Opera 10 on ubuntu save hangs
NewsGroup
wondering if any one can shed any light on why Opera 10 hangs when I try
to save a file?
This does not always happen but sooner or later during an opera session
this hang happens.
--
Kevin Pugh
TimC
was almost, but not quite, entirely unlike tea:
Are you using automounts, any fuse filesystems, or have a dodgy
filesystem?
--
TimC
"I often hear people claim they perform skills better slightly drunk if
they learned that skill drunk. I wonder if that applies to Perl. Get good
and liquored up, dash off a few scripts, see how you like it." -Rob Chanter
NewsGroup
> On 2009-09-27, NewsGroup (aka Bruce)
> was almost, but not quite, entirely unlike tea:
>> Hi folks,
>>
>> wondering if any one can shed any light on why Opera 10 hangs when I try
>> to save a file?
>>
>> This does not always happen but sooner or later during an opera session
>> this hang happens.
>
> Are you using automounts, any fuse filesystems, or have a dodgy
> filesystem?
>
partition is SWAP.
The XP partitions are mounted via fstab at startup.
Not too sure about fuse - how would I check for that?
--
Kevin Pugh
Jernej Simončič
> Not too sure about fuse - how would I check for that?
If you're using the (read-only) kernel driver to access NTFS partitions,
you aren't using FUSE. If you use ntfs-3g, you're using FUSE.
--
begin .sig
< Jernej Simončič ><>◊<>< jernej|s-ng at eternallybored.org >
end
TimC
was almost, but not quite, entirely unlike tea:
>
>> Not too sure about fuse - how would I check for that?
>
> If you're using the (read-only) kernel driver to access NTFS partitions,
> you aren't using FUSE. If you use ntfs-3g, you're using FUSE.
If you're trying to get opera to write to the ntfs partition to share
data, or if the last directory you saved to was in the ntfs partition,
then as unlikely as it is to screw up, it might still be worth forcing
a fsck (or whatever they call it in windows-land - a disk check) next
time you boot windows.
(the reason I mentioned fuse in the first place, was in my case when I
use sshfs, if the link goes down, then it can block applications that
I had long ago forgotten that had been anywhere near the blocked
directory. Just kill the sshfs process, and it returns. But just
killing the ntfs-3g process could be slightly (but shouldn't be) more
drastic to that filesystem).
--
TimC
Computer screens simply ooze buckets of yang.
To balance this, place some women around the corners of the room.
-- Kaz Cooke, Dumb Feng Shui
NewsGroup
> On 2009-09-28, Jernej Simončič (aka Bruce)
> was almost, but not quite, entirely unlike tea:
>> on Mon, 28 Sep 2009 23:15:08 +0100, NewsGroup wrote:
>>
>>> Not too sure about fuse - how would I check for that?
>> If you're using the (read-only) kernel driver to access NTFS partitions,
>> you aren't using FUSE. If you use ntfs-3g, you're using FUSE.
>
> If you're trying to get opera to write to the ntfs partition to share
> data, or if the last directory you saved to was in the ntfs partition,
> then as unlikely as it is to screw up, it might still be worth forcing
> a fsck (or whatever they call it in windows-land - a disk check) next
> time you boot windows.
>
>
>
> (the reason I mentioned fuse in the first place, was in my case when I
> use sshfs, if the link goes down, then it can block applications that
> I had long ago forgotten that had been anywhere near the blocked
> directory. Just kill the sshfs process, and it returns. But just
> killing the ntfs-3g process could be slightly (but shouldn't be) more
> drastic to that filesystem).
>
Thank you all for your comments, it does nopt seem to make any
difference whether I save to an ext3 or ntfs file system - although I
have noticed that the pcmia card might be behind the problem.
If I work it out I will let you know.
BTW: is there a debug mode for Opera that can be switched on?
--
Kevin Pugh
David W. Hodgins
> BTW: is there a debug mode for Opera that can be switched on?
From a terminal, run "opera --debughelp".
Regards, Dave Hodgins
--
Change nomail.afraid.org to ody.ca to reply by email.
(nomail.afraid.org has been set up specifically for
use in usenet. Feel free to use it yourself.)
NewsGroup
> On Tue, 29 Sep 2009 14:44:35 -0400, NewsGroup <no...@none.invalid> wrote:
>
>> BTW: is there a debug mode for Opera that can be switched on?
>
> From a terminal, run "opera --debughelp".
>
> Regards, Dave Hodgins
>
--
Kevin Pugh
Thomas Glanzmann
> wondering if any one can shed any light on why Opera 10 hangs when I
> try to save a file? This does not always happen but sooner or later
> during an opera session this hang happens.
