Zugg Software :: View topic

Posted: Mon Apr 07, 2008 5:39 pm

As most of you know, I am having Lasik eye surgery on Wednesday (4/9), and so I won't be able to do a lot of computer work for the next week or so. Then, next week Chiara and I are going to Austin TX for the wedding of one of Chiara's sisters. So that will take another week. This means that I won't be able to work on CMUD for at least 2 weeks, which means the 2.23 version is probably about a month away (end of April or early May).

Since 2.22 seems to be working pretty well for many people, I encourage you to use v2.22 as much as you can and report the bugs that you find in it. I would like to make 2.23 a Public version so that I can let people use that while I work on the MyMuds.com site.

I've been pretty happy to see that most of the bug reports in 2.22 are actually problems that already exist in the current 2.18 beta version. Given all of the changes that I have made in the Settings Editor during this round of beta versions, I'm happy to see that people are not reporting a lot of problems with it. But definitely play with the new settings editor to make sure it's really solid. Given the number of crash dumps I continue to get from the settings editor in the 2.18 Public version, making 2.23 a new public version should really help with a lot of those problems.

I should still be able to post to the forums after the surgery and while we are in Austin, but I probably won't be posting as much as normal. I'll update my Lasik Blob thread a couple of days after the surgery to let everyone know how it goes.

Posted: Mon Apr 07, 2008 5:55 pm

Good luck with your eye surgery. I've been thinking about getting it done myself. Can't wait to her how it goes.

Charneus

Posted: Tue Apr 08, 2008 1:52 am

Good luck with the surgery. Having your eyes operated on has to be a little nerve wracking given how vital a sensory organ they are.

Since you mentioned bug reports I was wondering if you might comment on their usefulness to you. At one point you had made a comment about bug reports version 2.18 volume being less than you'd seen in previous versions. At that point in time I had stopped sending them in assuming that they were just repeats of things you'd already seen in the past. I hadn't realized you were using the volume of them to gauge the stability of the program as well. After that comment I had started sending them in again as they came up. This brings me to my current curiosity. I've been very active in sending in crash reports. Just looking at the past weeks period I count roughly 18 or so and it was a pretty typical week.

I guess my questions are on how useful these are and whether I should continue to send them even though I think they might be a repeat on something I've sent before. Also, we rarely seen the connections between our reports and bugs fixed unless we draw our own dotted lines from what we suspect we crashed on and items mentioned in the release log. So I can't help but wonder how often the bug reports actually help fix the real bugs or whether we need to come up with a procedure and post it for it to be truly effective.

As a final thought, thanks for all your attention on the bugs we've brought up. It pains me when I see posts on my Mud's site of people giving up on Cmud in frustration because of the bugs they experience. These last few versions have gone a long way, at least for me, in making it a lot more stable and usable. So I'm glad that a lot of these bugs were addressed before further features were added in.

Posted: Tue Apr 22, 2008 7:24 pm

OK, back from the wedding now. I will go through all of the new bug reports in the Beta forum tomorrow and add them to my bug list. My current plan is to try to do a 2.23 public release this week and then try to get the MyMuds.com site written over the next few weeks.

At some point I need to get out of the "bug fix - release new version - goto 1" endless loop so that I can actually get the MyMuds.com stuff going. It's already a month past the time I hoped to have it ready to play with and I've barely started on the coding.

But I understand that it's important to get the 2.2x version of CMUD out to the public so they can stop being frustrated with some of the bugs in the current public version. I'm definitely seeing fewer crash dumps from 2.22 than in previous beta versions, so I'm hopeful that 2.23 will improve things even more. Although it's possible that this is also the "spring break effect". I always get less sales and less MUD playing this time of year.

ReedN: In your specific case, since you send the crash dumps via email instead of via the normal crash dump system, they don't necessary help as much. If I'm working on a specific problem that you report in the forums, then I'll look at your crash dump in more detail. But because they are being sent via email, they do not get entered automatically into our crash dump database. At some point you might want to look into your Internet Explorer proxy settings to see if you can get the crash dump system to work properly for you.

Right now the best way to report a bug is to post a procedure for reproducing it to the forum. Pasting the stack dump from the crash dump into the forum post can be more helpful than just sending it in a separate email because I often link the forum post into my bug database and use the information in the forum to reproduce and fix the bug. Also, by posting into the forums, other people can check to see if they can reproduce it too, and Taz can get it entered into the online bug list.

