Response to June 2014 Meeting Minutes (an outsiders view)

Here you can discuss ReactOS related topics.

Moderator: Moderator Team

Post Reply
danunder
Posts: 5
Joined: Sun Jun 09, 2013 4:51 am

Response to June 2014 Meeting Minutes (an outsiders view)

Post by danunder »

I do a lot of hardware development so I know exactly how it feels to have a piece of equipment 80-90% complete yet only able to do 5-10% of its final function. As we speak I have half a dozen seperate pieces all in development that need to be hooked together to form a fully functional prototype. But to an outside observer it still looks like a pile of junk: Even an engineer who understand its final structure will most likely have an attitude of 'meh', simply because the world is full of unfinished projects by many people who think they know what they are doing.

What outsiders need is a concrete piece. You don't need a functioning whole, but one piece needs to be absolutely complete. No only that, it needs to better than the original.

The obvious candidate for this is explorer new. If on an xp system you can kill explorer, run explorer new and have it perform flawlessly, demonstrate it does not have any bugs that may exist in the current explorer and do things that explorer cannot, you demonstrate 3 things:

1. Your code is compatible
2. Your code is superior
3. You are able to complete a finished project

Until you can perform such a concrete demonstration, this project will remain to most outside observers just hope and promise.

Take a converse view. Say there was another OS in direct competition with ROS for a large sponsor donation and they had to state why the money shouldn't go to ROS: if they could say "they don't have a single module working correctly", that's a pretty compelling argument.

I have been writing code on and off for over 30 years, so I know how the process works. I am also an outsider on this project so I can tell you exactly how its looks to an outsider. I don't really have the spare time to contribute significantly to the project but I have considered the idea. There will be many more people out there besides myself who do have the time and greater relevant coding skills who could contribute. Explorer makes up what, say 1% of the code base? If you can't get that right it really doesn't look good.

Now this is also a special case in that end users without any coding knowledge can still easily detect flaws. Most of these people will never file bug reports. Frankly its a PITA, especially when you could just make a forum post saying hey when I tried to do x, y happened. If you have a single official thread dedicated to it, you would make life a lot easier for the people who are trying to help you.

Think about it: If you have 100+ people all testing one spefic module and able to easily report issues, how long would it take to sort out all the bugs? It would be a lot faster than the rate you're going now.

Until you do this you will have difficulty drawing in more outside help and without those extra hands development will continue at a snail's pace.
User avatar
dizt3mp3r
Posts: 1874
Joined: Mon Jun 14, 2010 5:54 pm

Re: Response to June 2014 Meeting Minutes (an outsiders view

Post by dizt3mp3r »

I'd agree with that approach, I know it is difficult to 'compartmentalise' something like explorer as it has its hooks into everything but we do need to demo that ReactOS 'works' at some point. It will never work completely in all respects and that is understood, bugs will always arise in relation to Windows functionality, comparisons will be drawn but to have a fully useful component like the new ROS explorer could be useful to have in Windows too. Think of it as a good advert for ReactOS.

Ultimately, if we could replace windows XP components with the ReactOS version, 95% XP 5% ReactOS, the proportions increasing over time, we'd have an increasingly supported version of XP, it's native Microsoft components gradually diminishing. Perhaps that is a better direction for ReactOS, replacing XP bit by bit with fully working, supported nibbles/chunks/apps.
Skillset: VMS,DOS,Windows Sysadmin from 1985, fault-tolerance, VaxCluster, Alpha,Sparc. DCL,QB,VBDOS- VB6,.NET, PHP,NODE.JS, Graphic Design, Project Manager, CMS, Quad Electronics. classic cars & m'bikes. Artist in water & oils. Historian.
Frontier
Posts: 70
Joined: Fri Sep 20, 2013 10:29 am

Re: Response to June 2014 Meeting Minutes (an outsiders view

Post by Frontier »

You actually can kill explorer on xp and run explorer_new and have it run really, really well. The only 'glitch' I've noticed is the taskbar's resize is a bit dodge. Apart from that it's hard to tell the difference.
User avatar
dizt3mp3r
Posts: 1874
Joined: Mon Jun 14, 2010 5:54 pm

Re: Response to June 2014 Meeting Minutes (an outsiders view

Post by dizt3mp3r »

Then that's maybe what we need to do, offer parts of ReactOS as replacements for the XP versions, make their downloads more obvious and start distributing them as tools, offer support for those individual components. Give XP users a crumb of hope for the future.