I can confirm this, this is why I downgraded all my workstations to the
following very stable beta build of opera 10. This specific build
doesn't have the problem:
http://thomas.glanzmann.de/tmp/opera_10.00.4394.gcc4.qt3_i386.deb
Thomas
Happy Oyster
wrote:
Opera is messed up. One of my machines I have to restart and restart and
restart, until, at last, it MIGHT be able to save a download file.
The stupidity of the guys who call themselves programmers at Opera is a shame.
--
Der Schmock der Woche
Naruki Bigglesworth
Weird, I thought the Linux platform was for competent people who knew how to solve their own
problems. Didn't realize AOL had customers here.
Happy Oyster
wrote:
I do not know WHAT the Opera guys did, but for quite a time now the opera
browser is crashing at least 5 to 10 times a day. This I consider not
acceptable.
Eirik Byrkjeflot Anonsen
> On Tue, 6 Oct 2009 09:04:41 +0900, Naruki Bigglesworth <Nar...@nothankyou.com>
> wrote:
>
>>On Tue, 06 Oct 2009 01:46:03 +0200, Happy Oyster wrote:
>>
>>> On 5 Oct 2009 07:38:39 GMT, Thomas Glanzmann <sith...@stud.uni-erlangen.de>
>>> wrote:
>>>
>>>>Hello,
>>>>
>>>>> wondering if any one can shed any light on why Opera 10 hangs when I
>>>>> try to save a file? This does not always happen but sooner or later
>>>>> during an opera session this hang happens.
>>>>
>>>>I can confirm this, this is why I downgraded all my workstations to the
>>>>following very stable beta build of opera 10. This specific build
>>>>doesn't have the problem:
>>>>
>>>>http://thomas.glanzmann.de/tmp/opera_10.00.4394.gcc4.qt3_i386.deb
>>>>
>>>> Thomas
>>>
>>> Opera is messed up. One of my machines I have to restart and restart and
>>> restart, until, at last, it MIGHT be able to save a download file.
>>>
>>> The stupidity of the guys who call themselves programmers at Opera is a shame.
A piece of friendly advice: If you want someone to do something for you,
it is generally not a good idea to start out by insulting them.
>>Weird, I thought the Linux platform was for competent people who knew how to solve their own
>>problems. Didn't realize AOL had customers here.
Calm down. Fighting insults with insults generally only leads to hard
feelings and hardly ever resolves anything.
> I do not know WHAT the Opera guys did, but for quite a time now the opera
> browser is crashing at least 5 to 10 times a day. This I consider not
> acceptable.
I assume you have made good, relevant bug reports for these crashers?
(http://bugs.opera.com/wizard/)
eirik
Happy Oyster
wrote:
That would be occupational therapy. They are so damned stupid, that Jon Tezchner
BY WILL spoiled the design of Opera. He made one of the most idiotic mi8stakes
ever done in the history of EDP: he spoiled the file system.
--
Herausforderung an die Universit�t Leipzig
http://www.ariplex.com/ama/ama_ho12.htm
Frank Bell
<happy....@ariplex.com> wrote:
> I do not know WHAT the Opera guys did, but for quite a time now the opera
> browser is crashing at least 5 to 10 times a day. This I consider not
> acceptable.
Why do I not have these problems? I have saved numerous files, both text
and HTML (mostly from Project Gutenberg) directly from Opera without any
issues, (plus downloads).
Opera 10.0 on Ubuntu 9.04
--
Blogging from Pine View Farm (http://www.pineviewfarm.net/weblog)
Updates daily. Worthwhile updates occasionally.
Opera (http://www.opera.com), Linux (http://iso.linuxquestions.org/), and
Fluxbox (http://www.fluxbox.org)--the ultimate internet experience.
Check out Geekazine: http://www.geekazine.com
the ultimate internet experience.
Happy Oyster
>On Mon, 05 Oct 2009 21:51:09 -0400, Happy Oyster
><happy....@ariplex.com> wrote:
>
>> I do not know WHAT the Opera guys did, but for quite a time now the opera
>> browser is crashing at least 5 to 10 times a day. This I consider not
>> acceptable.
>
>Why do I not have these problems?
I don't know. I run Opera 9.6.4 on SuSE 11.1 - LOTS of crashes...
>I have saved numerous files, both text
>and HTML (mostly from Project Gutenberg) directly from Opera without any
>issues, (plus downloads).
