Skip to main content
Welcome guest. | Register | Login | Post

The future of GetGNULinux.org

33 replies [Last post]
ariadacapo's picture
Offline
Joined: 2006-07-13

Hello everyone,

Gustavo sent me an extremely interesting email in which he gives his views on the future of GGL.o
It's a pdf file you can download on the temp site.

I must say that I'm very enthusiastic about it - there are some parts I'm not sure about and some parts which definitely get me going.
Immediately I have very little time to comment properly but I will do so very soon, so it's not rushed.

I invite you to read through it ; please express your thoughts if you feel like it.

This is closely related to the thread on GGL.o and money which I have started out recently, but I have opened a different thread since this is quite broader.

Thank you gustavo for this... I will express myself soon.

libervisco's picture
Offline
Joined: 2006-05-04
It looks like Gustavo

It looks like Gustavo really gave this some time and thought it through. There are some really great ideas there which if implemented successfully would make GGL a prime example of a GGL advocacy organization, something I'm not sure exists at all yet.

Of course, to implement this fully there would have to be a good deal of people interested in getting actively involved, but something tells me that actually wont be a big problem considering the interest shown through this forum so far (you already have translators for example).

What Gustavo suggests is so broad, actually, that it covers some of the areas already covered to a point by others. He even mentioned LQs HCL and expanded on that idea for GGL. Another thing that might end up being a bit "outdated" by this new GGL are Nuxified.org forums, if GGL help center ends up working better for people.

In any case I'd be willing to get involved in some part of this. Assembling a Croatian translation team is one idea.

Another part where I could get involved, along with Nuxified.org as a whole is potentially working on having the GGL help center developed and hosted on Nuxified.org in which case there might be a sort of a merge between two projects, or just a continuous relationship based on one we already have, a cooperative partnership.

That's as much as I have to say at this point. I'm looking forward to comments of others.

Gustavo's picture
Offline
Joined: 2006-09-11
Great!

Hi, Olivier and Libervisco.

First of all, thank you Olivier for publishing this pdf.

I'm glad you both liked the document.

libervisco wrote:

It looks like Gustavo really gave this some time and thought it through.

Well, I've been with a little piece of paper in my pocket for a week in order to write down any idea that comes into my mind. But I started the document 3 days ago.

libervisco wrote:

Another thing that might end up being a bit "outdated" by this new GGL are Nuxified.org forums, if GGL help center ends up working better for people.
(...)
Another part where I could get involved, along with Nuxified.org as a whole is potentially working on having the GGL help center developed and hosted on Nuxified.org in which case there might be a sort of a merge between two projects, or just a continuous relationship based on one we already have, a cooperative partnership.

I like that idea. Moreover, if you don't mind, NXFD might be the Help Center itself, unless it's implemented inside GGL.

libervisco wrote:

That's as much as I have to say at this point. I'm looking forward to comments of others.

So do I.

Cheers!

tbuitenh's picture
Offline
Joined: 2005-12-21
Instead of creating a

Instead of creating a separate help center, I suggest requesting the features you want in a help center as improvements to nuxified. It's better to have one really good help site with a big community than it is to have two smaller less perfect sites. If people have questions, they shouldn't have to wonder if the question is more suitable for GGL or NXFD: they want answers, not more questions!

Gustavo's picture
Offline
Joined: 2006-09-11
That's what I mean
tbuitenh wrote:

If people have questions, they shouldn't have to wonder if the question is more suitable for GGL or NXFD: they want answers, not more questions!

That's what I mean. In the end, we should only suggest one site for people to look for help.

Cheers.

Offline
Joined: 2006-10-30
Hey guys, I have read the

Hey guys,

I have read the document Gustavo wrote about the future of GGL and want
to share my thoughts for future modifications.

Hardware Compatibility DB → Linux-Drivers is the first place someone could go (LQ's HCL is also there). A translation of that DB would be a really huge work, and even a project like that or Linuxprinting is currently maintained almost entirely by one person.

tbuitenh wrote:

It's better to have one really good help site with a big community than it is to have two smaller less perfect sites.

The main problem comes with distribution-specific questions, how can that be solved?

Joining efforts/Splitting up the work is the best way to go. This idea can be applied to all proposals.

Gustavo wrote:

"Actually, it is not necessary that your computer store provides you with driver CDs
when you buy a brand new computer"

Ideally everyone would install an OS without knowing that, for example, when you buy a ups the manufacturer provides an interface of his own.

Libervisco published something on another thread that should be reflected on the main site.

Open Drivers → if the manufacturer abandons a software, the user could be left with a brick and no support.
If source code is available, users of that software will be able to run it without troubles. Even other people could add more features that the original designer did not think.

Component Requirements → There is a great article in Free Electrons. I'm a computer user and components cost money, what cand I do about it? That slideshow explains that matter.

Tutorials/PreInstall → Olivier mentioned this, too (ubuntuvids, maybe?). There is a great work available on Ubuntuclips, video guides about how things are done for newcomers.

The idea of a self-profitable project is right, text ads or donations sound as the best options.

Greets.

Gustavo's picture
Offline
Joined: 2006-09-11
Alright

Hi, Miguel.

Kuan wrote:

Hardware Compatibility DB → Linux-Drivers is the first place someone could go (LQ's HCL is also there).

