1 (aka "localhost") unless we turn off the "localauth" setting.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] TH1: support for octal and hexadecimal numbers in expressions

[Ron Wilson]

On Thu, Apr 3, 2014 at 10:34 AM, Jan Nijtmans wrote:

> See: 
>
> > Newer languages have been abandoning the prefix 0, as decimal numbers
> > are often represented with leading zeroes. The prefix q was introduced to
> > avoid the prefix o being mistaken for a zero, while the prefix 0o was
> introduced
> > to avoid starting a numerical literal with an alphabetic character (like
> o or q),
> > since these might cause the literal to be confused with a variable name.
> The
> > prefix 0o also follows the model set by the prefix 0x used for
> hexadecimal
> > literals in the C language; it is supported by Haskell,[10] OCaml,[11]
> Perl 6,
> > [12] Python as of version 3.0,[13] Ruby,[14] Tcl as of version 9,[15]
> and it is
> > intended to be supported by ECMAScript 6[16] (the prefix 0 has been
> > discouraged in ECMAScript 3 and dropped in ECMAScript 5[17]).


Interesting. Never heard of this trend, but 90% of my coding for work is in
C/C++ as I develop SW for controlling electro-mechanical systems.

I use 0x (for hexadecimal) and 0b (for binary), but almost never use any
octal notation, as nearly all the integer values I work with are 8, 16 or
32 bits in length. (Occasionally, 24 bits)
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] New Fossil GUI: Requesting Feature

[Ron Wilson]

On Sun, Mar 30, 2014 at 12:41 PM, Stephan Beal wrote:

> Because you've already made the commit and finfo with a filename works on
> the current checkout version. That file isn't in your checkout, so it's
> unknown. That's not very intuitive, but i don't currently see how fossil
> could reasonably behave otherwise.
>

Perhaps a better error message, like say "file_a.ext does not exit in
current working space".

But then, why not look up the file in the repository even if it's not in
the current working space?


> History and full rename info is retained. i think you're just confused by
> the finfo behaviour described above (which is arguable but would introduce
> ambiguities and irregularities in results if it behaved differently).
>

Renamed files have been a problem in every VCS I've ever used, even hugely
expensive commercial systems. Moved files even more so. Unfortunately, the
only people I know who might be able to come up with a solution work for
companies who consider projects like Fossil or Git as competitors, so their
employment contracts prevent them from contributing.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] New Fossil GUI: Requesting Feature

[Ron Wilson]

On Sun, Mar 30, 2014 at 1:20 PM, Paulus Tuerah  wrote:

> I see, in the web ui the rename is working as expected.
> But how can I get that information from the command line parameter?
> fossil timeline only show the deleted file, not mention that it is renamed
> not deleted.
> I want to show the log all from my GUI, not using the fossil web ui at all.
> My plan is the web ui is only for changing settings, wiki, ticket, bug
> tracker, for the version control support everything is from my GUI.
>

Fossil also has a JSON API, which might be easier to use for information
requests from the Fossil repository. Information about your working space
would still need to come either from the Fossil command line interface or
from libfossil. But, you would need to research how to call a C library
from Java.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] command gpg not found - huh?

[Ron Wilson]

On Fri, Mar 28, 2014 at 9:16 AM, Stephan Beal  wrote:

> On Fri, Mar 28, 2014 at 2:04 PM, Michael Weise 
> wrote:
>
>> Can you give a hint (link) or background infomation on WHY the shell
>>  emits the gpg warning? What causes it?
>>
>
> because fossil internally uses system() to call gpg, but fossil itself
> does not know how to resolve the $PATH, so it does not check if gpg is
> available first.
>
...

> Reminder to self: once we prove that fsl_pathfinder works on Windows it
> might be interesting to port that in to fossil so that it can resolve the
> $PATH itself and skip signing if gpg isn't found.
>

No need to do that. The system functions execlp, execvp, etc. (_spawnlp,
_spawnvp, etc on MS Windows) already search the PATH environment variable
for the named executable. It is generally safer than using system() and is
more efficient. The downside is that you have supply the arguments
individually because you won't have the shell to split a command string
into arguments. You also won't have access to shell meta character
expansion nor other command line processing services.

FYI, for those who are curious, MS Windows also has the _exec family of
system functions, but lacks the fork() system function. To create a
subprocess under MS Windows, you use _spawn instead. Under Linux and other
Unix-like systems, you use fork and exec. The call to fork clones the
current process, returning the Process ID to the original ("parent")
process and 0 to the new ("child") process. the child process can then
either continue running as a clone of the parent or it can call exec to
"replace" the running program with another. Once started, the new loaded
program runs in the same process that called exec. Only failed calls to
exec ever return, in which case the variable errno will contain the ID of
the error that caused exec to fail.

Under MS Windows, _spawn works as if the child process immediately calls
exec, except no cloning of the original process is done. _spawn is called
exactly the same way _exec is called.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Please improve documentation for the "checkout" command

[Ron Wilson]

On Thu, Mar 20, 2014 at 1:16 PM, JR  wrote:

> However, I also merge to trunk fairly frequently so that I usually only
> have to fossil update trunk on a different computer
>

FYI, Fossil, unlike Git (and some others), always syncs the entire
repository (except for  private branches), so all of your (non-private)
branches will be up to date on all your computers.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Please improve documentation for the "checkout" command

[Ron Wilson]

On Thu, Mar 20, 2014 at 11:44 AM, Matt Welland  wrote:

> A comment to those who find they seldom branch: I use branching and
> merging extensively as a tool to organize work and enable flights of fancy
> and exploration. Obviously it depends highly on the problems you are
> solving but I suspect that if you aren't branching then either you are
> working on a relatively linear and straightforward problem or you aren't
> using fossil to its fullest.
>

When I first started using "feature" branches, it took me a while to get
used to "reseting" my working space to current baseline after finishing and
committing something. After a while, it became automatic. It is a very
useful technique, even for "solo" projects.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] G+ Fossil page?

[Ron Wilson]

On Thu, Mar 20, 2014 at 4:11 AM, Martijn Coppoolse <
li...@martijn.coppoolse.com> wrote:

> On 19-3-2014 18:28, Ron Wilson wrote:
>
>>   There is also freecode.com <http://freecode.com>. Fossil has a listing
>> there, but it is out of date.
>> (http://freecode.com/projects/fossil)
>>
>
> Does anybody know who the maintainer is?
>
> Freecode seems to have an API, and so do Google+ and Facebook. It should
> be feasible to write a script that parses the downloads page for new
> versions, and publish the new version (including "what's new") to those
> pages, no?


