Release QupZilla - Best WebBrowser for OP [IMHO] - BETA PND


Thanks for the port!


Sorry I got confused: is this app writing to the NAND or to the SD card? Since I want to preserve the NAND, I only use apps that use the SD card. Especially for a web browser that can heavily write files into its cache. ;)

Nope. QupZilla is NOT WRITING ANYTHING to NAND. Please take a look into qupzilla's option: -> HELP -> INFORMATIONS ABOUT APPLICATION ->


You should get all info where qupzilla writes everytings. Everything [profile/settings/session/data/themes/plugins] points to /mnt/utmp/qupzilla-app/
 
QupZilla is NOT WRITING ANYTHING to NAND.
Sadly it sound like your reputation is done :(


Since HF6, for a PND to write on NAND it have to do it _on purpose_ !


And no, Streak purpose is not to burn your NAND but help you with more awesome apps !
 
QupZilla is NOT WRITING ANYTHING to NAND.
Sadly it sound like your reputation is done :(

..so I decided yesterday to drop all quick compiled apps for OP users. There will be no updates anymore [for OP users since i have better things to do than listening questions about writing to NAND. ]. I'll maintain only my 4 polished apps/emus. CLI Love / DosBox EX / UAE EX .. so no new apps / quick ports anymore.
 
..so I decided yesterday to drop all quick compiled apps for OP users. There will be no updates anymore [for OP users since i have better things to do than listening questions about writing to NAND. ]. I'll maintain only my 4 polished apps/emus. CLI Love / DosBox EX / UAE EX .. so no new apps / quick ports anymore.
Once again :( that's sad. But if I try to wear you shoe I feel the pressure to take the same dispositions :(


It's sad to see that we have done changes in HF6 to stop these fears so you can continu porting stuff like mad without having to handle users fears about your works.


In the end we have :


- More complains that your apps _might_ write to NAND (without getting that it nearly cannot anymore...)


- You stop porting because of these complains


- Complains because the changes in HF6 _might_ broke some PNDs (even if nobody care that there is a _simple_ fix that nobody have cared enough to ask for it)


- Now ED feel the needs to justify our changes in HF6 and become stressed about all this.


So even if we tried to help you to help others, everyone lost here. It's bad to see how people are like sheeps or rats : doing the worst decision when not even getting what's going on :(


@Milkshake : every dev need a motivation. Good feedback motivate, constant bad feedback kill it.
 
Well, to be honest, I've seen more praise from people about StreaKs ports than I've seen complaints.


Some users still fear that an app writes to the NAND, but that's about it.


You will ALWAYS have a few users complain about everything, that's normal.


But you should always ignore them, unless there are more users complaining than loving what you do.


You can never suite everyone, that's just a fact.


If we would've listened about a few complaints from users while building the Pandora and ignored the rest of the community backing up with what we're doing, then the Pandora would've never seen the light of the day.


In the end it's always the decision of the creator of something (be it hardware or software), but I think you should never listen to a few complaints if the remaining users like your stuff.
 
Sad, that you let people annoy you with these things...


Why not just do something about the complaints?


Really, just start over and look at every app, you released to the public. Fix all nandwrites and upload to repo one by one.


@sebt3: Applications, that use g_get_home_dir() or the Qt equivalent still write to NAND with HF6.


Also Applications, that use XDG_CacheDir and QtCacheDir etc. (The names are from memory ;) )
 
It is a shame; StreaK, I appreciate everything you've done so far.


At the end of the day, it's all a matter of perspectives. For most users, they just want an application to work. They (not all, some are quite helpful) find that if it doesn't work exactly as they see fit, well, why would anyone upload or submit something that has a fault? (And as a user, I'm guilty of this, too).


But the users don't see the 20 other things you spend the last five nights working on to even get the application working as it does. Applications take time, and most developers do it out of fun, or as a hobby, so when the fun gets taken away, why develop?


So users: Appreciate your developers. There are a lot of behind the scenes coding and effort that you may never even notice that eats up a lot of time.


@sebt3; PyChess at first was writing to the NAND, too (but not anymore). It was pretty much going out of it's way to do it, oddly enough. Yay for hardcoding paths.
 
@sebt3: Applications, that use g_get_home_dir() or the Qt equivalent still write to NAND with HF6.
Also Applications, that use XDG_CacheDir and QtCacheDir etc. (The names are from memory ;) )

g_get_home_dir use pwent which is the last way to get the real home for a user. that's why I said "nearly".


pnd_run.sh should set all the XDG directory within appdata. If not that's a pnd_run.sh bug I would need to fix ;)


@Mr Rob most my PNDs wrote to NAND at a point or an other. I dont have an issue with that per see ;)
 
Why not just do something about the complaints?


Really, just start over and look at every app, you released to the public. Fix all nandwrites and upload to repo one by one.

If it could be THAT simple. Well, most of my sources lost few days back, because my HDD decided to pass away in a "nice" way: "Im going to take your sources with me to my grave, StreaK , muhahaha..".


As You can see, there was a lot of apps that should be re-done once again [maybe more than +100 apps]. I personally dont want to waste my time again on re-porting / re-compilation some older stuff anymore
 
That is understandable then. But as said, sometimes its as easy as changing some environmental variables in the runscript. Do you have a list with apps, that need reworking? I could do some of them, that are useful to me too.
 
@Mr Rob most my PNDs wrote to NAND at a point or an other. I dont have an issue with that per see ;)

I'd think the bigger concern would be filling up your NAND and causing Linux to weep when it runs out of space, most users wouldn't know how to recover that other than a reflash. But it is instinct to want to preserve the life of the hardware as much as you can, and it's hard to think otherwise. So things like NAND writing, overclocking the CPU, battery charging seem to get a bit more attention than they should.


