it was not given enough weight to;) while changing such mindsets is hard, i was suggesting to bring rational benefits to combat some irrational views.
there is one other reason, quite widely valid, though - code quality. i know of a company where i would advise against opensourcing some internal app simply because i know it's such a crap nobody would bother improving it...
wouldn't it be more efficient to build upon the existing solution and submit the improvements, thus getting a cheaper and better solution ? have you tried to inform your employer that such an approach would require less resources to implement and maintain - and that in the end it would result in a better quality software (with all respect to your coding skills, but there's always that one bug other coder will see instantly, and there's always a ton of bugs each additional tester will find;) )
as for links - maybe. but on the other hand that also makes documents much smaller if a user does not want to send them anywhere. it's not a sure choice...
This is precisely the point I am trying to make about usability. It is only "not a sure choice" because a developer understands the ramification of storing something as a small size. An end user wants to write the document to a floppy disk/thumb drive or attach it to an email. They don't want to think about breaking links; in fact, they are probably fairly clueless how the picture actually gets in the document in the first place.
well, it looks like oo.org developers agree with you - at least in the developer preview i have (dev300_m28) now picture insertion defaults to embedding, not linking;)
i think main reason is the choice of the old format
Again, this is an example of OSS community hubris. Instead of making it work like it should, blame the user.
hmm, i didn't want to blame anybody - all i did was trying to understand what caused the image size problems. i'm sorry if that came out as blaming you:)
Whoever modded him as a troll, I don't think he's trolling. This is the type of conversation OSS developers need to participate in to learn how to deliver usable software to the masses.
oh, i surely didn't. i don't troll actually, just sometimes joke or rant =) as for the mod, well, maybe that was some adobe pr person who didn't like my rant about flash;)
as for links - maybe. but on the other hand that also makes documents much smaller if a user does not want to send them anywhere. it's not a sure choice...
as for the thumbnail problem, i think main reason is the choice of the old format:) what i'd suggest - just break the links and save as mso 97/2000 - that practice should work well for you
* Try using open office and embedding pictures in a word processing document, only to find that Microsoft Word (which everyone else at the office uses) either can't load the pictures or the pictures come out scaled to thumbnail size. But, you have to export it in like Word 95 format to at least get the thumbnails.
can't comment on previous claims as i haven't experienced such situations. but i'll comment on these two:) this seems to be a very weird problem. is it really present in latest oo.org versions ? the only reasonable explanation for missing pictures that i can think of is pictures inserted as links, not embedded - hardly a format problem. and it's easy to solve - edit->links, select all, break. the "thumbnail" problem, never seen it. do you have a testcase that i could try ?
* He can continually wonder WHY THE HELL DOES FLASH KEEP LOCKING UP FIREFOX? Seriously, after a few LiveLeak or YouTube videos, you have to force-quite the browser and reload it. WTF?
ooooh yeah. wtf is a good question. but it should be asked to adobe. hi, adobe, you suck. flash plugin has been a joke on linux forever. in opera, it memleaked like hell. alsa support was late show for... a long time. latest versions indeed slow down and crash/hang firefox. what is it, adobe, can't you get your act together and fix the damn thing ? announce a public test sprint, ask for user input - i'[m sure you will get a lot of useful debugging information. if only you were interested. well, that's a classic example of reliance on a single vendor and closed source software.
exactly. in the "old" days, if i couldn't solve a puzzle for a day, i'd go to my neighbours and find out how long in the game they are. now and then we involved parents into puzzle solving - well, that rarely helped, but sometimes they indeed passed things we were stumbling upon:)
then internet appeared. at first it was like "can't solve for 2 days ? find an internet connection somewhere, search for a solution - it might exist".
later "try for a day, go to some internet connected computer, with some searching find the solution".
now it's just "can't solve in 3 minutes ? look up the solution".
1. Apparently no one but you was able to reproduce suspend/resume problem, and after your laptop was broken there was nothing to test.
actually, this problem comes up now and then, and as can be seen on the bugreport, there was another person reporting very similar symptoms. which can be observed by reading it.
