[chora] Re: Chora not seeing files

[Jay]

Chuck Hagenbuch wrote:
> Quoting Jay <j.javed at sherara.com>:
> 
>> Not bothered to tell you what I was using??? Not meaning to be rude but
>> I suppose simple english escapes you? I said "I'm a new user to Horde
>> and Chora, and have managed to get it working successfully (latest
>> releases)."
> 
> "latest releases" means so many different things to so many different 
> people
> that it doesn't really tell us. I meant a version number.
> 
Well, I take "latest release" as being the latest release. And since I 
also said I was a new user, it's hardly going to be an older one, is it? 
And I used the latest stable release, even if I used the alpha release, 
surely the bug I'm experiencing would be nonexistent there too, so in 
every way you look at it, your comment was just plain pointless. Can you 
please qualify and backup your post, what exactly is the range of 
possible values for "latest release" given that I posted my message 2 
days ago? I think you're talking out of your rear end, but it'd be 
amusing to see your response

>> That wasn't the only problem by the way, and contrary to what you say,
>> it's NOT fixed because the repercussions are still visible.
> 
> 
> Requiring a trailing slash on directories was fixed in HEAD a while ago. I
> didn't know what version number you were using. But this might not be 
> what you
> were seeing, since you still haven't really given a reproducible 
> description of
> what you're seeing.
> 
Well, obviously not in this case. As a matter of fact, let me turn on my 
sarcasm mode and say it works wonderfully and I don't actually get paths 
running around chora with double slashes in them. Unfortunately, that 
would be a lie. Maybe I'm just blind? No, I don't think so, I know my 
way around PHP like the back of my hand, and I do know what I'm chatting 
about. And I didn't say I don't have a reproducible description of what 
I'm seeing or doing. You pulled that out of the air of your own accord. 
I solved the problem and didn't see any need to discuss the point any 
further.

Oh, by the way, if this really was fixed, then why is Lutz mischa 
Heitmüller complaining about this very thing in his/her message titled 
"500 Internal Error, CVSRoot not found!"? I don't know if he/she is 
using the latest version, but it's not a bad presumption given that it 
was released in mid March and his/her post was only 2 weeks ago. It's 
not looking good for you, perhaps you'd better go back to your code and 
check if it really was fixed. It doesn't seem like it to me.

>> Well, when I get around to it, I might submit them to the bugs section,
>> but I'll do that as and when I'm ready to do it. And would you like to
>> point out exactly where I was complaining?
> 
> 
> I'm not sure what else to call a collection of "this is broken, that was 
> broken,
> I hacked this, this is out of coding conventions", with no details that'd
> enable someone else to fix them.
> 
A collection of this was broken, that was broken? My God, english really 
does escape you! All I said was I had to perform a couple of hacks, how 
you've managed to expand that into a collection of this was broken, that 
was broken, is beyond me. Tell me exactly how you came to this conclusion?

And what's amusing about the coding convention is that somewhere on the 
Horde site it going into quite explicit detail about abiding by the 
coding convention (good rule), and yet here's released software which 
clearly falls foul! It's not something accidental. At the end of the day 
if the convention says to write code which doesn't return errors when 
error reporting is turned on fully, how hard it is to write complient 
code? Not very, but it seems to be beyond you. And all I did was point 
it out. If you want to take that as being a complaint, then more fool 
you. How else do you expect to find out about bugs if no-one highlights 
them? I'm not some PHP newbie that doesn't know one thing from another, 
I took the code apart and put it back together again. Not too much hard 
work, but I made it more efficient along the way. And not a complaint 
either.

AND, I did ask Jan how to go about letting someone know about them, to 
which he/she (sorry) gave me the bug URL. Now you've hardly given me a 
chance to post the bugs yet, so how you can say I've not given any 
details on how to fix them is beyond me. A little patience wouldn't go 
amiss you know.

>> Besides, seeing as you have obviously not fallen foul of these bugs, why
>> do you care so much about what they are and what the fixes are?
> 
> 
> Uh, because I try to fix bugs in my software?
> 
Well, good job so far. Keep it up!

