Ability To Shorten File Names When Greater Than 256 Characters Long.
Posted 22 August 2008 - 12:05 PM
I use Vista (and before that XP) and there are many occasions when I have tried to extract an archive, only to be greeted with the error message that the 'path depth should not exceed 256 characters.' (Which I assume is a file system limitation - or perhaps a limitation of WinRar?)
In any case it would be a very useful feature to have. We perhaps can't make people behave sensibly, but maybe we could remind them somehow that their file name is too long and/or provide a simple facility to shorten file names, or remove illegal characters or spaces etc. upon completion?
Doing something like this would serve to clean up many of the file names that currently exist on the emule network.
Posted 23 August 2008 - 10:13 AM
A good example of this might be something like this:
(This is a virus BTW, so don't download it).
This is by no means the worst offender I have seen (in fact it's really quite tame compared to the length of many other file names I have encountered on the emule network through the years).
This may not be 256 characters long on it's own, but add to this the depth of the default location of my Downloads folder: C:\Users\user1\Downloads\eMule and the fact that my own habit is to right click on a compressed folder and select 'extract to' and accept the suggested folder name, this would result in a path as follows:
C:\Users\user1\Downloads\eMule\Mp3 Song Polyphonic Ringtones - Lionel Richie Christina Aguilera Snow Patrol Luny Tunes & Tainy Carrie Underwood Lupe Fiasco Various Artists\Mp3 Song Polyphonic Ringtones - Lionel Richie Christina Aguilera Snow Patrol Luny Tunes & Tainy Carrie Underwood Lupe Fiasco Various Artists
Which is well in excess of the 256 character limit. (Particularly when you add in the often quite long names of the individual files inside the zip or rar archives themselves).
You could say 'why not adjust our behaviour?' Well firstly because on 70% of occasions accepting the default folder name to extract to supplied by winrar (and other compression software) works perfectly fine. However it is that other 30% which is a royal pain in the butt. Secondly I think his is a good idea because many of the file names on the network are often completely stupidly long - and there would be no harm in cleaning them up. Crazy long file names that have been tagged by just about everyone who ever downloaded them are a real pet hate of mine, as I spend a large proportion of my time renaming them anyway. Thirdly, doing this might cause what clearly must be some quite unaware and uneducated kind of people to think much more carefully when naming (or renaming) their files in the future. So overall the network benefits too.
Anyway as I said, it's just a thought.
This post has been edited by raid517: 23 August 2008 - 12:14 PM
Posted 23 August 2008 - 11:36 AM
Posted 23 August 2008 - 12:08 PM
There is plenty of software out there that can and does read file names and that does rename according to a set pattern. In general this is usually highly successful and rarely makes file names unreadable - even if the software itself does not 'know' whether the symbol or letter or number in question is in some way significant to the file name itself.
Certainly it can't make file names any more unreadable than some of the crazily long file names that already exist out there.
Moreover the ability to rename files in this way would have two effects. First it would result in an automatic standardisation of file names on the network - so if file names were changed according to a set pattern, this would mean that in time all the file names for that particular file would be renamed. This means that there would be no confusion, since all the file names for that file would be the same. (At least for the files that shared the same hash).
Second, overall it would not affect the availability, or the ability for users to search for files or to find files that they are looking for - since files are both searched and indexed according to their hash values - and not their file names.
So a file that has a shortened name on a users PC, can still be searched for using the full/correct name on the network, since this name (and the ability to search for it) is derived from the hash value and not from the name itself.
In effect therefore, this would mean that there would be no negative impact, either for the user who may select to enable this option, or for the availability (and the ability to search) the files on the network itself.
The only outcome would be that the actual file names on the network would be cleaned up over time - and it might stop the apparent multitude of fools who appear to exist out there giving their files often obscenely long names. (Sorry to be so blunt, but it really is a pet hate of mine).
It doesn't seem like anyone can really loose if this were possible.
This post has been edited by raid517: 23 August 2008 - 12:13 PM
Posted 23 August 2008 - 12:25 PM
I would not trust any program for fully automatic renaming, as there's no strict pattern in names.
Posted 23 August 2008 - 04:23 PM
But surely I can't be the only person who finds this frustrating?