>
>Opera 10.0 on Ubuntu 9.04
--
Eirik Byrkjeflot Anonsen
> On Tue, 06 Oct 2009 08:48:07 +0200, Eirik Byrkjeflot Anonsen <ei...@opera.com>
> wrote:
>
>>Happy Oyster <happy....@ariplex.com> writes:
>>
>>>>On Tue, 06 Oct 2009 01:46:03 +0200, Happy Oyster wrote:
>>>>
[...]
>>>>>
>>>>> Opera is messed up. One of my machines I have to restart and restart and
>>>>> restart, until, at last, it MIGHT be able to save a download file.
>>>>>
[...]
>>> I do not know WHAT the Opera guys did, but for quite a time now the opera
>>> browser is crashing at least 5 to 10 times a day. This I consider not
>>> acceptable.
>>
>>I assume you have made good, relevant bug reports for these crashers?
>>(http://bugs.opera.com/wizard/)
>>
>>
>>eirik
>
> That would be occupational therapy. They are so damned stupid, that Jon Tezchner
> BY WILL spoiled the design of Opera. He made one of the most idiotic mi8stakes
> ever done in the history of EDP: he spoiled the file system.
I'm starting to suspect you are a troll rather than being serious, but
I'll bite anyway.
You're complaining that opera crashes frequently, but you're not willing
to report the problems to us so we can fix it. Then you're ranting
about Jon having designed opera in such a way that it breaks your file
system. But you're not saying anything about what sort of breakage or
what the design error is, so again, we can not help you in any way,
including fixing the problem.
From the above it would seem that you are complaining about three
different problems:
- Opera frequently fails to download files.
- Opera frequently crashes.
- Opera intentionally corrupts the file system.
Now the last one is certainly wrong. I would not be using opera if it
was designed to intentionally destroy all the data on my hard drives.
And I think I can confidently say that opera hasn't done so over the 9
years I've been using it (including quite a few "unstable", buggy
versions). So let's rephrase the last one:
- Opera frequently fails to download files.
- Opera frequently crashes.
- Opera frequently corrupts the file system.
I can assure you that this doesn't happen to everyone. We would never
release opera if any of those involved in developing or testing opera
during development had seen such problems.
A few unreproducible failures to downloads and a few unreproducible
crash bugs we might choose to live with. But if we had seen file system
corruption, we would never have released.
On the other hand, it could be something about your setup that is
different from what our testers use. And it is quite reasonable that
some of our "few, unreproducible" crashers and failures to download
becomes "frequent" for that reason. And it may even be that we have a
bug that on your system ends up corrupting the file system.
You can be absolutely sure that we want to fix these problems. But we
can't fix them unless we know what causes them. Since we don't have the
problem, we can't figure this out on our own. Thus we would very much
like to have your help in figuring out what the bug is, so we can fix
it. That's why I ask you to make bug reports.
And again you spend more effort on insulting those who might be able to
help you then to actually ask for help. That is just bad strategy.
eirik
Thomas Glanzmann
> That's why I ask you to make bug reports.
what is the easies way to make opera bug report?
Thomas
Eirik Byrkjeflot Anonsen
I think the recommended way is:
https://bugs.opera.com/wizard/
(That seems to be what https://bugs.opera.com/ suggests as well).
eirik
Jan van Gent
<happy....@ariplex.com> wrote:
> On Tue, 06 Oct 2009 21:19:11 -0400, "Frank Bell"
> <frank...@comcast.net> wrote:
>
>> On Mon, 05 Oct 2009 21:51:09 -0400, Happy Oyster
>> <happy....@ariplex.com> wrote:
>>
>>> I do not know WHAT the Opera guys did, but for quite a time now the
>>> opera
>>> browser is crashing at least 5 to 10 times a day. This I consider not
>>> acceptable.
>>
>> Why do I not have these problems?
>
> I don't know. I run Opera 9.6.4 on SuSE 11.1 - LOTS of crashes...
>
Update to 10. You're commenting in a thread about 10.
Opera 9.64 crashed daily for me, 10 nearly never.
I find Opera 10 much more stable then 9.*
Can't recognize other problems, downloading files is no problem at all.
And actually I find Opera employees concerned and helpful.
Build 4585
--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/
Happy Oyster
wrote:
>> I don't know. I run Opera 9.6.4 on SuSE 11.1 - LOTS of crashes...
>>
>
>Update to 10. You're commenting in a thread about 10.
>
>Opera 9.64 crashed daily for me, 10 nearly never.
>
>I find Opera 10 much more stable then 9.*
Well,
a) I am a hevy user, I am doing investigations whioch do take MUCH time.
b) I am very reluctant because the latest releases of Opera got worse instead of
better. Actually I prefer to use the Version 3.21 on Win 3.1 for most of the
real tough things.
