Hello guys, we have done some progress on the new web site project. We need your comments for the design of the front page. We have 3 proposals or the design of the frontpage.
Please look at them, we need your help :)
http://qaweb.dev.centos.org/websitever2/
Best regards, Marian
Marian,
Marian Marinov wrote:
Hello guys, we have done some progress on the new web site project. We need your comments for the design of the front page. We have 3 proposals or the design of the frontpage.
Please look at them, we need your help :)
I could only see one wireframe:
http://qaweb.dev.centos.org/websitever2/node/40
(and an HTML skeleton at http://hydra.azilian.net/centos/centos.html).
Could you post the precise URLs of the 3 proposals? If they are somewhere then I cannot find them, I cannot make heads or tails of the navigation on that site.
Thanks Sebastiano Pilla
On Saturday 16 April 2011 09:36:45 Sebastiano Pilla wrote:
Marian,
Marian Marinov wrote:
Hello guys, we have done some progress on the new web site project. We need your comments for the design of the front page. We have 3 proposals or the design of the frontpage.
Please look at them, we need your help :)
I could only see one wireframe:
http://qaweb.dev.centos.org/websitever2/node/40
(and an HTML skeleton at http://hydra.azilian.net/centos/centos.html).
Could you post the precise URLs of the 3 proposals? If they are somewhere then I cannot find them, I cannot make heads or tails of the navigation on that site.
I'm sorry, I thought that they were publicly visible from the atrium site. Here are all of the proposals: http://hydra.azilian.net/centos/
1. http://hydra.azilian.net/centos/centos_frontpage_design_option_1.png 2. http://hydra.azilian.net/centos/wireframe1.png 3. http://hydra.azilian.net/centos/wireframe2.png
Marian
Thanks Sebastiano Pilla _______________________________________________ CentOS-devel mailing list CentOS-devel@centos.org http://lists.centos.org/mailman/listinfo/centos-devel
Marian Marinov wrote:
Here are all of the proposals: http://hydra.azilian.net/centos/
Thanks for the URLs, I would personally choose option 1 since it looks the most organized and structured to me.
I would however suggest to dedicate the 450x270 box to explain in a few words what CentOS is, with a link to the About page. See for example the Debian home page at http://www.debian.org or the Ubuntu home page at http://www.ubuntu.com , they present very well what Debian or Ubuntu are in a very prominent place.
Best Regards Sebastiano Pilla
One word... Ugly.
Maybe we should get some real website designers or artists to create some mock ups for a new centos site?
On Sat, Apr 16, 2011 at 8:16 AM, Sebastiano Pilla sebastiano@datafaber.netwrote:
Marian Marinov wrote:
Here are all of the proposals: http://hydra.azilian.net/centos/
Thanks for the URLs, I would personally choose option 1 since it looks the most organized and structured to me.
I would however suggest to dedicate the 450x270 box to explain in a few words what CentOS is, with a link to the About page. See for example the Debian home page at http://www.debian.org or the Ubuntu home page at http://www.ubuntu.com , they present very well what Debian or Ubuntu are in a very prominent place.
Best Regards Sebastiano Pilla _______________________________________________ CentOS-devel mailing list CentOS-devel@centos.org http://lists.centos.org/mailman/listinfo/centos-devel
On Sat, 2011-04-16 at 09:15 -0400, Chris wrote:
One word... Ugly.
Maybe we should get some real website designers or artists to create some mock ups for a new centos site?
CHRIS! Come one, that's some ones time and effort right there! As far as I am concerned, it's *progress*, and if it's the eye candy that annoys you, if I recall correctly CSS fixes 'ugly' websites right up.
I personally liked they eye candy, I can't vouch for content and structure, I took a quick look, but my first thought was "Great! That's a modern look".
