Off-topic: is fossil 2.4 (windows binary) broken in some way?

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
8 messages Options
Reply | Threaded
Open this post in threaded view
|

Off-topic: is fossil 2.4 (windows binary) broken in some way?

Olivier Mascia
Dear all,

Could it be that something is broken with fossil 2.4 windows binary (as available from fossil-scm.org)?
When attempting fossil ui the pages served seem to be missing things right after opening <!doctype html> resulting in rendering missing styling.

For instance:

<!DOCTYPE html>
<div class="header">
  <div class="title"><h1>Fossil</h1>Branches</div>
    <div class="status">Olivier — <a href='/login'>Logout</a>
</div>
</div>
<div class="mainmenu">
...

I suppose I have something setup wrong, but it all looked so simple...
--
Best Regards, Meilleures salutations, Met vriendelijke groeten,
Olivier Mascia
_______________________________________________
sqlite-users mailing list
[hidden email]
http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
Reply | Threaded
Open this post in threaded view
|

Re: Off-topic: is fossil 2.4 (windows binary) broken in some way?

Olivier Mascia

> Le 26 déc. 2017 à 23:35, Olivier Mascia <[hidden email]> a écrit :
>
> Dear all,
>
> Could it be that something is broken with fossil 2.4 windows binary (as available from fossil-scm.org)?
> When attempting fossil ui the pages served seem to be missing things right after opening <!doctype html> resulting in rendering missing styling.
>
> For instance:
>
> <!DOCTYPE html>
> <div class="header">
>  <div class="title"><h1>Fossil</h1>Branches</div>
>    <div class="status">Olivier — <a href='/login'>Logout</a>
> </div>
> </div>
> <div class="mainmenu">
> ...
>
> I suppose I have something setup wrong, but it all looked so simple...

Hint: from reading things, bits of code, and files here and there and testing...

I can get things right if I'm doing:

        fossil ui --skin default

But it looks like "--skin default" is not the default if not specified.  I'm not sure if this was the intent.

--
Best Regards, Meilleures salutations, Met vriendelijke groeten,
Olivier Mascia



_______________________________________________
sqlite-users mailing list
[hidden email]
http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
Reply | Threaded
Open this post in threaded view
|

Re: Off-topic: is fossil 2.4 (windows binary) broken in some way?

Olivier Mascia
> Le 26 déc. 2017 à 23:46, Olivier Mascia <[hidden email]> a écrit :
>
>>
>> Le 26 déc. 2017 à 23:35, Olivier Mascia <[hidden email]> a écrit :
>>
>> Dear all,
>>
>> Could it be that something is broken with fossil 2.4 windows binary (as available from fossil-scm.org)?
>> When attempting fossil ui the pages served seem to be missing things right after opening <!doctype html> resulting in rendering missing styling.
>>
>> For instance:
>>
>> <!DOCTYPE html>
>> <div class="header">
>> <div class="title"><h1>Fossil</h1>Branches</div>
>>   <div class="status">Olivier — <a href='/login'>Logout</a>
>> </div>
>> </div>
>> <div class="mainmenu">
>> ...
>>
>> I suppose I have something setup wrong, but it all looked so simple...
>
> Hint: from reading things, bits of code, and files here and there and testing...
>
> I can get things right if I'm doing:
>
> fossil ui --skin default
>
> But it looks like "--skin default" is not the default if not specified.  I'm not sure if this was the intent.

Still looking and finding my way: exactly same behaviour with the prebuilt macOS binary.  So I guess it's something I misread or did not read.  Feels strange though that '--skin default' isn't the actual default unless explicitly specified on the command line.

(I'll look for the fossil appropriate mailing list for future posts).

--
Best Regards, Meilleures salutations, Met vriendelijke groeten,
Olivier Mascia


_______________________________________________
sqlite-users mailing list
[hidden email]
http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
Reply | Threaded
Open this post in threaded view
|

Re: Off-topic: is fossil 2.4 (windows binary) broken in some way?

Warren Young
In reply to this post by Olivier Mascia
On Dec 26, 2017, at 3:35 PM, Olivier Mascia <[hidden email]> wrote:
>
> Could it be that something is broken with fossil 2.4 windows binary (as available from fossil-scm.org)?

It works for me.

Here’s my test:

    c:\tmp> fossil new ..\x.fossil
    c:\tmp> fossil open ..\x.fossil
    c:\tmp> fossil add *
    c:\tmp> fossil ci
    c:\tmp> fossil ui

That creates a fresh Fossil repo, opens it in the project’s working directory (not optional, even for a new project!), adds a bunch of files, checks them in, and then shows Fossil UI for the result.

If you tried something else and it fails, reply with how your test differed, which will probably tell us *why* it didn’t do what you expected.
_______________________________________________
sqlite-users mailing list
[hidden email]
http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
Reply | Threaded
Open this post in threaded view
|

Re: Off-topic: is fossil 2.4 (windows binary) broken in some way?