--
F�r eine wehrhafte Medizin!
Happy Oyster
wrote:
>> That would be occupational therapy. They are so damned stupid, that Jon Tezchner
>> BY WILL spoiled the design of Opera. He made one of the most idiotic mi8stakes
>> ever done in the history of EDP: he spoiled the file system.
>
>I'm starting to suspect you are a troll rather than being serious, but
>I'll bite anyway.
>
>You're complaining that opera crashes frequently, but you're not willing
>to report the problems to us so we can fix it.
You write "we", so I assume that you work for Opera, and are not only q guest
using one of their email addresses.
>Then you're ranting
>about Jon having designed opera in such a way that it breaks your file
>system. But you're not saying anything about what sort of breakage or
>what the design error is, so again, we can not help you in any way,
>including fixing the problem.
I do know what I talk about. And I am fed up with talking to the wall.
In the time of Opera 3.21 I was in contact with Tezchner, wrote what I could
find out (at that time), but never got an answer, and later saw, that they were
working on completely different things, so that my whole work was useless.
So, as you work for Opera, you PERHAPS will understand what I now try to
describe. (I posted these things in another group weeks go.)
"One of the most idiotic things I encountered were freezes of the computer where
it refused keyboard input for longer than a half an hour. Log-in as root (on a
Linux machine) was NOT possible. The idiotic answer by the system was that
log-in response was longer than 60 (?) seconds, so it was refused.
The origin (still not solved) is Opera, which blocks off everything else when it
- using an utmost idiotic algorithm - reorganizes its indexes. The guys at Opera
have committed some of the most incredible messes I have seen since I started
with computers in about 1968."
"The point is that I MUST use Opera because no other browser has some of the
properties I use of it. The cache is set to maximum, which is lousy 400
MegaBytes. I have open 50 tabs or more. The cache easily contains >20,000 files.
The stupidity is (as do some zerobrainer programmers) to, say, delete 1 entry
READ the file list, delete the file, change the file list, write that to HDD.
For 1 file that is easy. But for 5000 files, they would read 5000 times, change
the list, write the list. Instead of ONE complete work, they mess up with
something easy to program, but insane to go through with larger number of
files."
"Ordinary users have 1 tab open, do some chat or whatever. I have 50 tabs open,
some of them containing .flv videos. That is not a bad habit by me, but some
brainless advertisers insert video clips. So, reading a magazine online includes
these videos. If a video is in one window and a second video in another window,
the videos get chopped - 1 second (or so) from this window, 1 second from that
window. These programs are not made for a load which exceeds a child's play."
"Like switching "referrer" and other parameters. Most browser can't change any
of these. They are vital when you are on an investigation trip. The "referrer"
variable hands of to the web serverthe URL you were before. This could leave
trace, and is a must to avoid."
"A very big item is zooming. Already Opera 3.21 could zoom. Other browsers only
could scale the size of the letters. But when scaling the text, the pictures
kept their size. This is VERY bad when the pictures are on tiny buttons for
navigation. Opera zooms up to 1000%, even Opera 3.21. If 2.12 could do it, I
forgot. It took Firefox 10 years to learn to zoom. That is a real shame."
--------------
"Some time ago, perhaps a year, I read that Tezchner had pushed his way through.
There obviously was an internal war about how the browser had to workd, etc. And
Tezchner won. THAT was the connection I needed to understand the most stupid
mess I have seen so far: messing up file names.
If you install opera 9.64 (I do not know at which version the idiocy started),
and have an URL and see a web-page, then save that web-page - and look at the
very name. The name is NOT the file name of the original, but the headline.
Now look at all those fine (brrrrrrrr) languages with all their fine
(brrrrrrrrrrrrrr) special characters: Do these charactes fit into a reasonable
filename? NO!!! So, with Linux I save a web-page, Opera takes the headline as
filename and writes that to the HDD. Superb if +~*?) and other special
characters are part of the headline. How do you write that? You might end with
things the OS oses for special tasks. Wonderful...
But THAT is the simple stuff. The real hammer comes when you want to write to CD
or DVD, say for a backup. The file system of CD and DVD have restrictive use of
special characters. You CANNOT write the stuff from HDD to CD and DVD. You need
some converter to make printable AND allowed characters.
But even that is simple stuff. Because: you must not let the converter run
twice. A converted name can be converted again. Then it is something else...
But even that is simple. Just try to handle that file by hand. How do you enter
a filename with 50 or more characters, preferrably with some special characters
which you even can't find on your keyboard...?