Maybe look for updates to download.html in the Fossil RSS feed.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] G+ Fossil page?

[Ron Wilson]

On Wed, Mar 19, 2014 at 12:59 PM, Stephan Beal wrote:

> i just came across this G+ post:
>
> https://plus.google.com/100252397521353238214/posts/hW7ErV11uNB
>
> (For non-googlers; it's a release announcement for Fossil 1.28)
>
> and thought, "why haven't we been doing that all along?" It simply never
> crossed my mind.
>
> Does fossil need a G+ page (or FB page, or AOL page, for that matter)
> where that type of thing can be posted? i, for one, wouldn't mind manning a
> G+ community.
>
> (Or is there perhaps one i've missed so far? A cursory search revealed
> none.)
>
> There is also freecode.com. Fossil has a listing there, but it is out of
date.
(http://freecode.com/projects/fossil)
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] ticket editing

[Ron Wilson]

On Mon, Mar 17, 2014 at 3:37 PM, Stephan Beal  wrote:

> To give you an idea of what the JSON is possible of, here are a couple
> demo apps for the JSON API:
>
> One of the first demos, a (very) primordial wiki editor:
>
> http://fossil.wanderinghorse.net/repos/fossil-sgb/json/wiki-editor.html
>
> (It can save pages if you have an account on that test repo. Try saving
> without them!)
>
> Some more demos:
>
> http://fossil.wanderinghorse.net/repos/fwiki/editor/
>
> And i actually host a few documentation-only wikis using Fossil+JSON as
> the back-end to store the wiki pages, but i store them in Google Code wiki
> format and do all the rendering client-side with JavaScript:
>
> http://fossil.wanderinghorse.net/wikis/cson/
> http://fossil.wanderinghorse.net/wikis/cpdo/
> http://fossil.wanderinghorse.net/wikis/whio/
>
> Cool
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] fossil humor: poor man's bug tracking...

[Ron Wilson]

On Mon, Mar 17, 2014 at 12:52 PM, j. van den hoff  wrote:

> nice. but unfortunately not quite what we need here (it'd be somewhat
> awkward to assign a jpeg image as the vaulue to the tag, e.g. ;-)).
>
> what do other people think: is someone else missing the ability to
> edit/modify ticket comments retrospectively?
>

In another thread, I mentioned the possibility of using Javascript to
"replace" the ticket comment field with a Fossil wiki page. the result
would likely be kind of ugly, but would still have all the features
provided by Fossil tickets.

Another possibility, if one is willing/able to give up all the
ticket-specific features, would be to simply use wiki pages as the tickets.

(Honestly, the biggest obstacle to using any ticket system where I work was
the project managers *love* spreadsheets. Fortunately, "fossil ticket show"
produces a tab separated listing based on the specified report. Except
during meetings, they update tickets through the web interface. During
meetings, they still just make the changes in the spreadsheet, I have to
export the updated spreadsheet to a CSV file and use a Perl program to
generate "fossil ticket change" commands.)
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] ticket editing

[Ron Wilson]

On Mon, Mar 17, 2014 at 5:53 AM, j. van den hoff
wrote:

> I would hope that making this (retrospective modification/correction of
> ticket comments) is not really impossible in the current setup and that one
> of the core developers would look into this if his time permits: it is
> really a nuisance to look at a seriously misleading/erroneous/misplaced
> ticket comment (the more so, if it happens not to be the very last one) and
> not being able to correct it. "append only" really is not optimal. I do not
> of course understand sufficiently why the present setup was chosen but
> superficially I would be more happy with a logic similar to the wiki pages:
> just allow editing of a single comment field and track all edits in the
> timeline (I understand from your mail that this was the previous setup
> here, right?).


Using Javascript, you could use a (Fossil) wiki page in place of the ticket
comment field. Would probably have to use the JSON features that Stephan
Beal added to Fossil a while ago. Still, it likely would not be as clean as
the way ticket comments used to work.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] copying a branch and history into a new repository

[Ron Wilson]

On Wed, Mar 12, 2014 at 1:13 PM, JR  wrote:

> I host my fossil repository on a private server.  As a little background,
> I do mostly scripting work and keep all my "projects" in separate branches.
>  This has worked great for me and I will continue to use this method.
>  However, I am almost finished with a project I would like to host on
> chiselapp so the greater Internet can reach it.
> 
> I have run , but the resulting file for my repository is
> too large for me to realistically manually edit down to just the branch I
> want.  If anybody has some scripting-fu for that, that would be awesome.
>

In theory, a script could use SQL to identify the artifacts for a specified
branch. then create a shun list of the artifacts for the other branches.

Alternately, SQL could be used to load those artifacts in a new repository.

Another possible option would be to use "fossil deconstruct" to extracts
the artifacts into files, then copy the desired artifacts to another
directory and use "fossil reconstruct".

I suspect there are some other artifacts that will also need to be copied
in order to create a valid repository.

Caveat: For all of these options, each repository created this way will
have the same repository ID as the original, so it would be possible to
accidentally push undesired content in to any of them.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] importing/forking from Git

[Ron Wilson]

On Wed, Mar 12, 2014 at 5:04 PM, Chad Perrin  wrote:

> I'm stuck in kind of a no-win situation with this, then.  It's not
> really appropriate to log a bug in an issue tracker without knowing it's
> a bug, but nobody will tell me if it's a bug, how I should judge whether
> it's a bug, or whatever else applies here.  I'm not even sure yet that
> what I believe to be expected behavior is expected behavior, or whether
> I'm just using it wrong.
>

You could put in an enhancement request.

What you want to do, import in to a repository with only the initial, empty
commit, seems reasonable.


> . . . so while I wait to see if someone will give me some guidance, it
> will fall off everyone's radar, and nothing will be done, so that my
> only option going forward for some of these projects will be to just use
> Git.
>

 If the only possible way to start a new repository on the server is
"fossil new repoName.fossil", then a possible work around would be to do a
"normal" import to Fossil, then create a "marks" file (see "fossil help
export") with the RID of the initial, empty commit, then use the
"--import-marks" option to the "fossil export" command to (hopefully)
create something usable with "fossil import --incremental".

