Please login or register.

Login with username, password and session length
Advanced search  

News:

You need/want an older version of sNews ? Download an older/unsupported version here.

Pages: 1 [2]

Author Topic: User Suggested Roadmap  (Read 10936 times)

Joost

  • Guest
Re: User Suggested Roadmap
« Reply #15 on: April 08, 2008, 03:33:11 PM »

Jason,

Of course, when building the next version, a description of new features can be very useful is a must!  :D
However, when talking about 1.8, 1.9, 2.0 we have look much further ahead. So more abstraction is needed, an overall plan.
Logged

funlw65

  • Hero Member
  • *****
  • Karma: 96
  • Posts: 771
    • Country Lab
Re: User Suggested Roadmap
« Reply #16 on: April 08, 2008, 05:05:18 PM »

Philippe is right. Roadmaps should not be about which features to add, but be more abstract.

As I see it, sNews has reached its limits. For it to "grow", it should be redesigned. Let the 1.x series live for a few years and meanwhile start off with a fresh 2.x series.
................................
When it was clear sNews must be faster and scalable, not a toy anymore, for small websites....I know personal blogs with more than 1000 visitors per day, using Wordpress, I mean, small blogs.... We can gain Wordpress users if sNews can handle what Wordpress handle already.

I saw  something on web, weeks ago... A web designer using Wordpress and I think, Patric comment on his blog: I see you now are using Wordpress (speaking from memory, It was a Patric joke about betrayal..)... and that web designer replied: I will always return to sNews for small projects...
I don`t remember what blog, what web designer....maybe Patric can... or it was not him....
Logged

Keyrocks

  • Doug
  • ULTIMATE member
  • ******
  • Karma: 449
  • Posts: 6019
  • Semantically Challenged
    • snews.ca
Re: User Suggested Roadmap
« Reply #17 on: April 08, 2008, 09:07:59 PM »

I suppose I'll slide into this discussion for a bit. Bear with me... I'm going to get into a story-mode to premise my eventual final point(s).  :)

Let's start by taking a look at the evolution of one of the best known, extremely versatile and most widely used little "engines" used by customizers and hot rodders in North America today. I'm talking (with experience) about the Small Block Chevy V8.

A way back... in the early 1950s... North American automakers began manufacturing "overhead-valve" V8 engines to replace the "flat-head" (valves-in-the-block) engines that had been used since the birth of first internal combustion engines back in the first decade of the 1900s. One of the first on the market was a huge, over-weight (720 pound) engine made by Chrysler... the fore-runner to the well-known muscle-car killer of the 1960s... the Hemi.

Around 1954 or '55... General Motors (GM - the bloated North American auto manufacturer) released its first version of the "Small Block V8". It had a displacement of 265 cubic inches (the volume of all 8 cylinders with their pistons at bottom-dead-center). By 1957... it had been increased in size to 283 cubic inches and a high-output version... initially fitted in the early Corvettes... came out as a 327. By 1967... GM had taken the same engine and increased its displacement to 350... and we saw it first in the SS Camaros and Corvettes. That same year they also produced a 307-inch version for everyday consumer cars like the Chevy Byscane (anyone remember that one?).

The 350 continued in use for many years... right in late 1990s. In late 1970s, a 400-inch version came out... mainly for use in light trucks and vans. In the 1980s... increased demand fo fuel economy led to development of a 4.3 litre V6, based on the exact same engine configuration... it was basically a small block V8 shortened up with 2 less cylinders. This little 4.3 litre V6 became the base engine for all of Mercury Marine's inboard-outboard installations. (I have one powering my 1987 24' Doral powerboat).

Now... every one of these engines all used the exact same valve-train components, the same oil pump system, cooling & water pump system, intake and exhaust manifold configurations. With the V8s... it was (and is) possible to replace the heads on a 1957 283 with those from a 307, a 327 (from the 60s) or a 350 (fron the 80s and 90s). It's also possible to do the same with the intake manifold and several other components. You can also take a complete intake manifold and fuel injection system, as well as the HEI (pointless) ingition system, from a 1970s to 1990s small block unit and install them on any other small block from the 1960s and 1970s.