Yes, it's good to think about, but my Pandora is going to get accidentally stepped on, accidentally dropped, or accidentally thrown across the room in a fit of rage long before the NAND wears out.
 
You have made some fantastic contributions to the Pandora's library and I really am sad to see that go away.


It would be great if you would continue to release your "beta" .pnds on your facebook pages. Maybe just don't create an announcement here to avoid these comments.


I have run almost all of your ports and there are many that I really enjoy, I know they represent a tremendous amount of time from you.


Thanks!
 
@sebt3 For evopedia I needed to set some XDG variables to prevent uzbl from writing to the Applications dir in home.


Check the startupscript for more info.


Btw... Yay, I may have found a bug ;)
 
I appreciate all good comments from You, but i feel that the "shit already hits the fan" and I shouldnt let people download such a stuff, even as BETA at first place.


First time, when there was complaining [which i understand that] about NAND writing, i was a little bit disturbed about comments, but i known then that i'll not polish it because i dont want to waste free time, more than i need [since my OS runs from SD card and i dont care about nand writtings, and if nobody tests it or use it, why i should polishing it more?] and decided to mark all my stuff as BETA. That didnt worked, as well. I thought because it is a beta, i'll receive a little help with testing this stuff, and i'll save my free time on this partially. [i already said somewhere, that for me as a porter its more important to just test the stuff or run stuff i compiled, than "say" thanks.] That didnt work out, because i still recived more and more questions about NAND writtings, so i presume that stuff wasnt tested [or started on OP] so I shouldnt let people download such a stuff, even as BETA at first place. Of course it's my fault, i made some bad decisions in the process.
 
if you want, you could always just release the stuff to your fb followers (and anyone who was interested in your stuff could always follow you on fb). That might save you some of the grief you receive on the forums.
 
Of course it's my fault, i made some bad decisions in the process.
Even if.


I agree that when you put a BETA flag, users should be aware that it havent gone throw a complete QA process. Jesus, when MS out a beta of IE. Only thoses willing to deal with beta quality software download. And then if they have issues, they report bug, they dont complain.


I agree that with such a limited Dev force the pandora community should accept that the QA process have to be splited between the DEVs and the users so we get more apps available for pandora in the end of the week.


I was pleased to know you had setup a script to catch those nasty wrote on NAND since last drama. So you were coming clean since the drama. I was somewhat expecting the drama to be over.


In my opinion, if you are guilty of anything (beside wanting to provide apps for the community) you already have paid for your past errors (if any).


Having this drama starting all over again is just sad. It feel as if some prefer complaining then doing usefull thing (like a correct bug report). Now they will be happy to be allowed to complain a little more : they have one dev less to harass (exhaust)....
 
I still don't get why every is PARANOID about writing to the NAND, its designed to be written to, regular use is not going to destroy it. I can see that it hurts the "everything is contained on the SD card" idea with the PND system and the appdata folder but really a few writes to the NAND is not going to kill the system. Occasional writes to the NAND is perfectly safe and will not harm the chip within a reasonable time frame. The numbers published by NAND manufactures are worst case scenarios, they will almost never be that bad. The chips are tested at extremes of their specs and then padded to keep the company away from lawsuits and warranty returns.


If your worried about the system folders being filled up with garbage then a format once every 6 months to a year will not kill you, you will probably be doing that for updates anyways.


For example, someone tested an AVR chip with its 10 000 write cycle maximum and it reached somewhere into the 100 000+ without issues at room temp.


Everyone just needs to calm down a bit, appreciate what Streak is doing for all of us and if you don't then just don't download or use the app, it's just that simple.
 
Last edited by a moderator:
Seems an easy summary is ..


When you post things publicly, theres a certain expectation (rightly so) from the public for things to work a certain way; the dev can do whatever he wants, but by going public, peopel will get cross if things aren't how they are expected.


ie: You want to do it for yourself and followers, do that; if you want to post public, people _will_ bitch if its not working a certain way. (ie: on repo instead of FB, using conventions like no NAND, etc and so on.)


So do it as you have been, post as you want, make it good enough for you and people, and keep it on your FB page; then if somethign gets pollished enough, post it in the forums and talk it up, or on the Repo, etc.


But the two can't meet.. if its half baked, keep it in beta forum or off the forum, as you evidently don't want the attention :)


Can't make it half way then go public, just doesn't work :/


jeff
 
I still don't get why every is PARANOID about writing to the NAND, its designed to be written to, regular use is not going to destroy it. I can see that it hurts the "everything is contained on the SD card" idea with the PND system and the appdata folder but really a few writes to the NAND is not going to kill the system. Occasional writes to the NAND is perfectly safe and will not harm the chip within a reasonable time frame. The numbers published by NAND manufactures are worst case scenarios, they will almost never be that bad. The chips are tested at extremes of their specs and then padded to keep the company away from lawsuits and warranty returns.


If your worried about the system folders being filled up with garbage then a format once every 6 months to a year will not kill you, you will probably be doing that for updates anyways.


For example, someone tested an AVR chip with its 10 000 write cycle maximum and it reached somewhere into the 100 000+ without issues at room temp.


Everyone just needs to calm down a bit, appreciate what Streak is doing for all of us and if you don't then just don't download or use the app, it's just that simple.
because if your NAND goes, 90% of your pandora hardware goes with it. and we have very limited amount of free space on NAND as it is, which can cause a lot of unnecessary problems if it fills up with stuff that doesnt need to fill it
 
Back
Top