Hey pugdog; what´s happen with your mod for searching only in cat and subcat written for the beta or 1.1.
Have you done it till now?
If not i will it try tomorrow.

Do you know something about the review-mod? I wont change to the new links, cause my system is perfect at the moment (for the small needs i have :-)

Have done this night another jumpd.cgi to count the clicks on detailed and see in, old and all. Its simple the same than the jump or rate;
just another track_table; -> build_update and some mods inside the build and all other cgis doing with the build_update.

Im not confident with counting hits and detailhits in one field. So you nothing, only the sum of both.
We have a system reviewing all links for ourself and so we could test, if users are ready to read first our description on the detail before jumping to the site or jumping immediately to it (so we could save time for an detailed-page and write our comments directly to a comment-field at the cat-page.

So i must sleep for a while; its 6:24 in the morning; it was a long and hard night, but im happy to fix some todos on my list :-)
Robert

Widgetz has already stated that he is holding off for the next release to complete the Review Mod.

Regards,

Eliot Lee

Im afraid of this; after the new version is out, nobody will do anything for the 1.11.
Cause of 1 Mio. Mods inside my links i would not update at all and so i ask for this.
As i understand Alex right, there would be an review build in.

Robert

Robert,

You've put a lot of effort into the 1.11 version, and you are using to LEARN perl. Alex also used Links SQL to learn mod_perl and other tricky features. He's learned SQL, MySQL, Oracle and others while working on the program. That is the way you learn computers and software -- especially.

Once the new version comes out, you make a new subdirectory (hidded, protected) and import your old data/database into it. You see how it works, what new features are there, and what you don't need to change. Perhaps you could make a simple change to your database (rename a table, add a field, etc) to do what Alex has added to the "base" or "core" release.

Once that happens, you can then add in the custom mods you want, and you'll have a lot of help! Once it's ready, you then shut down your server, move your directories over, and restart your server. The process can take anywhere from 1-2 minutes to a half hour depending on how well you've planned it. (I know, I do it all the time!!) Then, you are running the 2.0 (or whatever) version of Links.SQL

When I migrated from the 2.0 version to SQL, I did this, and ran the SQL version for over 2 months before I made it live. I had a major learning curve -- all this was new to me -- even perl programming, and especially mod_perl. SQL was an alien language I had heard about and avoided like the plague. Now I'm getting more and more comfortable with it all.

I've made some tough decisions -- I am no longer programming in Pascal, C, C++, Delphi, (X)BASIC, Visual or Windows anything. I'm doing Perl/SQL, dabling in PHP, Shell scripts, and just keeping up with C to do compiles where I have to.

Why? The web is designed for SCRIPTING languages. Webservers at their heart are scripting languages. Talking to the chip, or even the OS is becoming the realm of the "nerd" where it belongs! We are finally moving to the "4GL" and "hardware abstraction" and "platform independence" that has been talked about for years. It wasn't a "revolution" or any one product. It was an evolution, and a concerted effort of a lot of players and the success of the Internet itself.

The breakup of M$ whether it happens now (or not) will certainly have major impact. Already OPEN SOURCE is taking over, and companies are looking at the costs of upgrading to Windows2000 and seeing that there are MANY OTHER OPTIONS now, that don't lock them into proprietary formats -- and all the real "wiz kids" are using those other programs not M$. The kids coming up are not going to be able to afford to play on Windows 2000 systems -- but they all have access to Linux, BSD, etc. Even Solaris is now _free_ for developer and non-commercial/deployment use.

So, what does this mean? Once the next version of Links comes out, trying to stick with the 1.11 version will be like trying to use DOS in a Windows environment. Take what you have learned, and apply it to the NEXT GENERATION of the code. You will find that modifications you made in the old one you can make better and more useful in the new version (not only because the new version itself is better and "fixed up" but because YOU ARE BETTER and know more!).

Remember my posts on releasing the recommend-it mod I made. I was _really_ embarrased by the first code I made even though it worked. I just got it working, not knowing what I was doing. The revision was _much_ better, even though I made it as "educational" as I could (not using any tricks or shortcuts to help explain how to use the DBSQL.pm, etc.).