And... interestingly enough... the bolt patterns on the back of all these V8 engines (where the transmission or bell housing bolts on) as well as the engine mount bolt patterns on the sides and front of the engine... were in the exact same locations. The un-used holes were not always threaded (only those being used in a given installation were) but it takes less than an hour to thread the holes when needed. This makes these small blocks completely interchangable. You can install a 1990s 350-inch or 400-inch engine in a 1956 or 1757 Chevy and they fit like a glove. Even the starters and alternators (and their brackets) are interchangable.

The point here is that this was the most successful and long-lived engine design ever produced because it continued to use the same design and platform over almost 4 decades. This is what also makes the Chevy Small Block the most written about and most widely used powerplant for the majority of hot rods and customized autos running the roads in America (and Canada) today... not just GM cars and pickups but many old Fords and Dodge products as well. I even used them in two of my first car projects - a 1938 Ford Coupe and a 1939 For 2-dr sedan.

This... my friends... is the approach we need to take with sNews. Perhaps the following might be the first three "pillars" or guiding principles.

1) We continue to refine, improve and maintain the "core" sNews product as the best, most efficient and easy-to-improve CMS engine.
    A core improvement is something like adding a "Comments On/Off" checkbox to enable/disable comment use site-wide.

2) A "Plug-in" system should be in the form of an Administration Add-on.

3) All the mods and addons should be configured to work with the "Plug-in"... running as independently as possible from the "core" engine.

This leaves lots of room for growth for "forks" developed by those who enjoy creating their "Packages"... like an... EQNews... ?  :)
« Last Edit: April 08, 2008, 09:13:27 PM by Keyrocks »
Logged
Do it now... later may not come.
-------------------------------------------------------------------------------------------------
sNews 1.6 MESU | sNews 1.6 MEMU

funlw65

  • Hero Member
  • *****
  • Karma: 96
  • Posts: 771
    • Country Lab
Re: User Suggested Roadmap
« Reply #18 on: April 08, 2008, 10:13:24 PM »

I  think  I  want  that  sNews.... :)
Logged

centered

  • Guest
Re: User Suggested Roadmap
« Reply #19 on: April 09, 2008, 12:56:08 AM »

Thanks Doug for that.  Of course some might fancy the mighty 427 or 455 GM big blocks...

Anyway in a programming sense, what you are mentioning is an OOP or modular concept (from my understanding)

Quote
With the V8s... it was (and is) possible to replace the heads on a 1957 283 with those from a 307, a 327 (from the 60s) or a 350 (fron the 80s and 90s).

Let's take a cam/crank-shaft for instance and let's say I am a third party manufacturer wanting to better a crankshaft in the engine
// orignal specs
class crankShaft {
   Specs
   fucntions
}
// my specs
class thirdParty_Shaft extends crankShaft {
   (extending original specs but probably adding more compression to make it better)
   no need to add the functions, as it's already defined
}
 
Let's go a step further

How would GM create the Chevy Nova, Camaro, Olds cutlass, buick GSX, chevy Chevelle, etc (almost same body style and engine)?

It would be a muscle car base, then extending that to form the nova
For the camaro, take away the extending nova code and add the camaro code
etc etc..

though I could be wrong about the whole thing....
   
Logged

Keyrocks

  • Doug
  • ULTIMATE member
  • ******
  • Karma: 449
  • Posts: 6019
  • Semantically Challenged
    • snews.ca
Re: User Suggested Roadmap
« Reply #20 on: April 09, 2008, 01:58:52 AM »


1)   Thanks Doug for that.  Of course some might fancy the mighty 427 or 455 GM big blocks...

2)   Anyway in a programming sense, what you are mentioning is an OOP or modular concept (from my understanding)

3)   Let's go a step further... How would GM create the Chevy Nova, Camaro, Olds cutlass, buick GSX, chevy Chevelle, etc (almost same body style and engine)?
      It would be a muscle car base, then extending that to form the nova... For the camaro, take away the extending nova code
      and add the camaro code etc etc..

though I could be wrong about the whole thing....
   

1)  Or the 454... based on the 396/427 platform... in a '70 Impala ragtop with a 4-speed Muncie in the tunnel.  ;D
     But we're digressing.

2)  I know what the definition of OOP is but don't know (yet) how to work with it.

3)  Well... back between 1967 and 1972-3... the Novas, Camaros and Firebirds all had the same unibody shell with different skins... and front clips & drivelines, depending on the powerplant (big block or small). Today, all auto manufacturers base several models on one shell... eliminates the need to have different assembly lines for each model.

