I am a fairly new unRAID user.  I have also been writting software (mostly Windows) for a living for the past 16 years. 

 

So far, I have been really happy with the product and have recommended it to friends/coworkers that need media storage.  The active and very supportive user community was one of the reasons I decided to buy it. 

 

In the short time I have been here, I have seen no reason to believe that LimeTech is any bigger than 1 or 2 people.  I think that explains some of the frustration that people are having.

 

In my experience I have found that it is rare to find someone that is good at/interested in both system level programming and UI programming.  Obviously Tom has the system thing down.  A trait I have noticed in hard core system guys is that a lot of them don't even really care about an elegant UI.  Give them a command line and they are peachy.  Obviously I have never even met Tom but it is possible that this description fits him.  I don't think that is a good or a bad trait.  It is just what it is.  Hiring a UI developer, even on contract, isn't going to be cheap.  If UI isn't really something that Tom cares about, it may be a tough sell to convince him to spend that money.

 

As far as supporting add-ons goes, I think he has to tread that line carefully.  If it seems like he is "officially" supporting something, people may decide he is on the hook for providing support for it.  That could be a burden that he doesn't want to take on.  Especially if LimeTech is just a few people.  I can see that support matrix getting out of hand very quickly.  Even responding to support issues with a "that problem is caused by a third party widget" takes time and some people would feel like they were just getting the brush off.  I'm not saying these add-ons are a bad thing.  I am happily using some of them and I greatly appreciate the time it took to develop them.  I can just see why maybe Tom is reluctant to show support for them.

 

All that said, I would love to see a modernized UI put on unRAID.  Something CSS based so it would be easily skinnable.  At the least I would like to see that dreaded "restore" button renamed and explained a little better.  Some things like integrated email alerts and automated error/parity checking seem like obvious deficiencies.  I was pretty shocked when I realized they weren't part of the product.  You mean I have to remember ever week or month to go manually run a parity check?!?!

 

Oh well... those are just my 2 cents on the subject.  Like I said at the top, I am really happy with the core functionality of the product.  There just seem to be a few holes that it would be nice to plug.

Excellent first post. Welcome 

 

Im going to throw this one into the pot as well based on that last post;

 

"it is probably fair to say that the community does the majority of the commercial product first line support"

 

I was going to heavily caveat that statement but I dont think I have too, theres no way one or two people could be doing more support than this community unless they do nothing but answer emails all day (there is no phone number that i have ever seen so that adds more weight to that).

 

Thats another HUGE statement if you think about it.

 

I am also going to say again for the record I am huge unRAID faboy and Im not flaming the product in anyway.

CSS is not a solution.

 

A control port (i.e. http://tower:9101/?cmd=xxxxxxx) or a pseudo-device (as we already have in /proc/mdcmd) that exposes and API to unRAID (read and write) with a command set is the way to go.

 

I prefer expanding the existing pseudo-device /proc/mdcmd.  Give me that, and it would be simple to create any interface you want.

Every once in a while NAS likes to start a thread like this to complain and kick Tom about a little.  The last one is here.  It always generates lots of responses and other people complaining and moaning.  The questions seems to focus on ... why hasn't Tom done this and why hasn't Tom done that.  That would be so easy.  What is coming next and when.  Look at what this other product has done.  etc. etc.  These threads have, to date, not generated much of any response from Tom, and generally been pretty frustrating since Tom doesn't engage.

 

I want to remind everyone that Tom has agreed to a significant amount of user requested functionality in the next version. I can't remember everything, but it included read-only parity checks, and checking the spindown status of drives before spinning them up for a temperature check.  He also is considering a way to invoke granular unRAID functionality from another process in support of unmenu and other future projects.  He and agreed to a driver for a new 8-port controller card.  (I know there we others, just can't remember at the moment).  Although Tom has not come out and complimented particular projects, he has consistently complimented the community as a whole calling us key to his products success.  I wish he were more vocal, especially of RobJ and the TREMENDOUS job he has done with the wiki and FAQ, but don't think he has been totally uncomplimentary.  The offer to create the granular calls is certainly a step in the right direction in terms of supporting the community's GUI initiative.