In case you couldn't tell (given your past history, you probably can't), 
that was sarcasm. Bugs you say are not there clearly are, and there's 
more than just me mentioning (not complaining about) it.

>> you're part of the Horde development team, it's ironic to think that
>> some of these bugs are actually caused by code that violates your own
>> coding conventions. For shame.
> 
> 
> Okay, I'm shamed. 40 applications, and I'd rather fix bugs than go over 
> every
> single line of code to make sure that the code is clean. Actually, I 
> often do
> that when I get distracted, but especially doing it for old stable 
> versions,
> when a new major version has already been cleaned up and is close to 
> release,
> is kind of a waste of time.
> 
> -chuck
Well, I'm sorry, but some of the bugs are so poor, you should be ashamed 
for not spotting them. And not only that, you're trying to tell me that 
one particular bug has been fixed, yet it clearly hasn't. I've got the 
code right in front of me and the bug is definately there! Your testing 
procedure is obviously not very bust.

But let me give you an example of a pretty poor bug which clearly didn't 
undergo much testing or thought. In Chora\lib\CVSLib\Checkout.php, line 
52, you've got this code:
if (!preg_match('/^([\S ]+),v\s+-->\s+st(andar)?d ?out(put)?\s*$/', $co, 
$regs) || $regs[1] != $fullname) {

More specifically, I want to concentrate on $regs[1] != $fullname, which 
just checks that the file paths actually do match. Now, on a *nix based 
system where file paths are case sensitive, this is OK. But on Windows 
which is NOT case sensitive, it might fail just because a path is not 
case identical. And this was actually one of the causes of me getting an 
error. In my case, the OS returned the path as it actually is (a mixture 
of upper and lower case letters) as opposed to what I'd entered which 
was all lowercase.

Now, what might have been a clever thing to do would have been to check 
whether or not the OS was a Windows based one (or one which isn't 
case-sensitive), and allow a case-insensitive match. That's what I did, 
and voila, it worked.

The Horde coding convention also states that code should be able to run 
error free with error reporting fully turned on. Chora most definately 
does not. I was getting quite a few errors output by PHP before I found 
and fixed them. I can't remember what they were now, but they shouldn't 
be too hard to track back down again. All I need to do is install a 
fresh copy of Chora again ;)

Let's try something else too. Now, I don't know if you'd call this a bug 
or not, but when I want to see the contents of a HTML file, I'd like to 
see the source code, and NOT what it looks like in the browser. Of 
course, in this current release (still on 1.2.1 here, I know you get 
confused easily) it just spits out the code which means I've actually 
got to view the source of the page to see the source code for the file. 
So I put in a small hack to convert it into a HTML readable format as 
applicable.

Hell, I was on a role. I even put a little bit of code to highlight all 
the (built in) functions in the PHP output to link to the manual 
(checking the ini directives in case there's a local copy), so when I'm 
viewing my PHP source code, I can click on a php function to go directly 
to the manual page just in case for example I want to see what it 
returns on error or whatever. Granted that's not a bug, but still useful 
all the same, and took me hardly no time to implement.

Basically, for the most part it seems that as soon as Horde apps are 
implemented on Windows, they fall down. And they're not very good at 
capturing errors either. Now, don't take this as a complaint. It's not. 
I realise apps will always have bugs and whatnot, and I don't mind going 
through and fixing them. But you were the one with some attitude problem 
coming along demanding that I highlight these problems and their fixes 
(no, I'm not going to write the code for you or give you what I've 
written and tested, you can do that yourself now).

Next time someone posts a question, you might want to get off your high 
horse and address them properly, instead of being so miserable about it. 
Or maybe you thought you could patronise me under the assumption that 
I'm a PHP newb or something. To tell you the truth, I probably know more 
about PHP than you do (and I'm not exagerating), and at the very least, 
I know how to write far more user friendly and robust code! Now, time 
for you to eat some humble pie.