Anyway, DON'T even _think_ about not upgrading!!

Really. Once you see the new code I'm sure you'll want to, but don't feel left out by people (especially me) not wanting to deal with the 1.11 code any more. My reasons are time-related only. I just don't have the time to support older/off-versions when a newer version is out there.

You'll feel the same way too, once it arrives :)

Don't despair! SQL makes the data migration easy. OOP code and mod_perl means that the code itself is going to behave and not break any rules. Upgrades should be really easy now (especially with what was learned during the 3 betas upgrades from 1.0 to 1.11).

:)

http://www.postcards.com
FAQ: http://www.postcards.com/FAQ/LinkSQL/

You are absolutely right. BUT:
We have to do at least one new project in every month; cause we have planned a big network of different but related sites. For this we would license the next version for the next site(s), but the old one must stay as it is. There is no time to rebuild it. I have done all the needed mods from here or with help of forum for myself without the two mentioned in this thread. But it is not so important to have them there. Its more like a game to have all possibilities, like a collector for starpics :-)
So i have an intern links (in my intranet) with all ways links could go; i have thought about a screenshot to show it to you (counting hitsin, hitsinold, hitsinall, hitsout,hitsoutold, hitsoutall, detailhits, detailhitsold, detailhitsall and many other things), but probably its not so interesting for you.
But im really proud to have it :-) and so i would get the two mods somewhere, somehow, sometime.

Hope Alex would be ready with 2.0 at the end of the month; a real big project and the whole team for it wait only for me to do the heart of it :-(

I'm running a lot of sites as well, and what I do is "ripple" upgrades through them.

I've found it _much_ easier to take the time to slowly upgrade everything than to leave some sites older versions. The reason is you start to expect things to work a certain way. Trying to keep all the different "ways" in your head makes life much harder.

By putting all the really changable stuff in the templates and the HTML_Templates.pm file the real difference in all the sites is only that file, the /templates directory and the Links.pm file. (Also my jump.cgi since I've modified it so much varies a bit from site to site depending on whether it's an image or a links site.)

Also, I've found some upgrades are easier to make first to smaller, or certain sites, than to the "main" or busy sites. Once they are worked out on those sites, they move upwards.

To each their own, but don't plan to run different versions of Links (or any software) at this early stage! By the time you are done, you'll have a nightmare of code you'll have to try to support. One change to a business model or one site will end up causing weeks of frustration instead of a single file change.

I plan to set up the new code in a beta directory, get a site working, move my main site's development directory over to the new code, then upgrade each site from there. One at a time starting with a vanilla install. (I will probably deploy a new site based on the next generation code so I have a fully functional "fresh" install of a site to compare to.)

Everyone will do things differently, but this is how I plan to proceed. I've done it that way for each of the previous beta's and it's worked well.


BTW: by using the formula above, and always having a near-vanilla site that has the basic mods in it and my base template set I can set up a new Links site (including a starting import from DMOZ) in under 2 hours. That's on my server. ON someone else's server I'd probably estimate 4 hours just to be safe, but that's PDQ.

For example: I wanted to stress the code a little so I set up GIRLFRIENDS.COM to try to pick up some more advertising while the domain is up for sale. http://girlfriends.com That took about 8 hours to set up, since the import was fairly huge.

I set up PUGS.NET http://pugs.net to test out a new idea in the code I had, and that took me 2.5 hours, and I spent an hour working on the logo (which I think is pretty good for it's job!).

ORIGAMI.net http://origami.net took me under 2 hours (the import was really small) but I've been fiddling with moving categories and links around on that site.

LAPIDARYLINKS.COM (LapLinks.com)http://laplinks.com was set up to work on integrating different databases -- shopping, classifieds, catalog, yellow pages (store listings), encyclopaedia, etc into one site. I've stopped hacking on it until the next release since much of what I'm trying to do there will change. But, I've come up with some interesting ideas for tiered databses, that should allow site development without impinging on the Links license.