What the Hell did you expect? Very likely the problem is tied to specific BIOS, ACPI configuration or hardware release version, what you never bothered to test. Also no attempts to check how behavior changes if you force unloading driver before suspend/reloading after resume.
right, shame on me for not testing different hardware versions. besides, by reading the report one can see that it is a regression...
2. I have no idea why X11 is brought up in discussion -- by your description it looks like the problem does not happen in X.
rrright, because in this context we are talking only about "worksforme" things. this is a quite perfect example of the attitude this very same article/discussuion is about. if somebody reports a problem, a lot of energy is wasted to dismiss it, starting with blaming the person as just having googled the bits, then attempting to invent explanations why, you know, there is no problem actually !
i don't see such attitude that often, so i was a bit surprised about the amount of complaints - but i guess there are a lot of people who will never admit there is a problem. it all depends on what kind of a person user initially stumbles upon... being ridiculed for reporting problems will surely increase testers for the project.
Plenty of graphics adapters are unsupported by native framebuffer drivers, so everyone uses vesa or text mode for those. No modern general-purpose Linux program uses framebuffer for graphics directly, so unless your adapter is unsupported by X drivers, all you have lost is boot splash. Boo-hoo.
right. you didn't experience this problem, you just found about it on teh internets. ok, there is no problem. none. it just isn't there. ok, your problem just doesn't matter.
well, maybe this is a somewhat sophisticated semi-trolling to show that the article actually has a point:)
Only happens on NON-LAPTOP computers with analog LCD monitors (solution: press "Auto-Adjust" and "Save" on your monitor while console text is displayed).
Now, please, tell me, how did you encounter both "problems" at the same time, and if they aren't in fact the result of a google search for "linux problem"?
now, tell me, are you just trying to give us all an example of how arrogant and useless some persons can be ? as one can see from the above reports, i encountered both of these issues on a single machine. i reported them in the kernel tracker (as good as i could), i was ready to patch, compile and debug as much as required. and that's way more than the average user is willing or able to do, especially if met with such an attitude.
great, your answer shows exactly what this story is about:) "it's a wrong driver, i don't use that, it's your fault".
1. on some savage chipset framebuffer console was shifted one pixel to the left. i tried various fbset tricks and other methods, but it was not possible to make it look right. two years after the bug was submitted somebody asked to do some more tests, but the hardware had died in that time
2. as far as i know, it's related to some chipsets. unfortunately, nobody had a time to debug the issue, so i'm quite sure it's still there in the kernel somewhere
3. no, using keyboard shortcuts is no good. i'm talking about user friendly way (see, the thing this article is about !) that would give user a dialog showing all usable resolutions & colour depths, with confirmation dialog and so on. while this sort of works, it is plagued by many problems - not showing all possible resolutions (as a rule the oe use wants never is there), not allowing high enough refresh rate, not setting the refresh rate actually...
yes, _i_ usually can get the result i want, be it by editing xorg config or some other way. but that's not good enough.
things never get to the point of doing everything one could ever possibly want. they got to the level of being a commodity, they get to the point of being enough for large part of the community, but there's always room for improvement and innovation. take a look at washing machines, refirgerators, cars, operating systems. while good enough has been reached some time ago, that's just... not enough:)
i'm a linux user for quite some time. while i believe it is better in most cases than the other offers, it would be just plain silly to claim there's nothing to complain about. while most of these aren't that huge, it still can make an impression on a person who would be just looking for the negative aspects. it's things like slightly shifted terminal console, no sound after resuming from hibernation (both kernel issues, not resolved over a few years), pdf form support appearing very late in kde application stack (don't misunderstand me, i'm very grateful for all of the kde:) ), lack of seamless, robust on-the-fly resolution/colour depth switching... it is very important to identify such areas and not to claim there are none, because otherwise they would never be improved. this is where i mostly agree with allison (though i haven't read the blog in question).
well, i doubt that even for large distribution every single commit to every package is verified by somebody. that's a huge workload, and it duplicates efforts on the same software, where opensource models were supposed to promote collaboration and workload sharing (this does not include competing packages, which is a different aspect of collaboration/competition theme:) )
this assumes that only large distributions will exist in near future. imagine that somebody has to read full changelogs for ALL packages (included in the distro)... that's just not realistic and insane.