The automated crash dumps are useful, but only for a more statistical analysis. The automated dumps get grouped by the specific crash and I tend to only look at the automated dumps when lots of people are crashing on the same bug. Or they are useful when tied to a specific forum post, or if they contain a procedure for reproducing the bug. I'd say that more than 50% of the automated crash dumps don't have any comments or procedure mentioned at all and are impossible to reproduce, so those are only useful for statistics.

Also keep in mind that there *are* thousands of CMUD users out there, so I get plenty of crash dumps for statistical purposes. For people like you who are already involved in the beta forum, using the beta forum is still probably the best. The automated crash dumps are more for "normal" users who don't even know that there is a beta forum.

Anyway, back to the topic at hand. I'll update more tomorrow, but I'm hoping for this week or next week for the 2.23 public version.

Posted: Thu Apr 24, 2008 1:50 am

My personal opinion, for what it's worth, would be to continue with bug fixes for a while. In the past when I made this same comment it was dismissed saying most people don't see the issues I see and that Cmud is stable. So I've dug up some corroborating evidence. On my Mud's forums (Achaea) I see pretty regular posts on the bugginess of CMud and people getting frustrated with it and some giving up on it altogether. Many of the individuals making the points are not new users and they have a great effect on other uses with their opinions.

Here are some references to topics in which Cmud was either the primary topic or it was discussed:
http://forums.achaea.com/index.php?showtopic=31416&hl=Cmud
http://forums.achaea.com/index.php?showtopic=31110&hl=Cmud
http://forums.achaea.com/index.php?showtopic=27459&hl=Cmud

So between adding something new and getting what you already have working well, I'd vote for fixing the existing code. Bug fixes may be thankless and ungratifying work, but from my own experience and that of those I interact with, it is what is needed.

Wizard Joined: 03 Mar 2001 Posts: 1127 Location: London

ReedN: +1

The MyMuds project doesn't do anything for me personally, nor does the free 'use it anywhere' CMUD client idea. I used MUSH years ago, I think around zMUD 5.x days, and when I moved to zMUD I never went back. Interesting to see there's some opinion that it's faster these days - not that I do any mudding at the moment so it's all academic.
I think the guy saying CMUD is the Vista of the MUD client world is a bit harsh, but not a million miles away. CMUD needs to be generally accepted as being stable before MyMuds turns up.

Apprentice Joined: 26 Apr 2007 Posts: 113

I'm not so sure. I think Cmud is pretty solid right now. Most of the problems that others have come across on my MUD (Aetolia) are ones that I have been able to talk them through. (And I'm not exactly an expert)

I like the idea of the MyMuds project and can understand Zugg's frustration in bugfix > release and go back one space. However, I do see the demand for an easy to use MUD client (Be it bigfixing or whatever) as a real matter of concern.

Most of the problem seems to come from People bringing .MUD files to Cmud, though. It may just be more prudent just to remove the option altogether and let them copy/paste the TXT format into Cmud's command bar. I personally started off in Cmud with a full rewrite of my Zmud system (nothing imported) and never ran into many of the problems that others are reporting.

In short: I think MyMuds is a great idea, however, there would be nothing lost in making it wait for another few months.

Posted: Thu Apr 24, 2008 5:36 pm

Those kinds of posts are nothing new. You can always find posts like that if you look. Back in the prime zMUD days there were always posts like that too. The problem with those kind of posts is that nobody ever mentions what version of CMUD they were trying. For all we know they were still trying to use the 1.34 version (or some intermediate beta version). Especially if they are using "hacked" versions (of which there are always many). A hacked version often causes additional access violations because they don't usually hack it properly. But people use it anyway.

The main reason a lot of MUD players are so jacked up on MUSHClient is because it's free now. No offense to Nick (who reads these forums sometimes), but a lot of players are always going to pick a free client over a paid client.

Notice that they all talk about how great it is to use Lua in MUSHClient, but nobody ever mentions that you can use Lua just fine in CMUD too? MUSH isn't necessarily faster. It depends upon what you are doing. It is slightly faster when it comes to scrolling text, and the web site claims that it is the fastest client, and so people just believe whatever they read. Yes, Lua *is* faster. That's why I added it to CMUD. CMUD uses the standard PCRE library, which is the best available. So I don't know where anyone gets off saying that the regular expressions in MUSH are "better" in any way. But if you coded the same real-life complex script in Lua in CMUD and compared it to MUSH, I doubt there would really be much difference. It's all perception.