 

I can't answer the questions as to why Tom hasn't done some of the simple things.  It's unfortunate. 

 

But I think we have a great community here, and a tremendous amount of work and effort has gone into making community members more productive with unRAID.  These include the unmenu project (that Joe L. created and I and many other people including RobJ and bubbaQ here contributed to), preClear (another of Joe L.'s brainchildren), bubbaRAID (which bubbaQ has spent a TREMENDOUS amount of time and energy on), and smarthistory (which bubbaQ has also done with NLS's painstaking testing), powerdown (thanks to WeeboTech), and numerous packages and scripts developed by Biggy2872 and others.  These includes UPS support, email notifications, monthly parity checks and others.  hawihoney  contributed mySql.  prostuff1 has chipped in and helped with the motherboard compatibility matrix and other wiki updates.  I already mentioned RobJ and his work on the wki.  (Sorry if I left anyone out).  With all of this functionality it is hard to complain about a lack of capability.  Six months ago little of this existed.  But since then the community went from asking for enhancement to implementing them by themselves.  (This is not without precedent.  If you go over to the Sage forums you will find many tools created by forum members there.)  And now maybe we're moving towards V2 of the GUI enhancement - moving from awk to something else (PHP?) with PsyVision, a very talented Web graphic designer/programmer and new unRAID user.  I look forward to seeing the results, and hope that I may be able to help out and learn something new.

 

Realize and accept that Tom will do what he will do, but if history serves, it won't be fast.  I'm sure he knows that if he wants our opinion, he knows where to find it.

 

I urge the community to continue to move things forward.  Don't get drawn into complaining about Tom.  It is an exercise is frustration.

bjp999 please stay on topic ; no ones complaining except you, so please don't fork or flame in my thread. If you don't like it please refrain from posting or at least stay on topic.

 

 

So we have established the main problem is the GUI. We have also established that unless there is a extremely trivial way/official/portal of adding hacks its not likely to be that popular with commercial clients.

 

So let me ask this question. What is absolute bare minimum necessary that would be required to alter in unRAID official so that the community can develop properly? .... not nice to have... not better if done like this.... the absolute bare minimum?

 

I assume it is hooks in emHTTP. I also assume that we only need one for array started and array stopped.

No ones complaining please dont fork or flame in my thread.

 

Move along if you don't want to comment.

 

I have been wanting to comment along the lines of bjp999 several times in the past to posts by different people, but have always just `moved along'. But, this latest comment from you, NAS, is seriously uncalled for! Quite apart from the sheer offensiveness of it, it's also logically flawed: you, NAS, can start any old thread on someone else's forum, but fellow forum members cannot respond as they see fit within that thread because it's "yours"

 

Moving along ...

its just internet etiquette to not fork threads. can we not have a simple debate with falling into the "how dare you" club.

 

Just because something flies in the face establishment doesnt mean it cant be debated civilly.

 

please please can we get back on topic

 

So we have established the main problem is the GUI. We have also established that unless there is a extremely trivial way/official/portal of adding hacks its not likely to be that popular with commercial clients.

 

So let me ask this question. What is absolute bare minimum necessary that would be required to alter in unRAID official so that the community can develop properly? .... not nice to have... not better if done like this.... the absolute bare minimum?

 

I assume it is hooks in emHTTP. I also assume that we only need one for array started and array stopped.

I assume it is hooks in emHTTP. I also assume that we only need one for array started and array stopped.

The two event triggers that are probably most important to have hooks on are

• After Array Start - to be able to start add-on-processes once the array is available
  
• Before Array Stop - to be able to stop add-on-processes before the array is stopped
  

 

There are two API interfaces needed (at a minimum) to emhttp:

• The ability to send emhttp a signal to start the array.
  
• The ability to signal emhttp, to get it to re-read the actual array status is needed.  That could be as simple as a USR signal.  That way, if we stop the array, by sending mdcmd a "stop" command, we can signal emhttp, and it can get the mdcmd status and will correctly display the current array status.
  

 

With those few "hooks" the bulk of existing emhttp interface could be replaced with an alternative for most of the day-to-day operations.