On the flip side you list some of the (potential) limitations of (optionally) allowing files with very long file names to be renamed automatically and to continue to exist on the network. Perhaps you see some direct advantage in this that I can't see? If so please feel free to explain these to me? I am quite open to having my mind changed.
I will wager that everyone here who either reads, or comments on this topic has at one time or another (and probably quite frequently) encountered exactly the same issue. (Including no doubt, despite your objections, your own good self). Indeed I think it would be impossible to use he ED2K network right now even on a moderately rare basis without encountering this issue.
I have no idea what your objection to 000.part files would be - as firstly these would be unlikely to be renamed - since the stipulation would be to only rename files over given sensible limit - and so files like these and other files with file names shorter than this limit would never be likely to need to be renamed.
It would seem to be much more sensible to perhaps search the hash value of files first before searching file names - as I would assume that this would avoid searching and retrieving a great number of unnecessary duplicate files. (But then as I am not a programmer, I have no idea why this isn't the case). In this sense it shouldn't matter what a file is called, since all of the required information (including the original searchable file name) would ideally be contained within the hash value.
Even if all that happened was that the client issued a polite message to the user warning that a file name may be too long and that they should 'consider revising', at least it would be a start, at least it would be something. Maybe then people could be educated about the idea of file name size limitations on different kinds of operating systems and on their associated file systems?
I'm sorry I can't be kinder to people who give their files these insanely long file names - but as someone who has used the eMule client from more or less the outset and who has been a member of these boards for over 6 years and who has seen the quality of the network degrade over time as a result of this issue, my patience at last gave way and I thought it might finally be a good time to perhaps say something.
As for examples of software that can reliably rename files, I agree that none can 'understand' the concept of a file name (in perhaps a similar way that you and I may do), since what you are asking is for me to do is to provide you with an example of software that has some kind of innate sentient ability, which I think you know is impractical and not possible with current technology.
However there are plenty of examples of software that can reliably rename files and leave them in very understandable and readable condition. Software such as Tag & Rename, FATXRenamer (which I use for renaming files for XBMC on my XBox) and several FTP programs such as FlashFXP, Cute FTP and so on can all be set to automatically and reliably rename and limit file name sizes according both to the target file system and to a set of carefully predetermined rules.
So there is already a well established tradition of limiting file name size - particularly for files being transferred over a network.
It is also a practical problem - as I have encountered many examples where less proficient users have attempted to extract the contents of compressed archives in exactly the way I have described and have encountered this same issue. Since they were unaware of this limitation, they will often simply assume that the file is corrupt and delete it from their systems, thus limiting many potential new sources.
Still if you guys are happy with things the way they, it's all good with me.
As I said, I just thought I would put it out there for consideration.
I can't really see why anyone would really like files with such vastly long file names - but I admit that it wouldn't be the first time I was wrong if there were.
This post has been edited by raid517: 23 August 2008 - 08:45 PM
Posted 23 August 2008 - 08:11 PM
But even then there were problems with too deep nesting of subdirectories in archives.
So far, I had no troubles with file names being too long; but have seen a few incoherent abbreviations.
Yes, sometimes I use multi-rename tools (not with eMule's files, though). Mostly when I need to rearrange file names for batch processing.
But only when it is possible to create renaming pattern; which is hardly possible in general in eMule.
Unless you queue dozens of files daily, R-click on the file and change it's name to whatever you like. Even before you have the file: in eMule, in Transfers or in Shared files.
Posted 23 August 2008 - 08:42 PM
It's interesting that you say you have never encountered this limitation. Why should you never encounter it, while I encounter it at least once every time I use emule? (Particularly on compressed .rar archives). Do you only rarely download .rar archives by any chance?
I won't go into details about my downloading habits - I don't download porn and I am not referring to porn at all when I say this (so don't get them confused), but what I do download most of all is large entire 'collections.'
I don't download movies. photos, software, or individual mp3 files. At a guess the only way I can imagine that you would not encounter this issue (given that it took me nearly 7 years to actually bring it up and in that time I have encountered it maybe 100's if not thousands of times), is if somehow your downloading habits were vastly different from my own?