But even that is simple. Just try to have more than one file. How about saving a
dozen files from a web-forum thread? They all have the very same headline. The
original file name is a number. But Opera does not use that...
That is so idiotic, it is beyond description.
A small bunch of programmers made that - and hundreds of millions of users have
to deal with this god-damned shit.
Now back to the story about Tezchner, because that is most important here.
Tezchner, so the magazine article said, won the war on how the program is
designed. So it is him PERSONALLY who is responsible for this.
One might guess that Opera, the browser, when installed, has several options.
And if this one is shit, okay, then take an other one.
The answer is: There is no other option!
One might think that, okay, then this is a bug, happened unvoluntarily.
The answer is: NO!!! The braindead superidiots at Opera did that by will!
The key to the answer is... Opera 3.21.
Already in Opera 3.21 they had built in that mess. But it shows up under some
weird conditions. You close the program, you start it anew, and then the
exchange of file name and head line in the address line is gone. That means that
for about 10 years these idiots had this mess in the drawer and already in Opera
3.21 they had implemented parts of it.
They spoil the whole file system.
Do you think that such a stinkingly stupid bunch of idiots, including Tezchner,
who spoil the file system totally, would listen to what I tell them?
They DO KNOW the whole thing.
And this I do say. Because it is one of the biggest messes ever made in the
history of EDP. Because it was made by will. There is no excuse for that mess.
Eirik Byrkjeflot Anonsen
> On Wed, 07 Oct 2009 09:11:23 +0200, Eirik Byrkjeflot Anonsen <ei...@opera.com>
> wrote:
>
>>> That would be occupational therapy. They are so damned stupid, that Jon Tezchner
>>> BY WILL spoiled the design of Opera. He made one of the most idiotic mi8stakes
>>> ever done in the history of EDP: he spoiled the file system.
>>
>>I'm starting to suspect you are a troll rather than being serious, but
>>I'll bite anyway.
>>
>>You're complaining that opera crashes frequently, but you're not willing
>>to report the problems to us so we can fix it.
>
> You write "we", so I assume that you work for Opera, and are not only q guest
> using one of their email addresses.
Yes, I work for opera, though I haven't been working on the desktop
product for several years.
>>Then you're ranting
>>about Jon having designed opera in such a way that it breaks your file
>>system. But you're not saying anything about what sort of breakage or
>>what the design error is, so again, we can not help you in any way,
>>including fixing the problem.
>
> I do know what I talk about. And I am fed up with talking to the wall.
>
> In the time of Opera 3.21 I was in contact with Tezchner, wrote what I could
> find out (at that time), but never got an answer, and later saw, that they were
> working on completely different things, so that my whole work was useless.
That is a long time ago. We actually have a working bug system now.
(Well, mostly working anyway. Anyone can report bugs and all bugs are
looked at by real people. The main problem with the bug system is that
"outsiders" aren't allowed to look at the bugs.)
> So, as you work for Opera, you PERHAPS will understand what I now try to
> describe. (I posted these things in another group weeks go.)
>
>
> "One of the most idiotic things I encountered were freezes of the computer where
> it refused keyboard input for longer than a half an hour. Log-in as root (on a
> Linux machine) was NOT possible. The idiotic answer by the system was that
> log-in response was longer than 60 (?) seconds, so it was refused.
>
> The origin (still not solved) is Opera, which blocks off everything else when it
> - using an utmost idiotic algorithm - reorganizes its indexes. The guys at Opera
> have committed some of the most incredible messes I have seen since I started
> with computers in about 1968."
So basically, under certain circumstances opera uses so much resources
as to starve out the rest of the system. Yes, I can believe that can
happen. In particular, I often see heavy disk usage have this effect on
linux. Using git on a large working tree is a typical example.
On the other hand, such extreme problems as these we probably do want to
know about. We would want to figure out why it happens so we can either
fix it (if we're really doing something wrong) or try to alleviate the
problem. However, it is probably the sort of problem that is very hard
to fix unless we can test it ourselves.
> "The point is that I MUST use Opera because no other browser has some of the
> properties I use of it. The cache is set to maximum, which is lousy 400
> MegaBytes. I have open 50 tabs or more. The cache easily contains >20,000 files.
>
> The stupidity is (as do some zerobrainer programmers) to, say, delete 1 entry
> READ the file list, delete the file, change the file list, write that to HDD.
> For 1 file that is easy. But for 5000 files, they would read 5000 times, change
> the list, write the list. Instead of ONE complete work, they mess up with
> something easy to program, but insane to go through with larger number of
> files."
