Volunteers for appstore updating.

[Wrath Of Khan]

Craigix has stated several times that he needs help keeping the appstore up to date.Well he didnt start a thread so i thought i would.


Any volunteers.Perhaps someone could collaborate with craigix on upgrading the appstore also,adding more sub categories and additional features.


He has stated that hes working 12 hour days on icp and pandora production.I would help except i havent a clue and cant program.It was enough work for me to get my pandora up and running.Im no techie.Really ed and craig work their buts off.Edd goes to bed at 5.30 some nights.3 hours sleep.Mad.Maybe he


could do with some volunteers to help out with his huge workload too.

 

[craigix]

Just posting a list here of what needs updating/deleting would save me hours!

 

[dsleaf67]

Here's a same issue to look into.


When I first visit, I'm already logged in, but if I click on 'top rated' or 'new uploads' (that is the bar not a specify app in list) I will get taken to icontrolpad page that says 'page not found'


If I however, click on pandora logo (top left) first then it will work fine.


Also, probobaly related when it first comes up, I may have to click on pandora logo to have it show updated list on that main screen.

 

[Wally]

I've asked Craig several times and never got any response.

 

[richiz]

same.

 

[gruso]

Hmm. I've always been an advocate of Pandora Apps, but time has shown that the File Archive is the one that has stayed up to date, by virtue of the way it works. Getting Pandora Apps up to date is a good thing, but the fact that a new wave of help is being asked for / offered is not a good sign. It means the site isn't running itself, and this cycle is going to repeat.


A person only needs to visit Pandora Apps once and find a release missing, and they'll revert to the File Archive and never return.


Craig, I believe Pandora Apps has its place, but it's not working very well as a standalone entity relying on a small number of people to keep it current. I think it's time to look at a more integrated approach. One database of PNDs, with various portals that the community can use to access them.

 

[foxblock]

A person only needs to visit Pandora Apps once and find a release missing, and they'll revert to the File Archive and never return.

Or worse, assume that it has not been released yet. Some people (mostly those not much involved in the community and active on the boards) assume that apps.open-pandora is the only and definite place to get all PNDs, so they will be disappointed to see something missing when in fact it actually is not (one unsatisfied customer for unnecessary reasons).

Craig, I believe Pandora Apps has its place, but it's not working very well as a standalone entity relying on a small number of people to keep it current. I think it's time to look at a more integrated approach. One database of PNDs, with various portals that the community can use to access them.

JSON interface or similar, please


Also a different approach to the approval process is in order I guess as this is the most annoying thing that keep devs from uploading their PNDs.


In my eyes it would be really nice to have one centralized and official storage for all PNDs (kind of like a repo), but with an interface to access the data so anyone can build his own website or GUI around it and populate it with (for example) the JSON data. Whether that is dl.openhandhelds or whatever.com or apps.open-pandora does not really matter, but I would prefer the latter option, just so the connection to OP becomes clear.


Maybe that way we can get devs who are currently uploading their stuff to their own webspace to upload it to the central storage, too. That way people would not have to check dozens of different websites or somebody (or a group of people) has to keep track of everything and keep some lists or sites up to data (aka the lists on the wiki)


For example: a lot of people on the German board did not even know about PSXRearmed two months after its release as it did not show up on any of the major PND storage sites, but only on the forum (and wiki).


Oh well, I am derailing this thread, sorry for that, I just felt like that had to be stated (again).


foxblock out


Or just team-up with milkshake to make his site the (future) official one - it basically does everything I mention above.

 

[richiz]

milkshake is working on a sweet repo . Accepts pnds from upload or url.


Hopefully when its done craig will kill the app store but place milkshakes site in place. That way if anyone sees something missing they can just upload the url of where the pnd is. If milkshake finishes it i think he'll be the saviour of this project. I didnt even know pcsx rearmed had had 4 revisions until i trawled through gp2x forum (which i visit far less)

 

[Blue Protoman]

Ah yes, I've seen milkshake's site. Chalk me up as an advocate, though I personally don't mind using the wiki.

 

[Linux-SWAT]

Having two repositories is a bad idea.


The only way to make it a good idea is an "auto-link" between them, so there's no need to perform maintenance _twice_.


What i suggest is :


-remove one of them.


or


-make some php/rsync stuff, which enables the auto-link i spoken of, appstore being the slave site.

 

[craigix]

It will work because we will kill off the old archive when we're ready with pandora apps, which is obviously still in beta.


It's not taking care of itself because only a handful of people have upload rights.


That's the process we're talking about now, pushing it forward, giving upload rights to all users.


