You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@camel.apache.org by Francis Vila <fv...@gmail.com> on 2016/11/08 12:06:59 UTC

Time for an overhaul of the patterns symbols?

Hi,

I am new to both Camel and to open-source contributing, so forgive any
breaches of protocol...

The page on  http://camel.apache.org/enterprise-integration-patterns.html
<http://camel.apache.org/enterprise-integration-patterns.html>   shows an
impressive list of patterns. However the graphics don't all seem to be in
the same style, and some are missing. To me (although I don't have an
in-depth understanding of Camel), the diagrams symbolize the essence of what
Camel does. Maybe a homogeneous, updated styleset would be a good idea? 
I think I can contribute to such an task. I've done some tests of different
approaches for the /Message router /pattern. If this effort is deemed
worthwhile by the community, I can go on with the rest. The graphics were
done using Inkscape and are available in svg format.

This is the simplest version:
<http://camel.465427.n5.nabble.com/file/n5789901/MessageRouter_1.png> 

This version has added endpoints symbolizing the input and output locations
for the origin and destination of the messages :
<http://camel.465427.n5.nabble.com/file/n5789901/MessageRouter_2.png> 

This version encloses the pattern:
<http://camel.465427.n5.nabble.com/file/n5789901/MessageRouter_3.png> 

This version puts emphasis on the the most important part of the diagram,
namely the currently active path:
<http://camel.465427.n5.nabble.com/file/n5789901/MessageRouter_5b.png> 

This version targets the emphasis more closely, to just the "mobile" part of
the diagram:
<http://camel.465427.n5.nabble.com/file/n5789901/MessageRouter_6.png> 




--
View this message in context: http://camel.465427.n5.nabble.com/Time-for-an-overhaul-of-the-patterns-symbols-tp5789901.html
Sent from the Camel Development mailing list archive at Nabble.com.

Re: Suggestions for new set of integration pattern symbols

Posted by Francis Vila <fv...@gmail.com>.
Thanks for the feedback @whiskey  !
Not quite sure what you mean about the /need to ‘describe/name/reference’
major ‘conditions or logic predicates’ on message handling inside the box/
Is this the kind of thing you have in mind:
<http://camel.465427.n5.nabble.com/file/n5789957/MessageRouter_text_edit.png> 
This is an example in PowerPoint (maybe not the most popular software among
members of this forum?) with an export to .wmf, ungrouped then with text
edited. 
About formats: I think if we provide .svg files then everyone can convert to
the format most suitable to their editing software.
About text: maybe we can provide 2 versions with / without text




--
View this message in context: http://camel.465427.n5.nabble.com/Suggestions-for-new-set-of-integration-pattern-symbols-tp5789901p5789957.html
Sent from the Camel Development mailing list archive at Nabble.com.

Re: Suggestions for new set of integration pattern symbols

Posted by Vincent Nonnenmacher <vi...@gmail.com>.
I like the version 3 and 4 (with the nice round idea of the six version),
as there is always a need to name the box, so an outline is more often
quite mandatory.

The original lib (visio and omnigraffle) could benefit from a nice ‘design
refresh’.

One missing point however in using those blocks in diagrams, is the fact
that there is a need to ‘describe/name/reference’
major ‘conditions or logic predicates’ on message handling inside the box
(a route endpoint/node)

If you take your example, there is often a need to make an arrow pointing
inside
the ‘moving part of the switch’ to describe/reference/annotate what
destination will be chosen.

Having that in mind, will make the overall look and feel of the diagram
easier to grasp as there is no more ‘visual disparity’
in comprehension of the overall picture.



On Wed, Nov 9, 2016 at 8:26 AM, Francis Vila <fv...@gmail.com> wrote:

> This post contains an offer to contribute towards the diagrams symbolizing
> the Camel patters, with 6 different suggested formats
>
>
>
> --
> View this message in context: http://camel.465427.n5.nabble.
> com/Suggestions-for-new-set-of-integration-pattern-
> symbols-tp5789901p5789932.html
> Sent from the Camel Development mailing list archive at Nabble.com.
>

Re: Suggestions for new set of integration pattern symbols

Posted by Francis Vila <fv...@gmail.com>.
This post contains an offer to contribute towards the diagrams symbolizing
the Camel patters, with 6 different suggested formats 



--
View this message in context: http://camel.465427.n5.nabble.com/Suggestions-for-new-set-of-integration-pattern-symbols-tp5789901p5789932.html
Sent from the Camel Development mailing list archive at Nabble.com.