Ah, so by "indexes" you mean the list of files in the cache. I don't
think your analysis is correct, at least for the last couple of opera
versions. I believe opera only updates the cache index on disk at
(ir)regular intervals. At which point it dumps the current state to a
disk, rather than reading the old file, modify it and write it. I'm
quite sure opera only reads the cache index on startup.
> "Ordinary users have 1 tab open, do some chat or whatever. I have 50 tabs open,
> some of them containing .flv videos. That is not a bad habit by me, but some
> brainless advertisers insert video clips. So, reading a magazine online includes
> these videos. If a video is in one window and a second video in another window,
> the videos get chopped - 1 second (or so) from this window, 1 second from that
> window. These programs are not made for a load which exceeds a child's
> play."
I've currently got 50 tabs open, some of our testers have 300. So we
definitely test that sort of usage.
Admittedly, I always run with plug-ins turned off, both for security
reasons and just because the web is better that way.
When it comes to multiple instance of flash in multiple windows on
linux, I don't think we do anything to control what it does. It is
entirely up to flash itself how to allocate its resources to the
different instances.
> "Some time ago, perhaps a year, I read that Tezchner had pushed his way through.
> There obviously was an internal war about how the browser had to workd, etc. And
> Tezchner won. THAT was the connection I needed to understand the most stupid
> mess I have seen so far: messing up file names.
There was some noise in some big norwegian newspapers about a supposed
conflict between Jon and the opera board. Though I think that was a
couple of years ago by now.
If that is what you are referring to, the alleged conflict had nothing
to do with the technical details of the browser. It was all about the
business, and nothing about the product itself. (And that's assuming
the story was actually true in the first place. I have rather low
opinions of the big norwegian newspapers' ability to tell the truth from
a big headline.)
Jon is not involved in designing the technical details of the browser.
He doesn't really have much to do with designing any part of the
browser. That's not his job. As the founder and CEO his opinions on
what bugs and features are the most important to work on has some
weight, and certainly influences the people who actually make the
decisions. And in principle he could probably order us to do particular
things in particular ways. But he's not stupid enough to do that.
> If you install opera 9.64 (I do not know at which version the idiocy started),
> and have an URL and see a web-page, then save that web-page - and look at the
> very name. The name is NOT the file name of the original, but the
> headline.
You're evidently not an "ordinary" user. Most "ordinary" users would
prefer to have their data saved in files with names that actually goes
some way to describe their content. Certainly "Originalt Amundsen-foto
fra Sydpolen funnet i Australia - Nyheter - Innenriks - Aftenposten.no"
("Original Amundsen-photo from the south pole found in Australia - news
- domestic - Aftenposten.no") is far more informative than
"http:%2F%2Fwww.aftenposten.no%2Fnyheter%2Firiks%2Farticle3308708.ece"
(Remember, '/' is not an allowed character). Or even worse:
"article3308708.ece".
(Note that none of the above is "the file name of the original". The
article is most likely stored in a database, and as such it doesn't have
a file name at all. But I'll assume that by "file name" you really
meant "uri")
However, in my experience opera never saves files using the title as the
file name. Opera always asks me what file name I want to save the file
as. The title is just opera's suggestion. Obviously, opera wants to
suggest the name that is the most useful to the most users. And I
believe that more users find the title to be a useful file name than the
uri.
> Now look at all those fine (brrrrrrrr) languages with all their fine
> (brrrrrrrrrrrrrr) special characters: Do these charactes fit into a reasonable
> filename? NO!!! So, with Linux I save a web-page, Opera takes the headline as
> filename and writes that to the HDD. Superb if +~*?) and other special
> characters are part of the headline. How do you write that? You might end with
> things the OS oses for special tasks. Wonderful...
On unix systems, this is no problem. The only characters that the OS is
using for special things in file names are null bytes and '/'. So as
long as the text is valid text in the system encoding without any
slashes, there's no problem. If you have chosen utf-8 as your system
encoding (which I currently think is the most sane thing to do), the
only problem is with '/'.
If you're talking about the unix shell rather than the OS itself, use
single quotes and escape all single quotes. If you are a command line
user, you should do that anyway for unknown file names.
If you're not on a unix system, that's a different issue entirely, of
course.
> But THAT is the simple stuff. The real hammer comes when you want to write to CD
> or DVD, say for a backup. The file system of CD and DVD have restrictive use of
> special characters. You CANNOT write the stuff from HDD to CD and DVD. You need
> some converter to make printable AND allowed characters.
Yes, but you need to do that anyway. Unless you only ever burn things
you yourself create the file names of, and you are very careful never to
use "problematic" characters. But in that case you wouldn't be
downloading files from the internet without explicitly choosing the name
yourself, which makes the whole discussion moot.