If they know English, they'll have to follow several links in order to find out if their hardware is supported; then repeat the process for each piece of hardware. I suggest to use an HTML form with the following fields: Hardware type (select), vendor (select) and model (text/optional), instead of following several links.

On the other hand, if they don't know English... well, that's another story.

Kuan wrote:

A translation of that DB would be a really huge work, and even a project like that or Linuxprinting is currently maintained almost entirely by one person.

Translating such databases is not an issue, IMO. We only have to translate the hardware type: It's not necessary to translate the vendor nor the model name.

Kuan wrote:
tbuitenh wrote:

It's better to have one really good help site with a big community than it is to have two smaller less perfect sites.

The main problem comes with distribution-specific questions, how can that be solved?

That's why I said that, once the issue is solved, the staff should mark it as distro-independent or distro-specific, in case this datum is relevant.

Kuan wrote:

Tutorials/PreInstall → Olivier mentioned this, too (ubuntuvids, maybe?). There is a great work available on Ubuntuclips, video guides about how things are done for newcomers.

In English, once again. Also, it's only for *ubuntu.

If I had to summarize my proposal:

  • People must get their Linux OS up and running without googling so much. This is, they should not waste much time to do simple things, because unfortunately, this is something usual in Linux.
  • GGL must provide Linux newcomers with all the information they'll ever need in their own mother tongue, and in case we don't count with an specific information (yet), we should try to tell them where they might find a solution.

Once Linux have these non-technical features around, it will be the world's most used operating system.

Cheers.

PS: If the HC comes true, it would be great that we reported every bug found for it to be solved ASAP.

libervisco's picture
Offline
Joined: 2006-05-04
One way to resolve the

One way to resolve the issue of distro-specific and distro-independant issues is to tag topics as such. Nuxified.org already has tagging implemented so we only need to implement some tagging rules for moderators to make sure that topics are tagged properly.

Once we have the tagged we can easily provide links per those tags. For example, if you need a distinction between distro-specific and distro-independant topics a simple page with two of links to the corresponding tag would do it, and then in the distro-specific page same links would go for various distros.

With proper use of tags, categorizing things can be really made easy, and we already have a base for that on Nuxified.

ariadacapo's picture
Offline
Joined: 2006-07-13
An answer

Here are my thoughts on your thoughts Gustavo...
First I'd like to say I like the whole lot very much =) and thank you very much for speaking out. This is very constructive.

I will answer with my point of view, so with a different order than that you presented, but still try to address everything.

- Splitting up the work

This is by far the most important aspect of the future of GGL, IMO. Reading through the well-built document, I recognised many projects, ideas, prospects, etc that I had thought about and written down on my notebook already. The reason they are not started yet is... I do not have the 48 hours per day that I wish I had.
I am very, very willing to open up the development of GGL.o to others. I strive to do this and believe me, however poor the result, it takes a lot of time (setting up translation wikis for anyone asking, keeping a detailed changelog and to-date download file, discussing just every thought out in the forum, etc). Still so far no-one gets to write a page or anything on the site! Something has to change.
The danger is really that too much open-ness is not desirable. The best example I can think of is the Wikipedia Linux article, which is subject to continual disputes, and in the end turns off any potential newcomer. Judging only by the many suggestions I've received by email, if GGL.o was a fully-open wiki, it would not be successful at all.
This being said, we must install something to enable better management than now. Right at the start, when GGL.o hit digg in its first week and I was contacted by Danijel, I was thinking of a project-management system, and in the end we just went for this forum (rightly so, what a huge progress this has been so far!). Then later I set-up the develop wiki but it's not working very well, I believe, because of lack of "dynamism", priority sorting, and other things.

So, with a good project management system, like a bug tracking system, we can build many things much more efficiently.
The requirements for one are, IMO:
- the ability to prioritize tasks/to-dos/bugs
- the ability to add bugs/feature requests very easily for non-registered users
- the ability to build teams (basically permissions)
- (maybe) the ability to use SVN
I will give activecollab a try again, does anyone have proposals for other systems?
The tricky bit is to give members of the tracking system the means of actually doing something. This means access to the website files directly, so running in parallel with SVN (as the Spanish team has now developed) would be a real plus.

About the team organisation:
Well you have built a structure for a whole corporation! =) It's a nice structure but in my opinion we don't need a team such as QA. From my experience (I have worked in such a department in a car factory) it will be highly out of place in any organisation smaller than 100 people, and even then it's a highly delicate task improving quality without being a pain in the butt of everybody, inspecting whatever's wrong with what they do. The same thing applies to PR (we are basically all PR since GGL is an advocacy organisation and has its goals open).
Otherwise, then sure! teams sound good =) I however disagree with the "splitting of money", but more on that later on.