Not sure when I will be able to experiemnt with this more.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] libfossil milestone: first checkin

[Ron Wilson]

Congrats.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] option "--abs-path" on fossil extras/changes doesn't work

[Ron Wilson]

On Sun, Mar 9, 2014 at 3:00 AM,  wrote:

> I erroneously interpreted the CLI-help ("Display absolute pathnames") as:
> "show the complete path of the changed file as recorded by the Filesystem".
>
> I would have made the same assumption. this behavior is not intuitive.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] importing/forking from Git

[Ron Wilson]

On Sat, Mar 8, 2014 at 5:17 PM, Chad Perrin  wrote:

>
> Should I take it at this point that --incremental is deprecated and
> should not be used or expected to be present in future Fossil versions?
> So far, I have seen no responses that specifically addressed whether the
> behavior I expected from it was the intended behavior, whether anyone
> else might have the same problem indicating a bug in the software or
> whether this was confined to my situation and thus might be a local
> configuration issue, whether there was anything I could do to provide
> more information to help, and so on.  The closest thing to meaningfully
> addressing the issue that I have seen was akin to the old joke . . .
>
> patient: Doctor, doctor, it hurts when I do this!
> doctor: Don't do that.
>

I very much doubt that --incremental would be deprecated. I suspect that no
one else has tried your use case.

I still haven't been able to experiment more, but it occurs to me that
"fossil import --incremental" would assume that its input resulted from
"git --fast-export --incremental". Further, I would expect the
--incremental option to git to require that an initial non-incremental
export had already been done.

>From reading about Fossil export, it seems there might be a way to fake an
initial non-incremental export from Fossil. If that is so (or if this can
be done directly in git), then I would expect "fossil import --incremental"
to work as you want.

Of course, it would be useful for Fossil to support non-incremental import
to an existing repository, at least one with only the initial, empty commit.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] stashing of added files

[Ron Wilson]

On Sat, Mar 8, 2014 at 2:36 PM,  wrote:

> Hi,
>
> I'm new to fossil and not quite shure if this is a bug or a lack of
> understanding on my part.
>

Welcome to Fossil.


> I wanted to stash a newly added file, then apply the stash to another
> (feature-)branch.
> ...
> Fossil did record the addition of file "test" to the stash, it even said
> it applied it on both occasions (stash apply, stash goto).
> So why did the file remain "unmanaged"?
>

After you add files, you have to commit them.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] importing/forking from Git

[Ron Wilson]

On Wed, Mar 5, 2014 at 2:21 AM, Chad Perrin  wrote:

> I'm talking about a distributed team, where not everyone has SSH shell
> access or the ability to directly change anything on the filesystem, but
> they all have commit access to a specific set of repositories.  This
> means, for instance, that they could clone from, and sync with, an
> existing repository -- but not all could create or delete entire
> repositories.
>

Presumably the server admin (you?)  would be the one creating new
repositories on the server, so would be able to use "fossil clone" or
"fossil import" either via SSH or directly on the server.

I tried experimenting with the --incremental option. It appears to work for
keeping a Fossil repository up to date after the initial import. I don't
know if that is an intended limitation or a bug. Right now, I don't have
more time to experiment with it as it is not a feature I use.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] importing/forking from Git

[Ron Wilson]

On Tue, Mar 4, 2014 at 7:24 PM, Chad Perrin  wrote:

> Both of those solutions would surely work in this case (as long as the
> incremental option works on the server, but not in cases where the same
> would be done by someone with commit access but no SSH filesystem/shell
> access (which is likely in future cases).


If you mean using a hosting service, then I would think the service either
provides SSH access or a way transfer existing project repositories to the
service's hosts, possibly providing a means to clone from your local Fossil
repository to the server.

Indeed, making the server side repository a clone of your local repository
is probably better. Then there will be no question that you can push
further imports to the server repository.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] importing/forking from Git

[Ron Wilson]

On Tue, Mar 4, 2014 at 7:12 PM, Ron Wilson  wrote:

> Alternately, try:
>
>$ git fast-export --all |ssh user@server fossil import\
>   --git --incremental /path/to/projectname.fossil
>

Oops. Don't use the --incremental for the initial import.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] importing/forking from Git

[Ron Wilson]

On Tue, Mar 4, 2014 at 3:43 PM, Chad Perrin  wrote:

> Trying to fork/import from Git is kinda problematic.  I started by
> creating a new project on a server:
>
> $ fossil init projectname.fossil
>
> I then cloned locally:
>
> $ fossil clone  projectname.fossil
>

Is there some reason you can't do "fossil clone" on the server?

Alternately, try:

   $ git fast-export --all |ssh user@server fossil import\
  --git --incremental /path/to/projectname.fossil

I imported from Git:
>
> $ git fast-export --all | fossil import --git --incremental \
> /path/to/projectname.fossil
>
> I opened the repository:
>
> $ mkdir projectname; cd projectname
> fossil open /path/to/projectname.fossil
>
> No files appear.  I looked at the timeline:
>

That is surprising to me. Though I have not used this feature.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Could --ignore and ignore-glob be made non exclusive ?

[Ron Wilson]

On Fri, Feb 28, 2014 at 6:04 AM, Samuel Debionne <
samuel.debio...@ujf-grenoble.fr> wrote:

> > i can't say i've ever needed it. i always tell the SCM _exactly_ which
> > files it should add/commit. i rarely even rely on "fossil commit" with
> > no file arguments, simply out of principal (i don't like software trying
> > to be too clever on my behalf, and i also don't want to accidentally
> > commit extra files).
>
> Me too ! But priori to any "fossil add xyz", and because I don't trust
> my memory to decide what should be added, I usually do a "fossil extra".
> And that's where the ignore pattern is handy


Agreed. Mostly because there hasn't been a reason to make more than minor
tweaks to our build process. Until then, we live with the compiler leaving
the files it generates mixed in with the source files. True, there are
risks to using ignore, but our process quickly catches any missing files.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Emailing graphical diffs

[Ron Wilson]

On Thu, Feb 27, 2014 at 11:49 AM, Andreas Kupries
wrote:

> And here I thought (purely from the subject) that the diff would be
> pure HTML attached  to a mail, or *the* mail itself ;)
>

I haven't tried it, but the diff you get through the web UI looks likes,
other than the CSS, it might be self-contained (ignoring the links to other
pages, of course). If that's the case, then generating a self-contained
HTML document shouldn't be hard.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] libfossil script binding docs

[Ron Wilson]

On Wed, Feb 26, 2014 at 12:25 PM, Stephan Beal wrote:

> It's basically a JS-like data model (derived initially, in fact, from the
> same JSON code which runs fossil's JSON API), but provides a facility for
> clients to plug their own types into it (in a type-safe way - you don't
> have to blindly guess the type of the underlying void pointer!). libfossil
> uses that to bind, e.g., the Db and Db.Stmt classes.
>
> Quite interesting. Thanks. For a time, I was in to "systems languages",
but found my way back to my roots in electronics and "bare silicon"
software. Software that interacts more directly with the devices that
interact with the physical world. The example I mentioned was for a project
that needed an unusual level of reconfigurability. I headed off a scheme
for implementing our own shared library mechanism by offering to embed a
simple, low overhead scripting engine. though the language chosen was not
my first choice (for implementing the embedding), it still worked out.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] libfossil script binding docs

