30 replies to this topic

[Brito]

We shall see my Bob friend how this interesting chapter in wb history turns out.

 

Le wb est mort, vive le wb

 

[pscEx]

Le wb est mort, vive le wb

Hey Billy, you are right!

 

Bob

[Wonko the Sane]

We shall see my Bob friend how this interesting chapter in wb history turns out.

Very good
 
 
 

Que serà, serà
Whatever will be, will be
The future's not ours, to see

 


http://en.wikipedia....ll_Be,_Will_Be)
 

Weenko (the Seine)

[Baylink]

If you want to cling on to old names/products that never change and then give a brand new name for each new project that surfaces with the idea that this brings benefit to end-users, then do proceed with that approach for your works and we'll see which approach makes users understand what you are making available. 
 
The concept, spirit and purpose of Winbuilder don't change. Win7PE describes exactly what it does and what users get as a result.

Perhaps I'm confused.

That sounds like an argument for why the new Java Winbuilder *ought to have a different name*, and why the "win7pe" builder-control-interface project ought not to have the same name as *the object it produces* (that is: a "win7pe boot disc").

Naming is really important stuff, and to illustrate how that's true, I'll tell you a little story.

It's the story -- no, not of a man named Brady; my name is Jay -- who was trying to learn how to use a new tool to do an old job. I have lots of experience with UBCD4Win, both as a user, and as someone who made 3 or 4 end-user-disc builds for myself and the IT staff who worked for me, a few years back. That was a few years ago, and as you know, UBCD4Win is moribund now, it's fora having gone 404.

So I spent some of my free time yesterday whilst watching computers image looking for a replacement. That's harder than you'd think it would be, and part of the reason for that is that no one seems to be able to agree on the same common nouns to name the various layers of the stack. But no matter, I located WinbuilderJ and the associated win7pe "project", and did the requisite downloading... and it didn't work.

This is not my first rodeo, and the project is only a few months old; I'm ok with "didn't work on the first try". Figuring out why, though, has been more troublesome than I'd have preferred, and for reasons that branch out in several directions.

My ongoing efforts to get it figured out are in another thread, and I won't belabor all of them here, but the fundamental thing making that process unnecessarily difficult is two points, both of which Nuno will have to address for them to be fixed:

1) Error messages without enough (read: any) parameters

and

2) The Naming of Things.

Point 1 has to do with not simply telling the user "No project is defined for building", but taking the extra sentence to tell them what that means, in -- at the very least -- a paragraph of command help which explains what a project is, where they live, and how one "is defined for building".

Additionally, this assumes (see also "nerdview") that the user knows what the programmer means by "project".

That leads directly to point 2: having good, understandable common names for things, and good proper names for instances of those classes which don't conflict with other already extant objects is *critical*:

1) it reduces the support load of knowledge transfer from the devs and power users out to new users who need to know it -- especially while the program is in its early stages.

2) it does that at least in part because when one googles "winbuilder win7pe", what one will get will *not* be useless information about an earlier and completely different generation of the same or similar software.

It is somewhat habitual for programmers to short this part of the work on a project, because it "doesn't contribute as much to the goal".

Clearly, I think that opinion is shortsighted, especially -- as here -- in the case where the goal for the project is "public" use; it's slightly more tolerable for projects inside companies that are intended solely for the use of the coder... though even those often end up getting yanked up into the public eye, where they unnecessarily burden people for precisely the reasons I've mentioned.

===

Finally: an information architecture observation:

While fora are wonderful tools for user interaction on the medium time scale, they're not as good as IRC for quick easy answers, and they're perfectly miserable when compared to Wikis as *knowledge capture* resources.

Forums are good for questions.

Once you actually have answers, though, those should go where it is much easier to find them.

Whether that's a Winbuilder specific wiki, or a Windows PE/Boot Discs wiki, I'm not yet entirely sure.

Or perhaps there is one, and I have simply missed it in my googling.

But 30 years of working with open source software tells me that the bar to entry here is unnecessarily high. Perhaps that's purposeful; inept people should be protected from powerful tools.

But I don't consider myself inept, and I feel like I'm being "protected" from being able to get anything done as well.

Perhaps it's merely that it's Christmastime, and I would be being swarmed with answers on-point were it next month. :-)

I really, really want to like the WinbuilderJ/Win7PE-CP combo. I hope that I can.

[Baylink]

Well I see I can't edit that, so another thought by extension:

Perhaps the biggest problem I forsee on the Billy-Bob front -- and I didn't myself realize it until last night -- is this:

If I'm a newbie, looking for a replacement for UBCD4Win (which, amazingly, I am :-), one of the most important things I'm going to be looking for is what magazine writers like to call "a thriving ecosystem": in an environment where what you *need* is tools to get work done, and here, where those tools are apparently incorporated into your boot disc by "scripts", you're looking for the underlying system with the largest number of those scripts.

But this new tool called "Winbuilder" is appparently script-language-incompatible with the tool it replaces, called... "Winbuilder". And the people who created those older scripts certainly didn't say they worked with "old Winbuilder"; why would they? There was no "new Winbuilder" then. And no one has gone back in the taxonomy and properly reclassified them, either.

So as I look to determine with which tool to move forwards, I am inclined to think "Oh, look! This Winbuilder thing has lots of 'scripts' to let me include packages into it.

No, it doesn't.

I have no trouble with investing time and effort into bringing about the New Hawtness. But I'd sort of like to know in advance, y'know?

It's become more and more clear to me over the last 24 hours that *to get work done*, what I ought to be doing is installing the old, Delphi-based, graphical Winbuilder, since that is apparently what all the script packages in the repository *are for*.

And that's what I'm going to go do now.

[Brito]

Baylink, nobody here is paid to deliver you something ready and fully functional and (unfortunately) we don't have the resources to do better.

 

Would be nice to read a message of Baylink saying "I can help with something practical, where do I start?".

 

I'd just remember that we are working to improve the state of the art in boot disks. What was done before is quite nice, keep on using it in case this is what you need. What you see being developed addresses other needs (not yours likely) that we address today and were unable to address before. Development takes years to settle, we are moving within our possibilities.

 

On this new edition we support:

- Visually impaired users

- Cross-platform (Windows, Linux, OSX, Android)

- User interface abstraction (command line, Web, GUI)

- No administrative permissions

- Self contained (no dependencies to third-party tools)

 

Changes are needed. If you see room for improvement, jump aboard.

 

This is Winbuilder. Same situation as Microsoft Windows where you are now asking for software made for version Windows 3.11 to work on newer Windows versions like Windows 98 or Windows 8.

 

It is possible to add support for older scripts? Yes. Do we have the resources or the will to do this? No.

 

Time to close the topic. Nothing good is coming out from this discussion.