- The GNU/Linux newcomer center

The Pre-install, Install, Post-install sections sound brilliant. I had something like this in mind when I wrote the "Five things to know" article which is now on PolishLinux.org. We could mostly present these sections as to answer the "show me how it's done" feelings of newcomers.
I have doubts about the success of an Issue database, however. The task is just simply huge and the benefits would be rather small... let me explain.
There are two types of issues a newcomer could encounter.
The first is a common problem. For example, "how do I transfer my email from Outlook / Thunderbird on Windows, to Thunderbird on GNU/Linux?". This would work well and browsing such a database could really help a newcomer.
The second are very specific issues, like "I can't find my usb camera anymore". This is likely to be caused by a variety of factors, so usually
a) it's much more easily solved in a forum, where more advanced members can question further, give instructions, etc, this is what makes Nxfd so great!
b) It's not really relevant to many people, since the specific solution will not necessarily solve the problems of others.
So, keeping in mind that the amount of issues encountered everyday by GNU/Linux newcomers would probably keep 20 people busy on full-time, not withstanding translating each problem/solution, this sounds very ambitious.
In my opinion, partnering even further with Nxfd would be more efficient, as suggested by tbuitenh. I keep in mind that had GGL.o started with a standalone forum instead of benefiting from the hosting ("hosting" in the non-technical term) at Nuxified, then by no means would we have such a well-going project.
The same thing goes for the hardware database, which could be difficult to maintain.

But in any case, building a newcomer center where Windows users can be comforted with a detailed pictured description of, say, how a dual-boot is set-up, would be fantastic. Unfortunately I do not have the energy to build such a thing myself (I'd rather concentrate on improving the original GGL), but with a good project management, this is no problem at all =)

- A GGL.o planet website

I was already thinking of merging different places and pages into a single one: the "help us" page (which indeed needs to be more specific), the changelog, download page, the translation page, everything into one central community.getgnulinux.org place, where one would have the ability to oversee what is going on and find necessary resources.
This could go well together with your "planet GGL" idea... we could feature blog posts if we wanted (I very much welcome this, though I'm not a blogger myself), have a feed from the latest GGL forum posts, etc. This sounds very interesting. A very dynamic page would be required, ex something run by Drupal.
[An idea could be to also work together with blogs.libervis.com, so that any post on a libervis blog that has a GGL tag would display there.]
I agree that by now the site is too static. This is understandable (I wasn't expecting a need for anything else when I started, the thought was if I had to abandon GGL, it'd be better to have a static website than a dead dynamic one) but of course it needs to change.

- Money

First your proposals for generating revenue fit in nicely with what I had in mind... I would add: Partner with distribution CD vendors (the way distrowatch does it). For many people still (including me =) downloading even a CD is difficult.

As I expressed above, I disagree with the sharing of the profit in the way you describe it. I think it's pretty dangerous.
My experience with money and organisations in general (all in the physical world, large and small alike) is that it wouldn't work. Distributing money according to membership in a team is the wrong incentive. For example, there might be a lot more work to do in a team than the percentage attributed to it. Or one member might do most of the work in a team. Or people would enter teams mostly for the hope of profit. Or many other possibilities for it to go wrong.
Now I'm not saying: "I'm keeping it to myself" Smiling I simply think the most widely accepted and reliable way to distribute money is according to work time. My proposal is to first start and see how it goes. Maybe the site can pay for itself for a start (around 100/150$ a year). After that maybe we can constitute a small budget, say another €100, and then consider together the best way to spread it. The best solution in my opinion, would be that of "assignments" where projects could be proposed with detailed objectives and budget, and say €100 could be given once it is finished.
In any case before the first euro benefit is made, I want to have a policy well-laid and possibly some sort of official body set-up (see GGL and money thread). Some sort of transparency (ability for members to inspect account) in the revenues would be good.

- All other proposals for improvement

(all the "smaller" other ideas for improvement)
Unfortunately I run out of time again here.... But there are mostly very good ideas you suggested... I'll come back to this. I just nodded my way through the such a large number of suggestions!

Overall, I am mostly interested in the creation of a project management system, which if well-constructed, shall enable most of the other changes you propose, without having to go through my slow participation every time. This is the foremost "bug" in GGL.o that has to be solved Smiling
Thank you so much for that and I shall be back very soon (I also try to complete the French translation at the same time... and studies as always...)

Olivier.

Gustavo's picture
Offline
Joined: 2006-09-11
In response to Olivier

Hi, Olivier!

ariadacapo wrote:

- Splitting up the work
(...)
The danger is really that too much open-ness is not desirable. The best example I can think of is the Wikipedia Linux article, which is subject to continual disputes, and in the end turns off any potential newcomer. Judging only by the many suggestions I've received by email, if GGL.o was a fully-open wiki, it would not be successful at all.

That's absolutely right: Even the Spanish translation wiki received a lot of spam comments, among other things. Fortunately, we don't need it anymore because we have SVN.

ariadacapo wrote:

About the team organisation:
Well you have built a structure for a whole corporation! =) It's a nice structure but in my opinion we don't need a team such as QA. From my experience (I have worked in such a department in a car factory) it will be highly out of place in any organisation smaller than 100 people, and even then it's a highly delicate task improving quality without being a pain in the butt of everybody, inspecting whatever's wrong with what they do.

Well, I think the QA team is not really necessary in the end.

ariadacapo wrote:

The same thing applies to PR (we are basically all PR since GGL is an advocacy organisation and has its goals open).

We are, as a whole, a PR team, but it doesn't mean that we all, as individuals, wanted to talk/act in behalf of GGL. Hence I propose to delimit this group of people that is willing to be the GGL's mouth and ears.

