Beta Tinyfugue - A Mud Client

[Esn]

Let that be a lesson to everyone. Never, ever, ever, ever criticize a dev for whatever reason.

I disagree. Devs are no gods (well maybe notaz, ari and exo, but I digress)

Devs like any other person need to be told if he is wrong. If he cant handle critics, then just too bad.

Well, yes, "too bad" is the point. But "too bad" for the community from which someone criticized the dev, not "too bad" for the dev.

 

[Dunny]

Let that be a lesson to everyone. Never, ever, ever, ever criticize a dev for whatever reason.

I stand by my constructive criticism, and if StreaK wants to spit his dummy out then that's his problem. I also think it's a good thing that these apps are not on the repo - if StreaK can't find the time to finish them because he wants to work on his next ARM-build then that's all to the good. The repo is fast becoming the go-to place for Pandora software and if the vast majority of it ends up as unfinished, rushed builds it will hurt the Pandora's reputation (and that of the whole scene) very quickly.

Do you really want software uploaded that does such dangerous things as multiple NAND writes? It's not so bad for us at the moment because we know to stay away from such software, but if the Pandora becomes more widespread then less technically minded people might not understand what they're doing to their Pandoras by installing such software.

D.

 

[Esn]

Do you really want software uploaded that does such dangerous things as multiple NAND writes? It's not so bad for us at the moment because we know to stay away from such software, but if the Pandora becomes more widespread then less technically minded people might not understand what they're doing to their Pandoras by installing such software.

It's possible to designate a release as "beta" on the Repo.

I suppose I can agree that something that has a potentially dangerous behaviour such as frequent writes to the NAND can maybe be hosted elsewhere. But even so, a large WARNING: THIS WRITES TO NAND in the Repo description might do the trick. And let's face it, the Pandora is not becoming widespread any time soon.

I certainly do not agree with the notion that only polished ports should be on the Repo. I think that almost everything, except for programs that don't work at all or have dangerous behaviour, should be on the Repo. An unpolished port can still be extremely useful. The difference between "polished" and "unpolished" is pretty arbitrary anyway.

Anyway, StreaK explicitly stated that no port of his, in whatever state, will be on the Repo in the future, so it's not about that.

This is about a breach of contract between the dev and the community; the dev does charitable work in exchange for gratitude and attention. In this case, the dev perceived that the community wasn't holding up the "gratitude" part of the contract, so he raised the price for the "attention" part of the contract. All future updates being at the dev's site = increased attention, thus offsetting the loss of gratitude (if he uploaded to the Repo, people would be able to receive updates automatically via PNDstore; this way, they have to keep visiting his site). Meanwhile, making dozens of apps inaccessible (perhaps temporarily) = rebuke, reminder of power relationship.

Well, it's convincing to me. It doesn't cost much to be nice to the dev.

P.S. Did anybody manage to download WeeChat before it was removed? I was looking forward to using it...

 

[Dunny]

This is about a breach of contract between the dev and the community; the dev does charitable work in exchange for gratitude and attention. In this case, the dev perceived that the community wasn't holding up the "gratitude" part of the contract, so he raised the price for the "attention" part of the contract. All future updates being at the dev's site = increased attention, thus offsetting the loss of gratitude (if he uploaded to the Repo, people would be able to receive updates automatically via PNDstore; this way, they have to keep visiting his site). Meanwhile, making dozens of apps inaccessible (perhaps temporarily) = rebuke, reminder of power relationship.

This is so wrong it's unbelievable. I made several references to my appreciation of his work and my gratitude (ref: ReMood) but I was also voicing my concern that he was concentrating more on volume of "ports" than he was on making quality software available for the platform. If I'm not allowed to voice those concerns for fear that the dev might remove his software in a childish fit of temper then I'm afraid that this community is already on an irrevocable path down the proverbial toilet.

You seriously think it's not only cool for a dev to behave that way, but also to encourage that behaviour?

D.

 

[Esn]

You seriously think it's not only cool for a dev to behave that way, but also to encourage that behaviour?

As long as devs have the power to remove programs from circulation, yes. If there was a site mirroring Repo uploads, or a two-way PNDstore, then probably not.

I'm confident that most devs wouldn't use the situation to their advantage even if they could, anyway, because everything has a price. The squeaky wheel may get the grease, but it is also the first to get replaced.

 

[ivanovic]

You seriously think it's not only cool for a dev to behave that way, but also to encourage that behaviour?

As long as devs have the power to remove programs from circulation, yes. If there was a site mirroring Repo uploads, or a two-way PNDstore, then probably not.