[Ron Wilson]

On Wed, Feb 26, 2014 at 6:00 AM, Stephan Beal  wrote:

> On Wed, Feb 26, 2014 at 6:29 AM, Doug Franklin  > wrote:
>
>> So, I'm kinda late to the party, since I only started using Fossil a year
>> ago, or so.  But, I'd like to throw out a suggestion: something like
>> doxygen or javadoc so that the documentation and source are the same
>> artifact.  Knuth tried with tangle/weave, but maybe it's time to revisit
>> the concept?
>>
>
> That doesn't quite apply in this particular case. The C-level library APIs
> are indeed doxygen format directly in the header files, but the scripting
> engine has no native doc format (and there are no plans to
>
> I haven't had a chance to look at your scripting engine, but I would have
thought that the scripting bindings would have been implemented with
"primitives" written in C that map the scripting engine's data model to C's
data model.

Just curious..


PS: An example of what I mean: Years ago, I wrote bindings for a scripting
engine to use a library that had a C API. I wrote a set of primitives like
the following:

int x_Send(Context *cx)
{
int port;
int len;
char buf[1024];

port = e_GetInt(cx, 0); // get int parameter at index 0
len = e_GetString(cx, 1, sizeof(buf), buf); // get string at 1

if (mSend(port, len, buf) == 0)
{
e_ReturnInt(0); // set an int return value
return(e_OK);
}
strcpy(buf, "Send failed."); // error description - should provide
useful details
e_SetError(strlen(buf), buf);
return(e_ERR);
}
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] nice example of a fossil time warp

[Ron Wilson]

On Thu, Feb 20, 2014 at 6:57 AM, Stephan Beal  wrote:

> until last night i had never actually experienced a Time Warp in a fossil
> repo:
>
>
> http://fossil.wanderinghorse.net/repos/libfossil/index.cgi/timeline?c=2014-02-19+20:13:42
>
> Great to see that Fossil handles it correctly.

Show the this view to one of the project coordinators here. The look of
horror on his face was priceless. When I explained what happened, he
replied "I'm so glad we don't have time skew around here."

Thanks for sharing this.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Painful team interaction with fossil, how to improve?

[Ron Wilson]

On Thu, Feb 20, 2014 at 6:56 AM, Roy Marples  wrote:

> Here is a patch which adds a RSS feed icon after the ticket uuid.
> Probably not the best way of doing it, but I couldn't think of another way
> which would allow the user to use a different icon to match their skin.
>
> Maybe just "RSS Feed" as a menu item would be better?
>

>From my experience, the RSS icon is a very common way to provide a way to
subscribe to a feed.

Interesting that you need to make an actual source code change. I would
have expected it could be changed in the header.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] libfossil www interface: looking for ideas

[Ron Wilson]

On Wed, Feb 19, 2014 at 4:06 PM, David Given  wrote:

> On 2/17/14, 10:26 PM, Stephan Beal wrote:
> [...]
> > Can you recommend one in C? i'm self-imposed bound to C here. i "might"
> be
> > convinced to write it in C++.
>
> Unfortunately, no --- it's not something I'd really want to do in C, as
> C's strengths do not really match the kind of wholesale string shuffling
> you need for a servlet engine. (Although using sqlite *just for string
> manipulation* from C is surprisingly comfortable.)
>

Years ago, I used an "AWK lib" to code complex string handling in C.
Unfortunately, I lost the sources for those projects and have had no luck
finding the library, again.

While I have seem some CGIs written in C, I've never seen a web app frame
work for C.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] looking for interesting new fossil skins

[Ron Wilson]

On Fri, Feb 14, 2014 at 10:43 AM, Samuel Debionne <
samuel.debio...@ujf-grenoble.fr> wrote:

> Something I noticed while hacking the wikiformat code is the use of  id='foo'> which is not valid HTML AFAIK. Closing tags should
> not have attributes right ?
>

Technically, closing tags do not have attributes, but HTML parsers are
supposed to tolerate and ignore them. Practically, certain attributes, like
id, can increase readability when placed in closing tags. Much like putting
comments after }, #else and #endif to help identify the  the corresponding
{ or #if.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Painful team interaction with fossil, how to improve?

[Ron Wilson]

On Thu, Feb 13, 2014 at 4:06 PM, Mark Janssen  wrote:

> Some things that come to mind:
>
> * Ticket notification for admin, logger and followers (this is the big
> one).
> * Email verification to prevent spam.
>

I can think of ways of using a list server to mostly handle these 2 items.


> * Ability to edit your own comments.
>

It might be possible to do this by modifying the "edit ticket page".
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Painful team interaction with fossil, how to improve?

[Ron Wilson]

On Thu, Feb 13, 2014 at 2:46 PM, Mark Janssen  wrote:

> True with effort all of this could be retrieved from the underlying sqlite
> db. The reason I didn't do that is that you need opt out and email
> verification to prevent sending spam. I could add your email on every
> ticket and you would not be able to stop the notifications without admin
> interaction.
>
I had not considered that because, at work, everyone using Fossil is an
employee of the company. For my personal projects,  I don't have to worry
about this. For the OSS projects I contribute to, I either send patch files
or use github. (Mostly I send patch files.)

> Eventually you are really just rebuilding something like redmine. I really
> wanted to make it work, but in the end it just lacks too many features at
> the moment.
>
 I am curious what features are missing.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Painful team interaction with fossil, how to improve?

[Ron Wilson]

On Thu, Feb 13, 2014 at 1:13 PM, Mark Janssen  wrote:

> I have used the ticket hooks and they do work. However they will not allow
> automatic updates to be sent to whoever logged the ticket.
> As a result I have moved the bug tracker to redmine.
>
My impression, from reading comments on this email list, is that the hook
sends the ticket UUID in an HTTP request, then the recipient fetches the
details and forwards some of the details by whatever means.

Assuming that the recipient of the ticket-hook notification is allowed
access to the private_contact field, it could forward the notification to
the ticket originator. Likewise, access to any assigned-to and/or
subscribers fields would also be needed.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Painful team interaction with fossil, how to improve?

[Ron Wilson]

On Thu, Feb 13, 2014 at 10:24 AM, Baptiste Daroussin <
baptiste.darous...@gmail.com> wrote:

> In particular to interact with the submitter. I have found no way to:
> 1/ send a mail when new ticket is created (yes we can follow via rss feed,
> but a mail is way nicer)
> 2/ send a mail each time to ticket is modified to reporters and followers
>

A ticket hook was added recently. I have not used it, so I don't know how
it it works.


