Gossamer Forum
Home : Gossamer Threads Inc. : Discussion :

Gossamer Threads Development Survey

(Page 2 of 4)
> > > >
Quote Reply
Re: [GT Dev Team] Gossamer Threads Development Survey In reply to
Hi,

In most things I strongly agree with "Webmaster33".

I think there is a book and website GT should invest in and have a look at, which give answers and some very good tips towards more or less all the questions asked:

http://www.webstyleguide.com (Web Style Guide, 2nd Edition)

It will be a very good idea to have a read through it and keep many of things in that book / website in mind.

GT products templates i.e. xhtml and css should also try to comply to W3C (http://www.w3.org/)


The Questions:

1.

Using CSS and XHTML is a very good idea, since that would simply be to keep up to speed with developments on the rest of the net. No HTML or CSS should be left within the coding since it really messes up any customized looks. Going fully with CSS definitely is the right way to take.

Again as Webmaster33 says:

"GT should only use CSS as long the NS4+ and IE4+ compatibility is maintained!!! It should be kept backward compatible, and readable with 4.x browsers."

Compatibility and backward compatibility is extremely important!! Its important to remember that its the usability of a site that matters not how flashy it looks.

2.

Same again. Extremely important to maintain NS4+ and IE4+ compatibility. A user who cant view your site due to old browser or features not working due to some incompatibility, is a customer lost!

To many designers try to make their site look really cool and fancy, they might please themselves but end up creating websites that are useless and not very good for the people who matters, your customers / visitors.


3.

I would prefer to have no java scripts, or at least only for features that are of no importance. It comes back to Question 2 and compatibility. You really don't want to have something that might make the site look different, not work or make things problematic for any of your visitors.


4.

PDF format is good, but a HTML version should also be available.

5.

Installations seems to be fine, however sometimes it does feel a bit risky. To create backups of existing installations and maintain them should be made easier. Currently there are many different ways it seems to make backups, all of them not very straight forward and to be honest none of them seem to work very well.

I would feel much more confident doing installs and upgrades if I knew that any previous installation was backed up and easy to go back to.

6.

More examples and explanations of how different fields or columns can be used or what they are.

7.

As mentioned before. There is a constant feeling of a unstable product, this I think is mainly due to that for example in LinksSQL there is no 1 simple way of creating a full backup or part backup which then easily can be restored. Or a easy way to retrieve information from a backup file to a fresh installation. The times when something has gone wrong and I have tried do a fresh install.. only to find that whenever I use any backup file, it would tell me that some fields or data is missing.. I then find myself spending a long time trying to create the "correct" environment so that I can use a certain backup file.

There needs to be put in place a easy way to create a backup which will work on any installation, fresh or old. Something that works well and you can rely on. If I was confident that a backup existed which I easily could use then installing, maintain or simply playing around with a install wouldn't be so "scary".

Other features, I agree with some of the other guys have mentioned:

- Support LSQL to hide categories
- Support LSQL to set to exclude categories from link-submitting
- Less templates
- WAP Support (There will a lot of business (and money to make) in the next few years on the WAP / Mobile market.)
- More advanced Editors functions so that they can take over more work from the administrator, i.e. in validating etc. Maybe different levels of Admin access?!
- Easier to share or the possibility to share databases, i.e. 2 different sites (each with its own domain, licence and databases) being able to access and use each others databases. For example to be able to do searches on 2 different domains and database from one site. Basically easier to share and access information.

Quote Reply
Re: [maxpico] Gossamer Threads Development Survey In reply to
Quote:
It's useless that everytime one criticizes GT for their unmet timelines you reply with this same stuff. Nowadays in this fast paced world if you're not able to meet timelines you'd better change job or at least you're not successful. It could happen two, three times not to meet an already announced timeline and even so not with several months of delay.
I'm not saying this is easy but you really cannot say that "timetables are almost IMPOSSIBE to keep to...because we're humans". Sorry but this is a loser attitude

Dude... we're all entitled to our opinions. I'm just giving mine, from a developers point of view. Sorry if that offended you.

Cheers

Andy (mod)
andy@ultranerds.co.uk
Want to give me something back for my help? Please see my Amazon Wish List
GLinks ULTRA Package | GLinks ULTRA Package PRO
Links SQL Plugins | Website Design and SEO | UltraNerds | ULTRAGLobals Plugin | Pre-Made Template Sets | FREE GLinks Plugins!
Quote Reply
Re: [Andy] Gossamer Threads Development Survey In reply to
Sure Andy. Was just pointing out (in a direct manner) that your opinion is already clear.
Certainly don't want to start a flame on this. Peace Angelic
Max
The one with Mac OS X Server 10.4 :)
Quote Reply
Re: [maxpico] Gossamer Threads Development Survey In reply to
Hi Max,

The license interpretation put forth by Andy is correct, all Links SQL version 1 and 2 license holders will continue to get free upgrades for life.

