• Welcome to OS2World OLD-STATIC-BACKUP Forum.
 

News:

This is an old OS2World backup forum for reference only. IT IS READ ONLY!!!

If you need help with OS/2 - eComStation visit http://www.os2world.com/forum

Main Menu

eComStation v2.0 RC1 - Let`s install it!

Started by MikeG, 2007.07.05, 07:17:49

Previous topic - Next topic

obiwan

Scope of testing is based on what you want to achieve with the testing. It might seem on the surface like broader distribution must equate to more discovery of bugs and therefore better data, but that is not always the case. People can report many things as bugs that are just their own mistakes or misunderstanding of the product. Even if every bug report were 100% accurate, it can end up being an awful lot of information to have to deal with. You may have to help users along in order to uncover the root of the problem, and that eats up a lot of time. Even without that, you have to pick and choose what you will fix because you can't fix everything. Even if you are trying to solve only a few problems, only one problem, processing 100-1000 reports about it does not help, but only takes up valuable time. Even having 5 or 10 reports just makes it that 5-10 times more complicated. In order to actually solve the problem you need one single reliable test that you can reproduce. That way you can see if the work you do corrects it. Only after that test passes can you try again in a broader scope, and then you must reduce it again to a single test. There is no magic to turn bug reports into fixes. It is all work, so there is sense in limiting testing and trying to keep it orderly.

My experience with technical support SSI/Mensys over the years has been outstanding. Particularly considering the complexity and age of the product they are working with. Particularly in comparison with other desktop operating systems. Particularly in contrast with what a home user could expect from IBM for OS/2 back in the day. Does anyone else recall the sorts of experiences I had trying to get support from win16 software vendors for real problems when I mentioned I happened to be running it on OS/2? On the other hand, Joachim actually looked on Hobbes for a driver for me (which I should have thought to do myself) and talked me through enabling LAN PEER services (which I also should have just looked up). I really do not think "bad support" is at all a fair description, and I don't think describing it as such is at all helpful to anyone. If you want good support, there has to be a limit to it, and that means supporting the GA product, not processing endless bug reports for prereleases.

I don't know about these fixpaks, but I know there was a fix for JFS that is incompatible with bootable JFS, so applying it would mean breaking the product. I think if SSI chooses to delay application of fixpaks it is for a technical reason, not personal laziness.

kim

Another topic I think is a failure for quite a few of the new components within the future eComStation is that it's based on closed code and not open source. Since eComstation is based on a closed code base, ie OS2 from IBM, it would be far great if the new features and add on would be open source to get a wider usage.

I'm aware of that it's a hard decision for Serenity how to handle it; add new features that people can download without any cost and re-use on earlier version of OS2. But, taking the close code path once again I doubt will gain new users.

For example I'm really sad that Panorama drivers is closed code and as it looks now only will be distributed as part of eComstation. Why not at least have a shareware version of it that can be downloaded and tested on old system and fully unlocked. SNAP was a great product (and still is) - but we all know the story by now what happened there. Do we really want to take that path again?

And about the pricing for eComstation, it's not easy to get new users to jump on the train with the current price tags. The money that Serenity pays to IBM based on the OEM agreement is free money for IBM. Even when IBM actually sold OS2 as a products the OEM agreements was really lucrative for involved parties when it came to pricing. What I'm saying is that eComstation could have a far lower price. It's all about what path to take - high price -> few users or low price -> more users.

RobertM

What I'd like to see happen is IBM divest themselves of OS/2 by selling it to Serenity Systems.

The problems I see for that happening though are major:

  • Some code is still licensed from MS
  • At the last time I checked, IBM still relies on OS/2 for many of it's mainframe/mini lines [embedded OS to control the system - almost like a system BIOS... for instance the Z Series used to (and still may) use it in such a fashion]

I dont know how we as a group (OS/2 users, developers, distributors) could get around those issues short of re-writing OS/2... with the numerous inherent problems related to that (such as recreating the kernel and it's fantastic CPU, thread and process scheduling techniques - and duplicating the APIs, WPS, REXX subsystems, etc).

It may not be so much of a problem if IBM could at least lower their costs to Serenity, and provide more technical information (or source) to certain sections that they wholly have the rights to.

Either way, at least Serenity is making some pretty big steps...


|
|
Kirk's 5 Year Mission Continues at:
Star Trek New Voyages
|
|