I know that similar tools already exist as replacements for some core windows functions, there is no reason why the reactOS components could not be offered in the same way, the emphasis though from our point of view being simply as an advertisement for ReactOS.
Skillset: VMS,DOS,Windows Sysadmin from 1985, fault-tolerance, VaxCluster, Alpha,Sparc. DCL,QB,VBDOS- VB6,.NET, PHP,NODE.JS, Graphic Design, Project Manager, CMS, Quad Electronics. classic cars & m'bikes. Artist in water & oils. Historian.
florian
Posts: 510
Joined: Tue Nov 01, 2005 2:19 am
Location: Germany

Re: Response to June 2014 Meeting Minutes (an outsiders view

Post by florian »

I second danunder.
PurpleGurl
Posts: 1790
Joined: Fri Aug 07, 2009 5:11 am
Location: USA

Re: Response to June 2014 Meeting Minutes (an outsiders view

Post by PurpleGurl »

I was thinking in a similar way. Wouldn't it be good if the devs could do testing mainly in Windows and complete each part one at a time and gradually test with more and more? Then keep going until there is no Microsoft parts left. Of course, the main trouble would be with parts where we have no reference point for building and have no effective or accepted way of doing clean room, reverse engineering.
mrugiero
Posts: 482
Joined: Sun Feb 14, 2010 9:12 am

Re: Response to June 2014 Meeting Minutes (an outsiders view

Post by mrugiero »

danunder wrote: What outsiders need is a concrete piece. You don't need a functioning whole, but one piece needs to be absolutely complete. No only that, it needs to better than the original.
I disagree. Nobody needs it to be better than the original. Being on par and being free (both as in speech and as in beer) is already being better than the original, from a user POV. Not having to pay while obtaining the same quality is a merit on itself to choose it. Being able to fix things long after the one who wrote them had lost interest is a merit on itself.
Now this is also a special case in that end users without any coding knowledge can still easily detect flaws. Most of these people will never file bug reports. Frankly its a PITA, especially when you could just make a forum post saying hey when I tried to do x, y happened. If you have a single official thread dedicated to it, you would make life a lot easier for the people who are trying to help you.
And that would create a different set of problems, which are harder to deal with. Problem number one, bug triaging. Identifying the source of a bug takes usually more information than just the initial description of the bug. Tracking logs and different tries to trigger the bug would be a huge PITA in the forums. I'm not even considering the fact you would need to be able to easily find a specific bug, while the forum basically behaves as a stream. Searching with the search function would most likely introduce lots of noise. Bug reports exist because they are the most practical way to work with bugs we have. Forums wouldn't improve on that, because they don't really organize well this kind of data.
Think about it: If you have 100+ people all testing one spefic module and able to easily report issues, how long would it take to sort out all the bugs? It would be a lot faster than the rate you're going now.
The fact is it is not easier with the forums, specially if you focus them on a single thread, than with the bug tracker. Aside from the one or two clicks you save with the forum, you'll lose ten times the time of those clicks by being asked for the data the report you currently fill already asks from you, and developers will lose hours looking at the thread trying to make any sense from it, and then if someone new tries to look for a bug to get started he will have to wade through hundreds of unrelated posts to find the logs he needs.
Webunny
Posts: 1201
Joined: Sat Apr 28, 2012 1:30 pm

Re: Response to June 2014 Meeting Minutes (an outsiders view

Post by Webunny »

Hmpf. Sometimes I can understand the more pessimistic nature of z98 when dealing with forumposts. I mean, really. Look at this thread. No doubt it's well meant and all that, but is there *anybody* who really thinks this is going to change anything?
"What outsiders need is a concrete piece. You don't need a functioning whole, but one piece needs to be absolutely complete. No only that, it needs to better than the original."
"I agree."
"I disagree."

Ad infinitum. It's, of course, always dangerous to talk for someone else. Who are those outsiders? Can't they have, mayhaps, another opinion then stated?

Actually, while not an outsider, I prefer a 'working whole', rather than one piece being absolutely complete. Because, what is that 'piece' to me if I don't really need it? Changing explorer in winXP into explorer-new of reactos...who the f- cares? Only a very, very small minority, I would wager. (I'm not talking about specific testing purposes, but normal use). If you have to pay for winXP (or others) anyhow, no 'normal user' is ever going to bother to do that. It took more than a decade to pry them (those outsiders) loose of IE and try out some other webbrowser, even. And that was rather trivial, and because MS didn't bother to develop their browser for years, so the alternatives became vastly superior. None of this is true for changing explorer into explorer_new. Except for testing purposes and some tinkering by IT-minded people, this is not going to happen. Not until MS drops all development of it, and we get vastly superior in it. But then again, as someone already pointed out; there ARE already a lot of alternatives for most independent components. It's like trying to storm the market with yet another browser: there are already so many out there, that it's highly unlikely that you make any headway in marketshare, unless you have some drastic superior working or novelty (killer application) on there...which, if open source, will be rapidly adapted by the other browsers too, so...