So... yes... the same general approach works equally well with modular or "unitized" construction of residential housing units... and would apply to coding applications I would assume (though I'm not a pro in this field). Using your crankShaft class as an example... you would have several crankshafts that all fit the same engine block (main bearing journals are the same for all) but each performs a bit differently (longer stroke to increase compression) depending on "your specs".
Logged
Do it now... later may not come.
-------------------------------------------------------------------------------------------------
sNews 1.6 MESU | sNews 1.6 MEMU

philmoz

  • High flyer
  • ULTIMATE member
  • ******
  • Karma: 161
  • Posts: 1988
    • fiddle 'n fly
Re: User Suggested Roadmap
« Reply #21 on: April 09, 2008, 03:39:02 AM »

Unfortunately, when it comes to feature sets, the line between wants, needs and musts gets very blurred.
It becomes very subjective.

Someone may need multi-user environment. Someone else may need multi placement of content. Another would perhaps need unlimited subcategories. Meanwhile, a different person wants total modularity.
It is very possible that none of them want any/all of the other features, while all these features could be supplied as (complex) mods.

However, whatever is supplied in core must operate as expected, must be useable by greater than 50% of users, must be thought out very thoroughly, and must not cause bloat or drag.

...AND, ultimately, someone (individual or group) has to make the final decision as to what is to go in, and what is left out.

So, a more 'abstract' outline is also fundamental to ensure feature sets don't get totally out of hand in their implementation.

Just my thoughts.
Logged
Of all the things I have lost, it is my mind that I miss the most.

Dom

  • Full Member
  • ***
  • Karma: 5
  • Posts: 163
    • domdelimar.com
Re: User Suggested Roadmap
« Reply #22 on: April 13, 2008, 11:39:19 PM »

So, a more 'abstract' outline is also fundamental to ensure feature sets don't get totally out of hand in their implementation.
That couldn't be more true. "Abstract" is good, no... is needed here. First... and then we can talk about features.

And in line what Keyrocks said (kudos buddy!):
1) We continue to refine, improve and maintain the "core" sNews product as the best, most efficient and easy-to-improve CMS engine.
    A core improvement is something like adding a "Comments On/Off" checkbox to enable/disable comment use site-wide.

2) A "Plug-in" system should be in the form of an Administration Add-on.

3) All the mods and addons should be configured to work with the "Plug-in"... running as independently as possible from the "core" engine.

This leaves lots of room for growth for "forks" developed by those who enjoy creating their "Packages"... like an... EQNews... ?

I'd translate that into the following rules, pillars or vision of progress, however you want it...

1. sNews' core system will be the most efficient and easy-to-improve CMS engine and will not gain any new features easily.

2 .Except, first major new development in sNews should be adding a plugin system that will honor the above rule to the greatest extent possible (sNews core will remain efficient and easy-to-improve engine).

(bear in mind this is only my interpretation and a wish - I don't call the shots here, hehe...)

From this I'd say sNews will remain to be the most favorable choice for enthusiasts in web development who wish to learn and experts who value the simplicity and not bloatiness of a system but also experts who need a more robust solution for bigger projects of theirs, those sNews may not have been appropriate for so far.

My intention is not to set the rules and for that I value other people's initiatives like Keyrocks' one above and I would be happy if the people who know more than I know about sNews to agree on the direction sNews should go, set the rules (which can always be revised if the future demands it) and start coming up with the solutions...
Logged

Fred K

  • Still trying to learn stuff
  • ULTIMATE member
  • ******
  • Karma: 130
  • Posts: 2728
    • Personal
Re: User Suggested Roadmap
« Reply #23 on: April 14, 2008, 05:03:23 AM »

Personally I agree with what most of you have already said -- Dom sums it up quite nicely I think -- but there's one minor snag to all of this. Again, in my personal view.

We can discuss roadmaps and general principles and the painful beauty of women's neck lines ... erm ... until our faces turn blue, or red perhaps depending on blood pressures, but before our grand Master and project owner Luka officially hands over the baton or appoints a taskforce or similar, to actually ponder over and cast these details in stone, there's not much that we can do about the current state of affairs. sNews is at this point Luka's project. A bunch of us have had our hands in developing the 1.7 (RC), but it's still Luka's project.

Just thought I'd illuminate that point.
Logged
Pages: 1 [2]