Olivier Mascia
> Le 27 déc. 2017 à 00:13, Warren Young <[hidden email]> a écrit :
>
> On Dec 26, 2017, at 3:35 PM, Olivier Mascia <[hidden email]> wrote:
>>
>> Could it be that something is broken with fossil 2.4 windows binary (as available from fossil-scm.org)?
>
> It works for me.
>
> Here’s my test:
>
>    c:\tmp> fossil new ..\x.fossil
>    c:\tmp> fossil open ..\x.fossil
>    c:\tmp> fossil add *
>    c:\tmp> fossil ci
>    c:\tmp> fossil ui
>
> That creates a fresh Fossil repo, opens it in the project’s working directory (not optional, even for a new project!), adds a bunch of files, checks them in, and then shows Fossil UI for the result.
>
> If you tried something else and it fails, reply with how your test differed, which will probably tell us *why* it didn’t do what you expected.

Thanks Warren.
I have no issues or difficulties with fossil commands themselves and know how to, for instance, clone and properly open sqlite repository (which I used for my tests). My only issue is with the server or ui command which by default emits pages completely missing the top the page (head, style, even body).  Emitting pages like:

<!DOCTYPE html>
--> obviously missing content here <--
<div class="header">
 <div class="title"><h1>Fossil</h1>Branches</div>
   <div class="status">Olivier — <a href='/login'>Logout</a>
</div>
</div>
<div class="mainmenu">
...

Simply doing:

        fossil ui --skin default

instead of:

        fossil ui

fixes that, and that feels a bit unexpected (for a default).
I have since determined I have exactly the same behaviour on macOS, so it is not at all related to the Windows binary.

On Mac & Windows: This is fossil version 2.4 [a0001dcf57] 2017-11-03 09:29:29 UTC.

--
Best Regards, Meilleures salutations, Met vriendelijke groeten,
Olivier Mascia


_______________________________________________
sqlite-users mailing list
[hidden email]
http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
Reply | Threaded
Open this post in threaded view
|

Re: Off-topic: is fossil 2.4 (windows binary) broken in some way?

Warren Young
On Dec 26, 2017, at 4:22 PM, Olivier Mascia <[hidden email]> wrote:
>
> I have no issues or difficulties with fossil commands themselves and know how to, for instance, clone and properly open sqlite repository

Oh, I thought you were seeing this with your own new repository.

Now that I know what to test, I see it, too:

    fossil-native-2.4 clone https://sqlite.org/cgi/src x.x
    mkdir x
    cd x
    fossil-native-2.4 open ../x.x
    fossil-native-2.4 ui

This is with fossil-native-2.4.exe being the binary from fossil-scm.org, so-named to differentiate it from the fossil.exe binary I actually use here. :)

The problem is due to the improvements to skinning.  When you initially clone the repository, you get a copy of the remote skin configuration, but it will only work if you have a version of Fossil as of about a month ago, when all of this changed.

What you’re doing to fix it is override the skin that sqlite.org gave you with a configuration that works correctly with your local fossil.exe.

Until drh ships Fossil 2.5, you’ll have to either stick with that workaround or build your own fossil.exe against the trunk code.
_______________________________________________
sqlite-users mailing list
[hidden email]
http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
Reply | Threaded
Open this post in threaded view
|

Re: Off-topic: is fossil 2.4 (windows binary) broken in some way?

Olivier Mascia
> Le 27 déc. 2017 à 00:39, Warren Young <[hidden email]> a écrit :
>
> Until drh ships Fossil 2.5, you’ll have to either stick with that workaround or build your own fossil.exe against the trunk code.

Makes it clear.  Thanks.

--
Best Regards, Meilleures salutations, Met vriendelijke groeten,
Olivier Mascia


_______________________________________________
sqlite-users mailing list
[hidden email]
http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
Reply | Threaded
Open this post in threaded view
|

Re: Off-topic: is fossil 2.4 (windows binary) broken in some way?

Warren Young
On Dec 26, 2017, at 4:45 PM, Olivier Mascia <[hidden email]> wrote:
>
>> Le 27 déc. 2017 à 00:39, Warren Young <[hidden email]> a écrit :
>>
>> Until drh ships Fossil 2.5, you’ll have to either stick with that workaround or build your own fossil.exe against the trunk code.
>
> Makes it clear.  Thanks.

Well, “clear” is a charitable description.  Thank you for being nice. :)

I find my reply less than clear.  Let me make a few clarifications.

These improvements to the way Fossil does skinning are only about a month old, so you need a Fossil binary that’s newer than that to make use of the skin configuration you get when cloning the remote repository.

This does not mean that Fossil 2.4 cannot make use of the current sqlite.org repository.  It just means that sqlite.org is making use of unreleased features already.  (Again part of the “eat your own dog food” standard practice.)

There is a second workaround:

    c:\> fossil conf reset skin

That tells your local Fossil to throw away the skin configuration it got as part of the clone and just use what’s built into your local fossil.exe.

That’s one of the cool things about Fossil as compared to GitHub: not only can you skin it however you like, a clone needn’t necessarily have the same skin as the repository it cloned from.  The only time they’re synchronized automatically is on the initial clone.  After that, you have to *ask* for a re-sync:

    c:\> fossil conf pull skin

or:

   c:\> fossil conf pull all

to overwrite all the other locally-configurable bits.
_______________________________________________
sqlite-users mailing list
[hidden email]
http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users