This license change only affects clients who buy Gossamer Links version 3 licenses as their first dealings with GT.

As for development timelines, we're painfully aware of the problems in the past and have put in place several measures to help us address the issues. We shall see if our efforts have been in vain with this development push.

That said, even the largest software companies with their legions of coders have problems meeting timelines, so please don't be too harsh with your criticism should we stumble again, we're doing our best, working late into the night and giving up weekends.

Gossamer Threads Development Team

Last edited by:

GT Dev Team: Nov 17, 2004, 11:09 AM
Quote Reply
Re: [GT Dev Team] Gossamer Threads Development Survey In reply to
Hi,

I know that you said you don't want any product specific suggestions but I just thought I'd mention this as it is my biggest bugbear with LinksSQL. If you get a chance at some point it would be great if you could have a look at the internal search so that it would give the highest score to an exact phrase match, then all words, then any words and then substrings (if substrings is on). This would be a huge improvement.

Even if you could just give some pointers on how this might be done that would be helpful.

Thanks,
Laura.
The UK High Street
Quote Reply
Re: [demon] Gossamer Threads Development Survey In reply to
In Reply To:
http://www.webstyleguide.com (Web Style Guide, 2nd Edition)

It will be a very good idea to have a read through it and keep many of things in that book / website in mind.
We've recently been looking for a professional web application design firm to design our templates, and based on comments in this thread, we've now enlisted the help of a firm with expertise in XHTML + CSS to help us with the template design and UI improvements.

Thank you for your recommendation of the Web Style Guide book. We do have to take into consideration however that the book is now nearly three years old, and some of its contents might not be entirely applicable today as Netscape, prevalent then, has been
almost completely replaced by by Mozilla and derivatives.

In Reply To:
"GT should only use CSS as long the NS4+ and IE4+ compatibility is maintained!!! It should be kept backward compatible, and readable with 4.x browsers."

Compatibility and backward compatibility is extremely important!! Its important to remember that its the usability of a site that matters not how flashy it looks.
Old browsers like Netscape 4 have extremely limited and broken CSS support. It is impossible to write good/proper XHTML+CSS pages which look the same in new browsers as well as old browsers such as Netscape. However, well written XHTML+CSS will still be 'viewable' in these older browsers - look at Microsoft's webpage for an example. The user can still to use the site, but it won't look pretty. We believe this is an acceptable fallback for 4+ year old browsers.

In Reply To:
I would prefer to have no java scripts, or at least only for features that are of no importance. It comes back to Question 2 and compatibility. You really don't want to have something that might make the site look different, not work or make things problematic for any of your visitors.
As stated in the original post, javascript would only be used in providing extra non-essential features. For users without javascript, we will ensure that the UI won't be cluttered with non-working javascript links/buttons. A good example of this is the address book autocompletion that will be available in Gossamer Mail 3.x. If you have javascript enabled, you will be able to autocomplete addresses from your address book when composing an e-mail, otherwise, the text input acts like a normal text field.

In Reply To:
Installations seems to be fine, however sometimes it does feel a bit risky. To create backups of existing installations and maintain them should be made easier. Currently there are many different ways it seems to make backups, all of them not very straight forward and to be honest none of them seem to work very well.

I would feel much more confident doing installs and upgrades if I knew that any previous installation was backed up and easy to go back to.
Thank you for the suggestion, and we will be looking into improving this aspect of our programs.

Gossamer Threads Development Team
Quote Reply
Re: [GT Dev Team] Gossamer Threads Development Survey In reply to
Times are changing, but there are still really good reasons to have a "plain" template set. Maybe if you are using "plain" some features are disabled. It wouldn't really mean two full development trees.

Anyway, the big issues are using javascript to validate form input before sending to the server.

To have a javascript category box, so that you pick a top category, the next level opens, and so on. That would be the most major interface improvement of all.

Also, for "serious" or "dangerous" options, like deleting, rather than just a click on a yes or no, you have to type in "yes". One place this is important is in "gadmin" where many windows machines will now place the cursor over the "hot" answer, so if you accidentally (or on a slow machine) click twice, you will force a delete, without an abort.

Typing in a "yes" to delete a server, or MySQL database is much safer, and could be done in javascript.


PUGDOG� Enterprises, Inc.