Anyway, keep up the good work V2 website people :)
On Sat, Apr 16, 2011 at 8:16 AM, Sebastiano Pilla sebastiano@datafaber.net wrote:
Marian Marinov wrote:
Here are all of the proposals: http://hydra.azilian.net/centos/
Thanks for the URLs, I would personally choose option 1 since it looks the most organized and structured to me.
I would however suggest to dedicate the 450x270 box to explain in a few words what CentOS is, with a link to the About page. See for example the Debian home page at http://www.debian.org or the Ubuntu home page at http://www.ubuntu.com , they present very well what Debian or Ubuntu are in a very prominent place.
Best Regards Sebastiano Pilla
Yes, this is by far the most important thing that must always be on the front page of any project's web site. 1 or 2 sentences that says "CentOS is ..." [a Linux distribution that aims to be binary compatible with Redhat] ... or something like that. There's nothing more annoying than spending 5 minutes clicking around a project site trying to figure out what the project actually does.
// Brian Mathis
On Sat, 16 Apr 2011 13:01:27 +0300 Marian Marinov mm@yuhu.biz wrote:
On Saturday 16 April 2011 09:36:45 Sebastiano Pilla wrote:
Marian,
Marian Marinov wrote:
Hello guys, we have done some progress on the new web site project. We need your comments for the design of the front page. We have 3 proposals or the design of the frontpage.
Please look at them, we need your help :)
I could only see one wireframe:
http://qaweb.dev.centos.org/websitever2/node/40
(and an HTML skeleton at http://hydra.azilian.net/centos/centos.html).
Could you post the precise URLs of the 3 proposals? If they are somewhere then I cannot find them, I cannot make heads or tails of the navigation on that site.
I'm sorry, I thought that they were publicly visible from the atrium site. Here are all of the proposals: http://hydra.azilian.net/centos/
Hi Marian and all,
Thanks for posting those links. When I tried accessing the first batch I enabled in no script and unblocked with adblock plus and still got nothing, so I tried Opera and nothing there either. :-/
I'd vote for number 2. Number 1 looks like it would have large areas that rely on having flash enabled (slideshow) and I'm not a fan of that. Sometimes I'll land on a site that requires flash to see the page at all and if it's someplace I haven't been before I usually just go away. Not to mention TUV's desire to only include items that are 'free' vs 'proprietary' so flash isn't included in the sponsored repo's. I know it's 'what everyone is doing', but ...
I would also suggest variable width vs fixed width on whichever theme you end up with, if it's possible. I think fixed width is more prone to break in some cases. And I wanted to add a thanks to Yury that pointed to cmsmadesimple, I'd not heard of it before and I'm going to check it out.
Cia W
On Sat, 2011-04-16 at 08:08 -0700, Cia Watson wrote:
I'd vote for number 2. Number 1 looks like it would have large areas that rely on having flash enabled (slideshow) and I'm not a fan of that.
It has nothing to do with Flash per se. Modern slideshows are easily implemented in JavaScript (I bet jQuery already has a dozen of plugins for that alone) and degrade nicely to a single frame in the environments where JavaScript is disabled.
Sorry I am just going by the quality of work I am used to. Using an MVC framework things should be easily built and put online. As far as the layout goes it needs to be as high quality as the OS itself.
On Sat, Apr 16, 2011 at 11:45 AM, Yury V. Zaytsev yury@shurup.com wrote:
On Sat, 2011-04-16 at 08:08 -0700, Cia Watson wrote:
I'd vote for number 2. Number 1 looks like it would have large areas that rely on having flash enabled (slideshow) and I'm not a fan of that.
It has nothing to do with Flash per se. Modern slideshows are easily implemented in JavaScript (I bet jQuery already has a dozen of plugins for that alone) and degrade nicely to a single frame in the environments where JavaScript is disabled.