>
> But even that is simple stuff. Because: you must not let the converter run
> twice. A converted name can be converted again. Then it is something
> else...
That makes no sense. A converter that takes "bad" characters and
replaces them with "good" characters should be idempotent (i.e. produce
the same output when run multiple times).
If you are talking about a converter from one character encoding to
another character encoding, you are right. But this isn't the problem
we are discussing.
> But even that is simple. Just try to handle that file by hand. How do you enter
> a filename with 50 or more characters, preferrably with some special characters
> which you even can't find on your keyboard...?
I usually use tab completion. The only problem then arises when you the
decision tree branches on a character not on your keyboard. That does
happen to me sometimes, as I (try to) read japanese but haven't set up a
working input method in xterm yet.
However, in general I would expect very few people to download web pages
in languages they can't write. So I would expect this to be a small
problem, and one that can generally be solved with a little creativity
in the few cases where it does happen.
> But even that is simple. Just try to have more than one file. How about saving a
> dozen files from a web-forum thread? They all have the very same headline. The
> original file name is a number. But Opera does not use that...
A number is a very bad file name. It says nothing of the content. I am
sure most users prefer a title that says something about what the file
contains to a file name that is just a number. I believe you are in the
minority here. (So am I, for that matter, but I always choose my own
file names when I save files. How else would I be able to guess what
the file contains by just looking at the name?)
> A small bunch of programmers made that - and hundreds of millions of users have
> to deal with this god-damned shit.
Now here I do agree with you. All our users have to endure whatever
choice we make. So we have to try to make the choice that is most
useful to the most users. And I firmly believe that your suggestion (to
use the url as the initial suggestion instead) is worse than our current
behaviour (to use the title as the initial suggestion).
The arguments you make above have some merit, but as my responses show I
don't think they are very strong arguments. I'm sure we can do better
than we do, but switching to using the last component of the uri is
probably not the solution.
> Now back to the story about Tezchner, because that is most important here.
> Tezchner, so the magazine article said, won the war on how the program is
> designed. So it is him PERSONALLY who is responsible for this.
If that's what the article said, it was wrong. On the other hand, as
the CEO of the company he does carry some of the responsibility for what
the company as a whole does, even if he wasn't personally involved.
Besides, it is not important at all. What is important is that opera is
as good a program as possible. Who made good decisions and who made bad
decisions is really not important.
> One might guess that Opera, the browser, when installed, has several options.
> And if this one is shit, okay, then take an other one.
>
> The answer is: There is no other option!
Opera has too many options. And too few. Just like most moderately
complex programs. I'm not convinced that adding this option to opera
would make it better (i.e. be more good than bad).
> One might think that, okay, then this is a bug, happened unvoluntarily.
>
> The answer is: NO!!! The braindead superidiots at Opera did that by
> will!
Yes, I believe this is a conscious choice, and I believe it is the right
choice. I may be wrong, but your arguments so far does not convince me.
> The key to the answer is... Opera 3.21.
That does bring with it quite a few other problems. Not the least that
it doesn't support linux (opera 5 was the first one to do so, I believe.
If you don't count the 4 tech previews of opera 4).
> Already in Opera 3.21 they had built in that mess. But it shows up under some
> weird conditions. You close the program, you start it anew, and then the
> exchange of file name and head line in the address line is gone. That means that
> for about 10 years these idiots had this mess in the drawer and already in Opera
> 3.21 they had implemented parts of it.
>
> They spoil the whole file system.
I do not see how this "spoils the whole file system". Opera suggests a
file name that you don't like. If you choose to save it using the
suggested file name, you have a file name on your system that you don't
like. If your file system breaks down because there is one file in the
file system with a name that you don't like, then I'd say there's
something else that is seriously broken here.
Of course, if you are in fact using a system that gets into problems
when a single file contains some particular characters, and it was
reasonable for opera to know about this, we should probably not include
that character in our suggested file names. However, this does not
apply to standard unix file systems.
If there are characters that would be likely to be an inconvenience to
our users if they appeared in a file name, we should probably also avoid
putting them in the suggested file names. However, on a unix system I
can't really see what that should be.
> Do you think that such a stinkingly stupid bunch of idiots, including Tezchner,
> who spoil the file system totally, would listen to what I tell them?
Yes, we would listen. We may disagree with you (as I do), but if you
present good arguments for your case, there's a pretty good chance that
we'd consider it. But we will still have to choose the solution that we
believe will be the most useful to the most of our users. We will not
choose your solution just because you say it is the right thing to do.