This might not make much sense in the mean time because GGL is still a new and small project, but take into account that GGL is always growing.

ariadacapo wrote:

I have doubts about the success of an Issue database, however. The task is just simply huge and the benefits would be rather small... let me explain.
There are two types of issues a newcomer could encounter.
The first is a common problem. For example, "how do I transfer my email from Outlook / Thunderbird on Windows, to Thunderbird on GNU/Linux?". This would work well and browsing such a database could really help a newcomer.
The second are very specific issues, like "I can't find my usb camera anymore". This is likely to be caused by a variety of factors, so usually
a) it's much more easily solved in a forum, where more advanced members can question further, give instructions, etc, this is what makes Nxfd so great!
b) It's not really relevant to many people, since the specific solution will not necessarily solve the problems of others.
So, keeping in mind that the amount of issues encountered everyday by GNU/Linux newcomers would probably keep 20 people busy on full-time, not withstanding translating each problem/solution, this sounds very ambitious.

I now agree the Issues DB is not necessary either and we're better off using a web forum.

However, taking into account your comments, it would be great that we do this:
* For common doubts of Linux newcomers: We should have an small and useful article per common doubt, translated into each supported language.
* For hardware problems: In the supported hardware DB, each entry should also have links to those threads in our forum that are related to this hardware piece. This isn't hard to do.
* For weird/other problems: We should use a web forum, available in each of the supported languages.

ariadacapo wrote:

In my opinion, partnering even further with Nxfd would be more efficient, as suggested by tbuitenh. I keep in mind that had GGL.o started with a standalone forum instead of benefiting from the hosting ("hosting" in the non-technical term) at Nuxified, then by no means would we have such a well-going project.

I think it's a good idea, but then we have elaborate on the NXFD's role in this project. I mean, we need to identify the goals of getgnulinux.org and nuxified.org in this effort.

We might apply the approved proposals to NXFD and keep getgnulinux.org as a brief informative web site about the advantages of Linux.

ariadacapo wrote:

The same thing goes for the hardware database, which could be difficult to maintain.

I think it would be difficult to set-up and extremely easy to maintain, actually, what do we have to do once it's running?

Once again, this would be a one-time effort for a long-term benefit, IMO.

ariadacapo wrote:

But in any case, building a newcomer center where Windows users can be comforted with a detailed pictured description of, say, how a dual-boot is set-up, would be fantastic.

Mac users should also be taken into account, IMO.

ariadacapo wrote:

- A GGL.o planet website

I was already thinking of merging different places and pages into a single one: the "help us" page (which indeed needs to be more specific), the changelog, download page, the translation page, everything into one central community.getgnulinux.org place, where one would have the ability to oversee what is going on and find necessary resources.
This could go well together with your "planet GGL" idea... we could feature blog posts if we wanted (I very much welcome this, though I'm not a blogger myself), have a feed from the latest GGL forum posts, etc. This sounds very interesting. A very dynamic page would be required, ex something run by Drupal.
[An idea could be to also work together with blogs.libervis.com, so that any post on a libervis blog that has a GGL tag would display there.]
I agree that by now the site is too static. This is understandable (I wasn't expecting a need for anything else when I started, the thought was if I had to abandon GGL, it'd be better to have a static website than a dead dynamic one) but of course it needs to change.

I agree with you here.

ariadacapo wrote:

First your proposals for generating revenue fit in nicely with what I had in mind... I would add: Partner with distribution CD vendors (the way distrowatch does it). For many people still (including me =) downloading even a CD is difficult.

It's a good idea, IMO.

ariadacapo wrote:

As I expressed above, I disagree with the sharing of the profit in the way you describe it. I think it's pretty dangerous.
My experience with money and organisations in general (all in the physical world, large and small alike) is that it wouldn't work. Distributing money according to membership in a team is the wrong incentive. For example, there might be a lot more work to do in a team than the percentage attributed to it. Or one member might do most of the work in a team. Or people would enter teams mostly for the hope of profit. Or many other possibilities for it to go wrong.
Now I'm not saying: "I'm keeping it to myself" Smiling I simply think the most widely accepted and reliable way to distribute money is according to work time. My proposal is to first start and see how it goes. Maybe the site can pay for itself for a start (around 100/150$ a year). After that maybe we can constitute a small budget, say another €100, and then consider together the best way to spread it. The best solution in my opinion, would be that of "assignments" where projects could be proposed with detailed objectives and budget, and say €100 could be given once it is finished.
In any case before the first euro benefit is made, I want to have a policy well-laid and possibly some sort of official body set-up (see GGL and money thread). Some sort of transparency (ability for members to inspect account) in the revenues would be good.

I think your approach is the way to go.

By the way, what do you think about implementing "embassies" the way I said?

Cheers!

libervisco's picture
Offline
Joined: 2006-05-04
About the project

About the project management system maybe you could take a look at trac which I've seen used on quite a few sites. It seems to have all of the features you mentioned, including an issue tracker (basically a bugzilla) and interfaces to version control systems.

Olivier wrote:

This could go well together with your "planet GGL" idea... we could feature blog posts if we wanted (I very much welcome this, though I'm not a blogger myself), have a feed from the latest GGL forum posts, etc. This sounds very interesting. A very dynamic page would be required, ex something run by Drupal.

Indeed, and I just want to note that Drupal also has modules for project management. If you decide to use Drupal then it could cover both the project management and a dynamic "planet" section at once. This way you don't need to maintain two separate web systems installations.

Olivier wrote:

[An idea could be to also work together with blogs.libervis.com, so that any post on a libervis blog that has a GGL tag would display there.