first, i guess the reasoning here is to openly mark bugs already known to be security related as such.
second, really, this is about "us" being able to easily understand when we really, really should upgrade the kernel. even if did read full kernel changelogs, i wouldn't be able to understand which commits are security related. so i would rely on somebody to do that AND publicise it, at which point it gets more publicity than simply marking it in the changelog would have provided.
i'd argue that by not making them stand out will only create more hype - and for a good reason. one thing i can see the motivation behind - not marking a fix as a security related until it has been developed, somewhat tested and maybe even a new kernel version is released.
actually, it is happening. not instantly, but if you look at situation snapshots several years apart, it is quite visible. those who offered no support, offer binary drivers. those who offered binary drivers, offer hardware specs. those who offered hardware specs, write their own in-kernel drivers. and those who wrote in-kernel drivers now write in-kernel drivers before the hardware is even launched. with major vendors embracing linux even on desktop the trend should only accelerate.
With a wi-fi router you at least have the means available to you to (try to) prevent other people from using it, assuming you have the requisite knowledge.
but why should i ? that's a sharing. sharing some resource, some knowledge or whatever. i'd compare this to hitchhiking. if you take a hitchhiker who happens to be in the posession of something illegal, should you be held responsible ?
hmm... if those are images from camera, maybe new version does not use embedded thumbnail, instead always generating it's own ? in kde3, you can choose whether to use embedded thumbnail (if present) - i have no idea whether kde4 versions have such an option.
actually, most if not all scanner drivers are into sane (userspace application), most if not all printer drivers are into cups (userspace application). the reasons... those probably are a mix of historical and technical ones. but i wouldn't say drivers are neadlessly crammed into the kernel.
it also has quite a lot of features and is very, very robust. as a kde and oo.org user, i have no clear preference in which of these would be better for me if a company like google decided to actively work on it. koffice, despite lack of developers, has progressed quite nicely. oo.org lately has improved performance. and being huge... well, features don't work without code.
Slashdot Mirror
it was not given enough weight to ;)
while changing such mindsets is hard, i was suggesting to bring rational benefits to combat some irrational views.
there is one other reason, quite widely valid, though - code quality. i know of a company where i would advise against opensourcing some internal app simply because i know it's such a crap nobody would bother improving it...
you keep talking about binaries & compilation from the source all the time. so which is it, binary packages or compilation ?
wouldn't it be more efficient to build upon the existing solution and submit the improvements, thus getting a cheaper and better solution ? ;) )
have you tried to inform your employer that such an approach would require less resources to implement and maintain - and that in the end it would result in a better quality software (with all respect to your coding skills, but there's always that one bug other coder will see instantly, and there's always a ton of bugs each additional tester will find
as for links - maybe. but on the other hand that also makes documents much smaller if a user does not want to send them anywhere. it's not a sure choice...
This is precisely the point I am trying to make about usability. It is only "not a sure choice" because a developer understands the ramification of storing something as a small size. An end user wants to write the document to a floppy disk/thumb drive or attach it to an email. They don't want to think about breaking links; in fact, they are probably fairly clueless how the picture actually gets in the document in the first place.
well, it looks like oo.org developers agree with you - at least in the developer preview i have (dev300_m28) now picture insertion defaults to embedding, not linking ;)
i think main reason is the choice of the old format
Again, this is an example of OSS community hubris. Instead of making it work like it should, blame the user.
hmm, i didn't want to blame anybody - all i did was trying to understand what caused the image size problems. i'm sorry if that came out as blaming you :)
Whoever modded him as a troll, I don't think he's trolling. This is the type of conversation OSS developers need to participate in to learn how to deliver usable software to the masses.
oh, i surely didn't. i don't troll actually, just sometimes joke or rant =) ;)
as for the mod, well, maybe that was some adobe pr person who didn't like my rant about flash
as for links - maybe. but on the other hand that also makes documents much smaller if a user does not want to send them anywhere. it's not a sure choice...