Now, if you fail to convince us that your solution is better than ours,
a different strategy could be to report as bugs the actual problems it
creates. If our choice of default file names creates real problems, we
will probably want to try to improve our strategy to avoid those
problems.
Important hint: If you want us to fix the problems you see, you will
need to convince us that the problem is real and that it is something we
should prioritize. The best way to do that is probably to calmly and
clearly describe the problem. This will allow whoever reads your bug
report to evaluate how big the problem really is.
And again, insults does not strengthen your arguments. It may be
therapeutical, but it tends to make the listener lend less weight to
your actual arguments. The same thing applies to ranting and raving.
> They DO KNOW the whole thing.
>
> And this I do say. Because it is one of the biggest messes ever made in the
> history of EDP. Because it was made by will. There is no excuse for that mess.
If you think our default names of saved files is "one of the biggest
messes ever made in the history of " (Electronic data processing? So
"history of computers"), then I think you've led a very sheltered
computing life.
eirik
Happy Oyster
wrote:
I wrote:
I do know what I talk about. And I am fed up with talking to the wall.
**********************************************************************
You just - again - demonstrated, how right I am in what I wrote.
.
--
"Als h�tt' er nicht genug verbrochen, * Gott zwingt hungernde M�tter,
l��t Jahwe M�tter Kinder kochen. * ihre Kinder zu kochen
Es staunt nur noch der "'Atheist", * (Klagelieder Jeremias, Kap. 4)
was Christen alles heilig ist." http://www.reimbibel.de
Eirik Byrkjeflot Anonsen
[...]
>
>
> I wrote:
>
> I do know what I talk about. And I am fed up with talking to the wall.
> **********************************************************************
>
>
> You just - again - demonstrated, how right I am in what I wrote.
Ok, quick summary: You disagree with what we consider to be a good UI
but fail to provide convincing arguments for your position. I gave you
my reasons for disagreeing with you, and you respond with "I am fed up
with talking to the wall".
If you're not willing to argue your position, you can not expect anyone
else to agree with you.
(And, just in case you missed it: I can see how some of the issues you
raise can be problematic. But I strongly disagree that your solution
will actually be an improvement.)
eirik
Happy Oyster
wrote:
>Happy Oyster <happy....@ariplex.com> writes:
>
>[...]
>>
>>
>> I wrote:
>>
>> I do know what I talk about. And I am fed up with talking to the wall.
>> **********************************************************************
>>
>>
>> You just - again - demonstrated, how right I am in what I wrote.
>
>
>Ok, quick summary: You disagree with what we consider to be a good UI
>but fail to provide convincing arguments for your position. I gave you
>my reasons for disagreeing with you, and you respond with "I am fed up
>with talking to the wall".
>
>If you're not willing to argue your position, you can not expect anyone
>else to agree with you.
If you are not willing or even able to understand PLAIN FACTS, it is YOUR FAULT.
DO NOT DARE TO USE RHETORICAL WARFARE.
>(And, just in case you missed it: I can see how some of the issues you
>raise can be problematic. But I strongly disagree that your solution
>will actually be an improvement.)
>
>eirik
Now I have one more piece of proof for what I wrote about Opera.com.
--
"Syrer besetzten die St�dte,
frevelten dort um die Wette.
Wollten von Jahwe nichts wissen,
wurden von L�wen zerrissen." http://www.reimbibel.de
Jens Schuessler
>
> If you are not willing or even able to understand PLAIN FACTS, it is YOUR FAULT.
>
>
> DO NOT DARE TO USE RHETORICAL WARFARE.
>
>
>>(And, just in case you missed it: I can see how some of the issues you
>>raise can be problematic. But I strongly disagree that your solution
>>will actually be an improvement.)
>>
>>eirik
>
>
> Now I have one more piece of proof for what I wrote about Opera.com.
>
So you've proofed enough to be nothing else than a stupid troll.
Killfile adjusted.
Happy Oyster
I do not give a damn about such childish shit like "killfiles".
Fact is that Opera.com messed up.
--
Das Schrotgewehr Gottes
http://www.ariplex.com/ama/ama_gott.htm
NewsGroup
Well it seems that there is "something" going on with Opera, but from
the varied experience out there, it is not a bug that can be reproduced
easily. That is why I asked before raising a bug report in case it
was/is something to do with the particular configuration I have. It is
something to do with the Opera interface to the underlying file system,
that's about all I can say.
I never saw such a problem with the opera version 8's on Windows, but I
do get this periodic "Save / Save As" problem on both Windows Xp and on
Ubuntu.
--
Kevin Pugh