> The lacks of flexible hooks on server side is also a big problem to me
> What I do need on the hook side is:
> 1/ send a mail after each "push" of code with a diff:
> prior-push-tip/post-push-tip for each branch impacted to a given mailing
> list
>

Why not just links to the commits? For example,
http://fossil-scm.org/index.html/info/e327614047 appears to contain the
information you are asking for. This is a lot to put in to an email message.


> 2/ be able to get access to the diff and run random code on it and reject
> the push if not validated
>

Others on this list have set up automatic build servers. I don't think it
would be hard to add validation functionality.

May I ask what version control systems you have used before now?
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Scalability limits

[Ron Wilson]

On Fri, Feb 7, 2014 at 11:17 AM, Stephan Beal  wrote:

>
> On Fri, Feb 7, 2014 at 4:33 PM, Rich Neswold wrote:
>
>> I don't have any question; I just thought I'd document my experiences.
>>
>
> Thanks for your feedback! IMO (possibly a minority opinion), Fossil has
> never aspired to host repos quite as large as those. i remember the pkgsrc
> repo being mentioned before (but thought it was bigger than 2.7GB), and
> IIRC the delta manifest format was introduced to help support huge repos
> like that one and the core TCL repo. Fossil's original purpose was to host
> sqlite, and it works wonders for projects at that scale.
>

I am guessing this is a limitation of SQLite, which is designed to be
"light". It would be interesting to see how Fossil would perform when
"plugged in" to, for example, PostgreSQL, MariaSQL or other "heavy duty"
SQL server. Of course, that could require rewriting a lot of SQL queries.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] https-over-proxy

[Ron Wilson]

On Tue, Feb 4, 2014 at 12:20 PM, Jan Danielsson
wrote:

>So, the question is: Merge to trunk, or let it bitrot? I get the
> feeling that http proxies for outgoing connections is going away, making
> this feature slightly esoteric. Am I wrong? Is anyone using it?
>

Many of the out-of-office locations I occasionally go to transparently
route out going HTTP/HTTPS traffic through a proxy. My usual work-around is
to use an SSH tunnel to a friend's server, which also runs a private
HTTP/HTTPS proxy server. directly supporting HTTPS proxies in Fossil would
save steps and improve sync performance by reducing lag and increasing
through put.

thanks for your contributions to Fossil.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Command line option for HTTP Auth?

[Ron Wilson]

On Sat, Feb 1, 2014 at 1:29 PM, Andy Bradford wrote:

> Thus said "Andy Bradford" on 01 Feb 2014 10:04:23 -0700:
>
> > Thus said Kevin Martin on Sat, 01 Feb 2014 09:14:20 +:
> >
> > > How  does  fossil authenticate  with  a  server,  does it  send  the
> > > password plaintext? HTTP Basic Auth does!
> >
> > It's not encrypted, no, only base64 encoded.
>
> It seems I misunderstood what Mr.  Martin was asking here. Fossil uses a
> nonce and the password and sends only the SHA1 hash across the wire:
>
> http://www.fossil-scm.org/index.html/doc/trunk/www/password.wiki
>

There are some caveats mentioned in the documentAndy linked to, that you
should pay attention to:

1. When a user logs into Fossil using the web interface, the login name and
password are sent in the clear to the server.

2. If the USER.PW on the server holds a cleartext password, then the server
will also accept a login-card signature that is constructed using either
the cleartext password 

The second item relates to compatability with older clients. Although it is
unlikely anyone is still using Fossil versions that old (older than
010-01-11),
an old enough repository could possibly have cleartext passwords, so I
would recomend converting the passwords as described in the document.

Also, from reading the description of the sync protocol, I'm not sure how
safe the nounce is. The calculation appears to not make use of the current
time nor other value reasonably unique to the session.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Windows installer?

[Ron Wilson]

On Sat, Feb 1, 2014 at 11:39 AM,  wrote:

>  At work we only use the wiki portion of fossil. Our setup at work is to
> have NSSM (http://nssm.cc) create a windows service to start up fossil
> with a particular fossil repository. That might be a good thing to have
> with the installer.
>

FYI, Fossil has a "Windows Service" configuration command built-in (in the
Windows build, only). See "fossil help winsrv" for information.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Command line option for HTTP Auth?

[Ron Wilson]

On Wed, Jan 29, 2014 at 10:44 AM, Andy Bradford wrote:

> Thus said Kevin Martin on Wed, 29 Jan 2014 09:50:05 +:
> ...
> > I'd  like   to  use  basic   auth  and  fossil  auth   with  different
> > users/passwords.
>
> This currently  does not work.  I've actually done  some experimentation
> with separating sync user from the  transport user with SSH and while it
> did  work  quite well,  the  trick  is  getting  the interface  into  an
> intuitive flow.
>

What part seems to be the problems?

I've used github in the past. From what I know about SSH, I think github
uses the SSH server feature for associating an executable, and parameters,
with  a SSH key. With this, the github SSH server supplies the git command
with the github user id associated with the SSH key. Similar could be done
for Fossil and other services.

One possible problem I can see with this is the user id Fossil includes in
the transaction request.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Get a checkin diff from cli

[Ron Wilson]

On Wed, Jan 22, 2014 at 1:37 PM, Stephan Beal  wrote:

>
>
> On Wed, Jan 22, 2014 at 6:58 PM, Ron Wilson  wrote:
>
>> Expanding on my suggestion for relative tags like "+2" or "-3", the
>> capability to specify a commit relative to an arbitrary specified commit
>> would be useful as well.
>>
>
> There's a potential problem there: when a checkin has multiple parents,
> such an identifier is ambiguous.
> 
> That said, i like Lluis' idea of:
>
> fossil diff --checkin uuid
>
> as a shortcut for "diff the given uuid against its parent," with the
> caveat that a multi-parent checkin might need to toss an error there.
>

A commit with multiple parents would have a "primary parent" and one or
more "merge parents". I think that the "previous" (or "-1") parent would be
the primary/non-merge parent.

I have looked some at the code, but been working extra hours lately.

I hope your dog enjoys visiting his mom.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Get a checkin diff from cli

[Ron Wilson]

If you update your work space to the commit in question, I think you could
do::

fossil diff --from previous


Expanding on my suggestion for relative tags like "+2" or "-3", the
capability to specify a commit relative to an arbitrary specified commit
would be useful as well.