Honestly, it is *good* when the devs have control over the stuff they do. And yes, IMO constructive critics *are* important. For me stuff like "hey, there are problems on vanilla Pandora with the stock OS" as well as "could you improve the keymapping to make use of the normal keys?" and "please make sure not to write into the NAND whenever possible" clearly counts as constructive criticism. Without things things would never improve. Though of course devs can also easily state something along the lines of "sorry, I don't have the time/knowledge/whatever to look into this". And yeah, most will welcome patches, especially if it is just a case of simple default configs or run script adjustments (which might be the case for many issues like writing into NAND or button mappings).

 

[Dunny]

You seriously think it's not only cool for a dev to behave that way, but also to encourage that behaviour?

As long as devs have the power to remove programs from circulation, yes. If there was a site mirroring Repo uploads, or a two-way PNDstore, then probably not.

That is pretty much the worst way a dev could abuse their powers, and should not be encouraged in any way, shape or form.

I'm confident that most devs wouldn't use the situation to their advantage even if they could, anyway, because everything has a price.

Folks like Notaz make good software. They have the pandora in mind when they build and when they code. I see many, many threads of "complaints" that this game or that game doesn't work well in PSX-ReARMed, or that for example Homeworld crashes at certain points... and the devs in these threads haven't "taken their ball and gone home" in the way that StreaK has - and there is no difference between these complaints and those that I myself have voiced. The issues with StreaK's builds are glaring, and have nothing to do with the intended functionality of the app itself; they are fundamental problems with the porting process that StreaK employs.

Unfortunately, StreaK isn't the only one who does this and the official and non-official sources for pandora apps are littered with these. Many games that have been ported require that the user use non-obvious controls, from the Doom and Quake1/2/3 ports to homebrew that has appeared - most authors of those games looked at the Pandora and thought "cool! I can do a Pandora build" and basically set up a toolchain to make the build and then wrapped it up in a PND (hell, at least one of the RIOT compo entries didn't even use a PND and expected people to use a terminal to launch it!) with little to no modification to match the Pandora's unique qualities. This sort of porting results in sub-standard apps for our platform and acts as a very poor advertisement for the community.

Take the Quake3 ports that Pickle (very kindly) made - the Pandora has a mouse, right? It's whichever analog nub you assign it to. Hence, Q3 should work fine with a nub for the mouselook. It doesn't, because the nub accelerates the mouse rather than moving it - which is fine for the desktop, but results in an awful experience in an FPS. Having asked Pickle to address this (not just me, either - many have), he answers that he is currently learning more about OpenGL-ES and might get to it at some point, if his interest returns. It's not an attitude that will result in a better port, but at least he's honest and hasn't removed his apps and hidden them on a facebook page! Personally, I'm more inclined to have a look at it myself than I am to look at any of StreaK's code for just this reason.

So no, the fact that a dev can abuse his position of power is still no excuse for pandering to them when the work they have done shows that they don't really care for the platform they claim to be supporting.

D.

 

[Esn]

That is pretty much the worst way a dev could abuse their powers, and should not be encouraged in any way, shape or form.

The best discouragement is the preemptive sort. If that doesn't exist, why make the community suffer? What goes around, comes around. It'll come around eventually.

You seriously think it's not only cool for a dev to behave that way, but also to encourage that behaviour?

As long as devs have the power to remove programs from circulation, yes. If there was a site mirroring Repo uploads, or a two-way PNDstore, then probably not.

Honestly, it is *good* when the devs have control over the stuff they do.

Well, as a user, I can't fully agree. When a new version of a program is causing problems for me yet the group developing it refuses to provide links to an old version, I am glad that there are 3rd-party sites that have links to download the old versions of the software. Sometimes a software even sells out to some third party and the "official" up-to-date version becomes riddled with spyware. At times like those, I am glad that users have some power as well.

most authors of those games looked at the Pandora and thought "cool! I can do a Pandora build" and basically set up a toolchain to make the build and then wrapped it up in a PND (hell, at least one of the RIOT compo entries didn't even use a PND and expected people to use a terminal to launch it!) with little to no modification to match the Pandora's unique qualities. This sort of porting results in sub-standard apps for our platform and acts as a very poor advertisement for the community.

Usually better to have something than nothing at all. For example, sebt3's port of Pencil is a classic example of a very quick port that has a number of bugs - yet it is also pretty functional if you work around them. Same thing with slaeshjag's port of Audacity - it has problems saving to FAT file systems, yet it's very useful if you work around the problems.