At this point we don't have advanced tagging on blogs.libervis.com, but that site is still awaiting an overhaul and an upgrade. However, it would be quite simple to just open an official GetGNULinux.org blog there and feed its RSS to the planet on GGL.

Gustavo wrote:

I think it's a good idea, but then we have elaborate on the NXFD's role in this project. I mean, we need to identify the goals of getgnulinux.org and nuxified.org in this effort.

Indeed. What is necessary is to first determine what exactly needs to be accomplished that would best be accomplished through Nuxified.org and put it in bullet points so that we can clearly see how do these goals stand in relation to the current form of Nuxified.org. The catch is in figuring out how to achieve these specific goals by using what is already available on Nuxified.org or can be easily implemented without much disruption.

That said, one thing I am not sure about are language specific forums, simply because I'm afraid they wont have enough activity to be justified. If we on the other hand just try to keep translating english forums to other languages, that would require an enormous amount of effort by a rather large amount of people (if we want to support many languages).

Other things like categorizing topics in specific ways making it possible for people to easily search through already solved issues can be done through special tagging policies and pages with appropriately formated lists of appropriate tag links.

ariadacapo's picture
Offline
Joined: 2006-07-13
temporary

My fingers are burning for me to discuss this further, but I just would like to concentrate on the French translation + Fedora switch for the coming Saturday, so I will join this discussion again soon...
sorry - Olivier

libervisco's picture
Offline
Joined: 2006-05-04
No problem

No problem, sounds sensible.

Gustavo's picture
Offline
Joined: 2006-09-11
Don't worry, Olivier. I

Don't worry, Olivier. I understand you.

Moreover, this time out will enable us to think better about these and other things on the future of GGL.

Cheers.

ariadacapo's picture
Offline
Joined: 2006-07-13
Here are a few additional

Here are a few additional comments on all of Gustavo's suggestions... I tried to address every point.

- new "Tell a friend" tool
Great idea! This would fit well into the About page, maybe with this "4-page pdf download" idea I've had some time ago.
We could set this up rather easily I think, maybe with just a small anti-spam "text-image" field. The about page could be simplified if we build this "planet-GGL" website.
But anyway this comes in on the to-do list, no hesitation for me =)

- "Help us" section should be more precise
Absolutely right here. This can be done nicely if we move it on the "planet GGL" website where it can be used as a "top 5 priorities" page related to the bug tracking system.

- Post-Install article for beginners on GGL.o
I most definitely agree with you. This is needed. I sort of had this in mind when I wrote the "5 things to know when you switch" article published on PL.o.
The big problem here is that there isn't too much space. The "Switch to Linux" section already has four sub-sections, and with more we are risking overloading IMHO.
One possible solution to this would be to have such a section on the help center you mention. In short, one could define the missions of the two sites as:
- GGL: get people to know about GNU/Linux, know what can be done
- "Help Center": more concrete, direct tips as to installation and use.
Just an idea. But a beginner-friendly post-install place is needed.

- GGL.o web pages created from XML/XLST:
Ouch! Maybe it's possible to make life easier with this, but it's also possible to make it harder. I already thought about using PHP for the site (I even bought a book), in order to "call" different content & style "bits" for the dynamic construction of each page. But not only would the set-up be very time-consuming, also the risk is that we may obtain a slow and vulnerable website.
My personal point of view is that we should either stay with the actual website (I agree, not quite easy to manipulate, expand etc) or jump to some "human-compatible" system like Drupal or Wordpress, that would be dead easy to use.
But of course if someone can show that it's really a lot more efficient to use XML/XLST then I'm very willing to try.

- Suggest dual-boot
I am not quite sure what you refer to here specifically. The "try and install" page does explain both the liveCD and dual-boot notions. Is it this page you find not clear enough? How can we improve it? Otherwise I fully agree, dual-boot is the way to go.

- Mention that driver CDs are not necessary under GNU/Linux
You have a point here, but the other hand is that driver support is not perfect. I'd love to say "plug in whatever hardware, you don't need a cd", but if one has to say also "depending on the hardware, it might not quite work", it loses a lot of its value. Do you still think we should include it, maybe as an extra paragraph in the "Linux from Windows" page?

- Translation standards need to be set-up for neutrality
OK... so far I have had no problem with the French (there were Canadians and I'm continental French)...but of course there are many more variations of the Spanish language. If you wish to add things to the translation guidelines, before we move that thing to the planet GGL website, feel free to add.

- GGL.org non-profit, GGL.com for-profit
I think this has to depend on what we are going to do with the "status" of the GGL project.
Right now, I am leaning towards the idea of a non-profit organisation, that I would set-up in France (this "association de loi 1901" that was mentioned earlier), even if we might struggle later to have more personal earnings from it. I believe the "trust" from visitors to us is really important (thus the "About" section visible from each page) and so that will only help.
If such was the case, then I believe we could have for-profit things right in the middle of the existing website (with a link pointing to a page explaining what this money thing is about). This might make more sense, I think, since a direct link from the "try and install" page to some distro CD vendor, for example, would be highly relevant (more than having to transit.
But that's just quick thoughts. To sum up:
-if GGL becomes officially non-profit, then probably keep everything on the .org domain (?)
-if not, it is probably a good idea to separate .org and .com .