 

Thinking about it a tiny bit more, if we had only the two API interfaces to emhttp, we could use our own GUI to start and stop the array ourselves,  and the first two event triggers would not be needed as much.  We would be able to code our own.

 

To clarify:

• Event Triggers are needed if we continue to use the supplied unRAID management interface and want to be able to cleanly start and stop additional processes added locally.  They are not needed if the supplied unRAID management interface is not used, as they probably would not be involved at all.
  
• The API Interfaces are not involved unless an alternate user interface is developed.
  

 

Joe L.

 

I assume it is hooks in emHTTP. I also assume that we only need one for array started and array stopped.

The two event triggers that are probably most important to have hooks on are

• After Array Start - to be able to start add-on-processes once the array is available
  
• Before Array Stop - to be able to stop add-on-processes before the array is stopped
  

 

There are two API interfaces needed (at a minimum) to emhttp:

• The ability to send emhttp a signal to start the array.
  
• The ability to signal emhttp, to get it to re-read the actual array status is needed.  That could be as simple as a USR signal.  That way, if we stop the array, by sending mdcmd a "stop" command, we can signal emhttp, and it can get the mdcmd status and will correctly display the current array status.
  

 

Although not part of this message, but within the topic, Realize that Tom rolls things out slowly so there are not overwhelming support issues.

 

This is a perfect response to the "what is wanted/what is needed question"

Tom mentioned to me once before, the possibility of these hooks may occur because of the community support unRAID has been getting.

It's our collective assistance which helps.  Tom has mentioned to me, he thinks this community is one of the nicest around.

Debate on the features we need.

 

You criticize my post as being off topic, when in truth it is responding to the sentiments and statements of others in the thread, including you and posts your herald as excellent.  My opinion is that you’d just prefer to call me “out of order” than try to address the issues.  It is a way that you can ignore the facts and dismiss community initiatives that you have had absolutely no involvement in.   If Tom magically came forth with a beautiful GUI, you’d be the first in line trying to take credit as the though leader of such a redesign – while justifying dismissing the hard work of others.

 

I want to make sure to say this.  I am not being critical of the sentiments expressed here.  NLS, your post struck a chord with me and I hope that Tom reads it thoughtfully.   And it is constructive to get those sentiments on the table and make sure Tom hears them loud and clear.  But I’m not certain the open forum is the right venue (maybe it is).  And I certainly don’t think it should be done in a thread about why isn’t unRAID more successful.  If I were Tom and read even this thread name, I’d be offended (“just for fun”)

 

If you go to the SageTV forum, you will find many many community (and individual) projects to add functionality to the product.  There has been a whole new GUI developed that is not part of the product.    It is very popular there!  The parallels are not perfect, but most everyone installs many of these community components.  It contributes to the success of the product.

 

I disagree with you in dismissing community projects as hacks.  Are they perfect?  No.  Is there room for aesthetic and functional enhancement?  Yes.  Are they hacks?  No.  Should all the leaders here be supportive? …

 

And then you suggest we get back on topic only to begin a technical discussion about the desired features in unRAID to support improved GUI development by the community.   YOU are forking this thread.  ROFLMAO

 

I agree with what ReneV has said, and think that you don’t think twice about being offensive to others, as you brow-beat an absent Tom, and criticize community projects as “hacks” and ugly.  And all the while claiming to be doing it for the greater good.  And what have you done to make it better.  Despite  an appeal to Tom of taking on and improving the wiki and granting you admin access (which he gave you), you have done almost nothing there.  I bet you haven’t had a contribution there in 6 months or more, while RobJ and prostuff1 have picked up the slack.

 

But let’s get back to the main topic of discussion.  Enough of this off topic gibberish.  What stops unRAID from being a major contender?  That’s the topic, right?

 

I contend it has absolutely zero to do with the GUI.  The GUI is neat, professional, and functional.  Not a single one of us looked at it and ran for the door.  And no one that liked its feature set and functionality would run for the door after seeing the GUI.  Could it be better?  Yes.  Are there warts?  Yes.  Is it the #1 thing making it not a contender?  No.