The whole point of having an open system like this is that you're going to get all kinds of software, and some of it buggy. If everyone waited to release apps for the Pandora until they were perfect, we'd have 10 times fewer apps. And the only way to change this is to pay people, and nobody on the team has any money for that. So I'm thankful for any new port there is, and try to do my part to make things easier for users by maintaining those software lists which have links to discussion threads where any questions can be asked, issues raised and tutorials written.

 

[ckblackm]

Ok, so I've spent a while updating the wiki lists, painstakingly going through the list of Streak's ported apps and cross-referencing them with any links posted on his Facebook page.

All in all, 29 apps and 4 games have been rendered inaccessible (as in, the only place where they were uploaded was the Repo).

The following apps have been removed: BabyPanda, BeeDiff, BFRunner, Chiffer, ClipGrab, Cobras IDE, CScreenie, DJView4, EasyPaint, Extreme Image Converter, ImageViewer, JED Editor, Joe Editor, LXTask, LXTerminal, MikMod Player, MooEdit, muCommander, Nagaina, NetFleet, qPSP Manager, QtCumber, qtFM, QtMindMap, Quicky, Repository News, XHTML WYSIWYG Editor, WeeChat, zNotes.

The following games have been removed: FairSeaBattle, HexGlass, Pairs, QCheckers

Let that be a lesson to everyone. Never, ever, ever, ever criticize a dev for whatever reason.

By the way, if anyone still has any of those apps on their computer, I'd appreciate it if you could upload them somewhere (for example, MediaFire.com) and post a link.

the sinclair zx81 emulator was also removed (sz81). I had asked him to compile it for me specifically. I have the pnd still. I've also got several others on the above list. (CScreenie, BabyPanda, Repository News, WeeChat). I believe he also had a version/fork of Dosbox that's no longer there either.

Christopher.

 

[Esn]

the sinclair zx81 emulator was also removed (sz81). I had asked him to compile it for me specifically. I have the pnd still. I've also got several others on the above list. (CScreenie, BabyPanda, Repository News, WeeChat). I believe he also had a version/fork of Dosbox that's no longer there either.

Christopher.

I think all of the emulators still have working links for them. At least, I updated the emulator list with them. I hope they still work. I downloaded the DosBox emulator from this link, which seems is from August 16.

If you could upload those PNDs somewhere, it would be much appreciated. Well, or you can wait a bit; maybe StreaK will do it himself.

 

[streak]

I recall seeing a quote from you a while ago - a goal stated as something like "got to fill up the repo's front page with my ports" as if the point was to be the guy who has ported the most apps and games... And then falling back to the old OSS nugget of "well, you could just do that yourself you know, it's open source" when questioned as to why they just build an ARM binary rather than actually put any effort in.

This is the only reason for me to remove apps from repo. If You thinking this way that this is my real statement, and not a quote told for fun on IRC some time ago-*. If one person thinking in this way, then there might be others, and i dont want anyone to think it's true.

And other thing:
Im not really angry, just a little bit surprised that it turns into somehow Dunny's-Crusade for better ports [and it somehow focused on me]. In this way it'll never work, because every porter / developer had other points of view. My is: Its better to have 40 [different] functional apps than 1-app hyper-polished, and this it obviously different view from Dunny's view.

*- My exact quote on IRC was: "taking over the repo news page with my ports "

 

[sebt3]

Dunny is not after over-polished stuff. but at least correctly check to ffit the few standards we've setup between dev's and users, things like :
- Configurations files and all the other runtime generated files goes to the appdata folder.
So if the user wants to restart from scratch he dont haave to dig for a hidden folder in $HOME but just flush the appdata.
- Maps games controls to pandora's one
So the user don't have to use convulted control on the pandora but use the nicest buttons available
- Scale the game to fullscreen but dont harm the gfx for nothing
So the users can enjoy the game to it's fullest
- Do a comprehensive PXML.xml
So all the nice feature the repo provide are avail for us all.

Each of these requierement are a 5 minuts jobs at much. So all we are asking as a user is that you put these fews minuts to reach our commons standards.

 

[Dunny]

This is the only reason for me to remove apps from repo. If You thinking this way that this is my real statement, and not a quote told for fun on IRC some time ago-*. If one person thinking in this way, then there might be others, and i dont want anyone to think it's true.

This is a personal peeve of mine, yes - and maybe my original post(s) came across not quite as I intended - but it certainly looks that way. If someone ports something and then, before having made necessary changes to suit the Pandora (controls, correct PND categories etc) and then proceeds to the next job with nothing more than a "beta" tag and some handwaving kind of "I'll get to fixing stuff later" then it really paints quite a bad picture.

