Re: [fossil-users] Setting up an internet Fossil server

[Richard Hipp]

The SQLite.org website, and the Fossil-SCM.org website (which is the
same server) have always run off of the "althttpd" webserver.  I have
just put up documentation for this very simple but effective webserver
at

https://www.sqlite.org/docsrc/doc/trunk/misc/althttpd.md

If you prefer to run Fossil using Apache or Nginx or Lighttpd or
whatever else you are comfortable with, do so with my blessings.
Diversity is good in this context as it helps to flush out bugs in
Fossil.  Information about althttpd is provide only in case you are
interested.

-- 
D. Richard Hipp
d...@sqlite.org
___
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] Converting to/from symlinks

[Warren Young]

On Feb 27, 2018, at 2:42 PM, Svyatoslav Mishyn  
wrote:
> 
> So, if I want at some point to replace file by a symlink,
> first need to remove that file, commit changes,
> and only then add a symlink, right?

I did just that for a file here just days ago using the 2.5 release version.  
I’ve had no trouble due to the file -> symlink change, even when doing a 
“fossil update” on machines that haven’t had an update since before the change, 
so they’re effectively getting both changes at once.

If for some reason you had to have this change appear atomically on the trunk, 
you could remove the file, check it in on a branch, add the symlink, check that 
in on the branch, and then “fossil merge --integrate symlink-fix-branch” to 
cause both changes to appear on trunk together.
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

[fossil-users] Converting to/from symlinks

[Svyatoslav Mishyn]

Hi,

it seems that when `allow-symlinks` option is enabled,
file which was already added cannot be overwritten by a symlink,
see a test.


mkdir test && cd test
fossil init .fossil && fossil open .fossil
fossil set allow-symlinks 0

cp ~/.ssh/default.pub . && fossil add default.pub && fossil ci -m file

fossil set allow-symlinks 1
ln -fs ~/.ssh/default.pub default.pub

fossil ci -m symlink
New_Version: 163fd28f692b571070e1471fa3236ba46b0f121cbc11553c4203fc937831dca5
ERROR: [default.pub] is 27 bytes on disk but 726 in the repository
NOTICE: Repository version of [default.pub] stored in [file-f8c1d36b817714f6]
working checkout does not match what would have ended up in the repository:  
6810fe7cfca414a1c2f36cea0619a695 versus 8fa57811ea8fcd434443c61628f62d73


So, if I want at some point to replace file by a symlink,
first need to remove that file, commit changes,
and only then add a symlink, right?


(Tested with trunk version.)


-- 
https://www.juef.space/


signature.asc
Description: PGP signature
___
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 purge Command

[Warren Young]

On Feb 27, 2018, at 10:55 AM, Agrawal, Ritika  wrote:
> 
> - Need to clean up the files manually as well

That’s consistent with the current default behavior of “fossil rm” as well, so 
it’s correct until/unless that changes.

>-  Purging happens only in local clone copy and never syncs back. 
> Need to copy back the clone to central repo

That’s consistent with the behavior of the /shun command.  I’m uncertain how 
these two mechanisms interact and overlap, but it is at least clear to me that 
they should behave the same way.

I’m uncertain because I’ve never used either mechanism in all my years of 
Fossil use, on purpose, for philosophical reasons.

The reason these commands behave that way is that, once you distribute a given 
artifact to someone else, you should not be able to *make* them forget that you 
sent it to them.  It takes a collaborative effort to expunge an artifact from 
all repository clones, everywhere, on purpose.  This is because it is 
necessarily a social matter rather than a technical matter: you need to 
convince everyone they they should also forget the artifact.

This mailing list works much the same way.  I cannot un-send a message to the 
mailing list.  There is doubtless a way for drh to delete a message from the 
mailing list server’s archive, but that doesn’t un-distribute it to the mailing 
list members.  I’ve occasionally regretted sending out an email to a mailing 
list, but I wouldn’t presume to reach into your computer to delete the copy you 
received to “fix” my mistake.

See also: https://www.fossil-scm.org/index.html/doc/trunk/www/shunning.wiki
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

[fossil-users] Fossil purge Command

[Agrawal, Ritika]

> There may come a day when going to an HTTP-only web site will require 
> multiple affirmations asymptotically approaching ?Yes, I?m really quite 
> certain I want my face eaten by a rabid grue.  Just let me look at this one 
> web site first, please.?
> ___
> fossil-users mailing list
> fossil-users@lists.fossil-scm.org
> http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
>


--

Message: 3
Date: Tue, 27 Feb 2018 08:44:47 -0700
From: Warren Young <war...@etr-usa.com>
To: Fossil SCM user's discussion <fossil-users@lists.fossil-scm.org>
Subject: Re: [fossil-users] Setting up an internet Fossil server
Message-ID: <0765fe52-08d1-4b7f-84ab-9d1dda23f...@etr-usa.com>
Content-Type: text/plain;   charset=utf-8

On Feb 27, 2018, at 8:37 AM, Roy Keene <fos...@rkeene.org> wrote:
> 
> You don't lose support for TLS, since Apache supports TLS.  It's just running 
> Fossil as a CGI -- this is exactly how ChiselApp works.