The best way to contact me is to NOT use Email.
Please leave a PM here.
Quote Reply
Re: [GT Dev Team] Gossamer Threads Development Survey In reply to
Hi,

I' d just like to say that having worked on large scale both corporate and Business 2 Consumer web sites in Europe I disagree with the opinion that everything has to be done to keep the system compatible with browsers 4.X. I have spent hours in board rooms with directors explaining that 6% of the target population use 4.X browsers and therefore not only do five new interesting functions need to be dropped and the design needs to be reviewed in certain areas only for them to do a U-Turn when the stats drop by 2% four weeks before launch !

Does somebody have access to up to date stats on the 4.X browser usage (maybe by region as well : America/Europe/Asia...) ?

I don't believe in the whole "winner takes all", "think for yourself first" approach but dropping functionality and usability and design versus "basic usability" seems to me to be very open for discussion. Making decisions can't always be about pleasing everybody without exception to the detriment of over 95% of the people and sometimes that means leaving some people out. Using CSS and XHTML may not be the perfect solution but providing HTML that is XHTML compliant and adding CSS should be the way forward.

I have worked with ergonomics /user friendly specialists over the years and like marketing specialists you sometimes feel like saying well of course we should do that but when they're not there you sometimes forget to do things the right way. In this case we are talking about not providing new functions in certain cases or even not providing more user friendly templates as opposed to reducing functionality for a minute percentage of the population...

Thanks,

John
Significant Media
Quote Reply
Re: [Jag] Gossamer Threads Development Survey In reply to
Wanted to add this to my previous post but passed the edit time delay :

PS : I do think that demon's point about "a customer lost" is a valid argument there's no doubt. Even more so if people are selling. However from an economical point of view I suppose one needs to weigh the cost of a) losing functionality to the competition b) the extra cost of backward compatibility (that goes so far back IMHO) compared to losing customers that are using old browsers.

Potentially one could even discuss (but this gets very 'iffy' and I don't really like it) the buying power of people with very old equipment. I've already been in discussions where this argument has been used and for industrial companies it was explained that it's not always valid.

In my case the web site is funded through my own money and for free so I am not interested in spending extra cash (from my pocket) to deal with people that for different reasons cannot or do not want to install newer versions of browsers.

John
Significant Media
Quote Reply
Re: [GT Dev Team] Gossamer Threads Development Survey In reply to
In Reply To:
To provide more information on your issues regarding scalability, we just wanted to point out that we do have some sites running our programs in a clustered environment. For a couple of examples, please check out our new Clustered hosting page here:


Is this also a solution for people who don't host with you? Or just another ad for Gossamer hosting? Wink

My biggest suggestion thus far:

Cease selling Gossamer hosting as the answer to problems and feature requests, ebcause some people can't host with Gossamer(I, for one, have explained why over 17 times when getting the ole Gossamer hosting spiel).

I have great respect for Gossamer products and purchase many of them, along with custom coding.

But hosting with Gossamer doesn't make business sense (prices out of market range for fanatically managed servers) and doesn't make technological sense (the best servers G-Host offers have specs under the absolute worst I can get by with) for me, so when I get this spiel each time I contact Gossamer it gets irritating.

And note, I am a big Gossamer fan. Were I not I'd probably never buy Gossamer products again, it sometimes has the feel of dealing with used car salesmen who ignore what you need and harp on a sale.

cdkrg

Able2Know :: Ajooja Directory

Last edited by:

cdkrg: Nov 21, 2004, 2:16 PM
Quote Reply
Re: [GT Dev Team] Gossamer Threads Development Survey In reply to
In Reply To:
That said, even the largest software companies with their legions of coders have problems meeting timelines, so please don't be too harsh with your criticism should we stumble again, we're doing our best, working late into the night and giving up weekends.


As an example one could point to teh biggest development team of all, Microsoft.

Look at the Longhorn development cycle. Look at how file indexing changes were delayed and removed from Longhorn because they couldn't meet the schedule.

Development deadlines are diametrically opposed to development quality and sound releases.

Another suggestion for Gossamer: continue to get it right, instead of satisfying those who want deadlines.

cdkrg

Able2Know :: Ajooja Directory

Last edited by:

cdkrg: Nov 21, 2004, 2:20 PM
Quote Reply
Re: [cdkrg] Gossamer Threads Development Survey In reply to
Quote:
Is this also a solution for people who don't host with you? Or just another ad for Gossamer hosting?
It's a solution for people who don't have the expertise to handle setting up a clustered environment and/or don't want to manage everything, as well as taking advantage of our experience with mod_perl and our software. What we have done with the Gossamer Host cluster setup can also be done in your own setup if you have the time and experience/knowledge to set it all up. What we meant by the post was more of an example how our products can scale. Sorry about any confusion.

