[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: Renaming files on win32

From: Branko Čibej <brane_at_xbc.nu>
Date: 2004-12-23 05:13:51 CET

Gili wrote:

>On Thu, 23 Dec 2004 03:39:39 +0100, Branko ÄŒibej wrote:
>
>
>
>>Gili wrote:
>>
>>
>>
>>>On Wed, 22 Dec 2004 22:35:38 +0100, Julian Reschke wrote:
>>>
>>>
>>>
>>>
>>>
>>>>Gili wrote:
>>>>
>>>>
>>>>
>>>>
>>>>>On Wed, 22 Dec 2004 20:16:08 +0100, Erik Huelsmann wrote:
>>>>>
>>>>>
>>>>>...
>>>>> This covers all major cases off the top of my head. What do you
>>>>>think?
>>>>>
>>>>>
>>>>>
>>>>>
>>>>How do you handle other cases, such as
>>>>
>>>>1) Windows automativally loosing trailing dots and blanks,
>>>>
>>>>2) specific filenames that are illegal under Windows (con.*)?
>>>>
>>>>
>>>>
>>>>
>>> For case 1, you can store trailing dots or spaces by
>>>surrounding the filename with quotes. For case two, that's fine,
>>>Subversion already doesn't support this properly. We'd be stepping
>>>closer to the ideal behavior -- I don't claim it would be perfect.
>>>
>>>
>>>
>>>
>>Huh? You mean, in the filesystem? That won't work, because a) changing
>>the file name would create other sorts of conflicts, and b) Windows
>>filesystems don't allow quotes in file names.
>>
>>
>
> No. You asked how we deal with "Windows automatically loosing
>trailing dots and blanks". I stated it does not magically lose trailing
>dots and blanks if you surround the filename with quotes. I have no
>idea what else you're talking about.
>
>
I'm saying that you cannot create a file on Windows that has a trailing
dot or blank, no matter what you do.

>>To be brutally frank, you seem to think we overlooked something obvious.
>>We didn't. Sorry.
>>
>>
>
> I don't think you overlooked something obvious. I think that
>coming from a unix-centric view of the world, the developers simply
>placed this issue as a lower priority so it was not taken care of.
>
For the record, I've done all my SVN development on Windows for years,
so I don't have a particularly Unix-centric view of the world. And yes,
this issue is relatively low-priority compared to a fair number of other
issues.

> I
>don't think there should be a major difficulty in fixing this properly.
>
>
Then you've not been listening. Philip explained the issues quite well,
he's probably best equipped to comment on working-copy issues these days
so I'm inclined to trust his judgement.

>Also, I don't think it is fair to assume "Java must be taking
>shortcuts" as you stated in your previous email. You don't know that.
>
>
If your statement that Java has a locale-independent case folding
function for Unicode is true, then so sorry, my assumption is correct.
I've taken the trouble to learn quite a bit about localisation issues.

>If that was the case, you can be sure people would have complained
>about it not working as documented.
>
Sigh. As a counterexample, Java has had a broken threading API since day
one, and I don't hear very many people complaining (well, certainly
those who know about OO and threading, but they're a minority). I don't
accept "Java does this" as a good argument in general.

> I know I don't know for certain how
>it works, nor do I claim to know everything but I'm saying it warrants
>looking further into because whatever they did in
>Character.isLowercase(), we can do too.
>
>
-- Brane

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Thu Dec 23 05:13:48 2004

This is an archived mail posted to the Subversion Dev mailing list.

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.