-- Sincerely yours, Yury V. Zaytsev
CentOS-devel mailing list CentOS-devel@centos.org http://lists.centos.org/mailman/listinfo/centos-devel
Am 16.04.11 23:48, schrieb Chris:
Sorry I am just going by the quality of work I am used to. Using an MVC framework things should be easily built and put online. As far as the layout goes it needs to be as high quality as the OS itself.
Was that an offer for helping?
Ralph
Yes that was an offer to help centos ;)
On Mon, Apr 18, 2011 at 5:57 PM, Ralph Angenendt ralph.angenendt@gmail.comwrote:
Am 16.04.11 23:48, schrieb Chris:
Sorry I am just going by the quality of work I am used to. Using an MVC framework things should be easily built and put online. As far as the layout goes it needs to be as high quality as the OS
itself.
Was that an offer for helping?
Ralph _______________________________________________ CentOS-devel mailing list CentOS-devel@centos.org http://lists.centos.org/mailman/listinfo/centos-devel
Hi!
On Sat, 2011-04-16 at 03:12 +0300, Marian Marinov wrote:
we have done some progress on the new web site project. We need your comments for the design of the front page. We have 3 proposals or the design of the frontpage.
Unfortunately, I can't comment on Atrium, but there is something I can say out of the top of my head regarding the CMS.
Unfortunately, the review you have on Atrium is incomplete and somehow lacks consistent discussion of pro's and con's of each possible choice, so I can't see how you can make a non-random decision here, which to my mind is quite important to begin with.
Let me try to contribute to your effort. I guess three possibilities you should be considering:
1) Version control based static generator, i.e. jekyll or hyde 2) All-in-one CMS stack for typical SMB websites (Drupal, CMSMS) 3) Framework-based website (i.e. Django + Django-CMS etc.)
== Static generators
Those are really nice and easy to use. Basically your whole website represents a directory tree in a version control system of your choice (git, of course! ha-ha, ok, you might use any another DVCS if you really want to), where texts are written in some lightweight markup language, such as Markdown plus an extra directory with templates.
On every push, the generator will reconstruct the website on the server and upon successful processing switch the symlink to the new directory.
The pro's of course are that once it's setup it's easy as in "easy":
+ Your texts can be later exported or imported in whatever way + Resource consumption is truly minimal (static pages) + The web server doesn't need maintenance: static pages can't be hacked + All the texts are version-controlled by definition + Collaborating with team and external contributors is a bliss (think GitHub and pull requests) + Every member of the group has a backup of the history of the site (plus you can set up additional git mirrors)
The con's are also quite obvious:
- This only works for mostly non-interactive heavily text-based websites where you don't need server-side scripts; so no feedback forms, shops, searches, dynamic calendars etc., sorry! (you can pre-generate stuff, like "related pages", tag cloud etc., but it's limited)
- You can't add a comment system unless it's something like Disqus, but then mind you that you are outsourcing your comments to third parties (no backup, unless you do it via API, they can die at any moment)
- You can generate RSS feeds from the content that you have on every push, but if you want to aggregate you will need some kind of script to auto-commit stuff to your repo and repush the website; maybe better just to install a separate planet for this (well, you have it already)
So the crucial point to decide upon whether it's a viable strategy for you or not would be to consider how do you actually see the website that you are about to build:
* What kind of information it will contain? * Is this information going to be mostly static or dynamic? * Are you willing to accept contribution from outside world?
== Content management systems for SMB
This is the most common choice nowadays, which comes with pro's and also with con's (which most somehow don't even consider):
+ Generally free and easy to install & use (Drupal, CMSMS) + Administrative interface is mostly familiar to anyone you ask + Able to built any tree-like website structure as the generators + Able to augment the website with dynamic functionality (ratings, comments, calendars, user profiles etc.)
+/- Content versioning can be added via plug-ins +/- Static generation (cacheing) can be added via plug-ins
- Requires a scripting language interpreter and a database - Consequently, needs much more ressources under high load - You will need to keep the CMS up-to-date, and I really *mean* it
So basically, in addition to the first variant it adds extra administrative overhead, proprietary way of storing texts (in the database) and inconvenient versioning system.