On Wed, Jan 22, 2014 at 6:29 AM, Lluís Batlle i Rossell wrote:

> On Wed, Jan 22, 2014 at 06:25:14AM -0500, Richard Hipp wrote:
> > On Wed, Jan 22, 2014 at 3:12 AM, Lluís Batlle i Rossell <
> vi...@viric.name>wrote:
> >
> > > Hello all,
> > >
> > > knowing the checkin hash, how can I get in cli the diff introduced by
> of
> > > the
> > > checkin? Similar to what appears in the checkin page on the ui.
> > >
> >
> > I think you have to know both the checkin hash and the parent check-in
> > hash.  Then:
> >
> > fossil diff --from $parent --to $checkin
>
> Ok, thank you. That is how I was doing it until now, but I thought there
> should
> be "a comfortable way". :)
>
> It'd be nice to have a "fossil diff --checkin X" or so. As for git, I
> often use
> 'git show' to achieve that.
>
> Regards,
> Lluís.
> ___
> fossil-users mailing list
> fossil-users@lists.fossil-scm.org
> http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
>
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] www.fossil-scm.org: permuted documentation index, useful?

[Ron Wilson]

The permuted index does make it easier to find things on some websites.
Personally, I don't think the list is too long. Other people might.

For larger sites, or sites with out a permuted index, I usually use the web
browser's find-in-page feature (control-F in FireFox and a few others I
have used).
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Odd permissions issue

[Ron Wilson]

On Wed, Jan 15, 2014 at 7:58 AM, Richard Hipp  wrote:

> Under Admin/Access the "Public Pages" entry box allows you to specify a
> comma-separated list of GLOB patterns for file that will be visible to the
> public even if the source code access is turned off.  This is used, for
> example, to make the SQLite Encryption Extension documentation files in the
> www/ directory visible (
> http://www.sqlite.org/see/doc/trunk/www/readme.wiki) without making the
> source code visible to non-licensees.
>

Just for clarification,I noticed the description of the "Public Pages"
settings only talks about "anonymous" and not-logged-in users. I assume
this setting also applies in the case of a "named, logged-in" user who
otherwise lacks "check out" permission,

(I also assume the repository owner should also remove clone permission
from the defaults (at least in 1.27, "nobody" has "g" (clone) permission by
default).)
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Odd permissions issue

[Ron Wilson]

On Wed, Jan 15, 2014 at 8:04 AM, Ron Aaron  wrote:

> Yes, thanks; I know about that, but that's too coarse-grained for my needs.
>
> In this case, I have a project I want to be *non-public* altogether, but
> I also don't want some people to have access to everything, and I can't
> put everything in the wiki.
>

/doc/ is not just for wiki (or markdown) pages, you can put other types of
files there as well.

Also, the "Public Pages" GLOB can potentially specify individual files


>
> On 01/15/2014 02:58 PM, Richard Hipp wrote:
> > Under Admin/Access the "Public Pages" entry box allows you to specify
> > a comma-separated list of GLOB patterns for file that will be visible
> > to the public even if the source code access is turned off.
>
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Odd permissions issue

[Ron Wilson]

On Wed, Jan 15, 2014 at 6:54 AM, Ron Aaron  wrote:

>  As of now, a "reader" has the capabilities "bcfhjkmnoprtw"
>
> I guess you are right about the security difference, but I would like to
> have a more fine-grained ability to allow access to (say) documents in a
> certain folder to "readers", but no access to the source code (or whatever).
>

If you run Fossil as a CGI behind a web server, I think that the URL based
access rules will also work on the path information passed to CGI scripts.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Odd permissions issue

[Ron Wilson]

On Wed, Jan 15, 2014 at 6:44 AM, Ron Aaron  wrote:

> I set up the "reader" user so that (I thought) it could access things
> needing "read" access.
>
> When I put a link to "./doc/tip..." something on my main repo page, I
> found that having "hyperlinks" permission was not enough, but "check
> out" permission was also required.
>
> Is this correct behavior? It seems odd to me.
>

"check out" permission is read permission for files. There are also read
permission settings for (internal) wiki pages and for tickets, "hyperlink"
"permission" is just whether the links are shown.


In theory, it would make sense for the wiki read permission to apply to
embedded documents accessed via /doc/, but since they are otherwise
"files", they are currently covered by the check out permission.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] how to deal with files which became directories?

[Ron Wilson]

On Mon, Jan 13, 2014 at 3:18 PM, Christian Pekeler wrote:

> I have a Numbers document in Fossil. It used to be just a file, now it's a
> folder. How can I commit this change to Fossil?
>
>
> > fossil rm Documents/daycare_stats.numbers
> DELETED Documents/daycare_stats.numbers
> > fossil add Documents/daycare_stats.numbers/*
> > fossil commit -m "upgraded Numbers doc"
> ...
> not an ordinary file:
> /Users/pekeler/Projects/kdw/Documents/daycare_stats.numbers
> abort due to prior errors
>

I think you have to commit the file deletion before doing the add.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Problem with compilation under MINGW

[Ron Wilson]

On Sat, Jan 11, 2014 at 5:32 PM, Alek Paunov  wrote:

> ...
> (yum: The Fedora family (CentOS, RHEL, etc) package manager is SQLite
> based).
>

Richard,  Belated congratulations.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] Small issue with ticket hook script

[Ron Wilson]

On Thu, Jan 9, 2014 at 8:05 AM, Mark Janssen  wrote:

> ...
> The reflected information in the query is the info from before the ticket
> update.
> I suspect the ticket hook is fired before the actual ticket change
> transaction is commited. Would it be possible to reverse this so that the
> hook script will be executed after the ticket change has been commited?
>

Actually, how about 2 hooks, one before and one after?
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

[fossil-users] relative tags

[Ron Wilson]

Fossil has the special tag previous and next to refer to the 2 commits on
either sire of the current check out. It would be useful to be able to
refer to more distant commits. I am thinking numbers prefixed with either
'-' or '+', so '-r -1' would be equivalent to '-r previous', '-r -2' would
be the commit before that, etc. Likewise, '-r +1' == '-r next', '-r +2' the
one after that, etc.

Thanks
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] curious: one/multiple repos for a source-project?

[Ron Wilson]

On Wed, Jan 8, 2014 at 5:07 PM, Michai Ramakers wrote:

> w.r.t. a recent question, I'm (still) curious how other people do this:
>
> for a single source-project consisting perhaps of multiple
> applications (each resulting in an executable binary, say) with
> perhaps documentation, reference-material and other non-source-stuff,
> do you typically use 1 big repo with everything in subdirs, or
> multiple repos for each sub-part of the project?
>

I would say it depends on how tightly coupled the suite of applications
need to be, How do they interact? How does a change in one application
affect the others?