- Embassies
Yes wonderful idea!
This could really be a case where money from GGL.o could be well spent. Say, we could make a small budget for the print of posters and burning of CDs etc, if the ambassador has a nice project and motivation.
I am very ready to be the Toulouse-Rangueuil ambassador now that the website is in my language :-)

Overall, many things that really can get done. I get excited like a child when I think of it. Smiling
My opinion is that we need to do the following first:
1. set-up a bug tracking system with some SVN support. This way we can organise all of these ideas!
2. settle this money / declaration thing so as to have a clear policy for GGL
3. build a "planet-GGL" website

Olivier.

ariadacapo's picture
Offline
Joined: 2006-07-13
Miguel suggestions for tracking system

On the translation mailing-list, Miguel proposed the following:

Miguel wrote:

There are five possibilities:
* dotProject (http://www.dotproject.net/),
* phpCollab (http://www.php-collab.org/blog/),
* streberPM (http://www.streber-pm.org/).
* trac (http://trac.edgewall.org/)
* tracks (http://www.rousette.org.uk/projects/)

If I understand correctly he is trying out both Trac and strebberPM Smiling

I don't have a preference personally. I am just hoping for something simple to use...
Olivier

Gustavo's picture
Offline
Joined: 2006-09-11
In response to Olivier

Hi, Olivier!

ariadacapo wrote:

- GGL.o web pages created from XML/XLST:
Ouch! Maybe it's possible to make life easier with this, but it's also possible to make it harder. I already thought about using PHP for the site (I even bought a book), in order to "call" different content & style "bits" for the dynamic construction of each page. But not only would the set-up be very time-consuming, also the risk is that we may obtain a slow and vulnerable website.
My personal point of view is that we should either stay with the actual website (I agree, not quite easy to manipulate, expand etc) or jump to some "human-compatible" system like Drupal or Wordpress, that would be dead easy to use.
But of course if someone can show that it's really a lot more efficient to use XML/XLST then I'm very willing to try.

XML/XSLT is more suitable and easier to handle than PHP in this case, since we're using an static website. Moreover, because GGL is being translated into several languages, XSLT is the way to go; I'm pretty sure of that.

Anyway, I'm a PHP programmer and I'm willing to help with this as well. However, I'm not very proficient in XML/XSLT, thus, I'd set it up in XML/XSLT unless there's someone that knows better XLST.

ariadacapo wrote:

- Suggest dual-boot
I am not quite sure what you refer to here specifically. The "try and install" page does explain both the liveCD and dual-boot notions. Is it this page you find not clear enough? How can we improve it? Otherwise I fully agree, dual-boot is the way to go.

I meant that we should _encourage_ people to set up a dual-boot, unless they are pretty sure that they won't need their previous OS.

ariadacapo wrote:

- Mention that driver CDs are not necessary under GNU/Linux
You have a point here, but the other hand is that driver support is not perfect. I'd love to say "plug in whatever hardware, you don't need a cd", but if one has to say also "depending on the hardware, it might not quite work", it loses a lot of its value. Do you still think we should include it, maybe as an extra paragraph in the "Linux from Windows" page?

Maybe something like "Most common devices will work out of the box, but it doesn't depend on Linux, but on the manufacturer".

Since I'm using Linux, I've installed it on several different computers and I've just found two hardware incompatibility: A Canon printer and my father's medical devices (he's a doctor, not a sick person, just in case), which is of course manufacturer's fault. This is why I insist on this.

ariadacapo wrote:

- Translation standards need to be set-up for neutrality
OK... so far I have had no problem with the French (there were Canadians and I'm continental French)...but of course there are many more variations of the Spanish language. If you wish to add things to the translation guidelines, before we move that thing to the planet GGL website, feel free to add.

Alright, thanks.

ariadacapo wrote:

- GGL.org non-profit, GGL.com for-profit
I think this has to depend on what we are going to do with the "status" of the GGL project.
Right now, I am leaning towards the idea of a non-profit organisation, that I would set-up in France (this "association de loi 1901" that was mentioned earlier), even if we might struggle later to have more personal earnings from it. I believe the "trust" from visitors to us is really important (thus the "About" section visible from each page) and so that will only help.
If such was the case, then I believe we could have for-profit things right in the middle of the existing website (with a link pointing to a page explaining what this money thing is about). This might make more sense, I think, since a direct link from the "try and install" page to some distro CD vendor, for example, would be highly relevant (more than having to transit.
But that's just quick thoughts. To sum up:
-if GGL becomes officially non-profit, then probably keep everything on the .org domain (?)
-if not, it is probably a good idea to separate .org and .com .

I agree.

ariadacapo wrote:

- Embassies
Yes wonderful idea!
This could really be a case where money from GGL.o could be well spent. Say, we could make a small budget for the print of posters and burning of CDs etc, if the ambassador has a nice project and motivation.
I am very ready to be the Toulouse-Rangueuil ambassador now that the website is in my language :-)

Good idea.

ariadacapo wrote:

Overall, many things that really can get done. I get excited like a child when I think of it. Smiling
My opinion is that we need to do the following first:
1. set-up a bug tracking system with some SVN support. This way we can organise all of these ideas!
2. settle this money / declaration thing so as to have a clear policy for GGL
3. build a "planet-GGL" website

Yes, the bug tracking system is the most important thing.

Cheers!

kode's picture
Offline
Joined: 2006-04-19
I am very fluent with PHP

I am very fluent with PHP and proficient with XML/XSLT, so count me in for some development work if you need it.