 

Although not part of this message, but within the topic, Realize that Tom rolls things out slowly so there are not overwhelming support issues.

 

This is a perfect response to the "what is wanted/what is needed question"

Tom mentioned to me once before, the possibility of these hooks may occur because of the community support unRAID has been getting.

It's our collective assistance which helps.  Tom has mentioned to me, he thinks this community is one of the nicest around.

Debate on the features we need.

 

I 100% agree with this!

Let's keep it positive.

We have a laundry list, which is sort of "what has to be washed"

 

Perhaps a shopping list of what sort of chrome parts or "flair" do we want to see on our unRAID macihines LOL!!!!!!!! 

 

Perhaps the topic should be, what features would "make" unRAID a major contender.

 

You also need event triggers:

 

Before Shutdown:  When shutdown is initiated, to cleanly close applications and other things that bay have open file or which may prevent clean shutdwon.

After Shutdown:  To turn off UPS, send signals to other devices, set the NIC to WOL and enter s3suspend, etc.

 

On Correctable Error:  Self explanatory

On Uncorrectable Error:  Self explanatory

On Disk Failure: :  Self explanatory

 

(FWIW, I agree 100% with bjp999)

 

Imagine this pseudo-conversation:

 

    Users: Can you open up an API so we can do stuff with your product?

 

    Dev: What stuff?

 

    Users: All kinds of neat things!

 

At this point, what's in it for the Dev?  This is where we were six months ago.  We were in essence, offering vaporware.  We needed to put out money where I mouth was.

 

This is exactly why I did BubbaRaid -- proof of concept as to what is possible.  IMHO, we need to keep plugging along with enhancements, and through trial, error, and re-engineering, we will find what can be done w/o Lime needing to help us, and what can not.  I'm estimating about 6 more months of enhancement development, along different lines, and then we will have MUCH more experience and real-world examples of what works, and what was tried and abandoned.  In BubbaRaid right now, I have stuff that was "quick-and-dirty" to get it working (proof of concept) that when I re-engineer it, will be redone differently.  That is the way RAD worls.

 

We will have a much better picture of what is and is not doable and at what level of complexity.  We started with Awk, and even wrote a replacement for wget in awk!  We have wget in the stock distro now.  I've used Lighttpd and php (and still some awk).    Some others will use something else.  It is an evolution.

 

In a few months, I predict we will have several players, and they will get together and re-engineer the add-ins paradigm, and come up with a new, more permanent structure, with 1-click installation of add-ins.  My personal model is something like Firefox add-ins or the SMF forum extensions.

 

I'm working on a simple script, that when run on a stock unRAID server, will add a complete web server and user interface -- Lighttpd with a scriptable download engine built in (the BubbaRaid on-line upgrade uses this).  All with just one simple script.  This will bootstrap unlimited possibilities.  No more removing flash, copying files, etc.

 

 

Feel free to continue this thread however I no longer have any interest.

 

The "unRAID fanboys" who blindly see only what they want and call anyone who dares to have a slightly different opinion "offensive"  lessen the community for their own "Im more moral/i know better  than you" objectives.

 

Good luck with unRAID; I am reverting to being a simple customer and will act accordingly.

 

Shame it could have been so much more.

NAS -

 

Over the past year (at least) the community has tried very hard to influence Tom's actions and get him to address what we felt were the most pressing concerns.  Tom has focused on his priorities and not shown an interest in debating these in the forum or out.  Asking, demanding, or shaming him into action just hasn't worked, and not for lack of trying!

 

just to keep things more on topic rather than just another hack set...

 

no matter how good your gui looks (and your mock up looks excellent + I agree with your unMenu sentiment hence why i dont use it ... no offence meant)... unless there is a simple gui within the official product to add these bolt ons its still a hack... and hacks are what my tiny poll shows arent viable to improve the core sales/product.

 

OK so it might be easy to add but all these things need to be installable and upgradable via the default gui to be viable.

 

The implication here is clearly that Tom's lack of a feature (or set of features) make community actions “hacks”, “not viable”, and not worthy of your involvement.  You don't think this would be offensive to those involved in community projects?  But worse than being offensive – much worse – is that it discourages future innovation.  Innovation that benefits us all.  Innovation that might get Tom's attention.