Unless they need to be within a few revisions of each other (for whatever
reasonable value of "few"),  I would be inclined to split it up in to its
component applications and libraries to make the project more manageable -
and easier to reuse components in other projects. Of course, this opens up
the potential issue of tracking the compatibility of the components with
each other. While you can still have a single, over all package version,
each component will still have its own set of versions, independent of the
other components.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] ckout for diff in web interface

[Ron Wilson]

On Wed, Jan 8, 2014 at 10:42 AM, Remco Schoen  wrote:

> As a (silent) newcomer to Fossil, this is a feature, that also came to my
> mind.
>
> I do most of my work in Eclipse and as I was using mercurial, I had a good
> integration there. For Fossil I'm missing this integration, but the other
> features like the wiki, tickets and auto-sync make up for the loss in ease
> of use.
>
> I personally don't mind working on the command line, although in Windows
> that sometimes can be irritating. But for my co-workers I guess this lack
> of integration will be a bridge to far, as they are not used to working at
> the command line and probably will lose more in their view then they gain.
>

For a side project, I briefly used Eclipse since the vendor of the
microcontroller used provided a (free) development environment based on
Eclipse. (Normally, we use SlickEdit, which has an easy way to to integrate
any command line version control system.)

I noted that Excplipse does have an External Tools feature (at least in the
C/C++ perspective). I did not try it, but perhaps that would be a way to
partially integrate Fossil into Eclipse.

Probably, one of the Mercurial, Git or CVS plug-ins could be adapted for
use as a Fossil plug-in for Eclipse.

I have heard of a few attempts to create non-web based GUIs for
Fossil.Maybe one of those would fit your team's needs, even if it would not
be integrated with your DE.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

[fossil-users] ckout for diff in web interface

[Ron Wilson]

The embedded document feature allows for previewing the checked out copy of
the document in the web interface. How hard would it be to support similar
capability in the file diff feature?

Something like:

http://localhost:8080/fdiff?v1=c26cba673fdfa670&v2=ckout&sbs=1

Thanks
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] OT: TH1 shell

[Ron Wilson]

Thanks. It is useful for testing logic interactively.


On Sun, Jan 5, 2014 at 1:15 PM, Sergei Gavrikov
wrote:

> Hi
>
>   http://chiselapp.com/user/sg/repository/thsh/index
>
> it is a tiny TH shell. Main goal was to keep original Fossil TH core
> (th.c and th_lang.c) without any changes and provide TH1 shell with
> alone source (th_shell.c, 391 lines in C). Only new one command was
> added [source] to interact with system, and a few TCL-like globals,
> argv0, argv (if that specified), th_interactive, and th_version.
>
> The main function to read TH sources is Th_EvalFile(). Even interactive
> prompt uses it, `thsh' in interactive mode just does process an input
> from stdin with Th_EvalFile().
>
> The shell can report about errors on exit more verbosely like TCL does
>
>   $ echo foo | thsh
>   no such command:  foo
>   while executing
>   foo
>   (line 1)
>
> or
>
>   $ echo foo > foo
>   $ thsh foo
>   no such command:  foo
>   while executing
>   foo
>   (file "foo" line 1)
>
> As ::argv0, ::argv are supported by default, sha-bang scripts work as
> expected (!# /usr/bin/env thsh).
>
> Interactive shell knowns how to deal with two globals ::th_prompt1 and
> ::th_prompt2 if that set, default `thsh' prompt is "% "
>
>   % set ::th_prompt1 [list puts {th> }]
>   puts {th> }
>   th> proc hello args {
> set hello hello
>   }
>   th> hello
>   hello
>   th>
>
> About [info complete]. TH lacks this command, but if Th_Eval() returns
> (rc==TH_ERROR && nResult==0) this means that there was no close brace,
> quote, or bracket in a program and this fact was used to build simple
> interactive shell.
>
> In 'Th1 shell' project I also use `makeheaders' and `translate' *AS IS*.
> Those programs are neat code filters/generators! With `translate' you
> can easy to add new TH procedures or embed TH code in your C sources.
>
> Once again, my goal was to use TH core without any modifications, so it
> should be easy to keep the shell up to date:
>
>   cp ../fossil/src/th.[ch] ../fossil/src/th_lang.c src/
>   make
>
> TH1 shell is really tiny shell (static build with musl libc)
>
>   $ size -t bld/*.o
>  text  data bss dec hex filename
>  6128 0   0612817f0 bld/th_lang.o
> 13124   432   0   1355634f4 bld/th.o
>  2943 8   42955 b8b bld/th_shell.o
> 22195   440   4   22639586f (TOTALS)
>
>   $ size thsh
>  text  data bss dec hex filename
> 44693   8403552   49085bfbd thsh
>
> and you can use TH by its first appointment, test/glue code in embedded
> systems. At least `thsh' (IMHO) provides more comfortable way to run TH
> scripts on *nix than 'fossil test-th-eval', of course, I may be wrong.
>
> Please, excuse this OFF-TOPIC. I thought that `thsh' would help someone
> to learn or play with TH language.
>
> Sergei
> ___
> fossil-users mailing list
> fossil-users@lists.fossil-scm.org
> http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
>
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] New Tree-View feature requires newer sqlite.

[Ron Wilson]

On Mon, Jan 6, 2014 at 7:52 AM, Jan Nijtmans  wrote:

> at (1): Supporting every SQLite 3.7.x release is madness,
> SQLite 3.7.17 is more than 7 months old now, that should
> be enough time for whatever distribution to upgrade to it.
>

That depends on what other packages are using SQLite and how important the
distro maintainers think the the other packages are. Even when the upstream
devs of a library assert the ABI has not changed, the distro maintainers
tend to be quite conservative about when to update libraries. This is
because even if the ABI has not changed, the results returned may have
changed in a way that breaks something else.

The best way to encourage distro maintainers to stay more up to date with a
library is for the library devs to also provide an exhaustive regression
suite that the distro maintainers can use. And, when the results of any of
these tests change, provide detailed explanations as to why the results
changed.

(I have, in the past, talked to a few distro maintainers. Also, my
company's validation testers demand our (product engineering) test reports
so they can better understand the changes we made.)
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] New Tree-View feature requires newer sqlite.