I am aware of how to rename files - even when they are still in progress. However that is just manually renaming files - and does nothing at all to address the issue. (The issue being - at the risk of sounding repetitive - people giving files impractically long names).
This post has been edited by raid517: 23 August 2008 - 08:51 PM
Posted 23 August 2008 - 09:49 PM
Seven for the Dwarf-lords in their halls of stone,
Nine for Mortal Men doomed to die,
One for the Dark Lord on his dark throne
In the Land of Mordor where the Shadows lie.
One Ring to rule them all, One Ring to find them,
One Ring to bring them all and in the darkness bind them
In the Land of Mordor where the Shadows lie.
Dark Lord of the Forum
Morph your Mule
Need a little help with your MorphXT? Click here
Posted 23 August 2008 - 10:50 PM
Posted 24 August 2008 - 12:05 AM
This option automatically cleans up filenames whenever a new download is added to the Downloads list. Special characters like _ , . ,  are removed or replaced by a blank and the first letter of each word is capitalized. Further, special strings which can be defined using the Edit button are also removed from the file name. There are already strings defined which will be used but new ones can be easily added by using | as separator.
Images for CD recording software often consists of more than one file, e.g. *.cue/*.bin (CDRWin), *.mds/*.mdf (Alcohol120%), *.ccd/*.../files/help/1033/*.sub (CloneCD), *.bwt/*.bwi/*.bws (BlindWrite). The option should not be used when downloading such files. Changing the filename will result in error messages when trying to burn them as the correct filename is also saved within these files.
If you change the filename anyway, it has to be replaced in the according layout files, too. For *.cue files this can easily be done with Windows notepad but the other formats mostly require some hex editor like UltraEdit or WinHex.
Files in the transfer list can be cleaned manually by selecting them then pressing CTRL+F2.
This post has been edited by f_seger: 24 August 2008 - 12:06 AM
Posted 24 August 2008 - 08:25 AM
The can't be that vastly different.
There are four main categories of things that can be downloaded from eMule - and most other Peer2Peer applications. These are video, images, applications/software and music. I personally have little interest in the first three of these. But eMule is and has always been most effective for the latter of these, particularly when searching for collections by artists who may be either very obscure, or who may be difficult to find by any other means for another reason. (I like a lot of tribal and ethnic/world music too - so copyright is rarely an issue).
These collections generally come in compressed .rar format (and more rarely in .zip format).
If you wish we can compare notes in PM and see perhaps how, after so many years I have encountered this issue on such a regular basis and why it seems you have not?
This post has been edited by raid517: 24 August 2008 - 09:01 AM
Posted 24 August 2008 - 08:45 AM
Not sure what you are driving at there, but yes I do have an issue with people giving files impractically long names - and I don't feel there is any reason to apologise or make any kind of excuses for this. In an ideal world, people should know better.
Of course a lot of people don't know better - which is why it might be a good idea even (if nothing else) just to perhaps attempt to politely remind them that there are physical limitations on all file systems about exactly how long a file name can be.
As I said I'm sorry I can't be kinder to people who do this, but after 6 years and seeing the problem get a lot worse over the last year or so, I have rather just come to despair.
Most if it is just 'tagging' anyway. I mean most of the time the majority of the file name comes in the first two dozen or so characters - and the rest is just nonsense about who's hands the file passed through at any given point and who repackaged the file along with extended (and usually unneeded) detailed information about the package contents.
This is useful and I have used it for a long time - but it doesn't seem to help with limiting file name length - or in educating users about the limitations of file name/character length on the file system they may be using.
As I said, even a polite message saying something along the lines that the 'file name may be too long' and that they should 'consider revising', would be a vast improvement from the current state of affairs.
Posted 24 August 2008 - 01:41 PM
I ran into this problem a number of times myself, and all I did was to prevent Windows from screwing itself up by placing my incoming folder directly off the root.
I haven't had this problem since.
It might be nice to have eMule truncate file names if they are too long. But its far better to rename the files yourself. Any automated system of file renaming has the potential to destroy valuable info within the file name - ie: the index number to a series of files, etc. If eMule were to have some sort of AI that would save the differences in file names, I'd hate the see the results. Series of files you find in eMule searches tend to have varying file name formats even for files which are part of the same series (Users choose the file names they like). eMule would only make them look worse.