People also love to compare stuff that doesn't even really matter. Yes, speed is important, which is why CMUD is faster than zMUD. But I've seen some people who compare stuff like EXE size and memory usage. And you know what, that doesn't matter these days. Computers are faster, hard disks are bigger. That fact that one program is bigger than another causes people to start saying that it is "bloated". Somehow they think that a smaller EXE means it is more efficient, when, in fact, it usually means that it has less features.

Also, some people are still just pissed that CMUD wasn't a "free upgrade". Many people still think of CMUD as an "upgrade to zMUD" rather than a whole new product that attempts to be as compatible as possible. The comment about Vista was harsh, but reflected this same perception. Sometimes I think Asilient is correct and I should just take out all of the zMUD Importing features completely. But then nobody would ever switch from zMUD to CMUD.

Of course, I find it hilarious (in a sick way) that some people loudly complain about how CMUD doesn't run their zMUD scripts perfectly and they don't want to mess with it, and yet they will consider switching to a completely different client where they will have to completely rewrite their scripts anyway (in Lua). Converting from zMUD to CMUD usually means "tweaking" your scripts. Going from zMUD to Lua means completely rewritting your scripts from scratch. Go figure.

It's also just another indication of how bad zMUD was in the kind of bad scripts that it accepted. A script could be full of syntax errors and other problems and yet it could still work in zMUD because it was so kludged. Sort of like the old days of Internet Explorer where it accepted a lot of bad HTML and allowed web designers to be really lazy about their pages. A lot of pages only got fixed when Mozilla/Netscape came up and refused to kludge their system to handle bad HTML.

But most of the problem is that these people with problems are usually using an older, buggier, and possibly beta version. That's why it's even more important to release a more stable Public version.

But that's no reason to stop adding new features. New features also drive people to software. People are not going to go through the pain of updating their scripts to CMUD unless there are compelling reasons for it. For example, a lot of people want the mapper to be improved and fixed. That's not going to happen until the current beta cycle is finished. And I'm not doing the new mapper until after the MyMuds stuff is working.

Obviously few people in this forum are going to be very interested in MyMuds...you all already own full copies of CMUD.

Anyway, the bottom line is that "CMUD being stable" is a perception issue. The people who are using CMUD without problems are the "silent majority". People post to forums when they have a problem, not when they are happy. All it takes is one vocal person on a forum who is either using an old version or has some kind of grudge to bad-mouth any piece of software.

I've been doing this for 12 years now. This is nothing new. I used to deal with this *all* the time with zMUD too. Do you want to know when zMUD was finally considered "stable"? After I stopped working on it. zMUD still has all of the same bugs that it always has, but it's considered a stable client. The fact is that as long as I continue to have an open beta policy and release new versions every few weeks, some people are going to perceive that as being "unstable".

Releasing a *good* Public version and then taking a month to work on MyMuds is really what needs to be done here. That will give people the time to get the public version and stop using beta versions.

Posted: Thu Apr 24, 2008 8:59 pm

Btw, back to the ETA of the original post. I'm going to take some of this advice and not try to release a "quick" 2.23 public version. Our next real-world time issue is that we have to go to Chiara's niece's graduation in Albuquerque on May 9th for a week. I always need a couple of days after any release to do any needed quick fix (in case I do something stupid and cause a new bug...it happens ;) So, I am going to shoot for a Public release of 2.23 sometime around May 2nd or May 5th. That gives me an extra week or so to try and fix as many of the remaining bugs as I can. Then I can work on MyMuds after we get back from Albuquerque. And that's the end of our trips until WorldCon in Denver in August.

Posted: Tue Apr 29, 2008 10:32 pm

I have been working on several hard-to-reproduce bugs for the past couple of days (as posted in other threads in this forum).

I found several problems at a pretty low level part of CMUD. Changing that part of the code always makes me nervous, since there can be unintended side effects on scoping and other issues.

So, what I am currently planning is to release the 2.23 version on Thursday (I'm busy on Firday) and give everyone a 3-day weekend to look for any new problems caused by my "fixes". Then I'll fix any new side effects early next week and actually release 2.24 as a public version (besides, we all like even-numbered public versions anyway ;)

Posted: Wed Apr 30, 2008 12:47 am

Yes we do, they are the best! Twisted Evil

At least there is a possibility of 2 positives by that time depending upon replies elsewhere. Very Happy