I should also add, we're never going to use someone elses 'appstore' for the simple reason that if I or Ed are not the people maintaining it we will be totally screwed if the owner decides to vanish/he does not like us anymore/dies/it gets hacked/ etc.

 

[Linux-SWAT]

It will work because we will kill off the old archive when we're ready with pandora apps, which is obviously still in beta.

Sounds consistent.


Is a package manager on the way ?

That's the process we're talking about now, pushing it forward, giving upload rights to all users.

Hope someone will check there's no rm -rf / or other bad stuff inside .pnd before pushing it into...


Maybe it will be a good thing to force the upload of sourcecode (if any binaries are shipped) too.

 

[Granitehead]

I should also add, we're never going to use someone elses 'appstore' for the simple reason that if I or Ed are not the people maintaining it we will be totally screwed if the owner decides to vanish/he does not like us anymore/dies/it gets hacked/ etc.

If you hosted it it wouldn't matter if the dev disappeared, as you had the source code. Same for all other problems you mentioned.

 

[craigix]

I should also add, we're never going to use someone elses 'appstore' for the simple reason that if I or Ed are not the people maintaining it we will be totally screwed if the owner decides to vanish/he does not like us anymore/dies/it gets hacked/ etc.

If you hosted it it wouldn't matter if the dev disappeared, as you had the source code. Same for all other problems you mentioned.

There is no way I'm digging though someone elses source code to find their bugs. It could take weeks.


I've looked at the source for the file archive and it's a nightmare. We looked at modifying that before going ahead with the new one.


I'm also a bit disappointed that no one has pointed out anything which is out of date on pandora apps yet.


Is it out of date or not?


And the PSX emu seems to be up to date? What's all this talk of it being 3 versions out?

 

[skeezix]

Theres several repositories (and thats fine) ..


apps (craigs) -- pnd only, half the files (devs aren't all uploading there; some just don't know, some don't like it, etc.)


--> cannot be the one stop shop by definition, since it doesn't handle zip or pdf or non-pnds (ie: some pnds like zx81's are zip, so it has the pnd inside along with the source); but more pdfs and random files are handy.. stuff like SDKs, or documentation, or background images, or minimenu theme bundles, etc etc .. they're not pnds.


--> needs a CSS for full-machines as well perhaps; the pages aren't great for on 20" desktop, but great for using on Chromium on pandora


dl.openhandhelds -- historical for gp32 forum, and so popular; its some nasty softeware for sure, but a lot of devs like to use it instead of apps.


--> does support pnd and non-pnd


--> how does it fit on pandora itself? not designed for small screen, so maybe blows up I dunno


wiki -- some contest it is most complete; it definately lacks some apps in either of the other two repos imho


--> to me, wiki is not the best, since it requires much work; any time someone uploads to apps or dl.open, someone else has to come along update the wiki .. just a bad idea imho Just begs for synchronizatin/versioning issues.. esp since two main repos don't have APIs so can't auto-populate wiki


--> wiki would be good place for app-specific docs from community, though, but that doesn't depend on 'repo' approach


sebt3's -- his perosnl page of a hundred apps; he pushes them to dl.open as well much of the time, but not always (?)


zx81's and other peoples -- personal download pages


There _will never_ be a one stop shop; thats okay.


Question on the table is.. what do peoplw want from Craigs app page? He's not a miracle worker, he needs to know what folks need and want, and then figure out what he can deliver


For my part -- I'd say support of non-pnd files (if it wants to be a one stop shop), or just admit to being a pnd-shop.


My input:


1) The app store coudl be the _Default_ pnd stop; just pnds, and for on-device work.


2) We shoudl change links in the firmware so they default to hitting it. So newbs get their pandora and poof, theres a place to get pnds, nice and easy.


3) We add to the firmware (we mean to, just not done it yet) a 'SD preparetion' option, which just does a mkdir of the /pandora/menu and /pandora/desktop directories, so folks have some idea where to put things.


4) We add to firmware so trhat libpnd looks (non-recursively) in root of SDs for pnds. It currently looks recursively in the other locations (and I didn't want root on there since I didn't want it having to scan the users whole SD to find apps, that'd take hours for 32GB cards weith 20,000 files.) But if we make it non-recursive, then users who just drop files on root of SD will still have apps detected, and still be fast.


(and yes, since I do work on the firmware every day, I will do those changes; this is not throwing out for someone else to do)


5) Desktop users directed to pandora-page which lists off the common download locations listed above. Desktop users can find apps anywhere they like.


So..


- keep apps pnd-only and device focussed; no point in caig fighting uphill battle to make it like large screens etc etc. Just keep its focused.


