Is it just me...

[chris ashdown]

What about a complete rethink of the whole update process, all are done by downloading so no extra delivery costs to Actinic

I suggest the actinic picks the top 6 bugs and then the forum gives them a rating from one to six

Actinic then solves bugs one and two and releases a V0.0.X, then three and four etc

No need for new features of any kind after initial release, keep them for next version, No need for big multifix testing as small fixes easy to check and test and also keeps most users happy

Keep it simple, why release 20-90 bug fixes in one go it only leads to delays and new bugs that are harder to trace

[Goz]

Actinic News - - I should have remembered.

My excuse is it has only happened the once so far and I need repetitive things to stick in my brain. Must be getting old!

[cbarling]

Chris, the process around maintenance releases is something we think about a lot. When there aren't too many bugs around, it works pretty well. When there are lots, every process including our current one, suffers.

Pros of frequent small maintenance releases: can fix ultra critical problems very fast, each release represents lower risk; cons are that some people, especially web designers with multiple clients, don't want frequent releases

Pros of current system: can fairly comprehensively and end to end test the product (not possible for many small releases), less effort on the release cycle (documentation, QA, announcements etc); cons are that releases are bigger and riskier and we have to wait for fixes.

In practise we copmbine the apporoaches - we do put out a small emergency release if a new maintenance release has introduced a truly disastrous problem.

Chris

[parklifeclothes]

we do put out a small emergency release if a new maintenance release has introduced a truly disastrous problem.

Chris

So why has no fix been made for the empty snapshot problem that has been around for a considerable time and that I now suffer, personally I would quote this as a 'truly disastrous problem'

[Duncan Rounding]

...especially web designers with multiple clients, don't want frequent releases...

Very true, but primarily because the incremental versions can't coexist and all sites have to be upgraded which can get complex when you are supporting many clients on differing releases.

[pinbrook]

on the whole my vote goes for fewer but bigger releases to sort out bugs - of course i wish the bugs were diminishing rather than increasing, buggy software is a nightmare and such a waste of a developers time.

I'm all for small emergency releases for show stoppers, but these should then be integrated into the next big release.

We looked at v9 and felt that the upload was unacceptably slow for too many people. So we made the massive change to sort out this one problem, but the changes did introduce some additional issues.
Chris

The biggest disapointment for me is the compression which only works out of the box on a small subset of server configurations (it requires host intervention to make it work)- with no timescale for rolling it out to cover the majority of servers, not everyone has a friendly host

[leehack]

The smaller the releases the more duplicate work you do when testing etc. so I can see the point. It's hard to balance and get it right, you never know if you got it right until after each release i'd imagine.

I think things lean slightly towards too many things being done at once at the moment. I'd like to see maybe 1 or 2 more releases a year. We saw how fast the 903 image nuking was subsequently fixed and released within a week, that's the first time I have seen such urgency and from time to time this would be good to see. Snapshots could have been treated with the same urgency IMO, they are all we really have in that area and it's what anyone who comes to this forum is educated to do regularly.

I think a projected timetable would also be good, first 2 weeks in every quarter, expect an upgrade type affair. Knowing what's being fixed is also key. There is nothing more demoralising than waiting 2-3 months for the next release to find that the bugs you suffer will not be fixed.

[chris ashdown]

I can understand a few designers having problems keeping up to date, but why should the rest of us suffer for a few designers who cannot keep their records up to date

The snapshot and dmp problems are a massive problem two many just you don't get fixed

As already seen nobody is willing to test the new updates as it seems prity pointless waste of time, with little feedback and only partial tests, i.e. no MU. since about ver 7 the updates have been slow at coming and causing as many problems as they fix

It's basic common sense to fix a few at a time and consentrate on the major bugs, over the life of the version I doubt we would be talking more than a dozen updates of which the last half dozen would be very minor fixes

Pros of current system: can fairly comprehensively and end to end test the product (not possible for many small releases), less effort on the release cycle (documentation, QA, announcements etc); cons are that releases are bigger and riskier and we have to wait for fixes.