This post has been edited by Dick_Manitoba: 24 August 2008 - 01:42 PM
Visit PunkHistoryCanada.ca and enjoy it's music at PHC Radio. Listen now!
::How To:: Learn about IP filters and fake files :: Info & Lists of Real and Fake servers
Remove Fake Servers & get good ones :: Post pictures on the web, no registration req'd
::Helpful sites:: DVD tools, tutorials, etc :: VLC video without codecs :: Codec info & downloads
GSpot Codec IDer :: Can't delete Locked files? :: Freedb aware players
::Free Virus Scans:: Trend Micro Housecall, Panda ActiveScan, Kaspersky Virusscanner, BitDefender
::My Favorite Applications::
::Archives:: IZarc : 7-Zip : WinRAR : WinACE : Any-OS Zip Tools and Source code - Info-ZIP
::Audio:: EAC : LAME : MP3 Book Helper (Tags etc)
Posted 24 August 2008 - 02:10 PM
People somehow named files, packed an archive and give a name to it. They share it and you can download it.
Where they are breaking limitations?
Posted 24 August 2008 - 04:03 PM
You could look at it from the 'flip side' and say since the file was successfully packaged that no limits were broken. However it is at the other end when a user may try to unpack a file/compressed archive where the problem emerges.
I didn't just come here for me - as I have seen this problem before on multiple occasions on other users systems. In such cases in general a user (particularly if they are inexperienced) may often simply opt to delete a file believing it to be corrupt, rather than shorten the file name, thus as I have said limiting many potential new sources.
Unfortunately the /root drive may not be accessible to many non-admin users, so placing a downloads folder there might not always be practical.
But as I said even if there was just some kind of polite warning that encouraged people to be considerate when naming files about other users, this would still be much better than nothing.
Whether a person renames a file, or whether a file is renamed automatically hardly seems relevant - as in many cases file names can be so long (or short given the perspective it seems of some posters here) that they are probably much more garbled and unintelligible as any file name assigned automatically by the by the client could be (or indeed in any other way).
Even if it wasn't anything to do with WinRar, or any other package extraction software, there is (at least for me) the simple sheer annoyance factor of it. I simply don't see the point in giving files such epically long names and even if I don't encounter any problems unpacking the package, I still find myself using up a disproportionate amount of my time when downloading from emule in renaming files to something more sensible that in most cases more accurately reflects the file/archive's content.
Well anyway, as I said, it's just an idea. Probably those who say they haven't encountered this will encounter it at some point - particularly as, as I said, in my view this is a problem that is getting progressively much worse (especially over this last year or 18 months or so), rather than better.
This post has been edited by raid517: 24 August 2008 - 04:35 PM
Posted 25 August 2008 - 07:42 PM
Making maximum path length equal to maximum file name length, while allowing a few levels in subdirectory tree is not a smart move by any standards. Especially MS itself uses lengthy default file location like
\Documents and settings\Username\Application data\My documents\My pictures
\Program Files\Microsoft Visual Studio .NET 2003\JavaLanguageConversionAssistant\1033
And these are not the worst possible.
Having said this, I still do not understand why you ask eMule devs to rectify user's behaviour. You could ask MS. You could ask rarlab people.
eMule only distributes exact copies of files, not paying much attention to it's contents.
And I have a suspicion that devs would not be thrilled with those friendly warnings; same for full-featured rename functions or random truncation of file names.
Posted 25 August 2008 - 09:08 PM
Be courteous and constructive in your posting. Yes, some questions are just plain dumb but if you don't have the patience to type something useful just let someone with more patience address the matter. Note: if you read the Documentation, FAQ and sticky threads that say "Read Me First" you are a lot less likely to ask a dumb question in the first place.
Offensive or otherwise worthless posts may be editted or deleted at the discretion of the Forum Moderators. In general, threads will be locked or Binned rather than being outright deleted.
eMule requires a lot of patience and is a community project supported by many volunteers. So please read first, think a little and then post.
This post has been edited by torpon: 25 August 2008 - 09:09 PM