- accept that its a device only frontend; peopel are probabyl annoyed because its not great as a desktop experience.. but it was not meant to be


So there, my thoughts


jeff

 

[foxblock]

It will work because we will kill off the old archive when we're ready with pandora apps, which is obviously still in beta.

I think this is a misconception on your end, because for every (or most) dev to accept apps.open-pandora as the primary/only/definite place to upload his work you will have to at least have the functionality of milkshake's site (aka PXML parsing, better browsing functions, category cloud maybe, PXML verification maybe, JSON interface for local programs and alternate GUIs, etc.) and an alternate user-interface for devices with a higher resolution (aka PC), or even better: a dynamic interface.


(The user privileges/approval process was just _one_ example of why devs chose not use the site)


If such things are already on your to-do list then please excuse me and consider this post redundant.


On the other hand if you are just going to shut the filearchive down and don't implement the above mentioned features you will just make matters worse as more devs will upload their work on their own webspace and you will have to manually add it to apps.open-pandora. Or they just use some other site (like milkshake's one), which also syncs itself with apps.open-pandora to get a list of all PNDs.


In my eyes, the main problem with apps.open-pandora is that there never was a time when you sat down with the devs and talked about what functionality THEY want for the site, because at the end they are the ones doing the work of creating PNDs. Sure, you asked for opinions on single problems like security or the interface (with the limitation of it being a Pandora-centred interface), but there never was a talk about general functionality.


Granted, the devs could have contacted you and asked for functions to be implemented, but you can't expect them to when they are busy doing work and on the other hand expect you to implement it for the best.


I think, such a discussion can still happen (maybe not in this thread, but while we are at it...) so you get an idea of what the devs want, can implement it and at the end everyone will be happy (probably not, but the majority will do).


For starters (and I hate to bring it up so often, I really don't want to drag milkshake in this personally, but his work speaks for himself) milkshake did an open discussion with the devs, listened to their input and implemented functions they demanded and now everyone seems to be very happy how his site is progressing (some people even want it to replace apps.open-pandora).


If you don't want to use his code, the maybe just contact him and talk about what he did with the site and why he did it and after that implement your own version of it.


(Though in my eyes that would be doing the work twice and depending on how well written and commented his code is, why not use it directly...)


Well, that is the state as I see it, I am happy for any clarification and comment.


foxblock out

 

[bzar]

^ I agree with foxblock's post.

 

[mcobit]

I really don't want to drag milkshake in this personally, but his work speaks for himself) milkshake did an open discussion with the devs, listened to their input and implemented functions they demanded and now everyone seems to be very happy how his site is progressing (some people even want it to replace apps.open-pandora).
If you don't want to use his code, the maybe just contact him and talk about what he did with the site and why he did it and after that implement your own version of it.


(Though in my eyes that would be doing the work twice and depending on how well written and commented his code is, why not use it directly...)

+1

 

[skeezix]

craig -- you're the only one who ever thought to kill dl.openhandhelds -- remember, thats also non-pandora devices. But the pandora section alone has more stuff than your site can handle -- desktop sized display, non-pnd files, etc.


I think it is wise to define your scope --


1) go small - device only - you can do this


2) go big - aim to replace dl.open -- you can do this, but its a lot of work; as is, your site is _not designed_ for this task.


NOTE: There will _Always_ be more than one site (and personally, that is as it should be. One site is also known as 'single point of failure', or at best 'single point of control' which is never good for the user ; milkshake's site is the 4th or 5th attempt at a new site, and eventually there will be some.


There is also the wiki etc.


So do not _compete_, be part of the ecosystem.


Craig -- so what is your _Goal_ for apps site?


once that is published, people can argue and bikeshed over whethor they agree with that goal or not


jeff


As to milkshake's site, its future is up to milkshake, no one else. Craig and ED etc are the driving force behind the hardware, and a few of us are the main bodies on the firmware; but the community ecology is not ruled by iron fist

 

[richiz]

Hi craig


No I meant I just didn't know about the newer revisions (was responding to the germans not knowing about pcsx even being released until 2 months later.)


Why not use his code? Seems ludicrous that someone has basically done the work for you with an app store that will essentially update itself along with url upload capabilities. How often, when milkshake finishes, would you modify the source anyway? Maybe a few tweaks here and there and that doesn't require you to have an in depth understanding of it.


Basically its use someone elses source and have a good app store now and you can start to understand the source whilst we have a working one, or wait for you to write one which will take ages and you've said yourself you're busy with pandoras and icps.


All that will happen is when milkshake gets it up and running people will use it as the official app store and when people come on here going "wtf?! This doesn't work!"


They'll get the reply


"Use milkshakes app store and forget pandoras one".


Much like everyone does with the wiki right now.