And again I'll reiterate: It's not solely targeted at you per se; there's a lot of porters in the community that do the same thing and it's just your misfortune that I used your thread as my soapbox, which I truly apologise for. Your work on ReMood was nothing short of excellent and with a little more attention it could be excellent. It's already better than the other Doom ports out there - it's the only one that does transparencies properly on my Pandora.

But all that said, it's disappointing to me that we have an excellent platform whose main strengths appear to be ignored.

And other thing:
Im not really angry, just a little bit surprised that it turns into somehow Dunny's-Crusade for better ports [and it somehow focused on me]. In this way it'll never work, because every porter / developer had other points of view. My is: Its better to have 40 [different] functional apps than 1-app hyper-polished, and this it obviously different view from Dunny's view.

*- My exact quote on IRC was: "taking over the repo news page with my ports "

I'm not talking about "hyper-polishing" apps - I'm talking about taking a little more time and effort to make sure everything works to the strengths and weaknesses of the platform. Your desktop apps work fine, I'm sure - but if they write to NAND without explicitly needing to (ie, a simple path fix under a compiler flag won't sort it) then they're actually contributing to a dwindling lifespan of the Pandora. I'm sure none of us want that. So as Sebt3 says, get the controls sorted and a proper PXML and it's all golden - and an app or game that really shows off what the Pandora can do.

D.

 

[skeezix]

See, what we've got is really..

When you get a .deb from a Debian repository, you don't have to ask if its Debian-ified (putting things into /etc in the Debian stancdard layout), it _is_; if it isn't, the .deb is rejected or fixed, until it does follow 'standards' or 'conventions'.

Likewise, when a user gets a 'pnd', they just assuem it is following Pandora conventions.

We could do somethign in the PXML, like..

<port-conventions>
<target device='openpandora-v1' port-style='driveby' input-map='incomplete'>
</port-conventions>

This way the system and repos coudl say 'this app works, but has the following deficienies'.

That just seems overkill

But there is a contract between the user and the developer, the assumed conventions. If they're broken by a few apps, then the user should be able to decide if they wish to use the app, or offer code patches to 'fix' the missing bits. The problem we have is the user has been unaware, and thats not right.

In IRC I just noted.. its like a 'license'; if a dev doesn't have time to do the full port (or doesnt' care to, hey, its his free time!), thats fine, nothing wrong with it; he's not a pariah, and still is owed thanks for doing the code or port at all. But he should explain to the users what the pnd is. If the description is a copy/pasta from the source website and does not define the pandora changes, or the pandora conventions broken, the documentation is 'bugged' and in error.

Thats all, no stress or drama is needed; StreaK just needs to post his 'license' for peopel to use; they shoudl know that some ports are full ports, some have a few oddities due to how far he wants to go. Fine, thats all.

--

in email dev list, I've proposed the idea that pnd-run can in turn report apps that have 'side effects' (if the home dir contents are changed during the duration of a pnd.) Pnd-run.sh could pop up a box saying 'this app has changed the homedirectory' (neither good nor bad, just a statement) and also offer the option ('dont' bug me with this info anymore').

That would help.. so users can find out apps that produce changes, and decide what to do.

--

But I don't think going much furhter is wise; no need to belabour PXML or the like .. respponsible devs will make responsible pnds.

jeff

 

[streak]

skeezix .. And Your idea is very good problem-solver.


about my quickports:
Sometimes i dont care about to use ENTER instead of [Pandora's] START key in my ports. I still think that if theres no STANDARDS for OpenPandora [for example -> from my side i miss the proper "BACK". If we have SELECT and START i think it would be nice if pandora will have "BACK"]. In this case, im not really convict'd to use pandora keys to swap with ESC key. ESC key is still better.

 

[skeezix]

Where 'not well defined', its up to you of coruse (it always is, as a porter, but more so when its fuzzy.) If you hook up somethign to Enter or not, whatever, thats youe choice. Peopel can complain or suggest, but its not at the level of 'pandora convention', thats just 'user preference', a whole other story.

Convention is .. making users use awsd instead of d-pad, say. Just seems out of place, unless there is good reason. Likewise with appdata.. its well defined, its there, users expect it, so not doing it is considered 'bug' (and should be.)

But like we said in IRC..

Just put this single sentence in your description at the front:

'This particular pnd may not follow all Pandora conventions, as my time is limited and I might have missed one'; without saying that, the user rightly assumes pandora conventions followed. IF youre' not sure, put once sentence in, and everyone is happy.