I use each of these sites to help answer certain types of questions, since each site was set up to do something a bit different (or to solve a different code problem). Each is using the same basic template set, and the same code skelleton, with 99% of the differences in the HTML_Templates.pm, Links.pm and /templates areas. The rest of the differences are tweaks to the admin or build routines to do something special (or not do something) specific to that site.




http://www.postcards.com
FAQ: http://www.postcards.com/FAQ/LinkSQL/

Hmmm, good ideas to manage sites; but i think it makes much more work for me than my way.
I have an internal links with all mods and description how to do; so when i would build a new links i must only do one step after the next from a clean copy. Maybe this is more work than copy one site to another, but its only a half day or so and thats ok. For me the most time is spended on design, pics (i really hate it, but for good pictures sometimes two hours are needed), the css and the conception before starting. Most projects are not so big, that i couldnt do them in a snap. Another big point is that i try to build real finished sites; so i wont upgrade or improve them. They all have one goal: making money or traffic (its the same for me :-); cause i have so many projects inside my desktop i will start only these ones, i have all things ready for them. Cause this, i cant start the next one; some new features from the next version are needed (if they will come); if not and i mean shurely the user-management, i will use two links, one for the data and one for the users (must talk to Alex for an discount price :-)
So everyone has a way to do. But for both we are shurely happy to have links; inst it so?!

Have just visited girlfriends.com. You would get rid of it? We plan something like this as one of the next projects, but with some more design and functions and for a lot of languages around europe. We ll see.
Why do you drop it?
Robert

BTW: Have you any code for the serch in cat only or the Review, i could use to improve?

Girlfriends is actually making some advertising revenue now. FWIW -- Bfree paid us less (for all their various "affiliate" programs) in 12 months, than LinkShare paid us in the last 3 weeks with about 1/1000th of the banner exposures per month. Has anyone else had this sort of problem with Bfree/reporting.net? I've finally convinced myself to _not_ use them for anything any more, no matter what.

The problem with Girlfriends for us, is that it's somewhat tangential to what we are currently doing. It's probably worth more to a person with a direct target on that market segment, than what our target lines are.

It's an absolutely prime domain name, _if_ you have the product/service for it. It's easy to brand, easy to remember, easy to fit into a 15 second radio spot, just plain easy.

I have not advertised the domain in over 3 years (it used to be a pin-up site -- Girlfriends/GirliePix brand images) on any search engines, banners, or anything. Traffic has been steady, and you can see by the search terms who/what the most of the visitors are looking for. I offered the domain to them at a very good price, but they weren't interested. (Yes, I'm being cagey on that.)

The domain could swing in any of 3 major markets, or even a combination -- Lesbian, Dating/Romance, and Adult. (or, if you already have a brand name, it can reenforce it tremendously). With our warm-fuzzies, and free greetings, the only market that really fits us is the Dating/Romance, but we already have several sites we are developing (or have developed) for that, that fit our current plans better.

We've chosen "CyberSexy" over "Girlfriends" for our greetings, so we've put GirlFriends on the auction block if someone thinks they could make more profitable use of it than we can.

I know what it's worth. In the adult market, with the right backing the domain could take in 2-3 million a month. Done right, you could even funnel people through it to the x-rated material, and _still_ get radio play and air time on the "Girlfriends.com" Portal.

As the father of two little girls, I don't have the time or the inclination to do it any more -- but that doesn't mean it can't be done. Working from home, I'm much happier when my kids tell their friends "Daddy runs POSTCARDS.COM" rather than anything else :)

Girlfriends.com is just an advertising vehicle for us at this time, but we are fully aware of it's value, and if the right offer doesn't come along, we probably will re-develop it into a Dating/Romance site and fold it into the "Postcards.com"/"PUGDOG Enterprises, Inc" Network.




http://www.postcards.com
FAQ: http://www.postcards.com/FAQ/LinkSQL/

i already released review.cgi 2.0 for links sql 1.1..

Jerry Su
email@jerrysu.com
http://www.jerrysu.com/

Where can this be found?
Robert