A Hubsoft Rundown

Site Announcements
Locked
Toast

A Hubsoft Rundown

Post by Toast » 27 Aug 2009, 10:49

So i thought i do a review for our users here on the forum about functionality and hub software, its about 8 months since DSHub was discontinued by Pietry. So what has happened since has there been progress in other hub software or is the hub software still lacking functionality?

Well I'm gonna try and go threw the the currently available software out there for the hungry ADC hub owner so they can make the right choice for them. Lets start by analyzing whats out there for the normal user.

ADCH++

This hub software has basically been an empty shell so far the latest "stable" release is more then an year old and the gui that was in the works seems kinda dead since there is no trace of Sulan and the code but that doesn't mean that the project is dead cause things are brewing in the BZR as of late due to Poys effort to make it into a working hub software for the normal user.
from DCDev Public wrote:[2009-08-27 00:43] <poy> ADCH++ commands now:
- [00:39] <ADCH++> Available commands:
+ban nick [reason] [minutes] - ban an online user (aliases: +banuser)
+bancid CID [reason] [minutes]
+banip IP [reason] [minutes]
+banmsgre msg-reg-exp [reason] [minutes]
+bannick nick [reason] [minutes]
+bannickre nick-reg-exp [reason] [minutes]
+help
+history [lines] - display main chat messages logged by the hub (aliases: +hist)
+info [nick or CID or IP] - information about a user, or about the hub if no parameter given (aliases: +userinfo, +stats, +hubinfo)
+listbans (aliases: +showbans)
+loadbans (aliases: +reloadbans)
+mass message [level] (aliases: +massmessage)
+regme password
+regnick nick password level
+test
+topic [topic] (aliases: +changetopic, +hubtopic, +settopic)
This kinda looks promising but there are some stuff missing still like configuring hub from within cause most hub owners don't wanna open up XML files every time they wanna change stuff also a simple first time configuration tool would be nice for those that aren't that experienced with how to setup a hub software since GUI is still missing.

on the matter of GUI while im on the subject i know that most developers want to include a GUI for their software, i can understand this to a point a GUI takes performance away from the software etc.

But a standalone GUI for those that want a software would be nice that way users could have the best of two world non GUI and GUI like Pietry did with DSHub it can run with GUI or without and looking at the current market of ADC Hubs out there its pretty clear that DSHub still holds its ground after the discontinuation.

ADCH++ has 2 developer: Arnetheduck and Poy and updates are kinda irregular at best since they also maintain DC++, that has effected the popularity of ADCH++ due to lack of maintenance lets hope that this changes for the future if anyone wants to help out and provide missing stuff to ADCH++ such as scripts (lua/python) or ideas for the hub software don't hesitate check out the recent version in BZR and help out.

µHub

This project is one that i have high hopes for the project of Jan Vidar Krey this poor overworked opera developer has really put effort into stability his hub software currently holds the record for most users on an ADC hub (2k users).

Now looking at at this hub software is basically the same idea as ADCH++ no gui, xml files configuration not the ideal for a beginner either but it has good documentation thats always a big plus for anyone that wants to learn how to use the software.

The get started guide is for Linux it could use an windows version just for clarification for the beginners most of the commands works but not all, it also contains a how to compile guide for both Windows and Linux that way they can checkout the latest version on Github i know that this project plans on having a WebUI for interface that is certainly a nice idea if its done correctly we have seen this idea tried before and semi fail due to security reasons.

Updates for uHub has been really nice in the past and we here at ADCPortal hopes it will remain that way in the future.

Again if you the user want to help out then download a copy of the hub software and suggest ideas at the official tracker of uHub, that way Jan Vidar gets what users wants in the hub software.

The command list so far it might look thin but the basic things are in there.
from uHub Developement Hub wrote:[09-08-27][15:14:34] <uHub Central> *** help: Available commands:
!help - Show this help message.
!stats - Show hub statistics.
!version - Show hub version info.
!history - Show the last chat messages.
!uptime - Display hub uptime info.
!kick - Kick a user
!ban - Ban a user
!unban - Lift ban on a user
!reload - Reload configuration files.
!shutdown - Shutdown hub.
!myip - Show your own IP.
!getip - Show IP address for a user
!whoip - Show users matching IP range
!broadcast - Send a message to all users
!log - Display log
!crash - Crash the hub (DEBUG).
Luadch

Last years winner of the "ADCPortal hub software project of the year" went kinda quiet for a long time it recently got a second wind when Blastbeat put some effort into it changing the core to libevent instead of his custom core thus improving performance.

The project also abandoned the old forum for a new host app forum on Sourceforge.

The project could use a new release on Sourceforge to get the message out that it is alive again and the SVN could use a cleanup since there are 3 different versions in the SVN.

I recently talked to Blastbeat about a simple GUI to hide the DOS window and at the time he seemed up to the idea of doing this he was also planning a rewrite of the project if memory serves me correctly.

Another good thing about luadch is the script base thats growing on its forum since the re-launch both in german and in english language with scripts like google bot and fresh 'n' stuff.

Conclusion

Developing open source isn't always easy everyday life vs Internet time is always a factor to be taken in if updates doesn't come out in time its always up to the developer to inform the user if such problems should appear so that the user base of the problem doesn't go suffering.

Too maintain a software successfully it takes time and dedication to the users the project has too take in what the general mass wants in a hub software making it a user friendly software, GUI seems to be what the masses are screaming for in this day and age and the only hub software that has it is a discontinued one that says a lot about the user input in the ADC side.