Re: Time for an overhaul of the patterns symbols?

Posted by Francis Vila <fv...@gmail.com>.
Hi Tadayoshi,

Thank you so much for the link and the encouragements... I thought the
diagrams belonged to Camel. I was wondering how to find contacts in the
community to get more people interested, it looks like the Gregor Hohpe guy
who runs the EIP website would be the best entry point.

I am looking for feedback before I start getting to work for at least two
reasons: for one, technically I don't know the constraints and what people's
wish list for such a collection might include, and secondly getting people
involved and talking about it and giving their point of view makes it more
likely they will adopt it when it's done.



--
View this message in context: http://camel.465427.n5.nabble.com/Suggestions-for-new-set-of-integration-pattern-symbols-tp5789901p5790174.html
Sent from the Camel Development mailing list archive at Nabble.com.

Re: Time for an overhaul of the patterns symbols?

Posted by Tadayoshi Sato <sa...@gmail.com>.
Hi Francis,

I now fully understand your intention and personally like the idea. (I'm
one of the big fans of "flat design" :-) ) As no one else seems to have a
strong opinion about it, let me suggest one more idea.

As I said EIP diagrams are not proprietary to Camel, I think it's best to
provide your works to the broader EIP community. One of the great things
about EIP is the diagrams are provided by the authors as CC-BY license:
http://www.enterpriseintegrationpatterns.com/patterns/messaging/
so you can freely reuse and improve them so long as your works make the
original attribution clear.

So why don't you just go on your project and publish them on somewhere like
GitHub?  If your work is great and people in the EIP community like it,
then I'm sure your work will prevail!

Best regards,
Tadayoshi

On Thu, Nov 10, 2016 at 7:26 PM, Francis Vila <fv...@gmail.com> wrote:

> Hi Tadayoshi,
>
> Thanks for your feedback.
> Maybe I was hasty in implying there might be differences in style. Looking
> again I see the logic to the different forms in the patterns. Initially I
> saw differences in line widths and colors, differences in types of coloring
> (gradient vs flat, border vs no border, shadow vs no shadow), but these
> differences do seem to be justified by the semantics of the icons.
>
> What I really meant is that they look dated now in 2016. That's inevitable
> because people's expectations change very fast. For a technology that has
> such a future ahead of it, I think some investment in image is worthwhile.
>
> I looked for some sites offering icon sets :
> https://icomoon.io/#icons or search for icons on  Google
> <https://www.google.com/search?client=ubuntu&hs=gfo&
> channel=fs&biw=1376&bih=776&tbs=qdr:y&tbm=isch&q=
> information+technology+icon+set&oq=information+technology+
> icon+set&gs_l=img.3...25490.30110.0.30494.22.20.0.0.0.0.
> 92.824.19.19.0....0...1c.1.64.img..4.10.414...
> 0i7i30k1j0i7i5i30k1j0i8i30k1j0i5i30k1.5hSHYE1O_W0&bav=on.2,
> or.r_cp.&bvm=bv.138169073,d.d2s&dpr=1.4&gws_rd=cr&ei=K0ckWMWyLsS9aaGksYgG>
> (the icon sets appear black, but people using them can change their color)
>
> Most sets use either flat areas (no borders, single color) or lines with
> uniform width (giving the "twisted paperclip" feel). When gradients or
> shadows are used (mostly they are not), they are used uniformly throughout
> the icon set. Some sets are colorful or festive, but they mostly refer to
> Christmas or the like.
>
>
>
>
>
> --
> View this message in context: http://camel.465427.n5.nabble.
> com/Suggestions-for-new-set-of-integration-pattern-
> symbols-tp5789901p5789995.html
> Sent from the Camel Development mailing list archive at Nabble.com.
>

Re: Time for an overhaul of the patterns symbols?

Posted by Francis Vila <fv...@gmail.com>.
Hi Tadayoshi,

Thanks for your feedback.
Maybe I was hasty in implying there might be differences in style. Looking
again I see the logic to the different forms in the patterns. Initially I
saw differences in line widths and colors, differences in types of coloring
(gradient vs flat, border vs no border, shadow vs no shadow), but these
differences do seem to be justified by the semantics of the icons.

What I really meant is that they look dated now in 2016. That's inevitable
because people's expectations change very fast. For a technology that has
such a future ahead of it, I think some investment in image is worthwhile.