Thats a load of b-lls,testing only happens properly with a small number of fixes at a time and far easier for those willing to test and see the results, as with any complicated problem you break it down into managable sections, documentaion is easier as only a couple of things to write about and probably better explained as not overloading writter with 70 od items to write about

I myself although a support customer, will not be upgrading to V10 in the first year if ever as so pissed off with the larger problems not being sorted out as soon as possible and even more pissed off that new items are brought out to seemingly add to actinics coffers whilst necessities are put on slow burner

If I cannot just press a button and be sure I am getting a good snapshot just stinks and means I have to stay late to keep trying or just take a mdb home

[leehack]

I can understand a few designers having problems keeping up to date, but why should the rest of us suffer for a few designers who cannot keep their records up to date

You seem very against designers at the moment Chris, you don't seem to understand that a designer is just a front for numerous site owners like yourself, you are actually the minority, not the designer. It's not about us keeping up to date at all, it's about maintaining a system our end that can easily assist you the client - that's all.

Jonty recently completed your site, so one presumes you have a 903 site at the moment. Lets say for example that 904 is launched next week and Jonty upgrades so he is building in the latest version (as he should do ideally). 3 weeks down the line you want a little tweak done to the design and jonty will need a snapshot to do it for you. So his system is now 904, as are the sites he is currently building. Yours is 903 though. To work on yours he needs to upgrade it to 904, to send it back to you, you will then need to be on 904 also.

So the more times we have upgrades the more times this occurs. Are you happy to be forced to upgrade in this situation? Are you suggesting that designers should have one desktop for each actinic version that exists? maybe you think we should uninstall actinic for each build and install the version needed for each site?

I wish you'd drop this site owners vs designers bullshit, we are all after the same thing, a stable version of the software, work together FFS and stop creating barriers. Rest assured (on my son's life if you like) we as designers are shafted every bit as much as you are at the moment. I have spent as much time on nuked images, trashed snapshots, talking to actinic and suffering boat loads of DMP files as you will have on sync server and snapshot problems.

[Duncan Rounding]

Whilst I understand your gripe with MU I don't take kindly to the remark:

...why should the rest of us suffer for a few designers who cannot keep their records up to date...

- it's nothing to do with records and I don't believe designers are making others suffer.

There are arguments for and against large and small incremental updates and what they should or should not contain, whichever is the case it affects designers and all of their clients (store owners).

It's very difficult to justify the amount of time and effort that these upgrades can cause and it's made that much more difficult when you have some clients on one incremental release and others on a different one.

As you know v902 for example can't coexist on the same computer as v903. There are ways around this but they're not clean and tidy and often don't work well.

So it's nothing to do with designers keeping their records up to date, it's a constraint of Actinic. A frustration very similar to you have with MU.

Still friends Chris? I hope so.

[jont]

I really empathise with Chris as he is having a torrid time with things particularly the DMP warning he gets several times a day. Chris' main site was redesigned and the others were to be rolled out with the same concept dropped over the top. This has proven a nightmare with the MOTO getting borked each time. I am now almost having to redesign the others as if doing them from scratch which is a ludicrous state of play for both parties.

[chris ashdown]

Just to confirm I have no problems with designers and appreciate their efforts and skills on this forum( but sometimes they do appear to sound like they are talking for the majority and that may not actually be the case)

With the present system, I have no doubts that many users are very warey of updating software as their livelyhood depends on the sites working 24/7, but would be much happier to take the risk for smaller upgrades

I hope john's designs will be strong enough for me to do any updates myself as I have always done, but yes if there is a problem I will no doubt ask his advice or the forums

[jont]

If different patch versions could co-exist that would make life better for everyone - designers don't need to keep virtual (or spare) machines, site owners could install and test the upgrade is working before upgrading, Actinic can push out the patch for wild user testing knowing the previous versions can be let untouched. There would be a greater uptake in user testing if the latest release can sit happily alongside the others.

[jont]

I hope john's designs will be strong enough for me to do any updates myself as I have always done

Could be tempting fate ... but other than the multi line input hack in the order script file they should upgrade quite nicely ... that's just condemned your site. Sorry Chris

[leehack]

Chris rest assured as designers we understand and empathise with what you are going through, it's the same for most of us. I'm not aware of one designer not pulling their hair out also. Myself, Duncan and Jont probably have 100-150 actinic sites between us, we see warts and all I assure you. Lets not alienate each other, stick together we might be able to help each other out more.

I suffered DMP errors very badly but have found a way to avoid them mostly. Selectors in the design library, php layouts in the library and multiple site option or business setting changes at once caused a vast amount of mine. If you are using any of Norman's addons, get the new dpimagesupport file, Norman has addressed actinic's apparent problem with php.

The image nuking has been solved and the image paths have been fixed, so 904 should finally be a stable version. As for MU, i have no idea, i have one main client struggling as you are, he's at his wits end too, i just hope some long standing problems get solved in this version.