as for the thumbnail problem, i think main reason is the choice of the old format :)
what i'd suggest - just break the links and save as mso 97/2000 - that practice should work well for you
can't comment on previous claims as i haven't experienced such situations. but i'll comment on these two :)
this seems to be a very weird problem. is it really present in latest oo.org versions ?
the only reasonable explanation for missing pictures that i can think of is pictures inserted as links, not embedded - hardly a format problem. and it's easy to solve - edit->links, select all, break.
the "thumbnail" problem, never seen it. do you have a testcase that i could try ?
ooooh yeah. wtf is a good question. but it should be asked to adobe. hi, adobe, you suck.
flash plugin has been a joke on linux forever. in opera, it memleaked like hell. alsa support was late show for... a long time. latest versions indeed slow down and crash/hang firefox. what is it, adobe, can't you get your act together and fix the damn thing ? announce a public test sprint, ask for user input - i'[m sure you will get a lot of useful debugging information. if only you were interested.
well, that's a classic example of reliance on a single vendor and closed source software.
exactly. :)
in the "old" days, if i couldn't solve a puzzle for a day, i'd go to my neighbours and find out how long in the game they are. now and then we involved parents into puzzle solving - well, that rarely helped, but sometimes they indeed passed things we were stumbling upon
then internet appeared. at first it was like "can't solve for 2 days ? find an internet connection somewhere, search for a solution - it might exist".
later "try for a day, go to some internet connected computer, with some searching find the solution".
now it's just "can't solve in 3 minutes ? look up the solution".
1. Apparently no one but you was able to reproduce suspend/resume problem, and after your laptop was broken there was nothing to test.
actually, this problem comes up now and then, and as can be seen on the bugreport, there was another person reporting very similar symptoms. which can be observed by reading it.
What the Hell did you expect? Very likely the problem is tied to specific BIOS, ACPI configuration or hardware release version, what you never bothered to test. Also no attempts to check how behavior changes if you force unloading driver before suspend/reloading after resume.
right, shame on me for not testing different hardware versions. besides, by reading the report one can see that it is a regression...
2. I have no idea why X11 is brought up in discussion -- by your description it looks like the problem does not happen in X.
rrright, because in this context we are talking only about "worksforme" things. this is a quite perfect example of the attitude this very same article/discussuion is about. if somebody reports a problem, a lot of energy is wasted to dismiss it, starting with blaming the person as just having googled the bits, then attempting to invent explanations why, you know, there is no problem actually !
i don't see such attitude that often, so i was a bit surprised about the amount of complaints - but i guess there are a lot of people who will never admit there is a problem. it all depends on what kind of a person user initially stumbles upon... being ridiculed for reporting problems will surely increase testers for the project.
Plenty of graphics adapters are unsupported by native framebuffer drivers, so everyone uses vesa or text mode for those. No modern general-purpose Linux program uses framebuffer for graphics directly, so unless your adapter is unsupported by X drivers, all you have lost is boot splash. Boo-hoo.
right. you didn't experience this problem, you just found about it on teh internets. ok, there is no problem. none. it just isn't there. ok, your problem just doesn't matter.
well, maybe this is a somewhat sophisticated semi-trolling to show that the article actually has a point :)
Only happens on NON-LAPTOP computers with analog LCD monitors (solution: press "Auto-Adjust" and "Save" on your monitor while console text is displayed).
http://bugzilla.kernel.org/show_bug.cgi?id=6523
Only happens on LAPTOP computers (who hibernates a desktop?, solution: update BIOS to a non-broken version).
http://bugzilla.kernel.org/show_bug.cgi?id=5621
Now, please, tell me, how did you encounter both "problems" at the same time, and if they aren't in fact the result of a google search for "linux problem"?
now, tell me, are you just trying to give us all an example of how arrogant and useless some persons can be ?
as one can see from the above reports, i encountered both of these issues on a single machine. i reported them in the kernel tracker (as good as i could), i was ready to patch, compile and debug as much as required. and that's way more than the average user is willing or able to do, especially if met with such an attitude.
great, your answer shows exactly what this story is about :)
"it's a wrong driver, i don't use that, it's your fault".