The point is, there is no massive interest in such a thing. The interest for it, comes only from a small minority. Claiming that small minority speaks for the large majority is self-delusion at best. People who come to ROS, are interested in, indeed, a wholly working (windows-like) OS, that can replaces their (windows) OS (for free). The interest in this project isn't already all that great, compared to other systems, such as Linux. If one would only focus on some component of it, that doesn't do anything unless you already have windows...then what's the point? That's how most will reason, to be honest.

I do agree filing bugs are a PITA. For the casual user/tester, it's really annoying and bothersome; far more easy would it be, if one just could place it in the forum. That goes without saying. However, and this makes the thing more ambiguous: there is also no doubt JIRA makes things more easy to follow up on bugs, and to have it structured. So, in essence, two things play here: what is easy for the ones reporting bugs, and what is easy for the devs to note and follow up on those bugs. The forum would be easier for the first, but more difficult for the latter, and vice versa with JIRA. It's this dichotomy which is the cause for most discussion in this thread about it: "yes, it's easier / no, it's not easier". Well, it depends on who you're talking about. For the casual tester who wants to just report it, it is. For the devs or others that want to notice and follow up on it, it doesn't.


But anyway, what I was getting at at the beginning, in a way, this whole discussion is fruitless. Granted, I have now contributed to it, but I'm still aware it's fruitless. Is there anyone who truly believes the devs will change their focus from creating an OS to creating only a component of that OS? Please raise your hands.

It will never happen. For the simple reason they either don't see the incentive (as I do neither), or because of the fact that they just do what they like to do when coding, Which means; most of them have their field of expertise and preferred thing they work on. Since an OS has so many components, there is much choice to pick from. Reduce that to a single component, and a lot won't be interested anymore or know enough about it to be useful. It's that simple. So it just won't happen. It just won't. It's already doubtful any of them are even aware of this discussion, and even if they are, it's unlikely that on the next 'meeting minutes' we'll read they held a vote and all decided to focus on only one component, to make it absolutely perfectly working. Idem with JIRA. There is no chance in hell that they would drop it and go for bug-reports in the forum - for the sipmle reason that would make THEIR life (well, bugfixing) more difficult. Yes, not doing so might make it a bit more difficult for those casual testers reporting it, but frankly...they don't care about that.

The best you *could* hope for, in that respect, is that they maybe would make a thread for bugs on the forum, and they appoint one of the devs/PR who comes to the forum as a 'gatherer' to place the more interesting ones IN JIRA. Someone picking it up here thus, to place it there.

But even that is doubtful, seen one is always complaining about meager resources and being short-handed. I think it's still easier for them to say 5 times 'put it in JIRA' and shrug if it's not put in JIRA, than to put it in JIRA themselves. I think that'll never happen, unless it's a really important bug (with a possible bugfix), or if someone else decides it's worthwhile enough. Let's not forget: once this would become common, it would be a full-time task, to put all bugreports here into JIRA.


So...you know...it's pretty futile to go on and on about this. It's not going to change. At most you guys that are going on about it, could suggest YOU are offering your time up to put bugreports here into JIRA. (But I didn't see that happening neither, as of yet. So the whole point becomes a bit moot). That, or, you can contribute to something that really has a chance of being implemented, like this:

viewtopic.php?f=25&t=13466

I didn't see you guys vote yet! :)
User avatar
dizt3mp3r
Posts: 1874
Joined: Mon Jun 14, 2010 5:54 pm

Re: Response to June 2014 Meeting Minutes (an outsiders view

Post by dizt3mp3r »

Too many words and too much pessismism, often the case on this forum, not just this forum. Don't mean to be rude, really. Trying to keep it upbeat.

The point is that ReactOS has marketing opportunities to get people to start using bits of it, to get people involved and to encourage use/feedback/positivism.

If I had something in my hand that was marketable and usable then I'd say get it out there in order to push ReactOS to the wider world. That might be the case with ROS explorer new and possibly some other components as they mature.