Just because its the internetz, doesn't mean we need to make everyone angry at each other

jeff

 

[Dunny]

skeezix .. And Your idea is very good problem-solver.


about my quickports:
Sometimes i dont care about to use ENTER instead of [Pandora's] START key in my ports. I still think that if theres no STANDARDS for OpenPandora [for example -> from my side i miss the proper "BACK". If we have SELECT and START i think it would be nice if pandora will have "BACK"]. In this case, im not really convict'd to use pandora keys to swap with ESC key. ESC key is still better.

A fine rebuttal, so here's my 2 penn'orth on this particular issue:

The Pandora is seen by most as a gaming device. I know it's not just a gaming device, but people in my experience have always seen the D-Pad, Start/Select and ABXY buttons before they even look at the keyboard. This is in part to do with how you hold the Pandora, and also because the keyboard is very busy and hard to decipher at first glance.

Like the NDS/GBA et al, people expect therefore to have a button to select menu options (usually say, A) and one to back out of menus (X or B possibly) - the ESC key is very much non-obvious here because firstly it's hard to find (FN+Q) and secondly because it's non-standard outside of PC usage. And the Pandora, as a games machine, is not obviously a PC. We know better, but most see it as a device like the NDS or PSP, and even those that don't expect the same conventions as other handhelds; many of us own Canoes or Wiz/GP2X especially around here. None of them had a "BACK" key either.

Your decision to use ESC is up to you at the end of the day, but I will probably whine about it when you ask for feedback ;-)

D.

 

[Esn]

Well, as long as everything that was removed does get re-uploaded somewhere, I figure it should be okay... will your site have an RSS feed for software news, StreaK? That would make my job of keeping the wiki up to date so much easier.

Dunny is not after over-polished stuff. but at least correctly check to ffit the few standards we've setup between dev's and users, things like :
- Configurations files and all the other runtime generated files goes to the appdata folder.
So if the user wants to restart from scratch he dont haave to dig for a hidden folder in $HOME but just flush the appdata.
- Maps games controls to pandora's one
So the user don't have to use convulted control on the pandora but use the nicest buttons available
- Scale the game to fullscreen but dont harm the gfx for nothing
So the users can enjoy the game to it's fullest
- Do a comprehensive PXML.xml
So all the nice feature the repo provide are avail for us all.

Each of these requierement are a 5 minuts jobs at much. So all we are asking as a user is that you put these fews minuts to reach our commons standards.

I agree, this is a nice list of minimal requirements for Pandora software.

skeezix .. And Your idea is very good problem-solver.


about my quickports:
Sometimes i dont care about to use ENTER instead of [Pandora's] START key in my ports. I still think that if theres no STANDARDS for OpenPandora [for example -> from my side i miss the proper "BACK". If we have SELECT and START i think it would be nice if pandora will have "BACK"]. In this case, im not really convict'd to use pandora keys to swap with ESC key. ESC key is still better.

A fine rebuttal, so here's my 2 penn'orth on this particular issue:

The Pandora is seen by most as a gaming device. I know it's not just a gaming device, but people in my experience have always seen the D-Pad, Start/Select and ABXY buttons before they even look at the keyboard. This is in part to do with how you hold the Pandora, and also because the keyboard is very busy and hard to decipher at first glance.

Like the NDS/GBA et al, people expect therefore to have a button to select menu options (usually say, A) and one to back out of menus (X or B possibly) - the ESC key is very much non-obvious here because firstly it's hard to find (FN+Q) and secondly because it's non-standard outside of PC usage. And the Pandora, as a games machine, is not obviously a PC. We know better, but most see it as a device like the NDS or PSP, and even those that don't expect the same conventions as other handhelds; many of us own Canoes or Wiz/GP2X especially around here. None of them had a "BACK" key either.

Your decision to use ESC is up to you at the end of the day, but I will probably whine about it when you ask for feedback ;-)

D.

I must disagree with you on this, Dunny. Games controls should absolutely be remapped to the d-pad and game buttons, but for other software, I prefer it if the keys stay exactly as they are. For a few reasons:
1) That way any guides I find online will be accurate for the Pandora version
2) ABXY are mapped to PgUp/PgDown/Home/End; mapping something else to them may remove those functions from the software
3) The Pandora generally has enough keys in comfortable-enough spots that leaving things as they are is no trouble for the user.

Of course, I love handheld-friendly software like PepViewer and MinAmp as much as the next guy, but most ported software was just not designed for handhelds, so the least damage is done if the existing controls are kept.