On the other hand, you can easily link any additional dynamic functionality by installing ready-made modules or writing your own scripts / snippets (even in the template language).
== Framework-based website
I would say, this option should only be considered is there's a web developer among you that can sacrifice his time to develop (and maintain in the future) a tailor-made Django instance for CentOS.
The pro's of course are that it comes without the bloat, security and performance problems of a swiss-knife type CMS, because it's tailor-made to the project, but the con's are that it requires rather high qualification to develop and maintain if you want it to be done properly (which you do, trust me).
= Conclusion
If I were you, I would have seriously considered static generators and Drupal vs. CMSMS [1]. I haven't seen you even mentioning the last one, but I think you really should have had a look at it. It's one of the few free CMS systems for SMB that I personally would actually consider ok.
(Drupal is bloated to my liking, but you have to admit that it's flexible and gets the job done if you invest into it, Joomla I loathe with a passion for its brain-dead Nuke-alike design and bloat, modx might be ok, but I haven't used it so far).
CMSMS is a rather small (compared to teh monstrous Drupal) content management system, well maintained, often updated, easy to use & administrate and quite flexible.
It has a lot of batteries included, but the most notable feature is the prominent use of Smarty (maybe not *the* most brilliant templating language, but it's quite ok) across the whole CMS, which basically makes it able to morph into what you'd like to see without any extra PHP coding or touching the kernel & subsequent problems on upgrades related to this (which is why many administrators I know always fear to upgrade their Drupal instances).
I can set up a demo if you want, but I am under impression, that to the people on this list such an offer would sound like an insult, considering their qualifications :-)
[1]: http://www.cmsmadesimple.org/
This post is under CC-BY. Feel free to add it to the Atrium and make any use out of. Hope this helps.
Dne 16.4.2011 12:16, Yury V. Zaytsev napsal(a):
- Version control based static generator, i.e. jekyll or hyde
- All-in-one CMS stack for typical SMB websites (Drupal, CMSMS)
- Framework-based website (i.e. Django + Django-CMS etc.)
Yury, Wow, I see you have plenty of time :o). Great post. As to CMS, framework, static pages, etc. This is something very hard for me to decide. I don’t know actual page statistics on www.centos.org. This point is very important to consider. So we need to know actual traffic to seriously start thinking on technical part of implementation. You may be considering Drupal, but the actual traffic might be so high, that Drupal without cache, memcached, etc. won't be able do deliver pages smoothly and quickly. As to the three proposed designs. Sorry guys don't take it personally, I don't like either of them. DH
I will offer my services to the centos team to help give back to the community. I am the creator of the zend framework. Whoever is leading the website coding and design end of centos please send me an email if you want any additional help. On Apr 18, 2011 2:01 PM, "David Hrbáč" david-lists@hrbac.cz wrote:
Dne 16.4.2011 12:16, Yury V. Zaytsev napsal(a):
- Version control based static generator, i.e. jekyll or hyde
- All-in-one CMS stack for typical SMB websites (Drupal, CMSMS)
- Framework-based website (i.e. Django + Django-CMS etc.)
Yury, Wow, I see you have plenty of time :o). Great post. As to CMS, framework, static pages, etc. This is something very hard for me to decide. I don’t know actual page statistics on www.centos.org. This point is very important to consider. So we need to know actual traffic to seriously start thinking on technical part of implementation. You may be considering Drupal, but the actual traffic might be so high, that Drupal without cache, memcached, etc. won't be able do deliver pages smoothly and quickly. As to the three proposed designs. Sorry guys don't take it personally, I don't like either of them. DH _______________________________________________ CentOS-devel mailing list CentOS-devel@centos.org http://lists.centos.org/mailman/listinfo/centos-devel
Am 18.04.11 20:00, schrieb David Hrbáč:
Dne 16.4.2011 12:16, Yury V. Zaytsev napsal(a):
- Version control based static generator, i.e. jekyll or hyde
- All-in-one CMS stack for typical SMB websites (Drupal, CMSMS)
- Framework-based website (i.e. Django + Django-CMS etc.)