Gossamer Threads Development Team
Quote Reply
Re: [GT Dev Team] Gossamer Threads Development Survey In reply to
In Reply To:
As for development timelines, we're painfully aware of the problems in the past and have put in place several measures to help us address the issues. We shall see if our efforts have been in vain with this development push.

I'm glad of that. But still, you don't have an even rough timeline for us? Tongue

Btw, are you looking into Perl 6? I don't know when the final release will come out but it would be cool to have new GT products enhanced with the new features of this major release. http://dev.perl.org/perl6/
Max
The one with Mac OS X Server 10.4 :)

Last edited by:

maxpico: Nov 22, 2004, 4:04 AM
Quote Reply
Re: [cdkrg] Gossamer Threads Development Survey In reply to
In Reply To:
In Reply To:
That said, even the largest software companies with their legions of coders have problems meeting timelines, so please don't be too harsh with your criticism should we stumble again, we're doing our best, working late into the night and giving up weekends.


As an example one could point to teh biggest development team of all, Microsoft.

Look at the Longhorn development cycle. Look at how file indexing changes were delayed and removed from Longhorn because they couldn't meet the schedule.

Development deadlines are diametrically opposed to development quality and sound releases.

Another suggestion for Gossamer: continue to get it right, instead of satisfying those who want deadlines.


For me, it is not neccessarily the missed timelines that are frustrating. It is the lack of communication. A timeline will be announced and the announced released date passes, there is no product and no communication from GT.

Then a year later, the product will be released, again with no communication as to why the extra time was needed to get it "right" instead of on time.

I would appreciate more information like the Nov 15th post from the GT Development Team.

I can understand the reasoning for not providing timelines and expected release dates. On the other hand, I don't understand why GT would not want to provide progress reports. Something posted in the announcements forum like:

Links 3.0 (SQL) - Developing features lists, logic and design concepts.

GForum 3.0 - Writing perl code libraries

GMail 3.0 - Close to a (mostly) functional alpha version. Features G, N W have been dropped due to time/budget/technical constraint/lack of demand.

GList 3.0 - Completed internal beta testing, fixing bugs in preparation for non-GT beta testing. New features X,Y and Z have been added.

DBMan SQL 3.0 - Finished initial non-GT beta testing. Preparing for public beta release. If beta goes successfully, this will become the first official release of DBMan 3.0.

This would provide useful information without attempting to predict future timelines.
--
Rob

SW Montana's Online Community
Modular Model Railroading
Quote Reply
Re: [GT Dev Team] Gossamer Threads Development Survey In reply to
Just make the move from Gossamer Threads links SQL to the next version an easy one... PLEASE!!! it took me forever to work out all those templates - can't imagine what will happen if I'll have to start over!


Thanks
http://www.FindSP.com - Hosting Directory
Quote Reply
Re: [FindSP.com] Gossamer Threads Development Survey In reply to
Quote:
Just make the move from Gossamer Threads links SQL to the next version an easy one... PLEASE!!! it took me forever to work out all those templates - can't imagine what will happen if I'll have to start over!

Just the point I was going to make.

People have spent a lot of time developing templates, and a whole look/feel to their sites, which will, apparantly, be invalidated in the next release. Major work will have to be done to implement it.

I know how complicated the GForum templates are compared to Links, and they are daunting, if not scary.

Vertigo has set in. My head is spinning from trying to visualize the issues here.


PUGDOG� Enterprises, Inc.

The best way to contact me is to NOT use Email.
Please leave a PM here.
Quote Reply
Re: [GT Dev Team] Gossamer Threads Development Survey In reply to
 

Best regards from
Bremen/Germany

Lothar
Quote Reply
Re: [FindSP.com] Gossamer Threads Development Survey In reply to
In Reply To:
Just make the move from Gossamer Threads links SQL to the next version an easy one...
The upgrade from 2.x to 3.x should be the same as a minor 2.x to 2.(x+1) upgrade. Also, with Gossamer Update, future updates will be even easier.

In Reply To:
it took me forever to work out all those templates - can't imagine what will happen if I'll have to start over!
What problems did you have with the templates? We're looking of ways to improve the ease of template upgrades and any input on what problems you had would help.

Gossamer Threads Development Team
Quote Reply
Re: [BeaverheadRiver] Gossamer Threads Development Survey In reply to
Hi Rob,

Thanks for your feedback.

Our intention is to fulfill those points you laid out, witness the fact that we started this thread in the first place. We will be posting regular updates in this forum as work progresses, we'll try to provide bi-weekly updates, perhaps even shorter intervals should we accomplish more than expected.