Compare Thomas’ post to the HOWTO I linked in my first post in this thread.  
The largest part of the difference between them is that my HOWTO gives you a 
Let’s Encrypt setup as well as a Fossil server.  Since Thomas doesn’t describe 
how to configure TLS on Apache, I’d say that it’s fair to say that’s one big 
reason why Thomas’ configuration is simpler than mine.

A much smaller part of the delta is plain old CGI vs “fossil server --scgi”, 
which I think is well worth the minor complexity to avoid the CPU and disk hits 
of repeated Fossil launches.

--

Message: 4
Date: Tue, 27 Feb 2018 15:56:16 +
From: "Agrawal, Ritika" <ritika.agra...@intel.com>
To: "fossil-users@lists.fossil-scm.org"
<fossil-users@lists.fossil-scm.org>
Subject: [fossil-users] Fossil purge Command
Message-ID:
<d1b5b600ee5bf04ea33dd0a5e61a204c38f7b...@fmsmsx103.amr.corp.intel.com>

Content-Type: text/plain; charset="utf-8"

Hi,

I am trying to use the purge command available with fossil 2.5 release.

Steps performed :
1. fossil purge files  2. fossil purge list 3. fossil purge 
obliterate 

After obliteration, the size of the repo remains the same. After trying to 
commit the changes to the repo, I see the error :


  working checkout does not match manifest after commit: 
e60149b4a3df3f1328936051f4cc5f80 versus 0734caf2221c1050ab1cf07cf8616043



Both these ids are neither current checkout id nor parent id. Any ideas?



Thanks,

Ritika

-- next part --
An HTML attachment was scrubbed...
URL: 
<http://lists.fossil-scm.org:8080/pipermail/fossil-users/attachments/20180227/b46a39f5/attachment-0001.html>

--

Message: 5
Date: Tue, 27 Feb 2018 09:59:28 -0600 (CST)
From: Roy Keene <fos...@rkeene.org>
To: "Fossil SCM user's discussion" <fossil-users@lists.fossil-scm.org>
Subject: Re: [fossil-users] Setting up an internet Fossil server
Message-ID: <alpine.lnx.2.20.1802270958410.20...@railroad.oc9.org>
Content-Type: text/plain; charset=US-ASCII; format=flowed

ChiselApp also uses Let's Encrypt

On Tue, 27 Feb 2018, Warren Young wrote:

> On Feb 27, 2018, at 8:37 AM, Roy Keene <fos...@rkeene.org> wrote:
>>
>> You don't lose support for TLS, since Apache supports TLS.  It's just 
>> running Fossil as a CGI -- this is exactly how ChiselApp works.
>
> Compare Thomas? post to the HOWTO I linked in my first post in this thread.  
> The largest part of the difference between them is that my HOWTO gives you a 
> Let?s Encrypt setup as well as a Fossil server.  Since Thomas doesn?t 
> describe how to configure TLS on Apache, I?d say that it?s fair to say that?s 
> one big reason why Thomas? configuration is simpler than mine.
>
> A much smaller part of the delta is plain old CGI vs ?fossil server --scgi?, 
> which I think is well worth the minor complexity to avoid the CPU and disk 
> hits of repeated Fossil launches.
> ___
> fossil-users mailing list
> fossil-users@lists.fossil-scm.org
> http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
>


--

Message: 6
Date: Tue, 27 Feb 2018 11:44:42 -0500
From: Richard Hipp <d...@sqlite.org>
To: "Fossil SCM user's discussion" <fossil-users@lists.fossil-scm.org>
Subject: Re: [fossil-users] Fossil purge Command
Message-ID:

Re: [fossil-users] Fossil purge Command

[Richard Hipp]

On 2/27/18, Agrawal, Ritika  wrote:
> Hi,
>
> I am trying to use the purge command available with fossil 2.5 release.
>
> Steps performed :
> 1. fossil purge files 
> 2. fossil purge list
> 3. fossil purge obliterate 
>
> After obliteration, the size of the repo remains the same.

You probably need to run VACUUM on the database.

 fossil sql VACUUM

Or you can vacuum as part of a rebuild:

fossil rebuild --vacuum


> After trying to
> commit the changes to the repo, I see the error :
>
>   working checkout does not match manifest after commit:
> e60149b4a3df3f1328936051f4cc5f80 versus 0734caf2221c1050ab1cf07cf8616043
>

Background information:
https://www.fossil-scm.org/fossil/doc/trunk/www/selfcheck.wiki

Those are MD5 checksums over the entire content of the check-in as it
exists on disk versus what Fossil is trying to push into the
repository.  Something has gone wrong so that the two do not agree,
hence Fossil rolls back the check-in to avoid problems.

-- 
D. Richard Hipp
d...@sqlite.org
___
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] Setting up an internet Fossil server

[Roy Keene]

ChiselApp also uses Let's Encrypt

On Tue, 27 Feb 2018, Warren Young wrote:


On Feb 27, 2018, at 8:37 AM, Roy Keene  wrote:


You don't lose support for TLS, since Apache supports TLS.  It's just running 
Fossil as a CGI -- this is exactly how ChiselApp works.