These suggestions are good ideas (but not necessarily technically) and if the ReactOS organisation was more profit-driven, it would probably be forced down this sort of route. It isn't profit driven, and we accept that it will be driven by devs not wearing the 'profitability hat' but there are those that want to increase ReactOS profile out in the real world and so we come up with these little ideas that could work given some positivity and the right mindset.
Skillset: VMS,DOS,Windows Sysadmin from 1985, fault-tolerance, VaxCluster, Alpha,Sparc. DCL,QB,VBDOS- VB6,.NET, PHP,NODE.JS, Graphic Design, Project Manager, CMS, Quad Electronics. classic cars & m'bikes. Artist in water & oils. Historian.
mrugiero
Posts: 482
Joined: Sun Feb 14, 2010 9:12 am

Re: Response to June 2014 Meeting Minutes (an outsiders view

Post by mrugiero »

dizt3mp3r wrote: The point is that ReactOS has marketing opportunities to get people to start using bits of it, to get people involved and to encourage use/feedback/positivism.
These suggestions are good ideas (but not necessarily technically) and if the ReactOS organisation was more profit-driven, it would probably be forced down this sort of route. It isn't profit driven, and we accept that it will be driven by devs not wearing the 'profitability hat' but there are those that want to increase ReactOS profile out in the real world and so we come up with these little ideas that could work given some positivity and the right mindset.
Yes, but opportunities that come with risks. If the profile is increased but you lose the interest of the people currently working in ReactOS (all but one working for free), you end up hurting the project.

I don't disagree with the idea of distributing some pieces that currently work relatively well as independent packages, but I think making it a focus would be the wrong move, and even if it wasn't, what you see as "negativity" is just not jumping towards an opportunity without weighing the risks it comes with.


On the bug reports discussion, I could volunteer a bit of my time to report some of the ones that are currently reported in the forums, provided that I understand what's being described, but I don't think promoting that would be a good idea, as I said before, and it would most likely be more than I have the time to address. I didn't volunteer before because I disagree with the idea, and while I know now and then it will happen that someone posts in the forum and gets a "put it in JIRA" answer, I think promoting it is a bad thing in the medium and long term.
Webunny
Posts: 1201
Joined: Sat Apr 28, 2012 1:30 pm

Re: Response to June 2014 Meeting Minutes (an outsiders view

Post by Webunny »

dizt3mp3r wrote:Too many words and too much pessismism, often the case on this forum, not just this forum. Don't mean to be rude, really. Trying to keep it upbeat.

The point is that ReactOS has marketing opportunities to get people to start using bits of it, to get people involved and to encourage use/feedback/positivism.

If I had something in my hand that was marketable and usable then I'd say get it out there in order to push ReactOS to the wider world. That might be the case with ROS explorer new and possibly some other components as they mature.

These suggestions are good ideas (but not necessarily technically) and if the ReactOS organisation was more profit-driven, it would probably be forced down this sort of route. It isn't profit driven, and we accept that it will be driven by devs not wearing the 'profitability hat' but there are those that want to increase ReactOS profile out in the real world and so we come up with these little ideas that could work given some positivity and the right mindset.
I prefer to look at it as 'realism'. ;)


Optimism is all good and well, but, just as pessimism, it fails to address the observable reality in a more correct way. There is no doubt in my mind, that all this talk will turn out to be futile, and the devs will not change their focus on a sole 'completely working' part. They will continue with different parts, to get ROS working as a whole, just as they have been doing for years. and... I can't really complain about that this time, since it seems to be the most sensible thing to do. I simply do not concur with the idea that it would be better to concentrate on one component and try to get that working. Most people are just not interested in that. As always, it's an IT-mind thinking IT-things with IT-interests. I understand fully. Yet, it's not how the common users think about things. Most will not even bother testing and putting it on JIRA. :P This is no pessimism, it's just the way things are. It's also understandable, because most people just are not interested in the inner workings or IT-related tinkering. So I don't see the appeal FOR THEM, and thus I also don't see the 'bonus' of broad appeal which is mentioned here, let alone a commercial one. At most, it would be of interest to a minority with a small market.

While not agreeing, thus, I'm still rather positive in what the devs do, and, indeed, in that we should continue as we do, and try to make ROS working as a 'whole' OS. I also think that is what is going to happen (aka, continue), and I'm pretty sure I'm right in this. Wanna bet? 8-)

@mrugiero:

Yes, I know what you mean. While I understand people (well, the casual testers/normal users) complaining about JIRA and saying it would be simpler on the forum, it is true that, if you promote that, more and more would do it, simply because it's the most easy way. Before you know it, it's a full-time 'occupation', where even those that might have put it on JIRA before, now also place it in the forum. As said, it's a difficult issue. All in all, you do need a system that properly captures and follows up the bugs. So a forum can't replace that. Putting it in JIRA by someone else *could* be a solution to some particular posts or posters, but it's not really a good thing if everyone was going to use it; it then becomes a burden.

In first instance, I think JIRA should be promoted, to be frank. If it really isn't done otherwise because the poster finds it too difficult or just can't be bothered, and it's an interesting bug/fix, then, yes, I guess someone (else) should put it there.
Post Reply

Who is online

Users browsing this forum: Kornel and 42 guests