Yury, Wow, I see you have plenty of time :o). Great post. As to CMS, framework, static pages, etc. This is something very hard for me to decide. I don’t know actual page statistics on www.centos.org. This point is very important to consider. So we need to know actual traffic to seriously start thinking on technical part of implementation.
At the moment Xoops can cope, but I have no idea how effective that is (and if it does page caching somewhere). But yes, I'd like to keep it on one box, if possible :)
Yury: I need to read your mail in a more silent moment, thanks for spending your time to explain those options.
Regards,
Ralph
On 4/16/2011 5:16 AM, Yury V. Zaytsev wrote:
Let me try to contribute to your effort. I guess three possibilities you should be considering:
- Version control based static generator, i.e. jekyll or hyde
- All-in-one CMS stack for typical SMB websites (Drupal, CMSMS)
- Framework-based website (i.e. Django + Django-CMS etc.)
What's the point of having anything but a wiki, forum, and bug tracker? Is someone with special expertise going to create things that most people won't know how to change? Is there content that can't live in a wiki?
Am 19.04.11 00:29, schrieb Les Mikesell:
What's the point of having anything but a wiki, forum, and bug tracker? Is someone with special expertise going to create things that most people won't know how to change? Is there content that can't live in a wiki?
Yes. Our wiki at least cannot import things like RSS feeds. Same goes for sponsor's banners and a few other things.
Plus: I have no idea if the wiki can keep up with the hit rates the web site has.
http://lists.centos.org/pipermail/centos-devel/2011-February/006674.html has a few more reasons.
Ralph
On 4/18/2011 5:47 PM, Ralph Angenendt wrote:
Am 19.04.11 00:29, schrieb Les Mikesell:
What's the point of having anything but a wiki, forum, and bug tracker? Is someone with special expertise going to create things that most people won't know how to change? Is there content that can't live in a wiki?
Yes. Our wiki at least cannot import things like RSS feeds. Same goes for sponsor's banners and a few other things.
What the current wiki can/can't do shouldn't be a basis for ruling out using some wiki that does what you want or that has a plugin mechanism where you can add that functionality. For example this (very large) list might cover your needs: http://www.mediawiki.org/wiki/Extension_Matrix/AllExtensions
I have more experience with twiki myself but don't think I'd want to run it on a public site.
Plus: I have no idea if the wiki can keep up with the hit rates the web site has.
So how do you know that about the other products being considered? If you use some other large site using it as evidence, isn't wikipedia good enough? Besides, a reverse-proxy with squid/apache/nginx can fix any load problem you might have with stuff that looks mostly-static from the viewer's side.
http://lists.centos.org/pipermail/centos-devel/2011-February/006674.html has a few more reasons.
It's not clear there as to why the content proposed for the 'website' wouldn't be just as appropriate on a wiki (assuming different access permissions), making one less thing to maintain. I could see why you might want one very pretty home page without the wiki style, but you don't need a whole CMS for that and all of the links could lead into the wiki sections.
On Tue, 19 Apr 2011 00:47:00 +0200 Ralph Angenendt ralph.angenendt@gmail.com wrote:
Am 19.04.11 00:29, schrieb Les Mikesell:
Yes. Our wiki at least cannot import things like RSS feeds. Same goes for sponsor's banners and a few other things.
Plus: I have no idea if the wiki can keep up with the hit rates the web site has.
Just a note ref. traffic to the web site proper (centos.org). I know that when one visits the forums and wants to log in it will then bring you to the main landing page after logging in; and login isn't persistent so one has to log in every time one visits.