Compare Thomas? post to the HOWTO I linked in my first post in this thread.  
The largest part of the difference between them is that my HOWTO gives you a 
Let?s Encrypt setup as well as a Fossil server.  Since Thomas doesn?t describe 
how to configure TLS on Apache, I?d say that it?s fair to say that?s one big 
reason why Thomas? configuration is simpler than mine.

A much smaller part of the delta is plain old CGI vs ?fossil server --scgi?, 
which I think is well worth the minor complexity to avoid the CPU and disk hits 
of repeated Fossil launches.
___
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

[fossil-users] Fossil purge Command

[Agrawal, Ritika]

Hi,

I am trying to use the purge command available with fossil 2.5 release.

Steps performed :
1. fossil purge files 
2. fossil purge list
3. fossil purge obliterate 

After obliteration, the size of the repo remains the same. After trying to 
commit the changes to the repo, I see the error :


  working checkout does not match manifest after commit: 
e60149b4a3df3f1328936051f4cc5f80 versus 0734caf2221c1050ab1cf07cf8616043



Both these ids are neither current checkout id nor parent id. Any ideas?



Thanks,

Ritika

___
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] Setting up an internet Fossil server

[Warren Young]

On Feb 27, 2018, at 8:37 AM, Roy Keene  wrote:
> 
> You don't lose support for TLS, since Apache supports TLS.  It's just running 
> Fossil as a CGI -- this is exactly how ChiselApp works.

Compare Thomas’ post to the HOWTO I linked in my first post in this thread.  
The largest part of the difference between them is that my HOWTO gives you a 
Let’s Encrypt setup as well as a Fossil server.  Since Thomas doesn’t describe 
how to configure TLS on Apache, I’d say that it’s fair to say that’s one big 
reason why Thomas’ configuration is simpler than mine.

A much smaller part of the delta is plain old CGI vs “fossil server --scgi”, 
which I think is well worth the minor complexity to avoid the CPU and disk hits 
of repeated Fossil launches.
___
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] Setting up an internet Fossil server

[Roy Keene]

You don't lose support for TLS, since Apache supports TLS.  It's just 
running Fossil as a CGI -- this is exactly how ChiselApp works.


On Tue, 27 Feb 2018, Warren Young wrote:


On Feb 26, 2018, at 3:33 PM, Thomas Levine <_...@thomaslevine.com> wrote:


Since it seems that the only dynamic stuff is in PHP and fossil,
I suggest using Apache mod_php and mod_cgi (contrary to Warren's
suggestion), as I think the configuration will be easier.


Of course, but then you lose HTTPS, which is the only reason my configuration 
is difficult at all.  If all you wanted is reverse proxying, you?d do away with 
steps 1-6, simplifying the HOWTO considerably.

I don?t view TLS as optional for password-protected public web resources in 
these post-Firesheep days.

Even if you don?t care about your own Fossil repo?s security, Google has been 
punishing sites that are not available via HTTPS for a couple of years now, 
both through reduced rankings in the search engine and through increasingly 
strident warnings in Chrome.

That?s not speculation, Google?s been announcing these things publicly:

  https://webmasters.googleblog.com/2014/08/https-as-ranking-signal.html
  https://developers.google.com/web/updates/2016/10/avoid-not-secure-warn

There may come a day when going to an HTTP-only web site will require multiple 
affirmations asymptotically approaching ?Yes, I?m really quite certain I want 
my face eaten by a rabid grue.  Just let me look at this one web site first, 
please.?
___
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] Setting up an internet Fossil server

[Warren Young]

On Feb 26, 2018, at 3:33 PM, Thomas Levine <_...@thomaslevine.com> wrote:
> 
> Since it seems that the only dynamic stuff is in PHP and fossil,
> I suggest using Apache mod_php and mod_cgi (contrary to Warren's
> suggestion), as I think the configuration will be easier.

Of course, but then you lose HTTPS, which is the only reason my configuration 
is difficult at all.  If all you wanted is reverse proxying, you’d do away with 
steps 1-6, simplifying the HOWTO considerably.

I don’t view TLS as optional for password-protected public web resources in 
these post-Firesheep days.

Even if you don’t care about your own Fossil repo’s security, Google has been 
punishing sites that are not available via HTTPS for a couple of years now, 
both through reduced rankings in the search engine and through increasingly 
strident warnings in Chrome.

That’s not speculation, Google’s been announcing these things publicly:

   https://webmasters.googleblog.com/2014/08/https-as-ranking-signal.html
   https://developers.google.com/web/updates/2016/10/avoid-not-secure-warn

There may come a day when going to an HTTP-only web site will require multiple 
affirmations asymptotically approaching “Yes, I’m really quite certain I want 
my face eaten by a rabid grue.  Just let me look at this one web site first, 
please.”
___
fossil-users mailing list
fossil-users@lists.fossil-scm.org
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users