We're making efforts to change our ways and you will see a difference in the communications forthcoming from Gossamer Threads in the next little while.

Projected timelines WILL be posted once we have a concreate idea of what we're dealing with in the push towards the new template format.

Gossamer Threads Development Team
Quote Reply
Re: [GT Dev Team] Gossamer Threads Development Survey In reply to
Hi there,

I also spent much time stripping down the standard templates into a set of "clean" templates which only contained the tags required by links.
I then rebuilt all of the templates using Dreamweaver and lots of globals.
Although this took alot of time, I would love Gossamer to push the products into using more CSS to clean the templates up. I would prefer to see a set of standard templates with the absolute bare minmum of html code, so it becomes easier for us to customise.

As others have noted, the feedback on development and timeline of Gossamer products is currently poor.
For instance we all waited 16 months for the upgrade from Linkssql 2.1.2 to Links 2.2.0
This makes me actively search for other similar products who have far more regular releases, with new features.
I'm thinking particularly of photopost php which in 18 months has added so many features. I would now never consider using linkssql with a mod as a photo gallery.

Sorry to all those who spend ages tinkering with the code. I for one would rather see a new release with new features that moved the whole product miles forward rather then having a release "hamstrung" by trying to keep all of the "problems" from earlier releases.

Gossamer should be striving to move forward rather then backwards. I see this debate as being a very positive step forward.

I have looked for other competitor products but still haven't found anything with the power of Linkssql, and such a friendly and positive supporting forum.

Best wishes and keep up the fantastic work.

Piers

p.s. will linksSQL 3.0 be released in 2005?!!!
Quote Reply
Re: [eljot] Gossamer Threads Development Survey In reply to
Hi,

where is my text?

Best regards from
Bremen/Germany

Lothar

Last edited by:

eljot: Nov 23, 2004, 3:13 PM
Quote Reply
Re: [eljot] Gossamer Threads Development Survey In reply to
Hi Lothar,

We're not sure, we did not remove it, please do post again.

Gossamer Threads Development Team
Quote Reply
Re: [BeaverheadRiver] Gossamer Threads Development Survey In reply to
Quote:
I can understand the reasoning for not providing timelines and expected release dates. On the other hand, I don't understand why GT would not want to provide progress reports.


Sounds like you want development roadmaps with milestones.

cdkrg

Able2Know :: Ajooja Directory
Quote Reply
Re: [GT Dev Team] Gossamer Threads Development Survey In reply to
GT Dev Team could you reply about my question on Perl 6?
Thanks
Max
The one with Mac OS X Server 10.4 :)
Quote Reply
Re: [maxpico] Gossamer Threads Development Survey In reply to
Hi Max,

In Reply To:
GT Dev Team could you reply about my question on Perl 6?
Thanks

Sorry for this, I thought I already had. Basically, our stance is that before we can make Perl 6 a requirement, it not only has to be finished and a final version released, but widespread acceptance of Perl 6 has to occur. Perl 5 code will still work with Parrot (via Ponie), so before we can move from a Perl 5 code base to a Perl 6 code base, we need to know that we aren't cutting of clients who don't have access to Perl 6 on their servers.

To give you a frame of reference for adoptation times, our current code base works with Perl 5.004_04 or later - that target was set about 4 years ago, when 5.004_04 was still encountered on servers from time to time. Today 5.004_04 is pretty much gone, and although 5.005_02 and _03 are encountered from time to time, they are somewhat rare (this knowledge is based mainly on the fact that from time to time 5.6+ code sneaks in, and it usually takes quite some time before we get a bug report about it). We tend to feel that it would be appropriate to lift that requirement to 5.005_02, but we haven't made a final decision on whether or not to require 5.005 for v3 yet.

For reference on release dates (you can get the full list from perldoc perlhist): 5.004_04 was released 1997-Oct-15, 5.005_02 was released 1998-Aug-08, 5.005_03 was released 1999-Mar-28, 5.6.0 - 2000-Mar-22, 5.6.1 - 2001-Apr-08. So, even ignoring the 5.6.0 release date (5.6.0 was admittedly quite buggy), we're still looking at a 3-5+ year upgrade cycle. That said, the Perl 6 upgrade may happen more quickly than that, seeing as Perl 6 code is entirely different.

The best we can give you for now is a "wait and see." If rewriting some of our modules in Perl 6 can significantly improve their performance we may consider maintaining both a Perl 5 and Perl 6 version - but this certainly won't be the common case.

Gossamer Threads Development Team
> > > >