So my question to the hub owners out there what are your thoughts and needs cause i think i got most of it into this post but please comment

----------------------------------
Toast - ADCPortal Maintainer

Aqualung
Newbie
Posts: 8
Joined: 20 Feb 2008, 20:06

Re: A Hubsoft Rundown

Post by Aqualung » 27 Aug 2009, 19:33

Thanks Toast thats a very useful update to the current state of affairs with hubsofts.

Although not a hubowner ,i will comment from the view point of " if i were to be one ".

To me the most important addition has to be the Gui , rightly or wrongly hubowners want to be able to see whats going on with their server and connection and feel they have some semblance of control over what is going happening ,the simplest way is through a gui.Without this one addition i feel no present hubowner will be keen to change to adc.

Possibly the stance our governments are taking over file sharing might be another important reason, encrypted transfers are what will be looked for by new users, hub - client as well as client - client so that all information passed on DC++ is encrypted that should be the goal , i understand the limitation of encryption ie given enough time and need , anything can be read but lets not make it too easy for them to read my data streams without heavy investment in Blue Gene style supercomputing.

Within reason how many users a hub can hold is way down the list , looking at hublists a 500 user hub would cover more than 90% of hubs in the public domain so size is not particularily important in my view at least not until we all have 100/100 connections which will be long after i am gone and we will be transfering files by teleportation by then :) , for the long term health of DC new hubowners are always required and need as much help as they can get and the majority dont care about running a large hub they just want a group of like minded friends.

If what i read and am told is correct ADC has the ability to do all i ask and way way more than nmdc ever was , to prove its worth at least one hubsoft has to be able to match ynhub as hub in a box.
I am certain the people i see working on projects now can do that, perhaps they just need to put aside their own projects for a couple of weeks and all work together to produce one working,functional hubsoft then the ADC show can really get on the road.

Many thanks to all who contribute anything, no matter how small ,to dc software, without you we filesharers would have been stuck on many occasions in the past and i am certain in the future, the pay sucks but i for one and i am sure i speak for many many more,am extremely grateful for your contributions and long may they continue.

Spookie
Junior Member
Posts: 19
Joined: 05 Dec 2007, 16:40

Re: A Hubsoft Rundown

Post by Spookie » 27 Aug 2009, 19:54

5 cents from a hubowner.

I have run DSHub , I like the GUI, think almost every hubowners like gui.
Dshub have a very good chatcontrol that all other adc hubsoft is missing.
Other hubsoft is missing lot of commands that i´m used to have running Dshub.
I also think if hubowners should change to adc they would look after a "hubinabox" soft.

Spookie

Pietry
Senior Member
Posts: 328
Joined: 04 Dec 2007, 07:25
Location: Bucharest
Contact:

Re: A Hubsoft Rundown

Post by Pietry » 28 Aug 2009, 08:15

Because we are making open source software ( free of charge, only in spare time ), we can;t make everything people want ( gui, performance, stability, features and so on ). Devs need to concentrate on something otherwise we will have a mess , working on gui will surely degrade the time given for performance and so on, and viceversa. I think there should be hubsofts which focus on performance ( like verlihub on NMDC ), hubsofts which focus on easyness of use ( ynhub ), and hubsofts which focus on features and extensibility via scripts and etc. ( ptokax ).
Then the hub owner can pick what he needs more.
Just someone

Aqualung
Newbie
Posts: 8
Joined: 20 Feb 2008, 20:06

Re: A Hubsoft Rundown

Post by Aqualung » 28 Aug 2009, 12:38

We do understand the massive amount of time spent to produce and support software and as i say it is most appreciated.

My comments were not as criticism in any way just as a " what i would need " to tempt me to run a hub.There has to come a point in the development where a wider feedback is required from people running software on widely different configured machines,and until there is a gui that in my opinion is not going to happen.

Pietry
Senior Member
Posts: 328
Joined: 04 Dec 2007, 07:25
Location: Bucharest
Contact:

Re: A Hubsoft Rundown

Post by Pietry » 28 Aug 2009, 12:49

Verlihub per se has no gui ( one can use some http interface but still ) and it has a huge success, by focusing on performance and stability.
I agree that gui is important but for me ( example ) performance is way more important. i dont want a hub having a gui but crashes on 500 users or uses all my memory.
Just someone

Crise
Senior Member
Posts: 139
Joined: 10 Nov 2007, 21:34

Re: A Hubsoft Rundown

Post by Crise » 29 Aug 2009, 11:04

Pietry wrote:Verlihub per se has no gui ( one can use some http interface but still ) and it has a huge success, by focusing on performance and stability.
I agree that gui is important but for me ( example ) performance is way more important. i dont want a hub having a gui but crashes on 500 users or uses all my memory.
Pietry, no-one is saying performance is not important... it is very much the contrary but the entire point in the discussion regarding usability in general or (G)UI's etc. is that a protocol that is barely used will be wasted.

It is good to have different projects to fill different needs but all things considered it is not good to neglect the other needs either... Verlihub for instance while focusing on performance and stability still does not entierly neglect neither usability nor extensibility (web interface and scripting).

What I mean is... surely a project thriving for performance, such as Verlihub, could improve it even more by taking scripts out of the equation but then both usability and extensibility would suffer.

Also your 500 users example (or perhaps characterization would be a better word) is not very good one... for one-man projects the order of things is obvious and we all know that no-one in their right mind should even think of creating a robust UI (be it desktop GUI, WebUI or whatever) if the hub soft core is in a state such as you describe.

Locked