As for the rest of the PDF, I'm not really knowledgeable or experienced in the social aspects of communities, but the ideas being discussing seem to be well-informed. When it comes time, I may need to bribe Gustavo to do the same thing for my project. :-) I hereby nominate Gustavo as 'el arquitecto supremo de la comunidad'... I hope that translates correctly, it has been a long time since I've used Spanish.

libervisco's picture
Offline
Joined: 2006-05-04
Olivier wrote: You have a
Olivier wrote:

You have a point here, but the other hand is that driver support is not perfect. I'd love to say "plug in whatever hardware, you don't need a cd", but if one has to say also "depending on the hardware, it might not quite work", it loses a lot of its value. Do you still think we should include it, maybe as an extra paragraph in the "Linux from Windows" page?

I think it would have significant value if it'd say that most of the hardware works out of the box without installing a driver yourself and that what doesn't work out of the box are only hardware for which there are no free drivers available yet due to the manufacturer witholding necessary information. This way the message people get is that free drivers are a good thing to support because it results in better hardware support. Smiling

Actually I think this out of the box support thing is something that isn't emphasized enough about GNU/Linux and still too many people perceive GNU/Linux as inferior when it comes to hardware support, merely because some of the few popular devices don't yet work out of the box (like some graphics cards). Windows doesn't support anything out of the box.

ariadacapo's picture
Offline
Joined: 2006-07-13
Summing up

OK for the driver idea, it's now in the to-do list Smiling

Trying to sum up and get organised:

- Let's try to have the bug system up and running by Sun 17. Dec.
We can throw in there every idea mentioned here and before about improving the site, and prioritize them, which will help a lot.
I think we need to start another thread to discuss that in particular

- Let's try to have a draft for the planet website drawn by Sun 17. Dec.
So we can have it up and running by 31st of December.
I'll open another thread about that too.

- I'll try to have a clear view for the GGL project laid out by 31st Dec.
With a better idea of how to organise things for the future, we will be much more efficient. In the case I decide to declare it as an "association loi 1901" in France, it'll probably take another month before it's officially up, but at least by the start of January we'll have the structure ready.

These dates might sound dreadfully slow to some of you, but if we get this done early, all the best! I simply like to have objectives laid out, it helps a lot. It also does not mean nothing else will be done, but simply that these will be the priorities.
Unless someone objects, this will be the Master Plan for December Smiling
Olivier.

libervisco's picture
Offline
Joined: 2006-05-04
They don't seem so

They don't seem so dreadfully slow, just cautious and realistic. Smiling

The plan looks good. Of course, there is also a possibility, I think, to have the first two things (the bug tracker and the planet) in one site powered by one CMS rather than two separate. But I guess that can be discussed later in another thread.

ariadacapo's picture
Offline
Joined: 2006-07-13
XML - PHP - XHTML

I completely forgot this XML, PHP, XHTML theme, my apologies.

I appreciate that I am very ignorant in these matters.
My vision:

GGL.o has an unescapable amount of page-specific coding, which is going to be here, in whichever language. Try to compare the Windows, Choose a distribution or About or Linux FAQ pages, for example. Whichever way we code it we are going to have to put tags around the text in specific places.

However one area where a lot can be won is the non-page-specific parts. The header, principally. It's complicated to keep it without bugs for each page and with every translation or new page, lots and lots of time is spent making it work.
Some of that could be solved by using absolute paths rather than relative ones like today (I used that for the footer). But it wouldn't help with, say, the translation tabs which still need to be page-specific, and other items.

If one could generate that header dynamically so we don't have to worry about it, it would be a major improvement. I don't have the knowledge to make that myself. But the core of the text will always have to be not-quite-so easily editable if we want to keep the nice little styles around the website.

So do you think there's a way XML or PHP or else can help in this repect? I'm very willing to have this implemented (... by someone else ;-) )
Otherwise we can either keep the code the way it is, or completely move to a system like Drupal or Wordpress or else. That would break my heart somehow Smiling but I can certainly get talked into it if that's the way to go.

Gustavo's picture
Offline
Joined: 2006-09-11
Definitely, XML is the way

Definitely, XML is the way to go for the whole website. PHP is only necessary for the "tell a friend" tool, as far as I can see.

Cheers!

libervisco's picture
Offline
Joined: 2006-05-04
Olivier wrote:If one
Olivier wrote:

If one could generate that header dynamically so we don't have to worry about it, it would be a major improvement. I don't have the knowledge to make that myself. But the core of the text will always have to be not-quite-so easily editable if we want to keep the nice little styles around the website.

So do you think there's a way XML or PHP or else can help in this repect?

Of course there is. I am not a coder and am barely familiar with PHP, but I've done it on http://network.libervis.com . Basically you have a separate header.php file which generates the whole header and then specific content php pages which generate the content. The content pages would then have an include command at the beginning and the whole file would basically look like this.

<?php include 'header.php' ?>
<div id="content">

<p>Your content goes here. You can have any (X)HTML tags in here as php files are fully compatible with PHP. It is a language that usually generates HTML as output anyway.</p>
</div>
<?php include 'footer.php' ?>

The example header.php file could look like this:

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html xmlns="http://www.w3.org/1999/xhtml" lang="en" xml:lang="en">
<head>
<meta http-equiv="content-type" content="text/html; charset=ISO-8859-1" />
<title>Libervis Network</title>
<link rel="stylesheet" type="text/css" href="style.css" media="screen" />
</head>
<body>
<div id="header"><img src="images/header.png" alt="Libervis Network - For a Free World :: A networked world!" /></div>
<div id="headerback"><img src="images/headback.png" alt="Subheader decor"/></div>
<div id="left">
Your left menu can go here. It's not part of the header, but can be included in the same file if you wish. That's how I do it on network.libervis.com.
</div>

So yeah, you could even include a left (or right) menu in it. The styling needed for it is in the style.css.