1. on some savage chipset framebuffer console was shifted one pixel to the left. i tried various fbset tricks and other methods, but it was not possible to make it look right. two years after the bug was submitted somebody asked to do some more tests, but the hardware had died in that time
2. as far as i know, it's related to some chipsets. unfortunately, nobody had a time to debug the issue, so i'm quite sure it's still there in the kernel somewhere
3. no, using keyboard shortcuts is no good. i'm talking about user friendly way (see, the thing this article is about !) that would give user a dialog showing all usable resolutions & colour depths, with confirmation dialog and so on. while this sort of works, it is plagued by many problems - not showing all possible resolutions (as a rule the oe use wants never is there), not allowing high enough refresh rate, not setting the refresh rate actually...
yes, _i_ usually can get the result i want, be it by editing xorg config or some other way. but that's not good enough.
things never get to the point of doing everything one could ever possibly want. they got to the level of being a commodity, they get to the point of being enough for large part of the community, but there's always room for improvement and innovation. :)
take a look at washing machines, refirgerators, cars, operating systems. while good enough has been reached some time ago, that's just... not enough
i'm a linux user for quite some time. while i believe it is better in most cases than the other offers, it would be just plain silly to claim there's nothing to complain about. while most of these aren't that huge, it still can make an impression on a person who would be just looking for the negative aspects. :) ), lack of seamless, robust on-the-fly resolution/colour depth switching...
it's things like slightly shifted terminal console, no sound after resuming from hibernation (both kernel issues, not resolved over a few years), pdf form support appearing very late in kde application stack (don't misunderstand me, i'm very grateful for all of the kde
it is very important to identify such areas and not to claim there are none, because otherwise they would never be improved. this is where i mostly agree with allison (though i haven't read the blog in question).
well, i doubt that even for large distribution every single commit to every package is verified by somebody. that's a huge workload, and it duplicates efforts on the same software, where opensource models were supposed to promote collaboration and workload sharing (this does not include competing packages, which is a different aspect of collaboration/competition theme :) )
this assumes that only large distributions will exist in near future.
imagine that somebody has to read full changelogs for ALL packages (included in the distro)... that's just not realistic and insane.
first, i guess the reasoning here is to openly mark bugs already known to be security related as such.
second, really, this is about "us" being able to easily understand when we really, really should upgrade the kernel. even if did read full kernel changelogs, i wouldn't be able to understand which commits are security related. so i would rely on somebody to do that AND publicise it, at which point it gets more publicity than simply marking it in the changelog would have provided.
i'd argue that by not making them stand out will only create more hype - and for a good reason.
one thing i can see the motivation behind - not marking a fix as a security related until it has been developed, somewhat tested and maybe even a new kernel version is released.
actually, it is happening. not instantly, but if you look at situation snapshots several years apart, it is quite visible.
those who offered no support, offer binary drivers. those who offered binary drivers, offer hardware specs. those who offered hardware specs, write their own in-kernel drivers. and those who wrote in-kernel drivers now write in-kernel drivers before the hardware is even launched.
with major vendors embracing linux even on desktop the trend should only accelerate.
can't you run opera on it ? while not opensource, it should be quite a decent tool for the purpose.
but why should i ?
that's a sharing. sharing some resource, some knowledge or whatever.
i'd compare this to hitchhiking. if you take a hitchhiker who happens to be in the posession of something illegal, should you be held responsible ?
steganography
hmm... if those are images from camera, maybe new version does not use embedded thumbnail, instead always generating it's own ?
in kde3, you can choose whether to use embedded thumbnail (if present) - i have no idea whether kde4 versions have such an option.
i feel an urge to emit a sound that quickly moving objects in air do
zomg ! you were boiling. you gotta be a superhero !
actually, most if not all scanner drivers are into sane (userspace application), most if not all printer drivers are into cups (userspace application).
the reasons... those probably are a mix of historical and technical ones. but i wouldn't say drivers are neadlessly crammed into the kernel.
it also has quite a lot of features and is very, very robust.
as a kde and oo.org user, i have no clear preference in which of these would be better for me if a company like google decided to actively work on it.
koffice, despite lack of developers, has progressed quite nicely. oo.org lately has improved performance. and being huge... well, features don't work without code.
Ãnonymous Cowards ?