[Ron Wilson]

On Fri, Jan 3, 2014 at 1:50 PM, Joseph R. Justice wrote:

> Actually...  What I was thinking of here was not anything related to
> preparation of the software changes or release package, but instead the
> actual deployment of the software itself into the field, and particularly
> into devices which cannot be reasonably expected to be ubiquitously
> networked and remotely upgradeable.  How does one update embedded software
> in a VCR or DVD player, or a refrigerator, or a vehicle (car / ship /
> aircraft)?  Or, for that matter, update embedded software used in an
> important industrial process, such as a power plant (nuclear or otherwise),
> a chemical manufacturing or processing facility, etc where failure of the
> process can potentially be catastrophic so you don't *want* it to be easy
> to update the software (to avoid malicious changes by third parties)?
>

 Our products are actually components for products of other businesses, so
the details vary, but they are all designed to require a special program
running on a laptop or tablet to communicate with a "flash loader" program
that is nominally permanently resident in our devices. (It is possible to
update the flash loader, but requires extra steps.) This special program is
specific to the producer of the final product and is normally only
available to authorized service businesses. (And yes, we have variations of
our flash loader for each of the final product producers.)

To reduce the possibility of unauthorized SW being loaded into our
products, we encrypt the binary images. Our flash loader decrypts the image
as it writes the image to built in "flash ROM".

How a technician connects the laptop/tablet to the component devices varies
based on the type of the final product and the producers' design standards.
Many such products, especially modern automobiles, have (possibly hidden)
service connectors that allow the service program to communicate with the
products component devices via an internal diagnostic "network". Other
products require the technician to plug directly into the individual
components devices.

While I don't know much of the details, I do know that many consumer
devices like BluRay players, and even TVs, are Internet enabled, so can be
updated over the Internet. For example, my BluRay player at home has an
obscure menu option to start the player downloading and installing the
latest SW from the producer's website. If the download or installation
fails, it it will automatically retry. Since mine has an Ethernet port, I
can take it anywhere I can get Internet access (If need be, I can use
pocket sized Ethernet to WiFi adapter.)
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] New Tree-View feature requires newer sqlite.

[Ron Wilson]

On Thu, Jan 2, 2014 at 2:04 PM, Joseph R. Justice wrote:

> 
> Further, in the embedded software world at least, I expect it is often
> difficult or even impossible to update software in the first place to patch
> bugs or provide new functionality, so there is no need much less desire to
> worry about making it *easier* to update software; if you can't update it
> at all, or can only update it in the first place with the very greatest of
> effort for only the most critical of reasons (such as bugs which are
> potentially life-threatening or would cost untenable financial losses if
> they are not fixed), who cares if it's a only a little more difficult,
> relatively speaking, to update a library.
>

Speaking as an embedded SW developer, I can tell you that the actual SW
changes, and preparing the release package is easy. Actually making the
release, though, is a big deal. After preparing the release package, we do
regression testing. While a lot of this is automated, changes to physical
inputs (switches, sensor, etc) and checking of physical outputs
(indicators, motors, etc) is done manually, so these tests can take many
hours. Once those tests have passed, we turn the release package over to
the validation team, who preform their own (independently created) suite of
tests. Once validation gives a pass, we send copies of the package to
manufacturing and to our customer service organization. There is also
documentation to prepare, which we can mostly complete in parallel with
validation. (The validation team has access to the issue tickets, so we
only need to give them a list of the ticket numbers applicable to the
release.)

We do use 3rd party libraries. The difference is that we have full control
over what versions of each library are used. There is no one downstream
from us repackaging our SW. The binaries we release are exactly what goes
into the devices we created the SW for.

BTW, the singular of "data" is "datum". Over time, "data set" has been
shortened to "data", so "data", as a noun, has become like "sheep", which
is both singular and plural.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] password manager support

[Ron Wilson]

On Mon, Dec 30, 2013 at 5:10 PM, David Rush wrote:

> If ssh-agent can work it's probably the way to go. It seems to be the most
> standardized amongst all the popular pass-phrase managers, at least on
> non-[apple|microsoft] platforms. But, as far as I can tell it only stores
> decrypted private keys and not any pass phrases and I don't have the
> resources to investigate.
>

This is true, but assuming it actually delivers the  decrypted key to its
client (as opposed to performing the authentication based on information
received from its client), then pass phrases could be packaged to look like
keys.

I know this is a "sideway" approach, but as best I can determine, ssh-agent
is the most multi platform tool.

(Last week, I came across a link to a proposed standard protocol for client
applications to communicate with a pass phrase manager, but I lost it and
have not been able to find it, again - at least not in the few minutes I've
trying)

Otherwise, I'd say to try to find the least platform specific, open source
pass phrase manager and interface to that. Then, hopefully, anyone who
needs it on a not-yet-supported platform could reasonably port the manager
to that platform.

Happy New Year
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] password manager support

[Ron Wilson]

On Fri, Dec 27, 2013 at 4:30 PM, Ron Wilson  wrote:

> There is a platform independent tool that does this for SSH: SSH-Agent. In
> theory, it could store other credentials. It is open source, so the client
> API is readily available.
>
> One of the versions for OSX, SSHkeychain, integrates with Apple Keychain.
> I would be surprised if there are not versions that integrate with Gnome
> and/or KDE key rings.
>

Interestingly, http://en.wikipedia.org/wiki/Ssh-agent claims that Gnome
Keyring and KWallet  also provide SSH-Agent services. I have not (yet)
confirmed this.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] password manager support

[Ron Wilson]

On Fri, Dec 27, 2013 at 11:56 AM, David Rush wrote:

> Those are all good ideas but they address the other end of my issue. My
> immediate need is with respect to "fossil as a client" that implements a
> somewhat standardized, but platform specific, way to store and retrieve the
> credentials that fossil sends in the "Authorization" http header so the
> HTTP server will let me in.
>
>
There is a platform independent tool that does this for SSH: SSH-Agent. In
theory, it could store other credentials. It is open source, so the client
API is readily available.

One of the versions for OSX, SSHkeychain, integrates with Apple Keychain. I
would be surprised if there are not versions that integrate with Gnome
and/or KDE key rings.

I don't know how Fossil's SSH support works, but I'd be surprised if it
doesn't support use of SSH-Agent.

Basically, SSH-Agent supplies ready-to-use credentials to its clients from
a (presumably encrypted) file. It should be easy to create a simple utility
to format non-SSH credentials to look acceptable to SSH-Agent, and a
corresponding function to un-package the credentials as retrieved from
SSH-Agent.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Re: [fossil-users] password manager support

[Ron Wilson]

On 12/24/13, Andy Bradford  wrote:
> Rather  than  making  Fossil  aware of  different  user  databases,  why
> not  simply  make a  Fossil  interface  for passing  user/authentication
> information onto an external program.

That is what PAM is about. PAM provides a standardized API for
applications to perform user authentication. When an application
requests an authentication, PAM will handle it based on the host's
local configuration and installed authentication modules. The
configuration and available authentication modules can be changed
as needed without needing to reconfigure the client applications.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users