I'm sure that accounts for a lot of the traffic. Being able to log in and have it set a cookie so one doesn't have to log in every time you visit the forums would be a good thing. And/or have it drop you on the forum index / landing page instead of the centos.org page after logging in would also reduce traffic to the main site. CentOS is the only forum I visit (used to visit) frequently that doesn't remember me from day to day. fwiw.
Hello All,
Thanks in advance for all the feedback, it's all been very constructive. I am just going to try and address everything up to now in this thread. Sorry if this email gets a bit long.
Just some quick back-story. Please see Ralph's original thread "Website Version 2, next steps" [1] He concluded with three main tasks: 1. Go over existing web content (not in wiki or forums) 2. Unified Authentication design, centering around LDAP 3. Feedback or suggestions for the software to power the web site.
Since that thread began, an IRC meeting was held [2]. Official notes of that meeting should be put up somewhere, but here's what I can remember (others in attendance, chime in if I miss something):
1. Community members who are actively trying to help with the web site tasks agreed to use the Atrium[3] web site for the project management. 2. The first wireframe[4] was introduced. This was meant as a visual tool to look at the how the existing web pages[5] could be revised or re-ordered. It was also hoped that it would spur some additional designs and layout work. Additional designs have been submitted on the Atrium web site. 3. It was asked if the web site could be directly powered by the wiki software. There were enough issues to recommend not using the existing wiki. We also wanted to keep the scope small, so switching wiki software was not considered either. 4. At the time, only two options for powering the web site were brought up, version controlled html/css and Drupal. There were at least a few who seemed familiar enough with Drupal to suggest giving it a try. No concrete decision has been made about the software. Just some avenues to pursue.
And finally, after a couple of us did some additional work on the Atrium website, we wanted to post what was going on to the mailing lists. Here we are.
Here's my take home so far:
If there are any technical requirements for the web site, the should be defined and documented. I think there are a few I know of, so I will post that as a discussion point on the Atrium website.
The software stack is still up in the air. I personally believe there will be more then one viable solution. If that's the case, having a champion or some expertise in the group could end up being the deciding factor. That said, I gave my non-technical view of why it's worth trying out Drupal here[6]
There's more then a few people who seem willing to provide some input, opinions, suggestions, less then gentle critiques ;-). And that is all great, and exactly why the Atrium web site and mailing list discussions are so important. The project should be transparent. But we are going to need more. If this is something you are interested in, register on the Atrium account and be ready to contribute. In my mind, this project should be completed to coincide with CentOS 6. Now, keep the CentOS 6 release remarks out of this, and think about how long that should take. If we are going to hit that deadline, we will need all the help we can get.
Thanks,
Jim
[1] http://lists.centos.org/pipermail/centos-devel/2011-February/006674.html [2] http://lists.centos.org/pipermail/centos-devel/2011-April/007288.html [3] http://qaweb.dev.centos.org/websitever2/ [4] http://qaweb.dev.centos.org/websitever2/node/40 [5] http://qaweb.dev.centos.org/websitever2/node/41 [6] http://qaweb.dev.centos.org/websitever2/node/42
On 19 April 2011 00:15, Jim Woods redkilian@gmail.com wrote:
Hello All,
Thanks in advance for all the feedback, it's all been very constructive. I am just going to try and address everything up to now in this thread.
I have been watching this unfolding with interest. I have been using various different methods of trying to tie in various components under a unified website (mainly website, wiki, forum and help-desk) so the conversation on V2 of the CentOS website has been very useful to lurk in.
May i ask that once the process has completed and the decisions have been implemented could the packages used be made available in extras or CentOSplus. regards
mike
-
Brian Mathis -
Chris -
Cia Watson -
David Hrbáč -
Jim Woods -
Les Mikesell -
Marian Marinov -
Michael Simpson -
Mister IT Guru -
Ralph Angenendt -
Sebastiano Pilla -
Yury V. Zaytsev