Note also that there is no php in that header.php file. The file is .php simply so it could be more easily included in a content.php since, AFAIK, doing it all in HTML is a bit more inflexible and not all hosts allow such include tags. PHP is IMO the best way to go.

Gustavo's picture
Offline
Joined: 2006-09-11
Indeed PHP might be used,

Indeed PHP might be used, but:

1.- Is it necessary such a "load" to the server? I mean, there's no dynamic content being generated at all. It's just a bit of language-specific code. Every time a page is requested, the PHP script will have to be parsed and executed; yes, we can use a Cache, but it's a waste of time in our case, IMHO.
2.- We were mixing up HTML and PHP code, while many people (like me) consider appropiate to separate the "view" from the "business". But it's almost a different discussion.
3.- XSLT is specially made for these tasks: It's faster and easier to handle. Also, Kode is willing to help us out with XSLT, why don't we use it?

Just my thoughts...

kode's picture
Offline
Joined: 2006-04-19
Another option would be to

Another option would be to have a fully dynamic backend (PHP/MySQL for example), and have it generated purely static pages for the site... This makes management easy, and keeps the load down too.... just a thought.

libervisco's picture
Offline
Joined: 2006-05-04
In my humble opinion the

In my humble opinion the load caused by php if you just use it for including a common header to all pages is just insignificant.

But anyway, it's just a suggestion as an answer to Olivier's question on how could something like this be done. I am not really familiar with XSLT so I can't comment much on that. Whatever you in the end believe is the best should then be the best choice. Smiling

Cheers

tbuitenh's picture
Offline
Joined: 2005-12-21
Why do it in complicated

Why do it in complicated ways? Just use this makefile (untested, but something like this should work):

all: page1.html page2.html page3.html

upload: page1.upload page2.upload page3.upload

.html.upload:
[tab]scp $< user@host:/directory/$< ; touch $@

.content.html: header footer
[tab]cat header $< footer > $@

Just create these files:
page1.content
page2.content
page3.content
header
footer

Now type make and the html files will be generated for you. You could also type make upload, and whatever was modified will be uploaded. Yes, if you modified the header or footer, then everything will be regenerated or uploaded.

libervisco's picture
Offline
Joined: 2006-05-04
That's an interesting

That's an interesting approach. Smiling It does seem to make things easier and faster so maybe I try it, though I'm too used to my way of doing it now. I usually just use konqueror to open the file I want to edit in kate and then whenever I make even a small modification I just hit ctrl-s and it uploads the change directly back to the server (which is a neat feature I've seen only in kate-konqueror combination).

ariadacapo's picture
Offline
Joined: 2006-07-13
more thoughts

I have kept away from this discussion not because I am not interested, but because I needed a bit more time to think about the best way to organise everything. I believe this dynamic/static issue is also related to the coming "planet-GGL" website and the community organisation as a whole.

I say this because of my experience with the French translation.
I didn't participate in the set-up of the translation at all; I just noticed everything stalled when around 75% was done (90% was translated but a lot had to be proof-read and corrected). Despite trying to organise things a little through the wiki homepage, I was completely alone while finishing the translation.
And in fact, despite there being at least 8 contributors, I also noticed that most did not learn about the actual release until the news came through other sites.
This is essentially my fault, because there was not enough means to keep an existing community lively. I suppose this is largely the case for other languages (there are many wikis but not many are active). But a translation mailing-list is not attractive to everyone and a wiki is not a proper way to build a community.

Trying to learn from this, I make the following suggestions:
- The planet website should include individual language sections with its own forums. Just so there's at least a place a newcomer can see that is active. And as the Spanish translators show brilliantly, there's much that can "come back up" from translations to improve other languages too.
- There should be a way to translate the site that is more direct than wikis. Not only would it make editing more rewarding (looking at the result directly is more encouraging than just translated wiki text), but also much quicker (it takes a week for me to transfer the text from wiki to html every time).

This makes me think that we could move the whole website to a CMS like Drupal.
The planet and GGL.o would still remain separate, but we could maybe integrate the editing of GGL.o within the CMS.
Cons:
- Lots of work to transfer. Maybe difficult to obtain the same "very simple and clean" layout of GGL.o
- The whole planet will be hard to maintain. I'm still a student, there is absolutely no way I can maintain such a site with different language sections and integrated editing and everything. But I'm sure there will be people ready to do that Smiling

Pros:
- No worry about things like urls, internal links between translations, headers, and everything for translators. This would all be entirely dynamic.
- Members can join and participate more easily (would require signup, but much more rewarding than actual "blank" wikis)
- Flexibility. The GGL project has a lot of potential, we can grow/modulate/evolve quite easily with this.

I know that this idea might come a bit late, but I believe overall it would be a good improvement, something that would allow us to grow quite a lot. What is your opinion?

Gustavo's picture
Offline
Joined: 2006-09-11