I looked for some sites offering icon sets :
https://icomoon.io/#icons or search for icons on  Google
<https://www.google.com/search?client=ubuntu&hs=gfo&channel=fs&biw=1376&bih=776&tbs=qdr:y&tbm=isch&q=information+technology+icon+set&oq=information+technology+icon+set&gs_l=img.3...25490.30110.0.30494.22.20.0.0.0.0.92.824.19.19.0....0...1c.1.64.img..4.10.414...0i7i30k1j0i7i5i30k1j0i8i30k1j0i5i30k1.5hSHYE1O_W0&bav=on.2,or.r_cp.&bvm=bv.138169073,d.d2s&dpr=1.4&gws_rd=cr&ei=K0ckWMWyLsS9aaGksYgG>  
(the icon sets appear black, but people using them can change their color)

Most sets use either flat areas (no borders, single color) or lines with
uniform width (giving the "twisted paperclip" feel). When gradients or
shadows are used (mostly they are not), they are used uniformly throughout
the icon set. Some sets are colorful or festive, but they mostly refer to
Christmas or the like.





--
View this message in context: http://camel.465427.n5.nabble.com/Suggestions-for-new-set-of-integration-pattern-symbols-tp5789901p5789995.html
Sent from the Camel Development mailing list archive at Nabble.com.

Re: Time for an overhaul of the patterns symbols?

Posted by Tadayoshi Sato <sa...@gmail.com>.
Hello,

Let me offer my thoughts. I think it's a good idea to somehow refresh the
EIP icons/diagrams and contribution is always great. However:

> the diagrams symbolize the essence of what Camel does.

Yes, but it also has a broader context: They actually symbolize the essence
of what most of the current competitive ESB or integration frameworks do in
common. They are symbols of our shared wisdom and best practices on systems
integration.

> However the graphics don't all seem to be in the same style, and some are
missing.

I haven't thought they aren't in the same style. Which one do you think
so?  Please note that in the original EIP book there are good reasons why
some icons are rectangles and some are not (and maybe why some are missing
originally as well). For example, in my understanding the Control Bus
pattern doesn't have a rectangle icon because in the context of the
original book it's rather an architectural pattern, while most rectangle
patterns are a component-level one.

Note also that some are missing just because they are devised (or
discovered) after the EIP book. So it might be a good idea to fill the gap
and devise a new icon for those new patterns, in the same style as in the
EIP book.

So, refreshing the icons is a good idea but I hope they won't diverge too
far from the original ones, because identity is the key for software
patterns in general.

Best regards,
Tadayoshi

On Tue, Nov 8, 2016 at 9:06 PM, Francis Vila <fv...@gmail.com> wrote:

> Hi,
>
> I am new to both Camel and to open-source contributing, so forgive any
> breaches of protocol...
>
> The page on  http://camel.apache.org/enterprise-integration-patterns.html
> <http://camel.apache.org/enterprise-integration-patterns.html>   shows an
> impressive list of patterns. However the graphics don't all seem to be in
> the same style, and some are missing. To me (although I don't have an
> in-depth understanding of Camel), the diagrams symbolize the essence of
> what
> Camel does. Maybe a homogeneous, updated styleset would be a good idea?
> I think I can contribute to such an task. I've done some tests of different
> approaches for the /Message router /pattern. If this effort is deemed
> worthwhile by the community, I can go on with the rest. The graphics were
> done using Inkscape and are available in svg format.
>
> This is the simplest version:
> <http://camel.465427.n5.nabble.com/file/n5789901/MessageRouter_1.png>
>
> This version has added endpoints symbolizing the input and output locations
> for the origin and destination of the messages :
> <http://camel.465427.n5.nabble.com/file/n5789901/MessageRouter_2.png>
>
> This version encloses the pattern:
> <http://camel.465427.n5.nabble.com/file/n5789901/MessageRouter_3.png>
>
> This version puts emphasis on the the most important part of the diagram,
> namely the currently active path:
> <http://camel.465427.n5.nabble.com/file/n5789901/MessageRouter_5b.png>
>
> This version targets the emphasis more closely, to just the "mobile" part
> of
> the diagram:
> <http://camel.465427.n5.nabble.com/file/n5789901/MessageRouter_6.png>
>
>
>
>
> --
> View this message in context: http://camel.465427.n5.nabble.
> com/Time-for-an-overhaul-of-the-patterns-symbols-tp5789901.html
> Sent from the Camel Development mailing list archive at Nabble.com.
>