 

You may not see it, but POSITIVE community involvement is the best tool we have to get Tom to act.  Case in point - myMain's need to be able to control aspects of unRAID (spinup and spindown) has lead to TOM’S suggestion of a means to access more granular unRAID functionality.  This is what we want.  If our actions and tools renew enthusiasm about unRAID in Tom, and encourage him to support us – that is win-win.  And it's happening.

 

And even if it doesn't, or doesn't to the extent we would like, the truth is that unRAID is quite good already.  Tom continues to enhance it and provide updates based, in part, on our suggestions.  I'd prefer that we not move backwards from here.

 

NAS, I wish and hope that you can find productive outlets for your interest in unRAID, rather than walking away.  I rather enjoy lively debate with you.  But if you can't, I wish you well in other pursuits.

 

Case in point - myMain's need to be able to control aspects of unRAID (spinup and spindown) has lead to TOM’S suggestion of a means to access more granular unRAID functionality.

 

Which was the point of my admonition that we, as developers, need to do development first, to start a feedback loop with Tom.

Who really cares about a flashy GUI, anyways?

Unraid is a media server that works "behind the scenes". As long as it does what it says it will do reliably, be happy you've got what you paid for. Are people really upset when they perform a parity check, and aren't presented with a screen that looks like an iphone? Isn't it kind of like spending time painting the inside of a BMW's gas tank? People seem to get caught up on the silliest things sometimes.

...Isn't it kind of like spending time painting the inside of a BMW's gas tank? People seem to get caught up on the silliest things sometimes.

 

No, its like buying a BMW with a dashboard from a Skoda. Not going to happen and thats business.

Isn't it kind of like spending time painting the inside of a BMW's gas tank? People seem to get caught up on the silliest things sometimes.

 

I love the analogy!  ROFLMAO  ;D

Yes but NAS is closer to reality.

 

Compliments to mark-hc for the analogy it does focus the mind.

Ever seen the dashboard of a NASCAR car?  No bells and whistles.  Pure functionality.  There are no NASCAR drives asking for more icons and colorful information displays in their car.

 

The difference between that and a BMW is marketing and (a need) to appeal to the proletariat masses (i.e. purchasers).

 

That's why marketers, and not engineers, design dashboards for car companies.

 

Following that reasoning, you need a marketer to design the unRAID interface.

Your very likely spot on here.

 

And for laughs (which when i started this thread it was supposed to be)

 

 

 

 

 

 

 

I'll take #3 if you don't mind.

 

The thing is that BMW is not JUST marketing.

(and one thing I know is that it will be VERY hard for me to go to anything different than my BMW motorcycle in the future, so it's not what they "sold me", it's what I actually GOT, which is pretty damn close to what they "sold me")

 

Returning to unRAID, ok if Tom doesn't really care to sell and just want to sell to people that have done their research first, then fine be me. If Tom wants to sell larger numbers, the "face" matters.

 

There is also a contradiction in terms with unRAID. The interface is basic "on purpose" because it is functional and hides its power? I would swallow that easy... but the interface is not like that, it is LIMITED (along with ugly). Sorry, that's what it is. I bought it, I don't regret at all (and in fact Bubba REALLY helped to further exploit my unRAID machine), but the facts are facts.

 

Having trouble swallowing the dashboard analogy for unRAID's GUI.

 

Don't get me wrong, I think the new GUI project is great.  All of us hard core users will love it.  And it could even push a few new users towards making a purchase.

 

But most users don't look at the GUI more than once or twice a month.  I am a heavy user, I don't find myself there more than a couple times a week.  The flashy GUI will be lost on the average user, unlike a car dashboard that even the most basic driver uses every second they are driving their car.  The real promise of the new GUI for the average user, is that it will lower the bar in terms of Linux skills required to add new features.

 

A part of me still agrees with the sentiments expressed by mark-hc.  Users just want it to work on autopilot.  (And his analogy was priceless!)

 

I think it would add value if people could start posting images of other NAS user interfaces.

 

This will help put it into context.

 

As a start Naslite M2